ServerView Deployment Manager V6.20 - Fujitsu Technology ...

User Guide - English 

ServerView Suite 

ServerView Deployment Manager 

V6.20 

Edition September 2012

Comments… Suggestions… Corrections… 

The User Documentation Department would like to know your opinion of 

this manual. Your feedback helps us optimize our documentation to suit 

your individual needs. 

Feel free to send us your comments by e-mail to 

manuals@ts.fujitsu.com. 

Certified documentation according to DIN EN 

ISO 9001:2008 

To ensure a consistently high quality standard and user-friendliness, this 

documentation was created to meet the regulations of a quality management 

system which complies with the requirements of the standard 

DIN EN ISO 9001:2008. 

cognitas. Gesellschaft für Technik-Dokumentation mbH 

www.cognitas.de 

Copyright and trademarks 

Copyright © 1998 - 2012 Fujitsu Technology Solutions. 

All rights reserved. 

Delivery subject to availability; right of technical modifications reserved. 

All hardware and software names used are trademarks of their respective 

manufacturers.

Contents 

Contents 3 

1 Introduction 13 

1.1 Supported systems 14 

1.2 Target groups and purpose of this manual 15 

1.3 Changes since the previous manual 16 

1.4 System requirements 18 

1.4.1 Deployment Manager packages 18 

1.4.2 Software packages on the deployment server 19 

1.4.3 Supported operating systems for image creation/cloning 19 

1.5 ServerView Suite link collection 22 

1.6 Documentation for the ServerView Suite 24 

1.7 Typographic conventions 24 

2 Deployment - Overview 27 

2.1 Supported server systems 28 

2.1.1 Generic server 28 

2.1.2 PRIMERGY server 29 

2.1.3 PRIMERGY blade server 30 

2.1.4 Detection of target servers 31 

2.2 Cloning process 32 

2.2.1 Reference installation 32 

2.2.2 Image creation 35 

2.2.3 Cloning 37 

2.3 Mass remote installation 41 

2.4 System architecture 44 

2.4.1 Hardware architecture 44 

2.4.2 Software architecture 46 

2.5 First step guide for using Deployment Manager 48 

3 Installing Deployment Manager 51 

3.1 Software-Requirements 52 

3.1.1 JBoss web server 52 

3.1.1.1 Launching and ports used 53 

3.1.1.2 Role-based user administration 55 

3.1.1.3 Managing certificates 55 

3.1.2 Installation of ServerView Installation Manager 56 

3.1.3 Installation of ServerView Operations Manager 56 

Deployment Manager 3

Contents 

3.1.4 Installation of ServerView agents 57 

3.1.5 Deployment server 57 

3.1.6 Network configuration 59 

3.1.6.1 LAN connections 60 

3.1.6.2 LAN connections (blade server) 62 

3.1.6.3 LAN connection topology and IGMP settings (Switch 

Blade) 63 

3.1.6.4 Multiple segment cloning 64 

3.1.6.5 Multiple LAN ports of target systems 66 

3.1.7 Creating an image repository 66 

3.2 Installation 67 

3.2.1 Installation via the ServerView Suite DVD 1 67 

3.2.1.1 Installing the Deployment Manager package 68 

3.2.1.2 Installing the Deployment Services package 78 

3.2.2 Installation via the Fujitsu Technology Solutions web server 83 

3.3 Update installation 84 

3.4 Uninstalling/Removing Deployment Manager 84 

4 Working with Deployment Manager 87 

4.1 Opening Deployment Manager 87 

4.2 Closing Deployment Manager 88 

4.3 Actions after starting the Deployment Manager front-end 88 

4.3.1 Role-based permissions on accessing Deployment Manager90 

4.4 Deployment Manager main window 91 

4.4.1 Servers view 92 

4.4.2 Repositories view 93 

4.4.3 Registered Boot Images view 93 

4.4.4 Tabs 93 

4.4.4.1 Servers (tab) 94 

4.4.4.2 Information (tab) 95 

4.4.4.3 Tasks (tab) 97 

4.4.4.4 Operations on the Tasks tab 99 

4.4.5 Icons 102 

4.4.6 Wizards 105 

4.4.6.1 General buttons 105 

4.5 Repositories 106 

4.5.1 Add Repository dialog box 108 

4.5.2 Adding a new folder 110 

4.5.3 Deleting a repository or folder 111 

4 Deployment Manager

Contents 

4.6 Registered Boot Images 111 

4.6.1 Register Boot Image wizard 113 

4.6.1.1 Image ID step (Register Boot Image wizard) 113 

4.6.1.2 Boot Image step (Register Boot Image wizard) 114 

4.6.1.3 Bootstrap Loader step (Register Boot Image wizard) 115 

4.6.2 Unregistering a boot image 117 

4.6.3 Unregistering a client 117 

4.7 Configuration before cloning/installation tasks 118 

4.7.1 Deployment Configuration wizard 118 

4.7.1.1 General Settings step (Deployment Configuration wizard) 

118 

4.7.1.2 LAN Ports step (Deployment Configuration wizard) 121 

4.7.1.3 DNS Search Suffixes step (Deployment Configuration 

wizard) 124 

4.7.1.4 Remote Management Ports step (Deployment Configuration 

wizard) 125 

4.7.1.5 Notes step (Deployment Configuration wizard) 129 

4.7.2 Export Deployment Configuration dialog box 130 

4.7.3 Import Deployment Configuration dialog box 131 

4.7.4 Deleting an exported deployment configuration 132 

4.8 Adding new servers 132 

4.8.1 New Server wizard 132 

4.8.1.1 General Settings step (New Server wizard) 133 

4.8.1.2 LAN Ports step (New Server wizard) 135 

4.8.1.3 DNS Search Suffixes step (New Server wizard) 138 

4.8.1.4 Remote Management Ports step (New Server wizard) 139 

4.8.1.5 Notes step (New Server wizard) 143 

4.8.1.6 Found MAC Addresses dialog box 144 

4.9 IP addresses 146 

4.9.1 IPv4 IP addresses 146 

4.9.1.1 Add IPv4 Address dialog box 146 

4.9.1.2 Edit IPv4 Address dialog box 147 

4.9.2 IPv6 IP addresses 148 

4.9.2.1 Add IPv6 Address dialog box 148 

4.9.2.2 Edit IPv6 Address dialog box 149 

4.10 Managing the power state of a server 150 

4.10.1 Manage Power State dialog box 150 

4.11 Collecting diagnostic information 151 


Contents 

5 Image creation 153 

5.1 File-System-Dependent image creation 154 

5.2 File-System-Independent image creation (raw image creation) 155 

5.3 Supported operating systems for image creation 156 

5.4 Image creation of a Windows reference system 158 

5.4.1 Manual rollback of changes done during system preparation159 

5.5 Windows 2008 and Windows 2012 reference systems 160 

5.5.1 Configuring the Windows firewall on the reference system 160 

5.5.2 Sysprep restrictions 162 

6 Mass Cloning 165 

6.1 Creating a cloning image 165 

6.1.1 Create Cloning Image wizard 166 

6.1.1.1 Task Name step (Create Cloning Image wizard) 166 

6.1.1.2 Deployment Server step (Create Cloning Image wizard) 

167 

6.1.1.3 Image Path and Name step (Create Cloning Image wizard) 

168 

6.1.1.4 Options step (Create Cloning Image wizard) 170 

6.1.1.5 Disks step (Create Cloning Image wizard) 172 

6.1.1.6 Bios Boot Type step (Create Cloning Image wizard) 178 

6.1.1.7 Scheduling step (Create Cloning Image wizard) 179 

6.1.1.8 Action after starting the "Create Cloning Image" task 182 

6.2 Cloning groups 185 

6.2.1 Adding a cloning group 185 

6.2.1.1 Group Name step (Add Cloning Group wizard) 185 

6.2.1.2 Cloning Image step (Add Cloning Group wizard) 186 

6.2.1.3 Group Members step (Add Cloning Group wizard) 188 

6.2.2 Copying a cloning group 189 

6.2.2.1 Group Name step (Copy Cloning Group wizard) 189 

6.2.2.2 Cloning Image step (Copy Cloning Group wizard) 190 

6.2.2.3 Group Members step (Copy Cloning Group wizard) 192 

6.2.3 Editing a cloning group 193 

6.2.3.1 Group Name step (Edit Cloning Group wizard) 193 

6.2.3.2 Cloning Image step (Edit Cloning Group wizard) 194 

6.2.3.3 Group Members step (Edit Cloning Group wizard) 196 

6.2.4 Deleting cloning groups 197 

6.3 Cloning an image 197 

6.3.1 Clone wizard 198 


Contents 

6.3.1.1 Task Name step (Clone wizard) 198 

6.3.1.2 Deployment Server step (Clone wizard) 199 

6.3.1.3 Disks step (Clone wizard) 199 

6.3.1.4 System Preparation step (Clone wizard) 201 

6.3.1.5 Settings step (Clone wizard) 204 

6.3.1.6 Post Deployment step (Clone wizard) 207 

6.3.1.7 Bios Boot Type step (Clone wizard) 211 

6.3.1.8 Scheduling step (Clone wizard) 212 

6.3.1.9 Action after starting the "Clone" task 215 

6.3.2 Clone with Image wizard 217 

6.3.2.1 Task Name step (Clone with Image wizard) 217 

6.3.2.2 Deployment Server step (Clone with Image wizard) 218 

6.3.2.3 Disk Image step (Clone with Image wizard) 218 

6.3.2.4 Disks step (Clone with Image wizard) 220 

6.3.2.5 System Preparation step (Clone with Image wizard) 222 

6.3.2.6 Settings step (Clone with Image wizard) 225 

6.3.2.7 Post Deployment step (Clone with Image wizard) 228 

6.3.2.8 Bios Boot Type step (Clone with Image wizard) 232 

6.3.2.9 Scheduling step (Clone with Image wizard) 233 

6.3.2.10 Action after starting the "Clone with Image" task 236 

6.4 Cloning/Installing Baseboard Management Controllers (BMCs) 238 

6.4.1 Displaying BMCs 239 

6.4.2 Changing the deployment configuration 240 

6.4.3 Actions after cloning/installation 242 

7 Mass Installation 243 

7.1 Installation Groups 243 

7.1.1 Adding an installation group 244 

7.1.1.1 Group Name step (Add Installation Group wizard) 244 

7.1.1.2 Group Members step (Add Installation Group wizard) 245 

7.1.1.3 Installation Configuration step (Add Installation Group 

wizard) 247 

7.1.2 Copying an installation group 248 

7.1.2.1 Group Name step (Copy Installation Group wizard) 248 

7.1.2.2 Group Members step (Copy Installation Group wizard) 249 

7.1.2.3 Installation Configuration step (Copy Installation Group 

wizard) 251 

7.1.3 Editing an installation group 252 

7.1.3.1 Group Name step (Edit Installation Group wizard) 252 


Contents 

7.1.3.2 Group Members step (Edit Installation Group wizard) 253 

7.1.3.3 Installation Configuration step (Edit Installation Group 

wizard) 255 

7.1.4 Deleting an installation group 256 

7.2 Installing servers 256 

7.2.1 Install wizard 257 

7.2.1.1 Task Name step (Install wizard) 257 

7.2.1.2 Installation Configuration step (Install wizard) 258 

7.2.1.3 Settings step (Install wizard) 258 

7.2.1.4 Bios Boot Type step (Install wizard) 261 

7.2.1.5 Scheduling step (Install wizard) 262 

7.2.1.6 Actions after starting the "Install" task 265 

7.3 Boot Groups 267 

7.3.1 Adding a boot group 267 

7.3.1.1 Group Name step (Add Boot Group wizard) 267 

7.3.1.2 Boot Image step (Add Boot Group wizard) 268 

7.3.1.3 Group Members step (Add Boot Group wizard) 269 

7.3.2 Copying a boot group 270 

7.3.2.1 Group Name step (Copy Boot Group wizard) 270 

7.3.2.2 Boot Image step (Copy Boot Group wizard) 271 

7.3.2.3 Group Members step (Copy Boot Group wizard) 272 

7.3.3 Editing a boot group 274 

7.3.3.1 Group Name step (Edit Boot Group wizard) 274 

7.3.3.2 Boot Image step (Edit Boot Group wizard) 275 

7.3.3.3 Group Members step (Edit Boot Group wizard) 276 

7.3.4 Deleting a boot group 277 

7.4 MDP booting of servers 278 

7.4.1 MDP Boot wizard 278 

7.4.1.1 Task Name step (MDP Boot wizard) 278 

7.4.1.2 MDP Application step (MDP Boot wizard) 279 

7.4.1.3 Settings step (MDP Boot wizard) 280 

7.4.1.4 Bios Boot Type step (MDP Boot wizard) 283 

7.4.1.5 Scheduling step (MDP Boot wizard) 284 

7.5 Generic booting of servers 287 

7.5.1 Generic Boot wizard 287 

7.5.1.1 Task Name step (Generic Boot wizard) 287 

7.5.1.2 Settings step (Generic Boot wizard) 288 

7.5.1.3 Bios Boot Type step (Generic Boot wizard) 291 


Contents 

7.5.1.4 Scheduling step (Generic Boot wizard) 292 

8 Crash Recovery 295 

8.1 Creating a snapshot image 295 

8.1.1 Create Snapshot Image Wizard 295 

8.1.1.1 Task Name step (Create Snapshot Image wizard) 296 

8.1.1.2 Deployment Server step (Create Snapshot Image wizard) 

297 

8.1.1.3 Image Path and Name step (Create Snapshot Image 

wizard) 297 

8.1.1.4 Options step (Create Snapshot Image wizard) 299 

8.1.1.5 Disks step (Create Snapshot Image wizard) 301 

8.1.1.6 Bios Boot Type step (Create Snapshot Image wizard) 304 

8.1.1.7 Scheduling step (Create Snapshot Image wizard) 305 

8.1.1.8 Action after starting the "Create Snapshot Image" task 308 

8.2 Restoring a snapshot image 309 

8.2.1 Restore Snapshot Image Wizard 309 

8.2.1.1 Task Name step (Restore Snapshot Image wizard) 309 

8.2.1.2 Deployment Server step (Restore Snapshot Image wizard) 

310 

8.2.1.3 Disk Image step (Restore Snapshot Image wizard) 311 

8.2.1.4 Disks step (Restore Snapshot Image wizard) 312 

8.2.1.5 System Preparation step (Restore Snapshot Image 

wizard) 314 

8.2.1.6 Settings step (Restore Snapshot Image wizard) 317 

8.2.1.7 Post Deployment step (Restore Snapshot Image wizard) 

320 

8.2.1.8 Bios Boot Type step (Restore Snapshot Image wizard) 321 

8.2.1.9 Scheduling step (Restore Snapshot Image wizard) 323 

8.2.1.10 Action after starting the "Restore Snapshot Image" 

task 325 

9 Messages 327 

9.1 Messages List dialog box 327 

10 Settings 329 

10.1 Deployment Server step (Settings wizard) 329 

10.1.1 Log On Properties dialog box 331 

10.2 Repositories step (Settings wizard) 332 


Contents 

10.3 Licenses Installed step (Settings wizard) 332 

10.3.1 Add New License Key dialog box 334 

10.4 Licenses Used step (Settings wizard) 335 

10.5 Global Options step (Settings wizard) 336 

11 High-Availability support 339 

11.1 Installation 340 

11.1.1 ServerView Operations Manager installation 340 

11.1.2 Deployment Manager installation 341 

11.1.3 Deployment Services installation 342 

11.2 Hints 345 

11.2.1 User Accounts 345 

11.2.2 Repositories 345 

11.2.3 Actions Needed When No Cluster Software is Available 345 

11.3 Failover scenarios 346 

11.3.1 Deployment Manager and Deployment Services Packages 

are Installed on Same Server 346 


are Installed on Different Servers 346 

12 Cloning deployment process 349 

12.1 Power control via remote control 353 

12.1.1 Management blade power control 353 

12.2 System preparation phase 354 

12.2.1 Manual preparation 355 

12.2.2 Automatic preparation as part of a DOS session 355 

12.2.3 Preparation based on WinPE 356 

12.2.4 Customer system preparation image 357 

12.3 Supported storage devices 360 

12.3.1 SCSI/IDE drives 360 

12.3.2 RAID devices 360 

12.3.3 FC and iSCSI Devices 361 

12.3.4 Partitioning and File System Formatting 361 

12.3.5 Multi-boot Operating System Partitioning 363 

12.3.6 Adding NDIS Driver to Deployment Manager 363 

12.4 Image creation 364 

12.5 Tag file handling 365 

12.5.1 Directory and tag file 365 

12.5.2 Creating a tag file on Windows systems 366 


Contents 

12.5.3 Creating a tag file on LINUX systems 366 

12.5.4 Removing the tag file 366 

12.6 PXE protocol 366 

12.7 LAN traffic and deployment methods 368 

12.7.1 Unicast and Multicast 368 

12.7.2 Switch, Hub and Bridge Configuration 369 

13 Appendix - network techniques 371 

13.1 MAC address handling 371 

13.2 PXE 371 

13.3 DHCP 373 

13.4 VLAN (Virtual Local Area Network) 373 

13.5 Example: Deployment over VLAN 375 



1 Introduction 

ServerView Deployment Manager (called simply Deployment Manager 

below) enables individual deployment, such as duplication and installation, of 

servers. To reduce data center administration costs, full parallel mass 

remote cloning and mass remote installation on a group of servers selected 

from the ServerView database are supported. An automated individualization 

process allows each server instance to be put into operation quickly. Automated 

and time-controlled installation or cloning of servers is supported using 

centrally held boot images in an own image repository database. 

Deployment Manager is a professional addition to the standard ServerView 

Suite. As part of this suite, Deployment Manager is integrated in the user 

interface of ServerView Operations Manager (called simply Operations Manager 

below). It provides a central overview of installed or additionally available 

servers, existing images and assignment of servers to deployment or 

installation groups. 

Server installation comprises a number of individual steps: 

l Hardware preparation, BIOS settings, RAID configuration, etc. 

l Installation of the operating system and parameterization (licensing, 

network configuration) 

l Application environment setup 

l User and service setup 

Deployment Manager and ServerView Installation Manager (called simply 

Installation Manager below) support these steps. Automated installation 

ensures that more servers can be put into error-free operation with minimum 

effort. 

Deployment Manager has the further advantage that servers can be quickly 

loaded with an operating system/application software image so that other 

tasks can be assigned to them. For example, by exchanging a hard disk 

image, you can allow a server that previously ran under Linux to perform a different 

task in Windows. 



1.1 Supported systems 

Deployment Manager provides generic methods to enable system preparation 

before cloning for all kinds of IA32-based servers. Deployment Manager 

is also certified especially for the following PRIMERGY servers. 

l BX600 Blade Server 

l Blade BX620 S4, BX620 S5, BX620 S6 

l Blade BX630 S2 



l Blade BX920 S1, BX920 S2, configurable with Storage Blade SX940 

S1 and SX960 S1 

l Blade BX920 S3, configurable with Storage Blade SX960 S1 and 

SX980 S1 

l Blade BX922 S2, BX924 S2, BX924 S3, BX960 S1 

l PRIMEQUEST 1400, PRIMEQUEST 1800 

l CX122 S1 

l CX400 S1 Multi-Node server system configurable with CX210 S1, 

CX250 S1 and CX270 S1 

l Econel 100 S2, Econel 200 S2 

l RX100 S3, RX100 S4, RX100 S5, RX100 S6, RX100 S7, RX100 S7p 

l RX200 S4, RX200 S5, RX200 S6, RX200 S7 

l RX300 S4, RX300 S5, RX300 S6, RX300 S6 (Nebs), RX300 S7 

l RX350 S7 

l RX500 S7 

l RX600 S4, RX600 S5, RX600 S6 

l RX800 S3 

l RX900 S1, RX900 S2 


l TX100 S1, TX100 S2, TX100 S3, TX100 S3p 

l TX120, TX120 S2, TX120 S3, TX120 S3p 

l TX140 S1, TX140 S1p 

l TX150 S6, TX150 S7, TX150 S8 

l TX200 S4, TX200 S5, TX200 S6, TX200 S7 

l TX300 S4, TX300 S5, TX300 S6, TX300 S7 

l TX600 S3 

1.2 Target groups and purpose of this manual 

Any last-minute changes/corrections of the supported systems can be found 

in the Deployment Manager data sheet 

(http://ts.fujitsu.com/products/standard_servers/system_management/svs_ 

deploy.html). 

1.2 Target groups and purpose of this manual 

This manual is intended for system administrators, network administrators 

and service technicians who have a basic knowledge of hardware and software. 

The manual explains how to create and manage image files with the Deployment 

Manager software and how to deploy these files. It also describes how 

to install PRIMERGY servers using Installation Manager configuration files 

(if Installation Manager is installed on the same system as Deployment Manager). 



1.3 Changes since the previous manual 

This edition of the manual applies to ServerView Deployment Manager V6.20 

and replaces the following online manual: “ServerView Suite Deployment 

Manager V6.10”, edition March 2012. 

This manual has been updated to reflect the latest software status and offers 

the following new features: 

l The list of supported systems has been updated, see "Supported systems" 

on page 14. 

l The Deployment Manager packages Deployment Manager and 

Deployment Services can be installed under Microsoft Windows 

Server 2012 Datacenter/Standard/Foundation, Windows Storage 

Server 2012 Standard. Microsoft Windows Server 2003 R2 Standard/Enterprise 

(x86, x64) is no longer supported. 

l The software package Microsoft Windows Server 2008 or Windows 

Server 2012 must be available on the deployment server. 

l Microsoft Windows Server 2012 Datacenter/Standard/Foundation, 

Windows Storage Server 2012 Standard is supported for image creation/cloning 

with personalization. 

l The preparation method Generic Boot Image has been changed to 

Preparation Boot Image. 

l The preparation boot image is provided with the Deployment Manager 

software in the Sample_PreparationBootImages directory. 

l The section "JBoss web server " on page 52 has been updated. 

l New: Servers by Boot Image group in the tree view of the Servers 

view, see "Servers view " on page 92. 

l You can now also perform the operation Unregistering all clients at 

the PXE service on the Tasks tab, see "Operations on the Tasks tab" 

on page 99. 


l In the Add Repository dialog box you can also add repositories for 

Generic Boot Images and MDP Applications, see "Add Repository 

dialog box" on page 108. 

l New: Registered Boot Images view, which displays a list of all registered 

boot images, see "Registered Boot Images" on page 111. 

This view allows you to 

1.3 Changes since the previous manual 

o register a boot image via the Register Boot Image wizard 

(see "Register Boot Image wizard" on page 113), 

o unregister a boot image (see "Unregistering a boot image" on 

page 117, 

o unregister a client from the PXE service (see "Unregistering a 

client" on page 117). 

l New: You can now add, copy or edit a boot group via the context 

menu in the Server by Boot Image group, see "Boot Groups" on 

page 267. 

l New: MDP booting of servers. Via the MDP Boot wizard, you can 

execute an MDP application on one or more servers, see "MDP Boot 

wizard" on page 278. 

l New: Generic booting of servers. Via the Generic Boot wizard, you 

can boot one or more servers with a generic image, see "Generic Boot 

wizard" on page 287. 



1.4 System requirements 

Please check the system requirements before you start installing Deployment 

Manager. 

1.4.1 Deployment Manager packages 

Deployment Manager consists of two packages which can be installed on 

the same system or on different systems, see also section "Installing Deployment 

Manager" on page 51. 

Deployment Manager package 

This package consists of the Web-based graphical user interface including 

the ServerView Deployment Manager service. 

This package can be installed under: 

l Microsoft Windows Server 2008 Standard/Enterprise/Datacenter/Foundation 

(x86, x64) 

l Microsoft Windows Server 2008 R2 Standard/Enterprise/Datacenter/Foundation 

(x64) 


Windows Storage Server 2012 Standard 

Deployment Services package 

This package consists of the deployment services including PXE service, 

TFTP service and deployment service (= cloning module). 

This package can be installed under: 


(x86, x64) 

l Microsoft Windows Server 2008 R2 Standard/Enterprise/Datacenter/Foundation 

(x64) 




Windows Script Host (WSH) as of version 5.6 is required for the 

deployment services. 

1.4.2 Software packages on the deployment server 

The following software packages must be available on the deployment 

server: 

l Microsoft Windows Server 2008 or Windows Server 2012. 

l Installation Manager as of version 7.804 for PXE-based remote installation. 

l Installation Manager as of version 6.711 and deployment service (cloning 

module) for PXE-based remote image creation and cloning. 

l A DHCP service (if none is already present in the LAN segment). 

l A PXE service must run on the deployment server. It can be installed 

using either Deployment Manager or Installation Manager. 

Installing the deployment server to PRIMEQUEST is not supported. 

PRIMEQUEST is supported only as a target server. 

Older Installation Manager versions (up to V10.10.09) can only be 

installed with the IIS Web Server. The ServerView Apache Web 

Server (that is delivered with these older Installation Manager versions) 

and the JBoss Application Server both listen on ports 3169 

and 3170 and therefore cannot coexist. 

1.4.3 Supported operating systems for image creation/cloning 

Deployment Manager supports the following operating systems for image creation/cloning 

with personalization: 

In Japan, Linux SuSE is not supported. 




l Microsoft Windows Server 2003 Standard/Enterprise (x86 and x64) 

l Microsoft Windows Server 2003 R2 Standard/Enterprise (x86 and 

x64) 


(x86 and x64) 

l Microsoft Windows Server 2008 R2 Standard/Enterprise/Datacenter/Foundation/Web 

(x64) 

Mass Cloning of Windows Domain Controllers (e.g. 

Windows Small Business Server) is not supported. 



l Resilient file system is not supported. 

l Storage Spaces and Virtual Disks created 

from Storage Spaces are not supported. 

l Before starting an image creation process for 

a Windows 2008 or Windows 2012 reference 

system, you must configure the Windows firewall 

on the reference system, see "Configuring 

the Windows firewall on the reference 

system" on page 160. 

l In cloning with personalization, only "full" Windows 

2008 or Windows 2012 installations are 

supported. Saving an image of a Windows 

2008 or Windows 2012 "core" installation is 

not supported. 

l Linux RedHat (AS/ES) 3.0 - not supported in Japan 

l Linux RedHat (AS/ES) 4.0 (x86 and EM64T) - not supported in Japan 

l Linux RedHat 5.0 (x86 and EM64T) 



Before image creation, a specific package must be 

installed on the reference system. This package is 

provided with the Deployment Manager software in 

the directory LINUX_support\RHEL6_Cloning_ 

Support. 

Before image creation, the NetworkManager should 

be disabled - otherwise the personalization may fail 

during cloning. Please set NM_CONTROLLED=no 

in the /etc/sysconfig/network-scripts/ifcfg-eth* 

files and issue the following commands: 

/sbin/service NetworkManager stop 

/sbin/chkconfig --del NetworkManager 

/sbin/chkconfig --add network 

/sbin/service network start 


l Linux SuSE Enterprise Server SLES 8 - not supported in Japan 

l Linux SuSE SLES 9 (x86 and EM64T) - not supported in Japan 

l Linux SuSE United Linux (with SLES 8) - not supported in Japan 

l Linux SuSE Enterprise Server SLES 10 (x86 and EM64T) - not supported 

in Japan 


in Japan 

l Reiser file system versions 3.5 and 3.6 on SuSE SLES 9 and SuSE 

SLES 10 

Before image creation, a specific package must be installed on the reference 

system if Reiser is used as the root file system. This package 

is provided with the Deployment Manager software in the directory 

LINUX_support\ReiserFS_Cloning_Support. 

l All other operating system types can always be supported without 

personalization in raw snapshot mode. See also the Read- 

Me file. 

l Creating a cloning image of a disk with multiple operating systems 

is not supported. Creating a snapshot image is supported. 



For restrictions see section "Supported operating systems for image creation" 

on page 156. 

1.5 ServerView Suite link collection 

Via the link collection, Fujitsu Technology Solutions provides you with numerous 

downloads and further information on the ServerView Suite and PRIM- 

ERGY servers. 

For ServerView Suite, links are offered on the following topics: 

l Forum 

l Service Desk 

l Manuals 

l Product information 

l Security information 

l Software downloads 

l Training 

The downloads include the following: 

o Current software statuses for the ServerView Suite as well as 

additional Readme files. 

o Information files and update sets for system software components 

(BIOS, firmware, drivers, ServerView agents and 

ServerView update agents) for updating the PRIMERGY 

servers via ServerView Update Manager or for locally updating 

individual servers via ServerView Update Manager Express. 

o The current versions of all documentation on the ServerView 

Suite. 

You can retrieve the downloads free of charge from the Fujitsu Technology 

Solutions Web server. 

For PRIMERGY servers, links are offered on the following topics: 


l Service Desk 

l Manuals 

l Product information 

l Spare parts catalogue 

Access to the link collection 

You can reach the link collection of the ServerView Suite in various ways: 

1. Via ServerView Operations Manager. 

l Select Help – Links on the start page or on the menu bar. 

This opens the start page of the ServerView link collection. 

2. Via the ServerView Suite DVD 2 or via the start page of the online documentation 

for the ServerView Suite on the Fujitsu Technology Solutions 

manual server. 

You access the start page of the online documentation via the 

following link: 

http://manuals.ts.fujitsu.com 

l In the selection list on the left, select Industry standard servers. 

l Click the menu item PRIMERGY ServerView Links. 


3. Via the ServerView Suite DVD 1. 

l In the start window of the ServerView Suite DVD 1, select the 

option Select ServerView Software Products. 

l Click Start. This takes you to the page with the software products 

of the ServerView Suite. 

l On the menu bar select Links. 

1.5 ServerView Suite link collection 




1.6 Documentation for the ServerView Suite 

The documentation for the ServerView Suite can be found on the ServerView 

Suite DVD 2 supplied with each server system. 

The documentation can also be downloaded free of charge from the Internet. 

You will find the online documentation at http://manuals.ts.fujitsu.com under 

the link Industry standard servers. 

1.7 Typographic conventions 

The following typographic conventions are used: 

Convention Explanation 

Indicates various types of risk, namely health risks, risk of 

data loss and risk of damage to devices. 

Indicates additional relevant information and tips. 

bold Indicates references to names of interface elements. 

monospace Indicates system output and system elements, e.g., file 

names and paths. 

monospace 

semibold 

blue continuous 

text 

pink continuous 

text 

Indicates statements that are to be entered using the keyboard. 

Indicates a link to a related topic. 

Indicates a link to a location you have already visited. 

Indicates variables which must be replaced with real values. 

[abc] Indicates options that can be specified (syntax). 

[key] Indicates a key on your keyboard. If you need to enter text in 

uppercase, the Shift key is specified, for example, [SHIFT] + 

[A] for A. If you need to press two keys at the same time, this 

is indicated by a plus sign between the two key symbols. 


Screenshots 

1.7 Typographic conventions 

Some of the screenshots are system-dependent, so some of the details 

shown may differ from your system. There may also be system-specific differences 

in menu options and commands. 



2 Deployment - Overview 

Today’s standard server environment has significantly changed the requirements 

of server administration. In a typical IT environment a large number of 

servers are used and must be managed remotely and without consoles. 

There are server farms, for example, where hundreds of blade servers are 

installed and must be administered. 

Scalability and ease of deployment or support for different tasks must be 

offered to fulfill the increasing demands on flexibility in IT systems. While a 

system may be used as a Web server for the Internet presence of a company 

during the day, it could be used as an FTP server at night with a different IP 

address. This must be achieved simply by remote administration rather than 

through physical operation. Administrators often have to set up the same software 

stack, such as the operating system and application software, on hundreds 

of identical servers in a server farm. In this case, manual installation is 

not feasible. 

New strategies must be developed to deploy the boot image of a server over 

the LAN without any interaction. The basic idea of deployment is to create 

one or more central deployment servers to administer this enormous number 

of clients over the LAN. To manage the servers on the fly, special base management 

controllers (such as management blades for blade servers, or Kalypso 

service processors) are assigned inside of a server chassis to guide the 

servers during their boot and working phases. 

Deployment is the convenient method of preparing server systems to start 

their work immediately. Deployment servers are responsible for preparing 

multiple servers and their environment from a central instance over the LAN. 

The servers can therefore start their work immediately. 

Deployment Manager provides two procedures for deployment: 

l Cloning 

Cloning means to install a reference system, create an image file of 

that reference installation, and clone this image file to a group of 

servers to be installed with the same configuration parameters. This 

procedure is described in more detail in the section "Cloning process" 

on page 32. 



l Mass remote installation 

Mass remote installation means to use configuration files created with 

Installation Manager, assign them to servers or server groups, and 

then start the installation for all servers. This procedure is described in 

more detail in the section "Mass remote installation" on page 41. 

2.1 Supported server systems 

Deployment Manager supports the following types of server: 

l Generic IA32-based server 

l PRIMERGY server 

l PRIMERGY blade server 

l PRIMEQUEST server (only as target server) 

2.1.1 Generic server 

From the deployment point of view, a generic server is a server that has been 

ideally prepared, so that each device access can be handled via a standardized 

generic device or BIOS APIs. 

Deployment Manager provides generic methods to support almost any kind 

of IA32-based server system. Cloning with Deployment Manager requires 

the following standard APIs: 

l PXE boot capability as part of the BIOS code based on PXE specification 

V2.1 

l A standard onboard NIC with full UNDI API support as required for 

PXE specification V1.2 

l A fully prepared bootable storage device accessible via DOS Int13h 

BIOS API 

For a generic server Deployment Manager expects a fully prepared storage 

device, e.g. with complete RAID configuration. This must be done by the 

user manually or by booting a PXE image in floppy emulation. This image can 

be a raw copy of a bootable DOS floppy providing the full knowledge to perform 

an unattended system preparation of a target server. The requirements 


for this Deployment Manager preparation image are similar to the MS ADS 

(Automated Deployment Services) required preparation image. For the creation 

of MS ADS compatible PXE preparation images, each system vendor 

offering ADS-compatible server systems provides a Web page containing a 

prescription how to prepare such a system preparation image for each server 

type together with the required command line tools. 

Finally, each image must initiate a reboot of its target server automatically in 

PXE mode (enabled static in the BIOS setting) to continue the Deployment 

Manager generic cloning phase. These kinds of images can be administered 

via the Deployment Manager repository management by a separate preparation 

boot image repository. 

An example of such a preparation boot image is provided with the Deployment 

Manager software in the Sample_PreparationBootImages directory. 

See also the ReadMe file in this directory for further information. 

Once a server fulfills the generic server requirements as described above, it 

can be cloned by Deployment Manager with the system preparation method 

Preparation Boot Image. This type is seen as a common classification for 

all types of server prepared in the appropriate way in the system preparation 

phase, either by Deployment Manager or by the users themselves. 

It is not necessary for a generic server to support remote management 

besides a static PXE boot mode, as configured once in the BIOS settings. 

PXE boots are initiated manually or just by a simple reboot driven by the 

locally running preparation image. A remote power on of a generic server can 

be done by "Wake On LAN" functionality integrated into Deployment Manager 

if the server supports "Wake On LAN". 

Remote identification of a generic server is not possible. The initial MAC 

address must be added manually. 

2.1.2 PRIMERGY server 


PRIMERGY servers are developed by Fujitsu Technology Solutions and can 

be supported in a much more detailed and convenient way than generic 

servers. 



PRIMERGY servers are equipped with: 

l Onboard or plugged-in service processor for remote management 

such as 

o automatic identification of MAC address by Operations Manager 

o automatic remote PXE boot 

o automatic modification of remote boot device table 

l Remote management via SNMP or IPMI (Intelligent Platform Management 

Interface (Kalypso- or Kronos-supported)) 

l High-level system preparation, e.g. RAID controller and attached storage 

devices 

l Fully supported by Fujitsu Technology Solutions service 

Based on these services, Deployment Manager can directly check whether 

or not a cloning image is compatible with a given PRIMERGY server. After 

cloning, this server can be managed directly by Operations Manager and 

Remote Management. 

PRIMERGY servers have at least 512 MB RAM and an onboard 100 Mbit 

LAN controller with one or more ports. The storage device may be a simple 

IDE or SCSI drive, an onboard RAID controller based on IDE, SATA, SCSI, 

or SAS and/or a SAN-attached storage device with an additional administration 

LAN port. These systems may exist in a floor-standing or rack housing 

with or without a local console. 

2.1.3 PRIMERGY blade server 

A blade server system is a rack system with server plug-in cards of up to 18 

(BX900), 9 (BX400) or 10 (BX600) server blades. Each server blade consists 

of one or two CPUs, one or two hard disks (IDE, IDE-RAID, SATA, SATA- 

RAID, SCSI, SCSI-RAID, SAS, SAS-RAID), at least 512 MB memory, at 

least 2 x 1Gbit LAN ports, and the system chip set. 

To manage these server blades remotely both online and offline, two management 

blades are assigned inside of these racks with a local IPMI-based 

communication path to each server blade at each significant point in its life 

cycle, e.g. power off, BIOS boot phase, operating system boot phase. The 


hard disk may be controlled by a SCSI/IDE or IDE/SCSI RAID controller. 

Management blades have their own type of hardware to administer and control 

the server blades in respect of power, temperature, BIOS and so on. 

To provide fast and stable connectivity to the server blades over the LAN, 

two or four switch blades are available in the same rack. 1 Gbit Ethernet channels 

or Fibre Channel connections are used. These consist of real 1 Gbit or 2 

Gbit switch hardware with external connectors to the external LAN, and are 

hard-wired with internal port connections to each LAN chip port of each 

server blade in a chassis. Two LAN ports per blade are connected to these 

switch blades for redundancy purposes. Each LAN port has its own MAC 

address. It is important that the PXE BIOS uses port 0 by default to correctly 

handle the PXE protocol. In the CPU blade BIOS, you can define which port 

is used as the default PXE boot port. 

A blade server can have a VGA connector and two USB 2.0 connectors for a 

USB floppy or USB CD-ROM drive for service purposes only. This is useful 

for native operating system installations on a reference server. 

The racks may be assembled into a rack tower to concentrate the number of 

CPU blades per square meter. A server farm may consist of many of these 

rack towers. 

2.1.4 Detection of target servers 


Deployment Manager obtains the list of server systems from Operations Manager 

front-end by reading the ServerView database. 

Deployment Manager also offers an Add Server dialog in the New Server 

wizard to create server list entries for servers which are initially not detected 

by Operations Manager. In this case the MAC address, IP address and host 

name must be added manually. These servers are called bare servers. If a 

server is deployed and a ServerView agent is active and running on the 

clone, it will automatically be detected by the discovery cycle running during 

the lifetime of the Operations Manager front-end. Operations Manager will 

take over responsibility for that server entry and update all parameters 

received from the agent (including LAN port settings). 



2.2 Cloning process 

The cloning process consists of the following steps: 

l installation of a reference system 

l image creation from that reference system 

l cloning of an image to one or more target server systems 

Figure 1: Cloning deployment cycle 

The following sections describe these phases in more detail. 

2.2.1 Reference installation 

First you must install and configure a server system (the reference system) 

manually. This so-called reference installation on the reference system is 

used to install further servers in the same way. For each kind of server to be 

installed via Deployment Manager, there must be one reference system. 


Figure 2: Deployment cycle - reference installation 


A reference system is basically a simple server of a specific type. It may 

have additional devices attached, such as a monitor, mouse and keyboard, 

mass storage devices (e.g. RAID), USB CD-ROM and USB floppy drive in 

the case of local operating system installation via CD and floppy. 

The hardware configuration must be detected by several agents and the system 

resources must be assigned by the BIOS and operating system 

instances. A large number of parameters of BIOS, storage systems and operating 

system must be defined previously and interactively in a separate configuration 

session or interactively in parallel with the installation process. 

Detailed information on the installation process of the individual reference 

systems can be found in the corresponding installation documentation. 

Installed applications on the reference system possibly may not run successfully 

on a target system after cloning without adaptation. In that case 

user-specific scripts are to be performed after the cloning process to ensure 

the successful operation. 



After each basic operating system installation, you may install and configure 

appropriate applications which will be started on each cloned server immediately 

after reboot. Image creation software which is part of the deployment 

server software generates a copy of this installed hard disk as an image file 

and copies it to the deployment server repository. It is used as a reference 

(master) image for further cloning sessions. The server management frontend 

of the deployment software allows you to define assignments between 

clients and reference images. 

Many parameters of the BIOS, storage systems (RAID) and operating systems 

can be defined interactively in a separate configuration session of Installation 

Manager before the installation, or interactively in parallel with the 

installation process using native operating system installation via local 

devices. You can use this installation as a reference installation for other 

servers with the same hardware and BIOS configuration. The other systems 

may differ in memory size, hard disk size and RAID level. 

The following methods of reference installation are supported: 

l Local installation with Installation Manager 

l Multiple remote installation with Installation Manager (maximum five 

clients) 

A detailed description of Installation Manager can be found in the Installation 

Manager user guide which is available on the ServerView Suite DVD 2. 

Local installation with Installation Manager 

Local installation with Installation Manager requires a bootable USB DVD- 

ROM and a graphical console with mouse, monitor and keyboard connected 

to the reference server for the master image generation. For blade servers the 

graphical console with mouse and keyboard is available using the KVM 

switch functionality via the management blade. 

With this local installation method you can use the Guided Mode of Installation 

Manager which provides an online check of the defined configuration 

parameters with the target hardware. 

Multiple remote installation with Installation Manager and PXE 

You can use Installation Manager Remote Installation which you will find on 

the Installation Manager CD/DVD version 6.605 (or higher) or on the 


ServerView Suite DVD 1 as of version 7.810. To do this, you must install the 

Installation Manager CD/DVD or the Installation Manager from the Server- 

View Suite DVD 1 on the PC from which you want to perform the installation. 

This PC acts as the deployment server for the remote installation. It can be a 

notebook or a server. 

Installation Manager uses WinPE as the boot platform and is booted via PXE 

in the RAM of each target server. This allows any kind of remote access during 

the installation process. 

Multiple remote installation is recommended for installing up to five clients. 

Detailed information on the reference installation with Installation Manager 

can be found in the Installation Manager user guide. 

2.2.2 Image creation 


From the reference system an image file is created which contains the configuration 

information on the reference system, e.g. on the hard disk, the configuration 

of the partitions, and all data of the operating system. It is independent 

of the RAID level, the BIOS settings and the memory size of the 

target systems. 



Figure 3: Deployment cycle - master image creation 

The following methods of image creation are supported: 

l File-system-dependent image creation 

This method is only available for supported file systems which only 

come with the supported operating systems (see section "Partitioning 

and File System Formatting" on page 361). In this case only used 

areas are copied to the image. This results in small image sizes and 

reduced cloning time. The target partition may be smaller than the partition 

on the reference system. 

l File-system-independent image creation (raw image creation) 

This method is used for unsupported file systems. In this case the 

complete partition information is copied to the image file and must be 

cloned to a partition with the same size. The image may be larger and 

the cloning process takes more time to recreate the partition. 

l Snapshot image 

A “snapshot image” can be used to restore the server after a system 

crash. 


The images are stored in an image repository. This is located in a shared 

folder on any file server in the network. The image is used as a reference 

image for further cloning sessions. Each image object in the repository refers 

to a set of files with identical names but different extensions: Image documentation 

(e.g. reference to a text document which is created during the 

image creation) could be used for describing an image in a Deployment Manager 

frame. 

*.img Reference to binary cloning image file (URL of a storage location 

somewhere in the network) 

*.txt Image documentation (e.g. reference to a text document which 

is created during the image creation) could be used for describing 

an image in a Deployment Manager frame. 

*.cfg A hardware and operating system parameter file. This information 

is used for later modification of an existing image and for 

a compatibility check against the assigned target hardware. 

The image creation process can be started from Deployment Manager. 

Detailed information on image creation can be found in the chapter "Image 

creation" on page 153. 

How to use Deployment Manager is described in the chapter "Working with 

Deployment Manager" on page 87. 

2.2.3 Cloning 


The cloning process deploys an image to the servers. It prepares the hardware, 

BIOS and storage devices, copies the image onto the hard disks, and 

initiates a reboot of the target server. No configuration session is required during 

the cloning process or later. 

The result is an identical system with the exception of variable parameters 

such as IP address, host name and SecureID (Windows systems only) of 

the operating system, which must be unique to each server. 



Figure 4: Deployment cycle - cloning 

The hardware configurations of clone and reference system must be identical, 

except for the RAID level, memory size, CPU clock and hard disk 

capacity. 

It is possible to define assignments between reference images and servers 

by setting up logical groups, called deployment groups, in the Deployment 

Manager front-end. Cloning images can be assigned to deployment groups 

and can be cloned simultaneously. 

Deployment groups must be created manually. Each group has its own set of 

group attributes (parameters) relevant for all group members, e.g. 

l Network path to master image in the Deployment Manager repository 

l File system formatting type 

l Final power status after post-cloning phase has finished 


All physical clients of that deployment group will be cloned with that referring 

image, and the target storage device will be configured in the specified way 

for each server. 

The attributes for a member of a physical group are basically defined in the 

deployment table, which contains an entry for each server listed in the physical 

server list. This deployment table is always the definitive reference for 

the current deployment status of the server indicating the last operation performed 

by Deployment Manager for that server or the currently running operation 

(e.g. cloned or installed). Deployment Manager always uses the 

parameters from this table. This helps to show the same status of each 

server if more than one Deployment Manager front-end has been started. If 

one front-end has initiated a job on a particular server, other front-ends are not 

able to access this server as long as the jobs are running. Each group 

member may have its own set of group parameters which may overrule the 

group settings for this server. Finally, each physical server, regardless of 

whether it is a member of a group or not, has its own set of individual physical 

settings which are stored server-specifically in the deployment table. 

A cloning session consists of several phases: 

1. The target server is powered on. 

2. The system is prepared. 

3. The image is cloned to the server. 

4. The post-preparation tasks are performed. 

Each phase offers different alternatives depending on the types of servers to 

be cloned. Details can be found in chapter "Cloning deployment process " on 

page 349. 

The cloning process with Deployment Manager works as follows: 

Preparations 


l The administrator defines a deployment group, assigns an appropriate 

image file, and adds the servers to the deployment group for deployment 

on the basis of the assigned image file. 



l The deployment table is set up with the individual parameters for the 

servers: IP settings, host name etc. For each server in the group, a 

separate entry is created in the deployment table. 

Starting the cloning process 

l The administrator defines the cloning parameters and selects the 

servers for the deployment job. 

l He/she initiates the cloning directly or configures the scheduler once 

per server in the deployment group. 

Cloning process 

l Deployment Manager prepares the system based on the preparation 

mode selected by the administrator: 

o unchanged: 

For already manually prepared target servers. No additional 

preparation tasks are performed. 

o Preparation Boot Image: 

For target systems to be booted via a preparation boot image 

created by the users themselves. The operations are performed 

via PXE boot from a 1.44 MB floppy image or an external 

bootstrap image. (See also "Generic server" on page 28.) 

o All Primergy: 

For target servers based on WinPE PXE boot (e.g. PRIM- 

ERGY servers). The RAID configuration is done via Installation 


l Deployment Manager performs the cloning task. It copies the image 

to the target system. 

l Deployment Manager performs the post-preparation tasks defined by 

the administrator, e.g. executing user-defined installation scripts and 

changing the power status. 

Deployment Manager stores the final deployment status of the server 

in the central deployment table. 

The cloning process is described in detail in the chapter "Cloning deployment 

process " on page 349. 


Information on handling the cloning process via the Deployment Manager 

front-end can be found in the chapter "Mass Cloning" on page 165. 

2.3 Mass remote installation 

With Deployment Manager the unattended installation process of Installation 

Manager can be initiated for mass installation via the Deployment Manager 

console. Appropriate configuration files created with Installation Manager 

must exist. 

The maximum number of supported target servers is basically unlimited but 

is influenced by the bandwidth of the network topology used. The final value 

must be determined for each environment. For example, in a 1 Gbit LAN topology 

about 20 to 25 concurrent installation jobs are a sensible limit. 

The mass remote installation process consists of the following steps: 

l Create installation groups 

l Assign configuration files to installation groups 

l Start mass installation process 

Figure 5: The deployment cycle for mass remote installation 




Installation groups are formed in order to group servers of a specific type with 

specific characteristics. A different configuration file is assigned to each 

server of the installation group. 

A server can be assigned to different installation groups, so it can take on different 

roles at different times. 

The configuration files created by the Installation Manager wizards during the 

configuration phase are stored in a repository folder and can be assigned to 

servers or server groups for mass installation. The administrator assigns an 

appropriate configuration file to the members of an installation group. A specific 

configuration file can be assigned to each server in the group. 

At the start of the installation, the configuration files are merged in each 

individual server information contained in the Deployment Manager deployment 

table and the remote installation processes are started parallel. 

The remote installation process with Deployment Manager works as follows: 

Preparations 

l The administrator defines a remote installation group and decides 

whether all servers in the group will have the same configuration file 

assigned or each server will have its own configuration file. 

l The administrator adds servers to the installation group and assigns 

the appropriate configuration file to the group or different configuration 

files to individual servers. 

l The deployment table is set up with the individual parameters for the 

servers: IP settings, host name and administrator password. 

Starting the mass installation process 

l The administrator sets the power control status for the group or for 

individual servers (power off or power on). 

l The administrator initiates the remote installation directly or configures 

a scheduled installation job. 

Mass installation process 

l A separate configuration file is created for each server in the group. 

l Deployment Manager performs the installation task. 


l Deployment Manager stores the final server status in the central 

deployment table. 

Figure 6: Mass installation with Deployment Manager 


The installation process with Deployment Manager is described in detail in 

chapter "Installation" on page 67. 

How to use the Deployment Manager front-end is described in the chapter 

"Working with Deployment Manager" on page 87. 



2.4 System architecture 

2.4.1 Hardware architecture 

Deployment Manager is designed for a server-client architecture. It supports 

Windows 2008 Server or Windows 2012 Server operating systems as management 

platforms. 

The clients, i.e. the Deployment Manager front-end user interfaces, can be 

installed on a central management station or can be accessed via a Web 

interface from any workstation with Web browser in the network. 

Figure 7: Topological overview of a deployment environment 


Deployment server 

The deployment server contains the following software packages: 

l a Windows 2008 Server or a Windows 2012 Server platform 

l Installation Manager v6.605 (or higher) for mass remote installation 

and for the support of mass cloning for all PRIMERGY servers 

l a DHCP (Dynamic Host Configuration Protocol) service (if there is 

none present in the LAN segment). PXE requires an existing DHCP 

server. 

l a PXE service coming from Deployment Manager or from Installation 

Manager or from the Update Manager tool. 

Only one deployment server can be in use by a Deployment Manager frontend 

per LAN segment, but several deployment servers can exist per LAN 

segment. If PXE and the DHCP service are installed on one machine, only 

one deployment server per segment is allowed. 

Even if it is allowed to have more than one DHCP service and one PXE service 

in one LAN segment, it is recommended to have only one of each service. 

Otherwise, it can only be guaranteed which service will be used by the client, 

if the DHCP and PXE services are configured carefully. The DHCP service 

may be installed on the same server as the PXE service but this is not necessary. 

The Deployment Manager cloning module requires Windows 2008 or Windows 

2012 (for the PXE service) installed on the deployment server to support 

DHCP proxy functionality. 

For fast image creation handling, the partition where the cloning module is 

installed should always have as much free space left as the currently used 

image size will be. Otherwise the image creation process may stop. 

Using the raw mode for image creation, the size could be up to 80% of the 

disk size (normally 2/3 of the disk size). 

LAN environment 


For administrative purposes, the LAN may be organized hierarchically with 

external switches, hubs and gateways at each node point. For switches 



Virtual LAN (VLAN) software may be used to simulate hub or bridge functionality. 

The deployment server must be placed inside of each LAN segment or on a 

higher level in the hierarchy behind switches or hubs. In these cases, the 

switch or hub ports must be configured to let through broadcasts coming from 

the PXE BIOS to the PXE boot server, as well as multicast broadcasts. 

2.4.2 Software architecture 

Figure 8: Architecture - overview 


Deployment Manager obtains the list of server systems from Operations Manager 

by reading the ServerView database. Both the Deployment Manager 

front-end and Operations Manager can run on the same machine. This 

machine does not have to act as a deployment server on which the Deployment 

Manager cloning module and the PXE service are running. 

But for easier installation and administration it is advisable to install all components 

(Deployment Manager (package Deployment Manager and package 

Deployment Services) and Operations Manager) on a deployment 

server. 

ServerView Deployment Manager service 

The ServerView Deployment Manager service is the central node which 

manages jobs for deployment, provides status and parameters to the Deployment 

Manager front-end, and collects information from the ServerView database 

and Installation Manager. The ServerView Deployment Manager service 

itself consists of different building blocks for the different actions. One 

important block is the module which is responsible for the communication 

with the cloning module. 

The cloning module may be located on a separate server, the deployment 

server. 

Both the ServerView Deployment Manager service and the cloning module 

have access to the image repository which contains the images for cloning 

jobs. 

Remote control 

Remote control is used to initiate requests to the target server especially for 

power control and server detection. Access to the service processors such 

as the management blade (via SNMP) and Kalypso BMC (via IPMI) is supported 

via remote control. PRIMERGY servers supporting none of these services 

can be rerouted via a message box to the Deployment Manager frontend 

to get the user to initiate tasks manually instead of remotely. 

Installation Manager Extension 


The Installation Manager Extension is used to communicate with Installation 

Manager. It is session-oriented based on a ServerStart agent concept. Starting 

and stopping an agent identifies the start and the end of a session. In 



between, certain commands can be sent down to the running agent. Each 

agent has its session ID, based on which the ServerView Deployment Manager 

service can manage many tasks to different agents in parallel, as 

required, for example, for mass remote installation. 

2.5 First step guide for using Deployment Manager 

The following steps describe the procedures from installation of Deployment 

Manager packages to starting the deployment task. 

You might need additional procedures which aren’t described below 

depending on your deployment/installation purpose. Therefore, you 

need to read each relevant chapter in the manual for details. 

Step 1: Configuring the DHCP server 

If the DHCP server doesn’t exist on the network, you need to configure 

the DHCP server. For details, please refer to section"Deployment server" 

on page 57. 

Step 2: Installing the necessary software on the deployment server 

Install the necessary software on the deployment server. For details, 

please refer to section "Installing Deployment Manager" on page 51. In 

most cases, the following software is needed at least. 

1. ServerView Operations Manager (local or remote) 

2. Java Runtime Environment package 

3. Deployment Manager package 

4. Deployment Services package 

5. ServerView Installation Manager (mandatory with Deployment 

Manager as of version 5.50) 

Step 3: Creating the image repository 

Deployment Manager needs a shared network directory as the repository 

in order to save the image/installation configuration file and so on. For 

details, please refer to section "Creating an image repository" on page 66 

and "Repositories" on page 106. 


2.5 First step guide for using Deployment Manager 

Step 4: Registering target servers 

Deployment Manager can automatically get the server list from Server- 

View Operations Manager database. Therefore, if the server you want to 

deploy or install is being managed in Operations Manager already, you 

don’t need to manually register the target server in Deployment Manager 

front-end. 

However, if you want to add new servers which aren’t being managed in 

Operations Manager, you can add new servers in the Deployment Manager 

front-end. In order to register new servers, the MAC address of the 

server is needed. For details, please refer to section "Adding new 

servers" on page 132. 

Step 5: Configuring the deployment configuration 

Before deploying the image to servers or before installing the servers you 

must configure the deployment configuration of the target system. For 

detail, please refer to section "Configuration before cloning/installation 

tasks" on page 118. 

Step 6: Starting the task 

You can create the cloning/installation/crash recovery task after finishing 

all steps from step 1 to step 5. For details about each task, please refer to 

the following section: 

l Mass Cloning: section "Mass Cloning" on page 165 and "Image creation" 

on page 153. 

l Mass Installation: section "Mass Installation" on page 243 

l Crash Recovery: section "Crash Recovery" on page 295 and "Image 




3 Installing Deployment Manager 

Deployment Manager is available on the ServerView Suite DVD 1. It is also 

possible to download the software and the latest updates from the download 

area of the Fujitsu Technology Solutions web server 

(http://support.ts.fujitsu.com/com/support/downloads.html). 

In Japan, it is also possible to download the software from the web site 

http://www.fmworld.net/cgi-bin/drviasearch/drviaindex.cgi. 

Deployment Manager consists of two installation packages: 

l Deployment Manager package 

Web-based graphical user interface including the ServerView Deployment 

Manager service. 

l Deployment Services package 

The deployment services including PXE service, TFTP service and 

deployment service (= cloning module). 

You can install both packages at the same time or separately from each 

other. The packages can be installed on different systems. 

If an image repository is used by more than one computer in a LAN, you must 

release the repository for the user account that was specified during the 

installation process. 

Deployment Manager must access the ServerView database. Both modules 

can be installed on one system. During the installation you will be asked 

where Operations Manager is installed. 

The following add-on packages are provided with the Deployment Manager 

software: 

l JavaRuntimeEnvironment package 

Package for the Java runtime environment. This package is required 

for the graphical user interface of Deployment Manager. 

l Installation Manager package 

Installs Installation Manager from the ServerView Suite DVD 1 as of 

version 7.810. 



3.1 Software-Requirements 

l Java Runtime Environment (ORACLE (formerly SUN) JRE as of version 

1.6.0_18). 

l ServerView Operations Manager as of version 5.00 (local or remote 

installation) 

l Installed DHCP server 

For a detailed description of system requirements see section "System 

requirements" on page 18. 

3.1.1 JBoss web server 

As of ServerView Operations Manager version 5.0, the Microsoft 

web server (MS Internet Information Server) and the ServerView 

web server (based on Apache for Windows) are no longer supported. 

As of ServerView Operations Manager version 5.0, the web server used is 

JBoss. 

The necessary files for JBoss are automatically installed when ServerView 

Operations Manager is installed. JBoss is configured as a standalone service 

ServerView JBoss Application Server 5.1, which you can start and stop 

via the Windows start menu: 

l Select Start – Administrative Tools – Services. 

You can also start and stop the service via the following command: 

%WINDIR%\system32\net.exe start "ServerView JBoss Application 

Server 5.1" 

%WINDIR%\system32\net.exe stop "ServerView JBoss Application 

Server 5.1" 

Automatic deletion of JBoss access logs 

Log files localhost_access_log..log are written in the 

…/jboss/server/serverview/log access directory. A separate log is created 


for each day. In previous versions they were never deleted. 

As of version Operations Manager version 5.10, JBoss now includes automatic 

deletion. In the …/jboss/server/serverview/conf directory the configuration 

file sv-com-config.xml describes the parameters: 

localhost_access_log. 

List of file name prefixes, qualifying the files that are to be checked for 

deletion. The default is a list with the single entry "localhost_access_ 

log.". 

12:00 

Test interval, measured in minutes. The test interval can range between 1 

minute and 24 hours. The behavior of the automated file deletion depends 

on the specified value: 

Number [1...1439] 

File checking begins when JBoss starts up, and is repeated at the 

specified time interval. 

Time [hh:mm] 

The test interval is 24 hours and the file checking occurs every day at 

the specified time. (The values for hh range from 00 till 23.) The 

default is 12:00, i.e. files are checked every day at noon. 

10080 

Maximum age of a log file, measured in minutes. Any file matching an 

entry in the fileNamePrefix list which is older than this value is deleted. 

The default is 10080, i.e. seven days. If no value is specified in the configuration 

file, the value 4320, i.e. three days, is used. 

3.1.1.1 Launching and ports used 

The launch address for Operations Manager begins with the prefix https. 

Port Used for 

3170 https (The port must be 

unlocked in the firewall.) 




Port Used for 

3171 Update Management (The port 

must be unlocked in the firewall 

to allow Update Management.) 

3172 Remote Connector Service 

(Server Configuration Manager, 

Power Monitor, etc.) 

(The port must be unlocked in 

the firewall.) 

3173 ServerView RAID Manager 

1072, 1111, 1149, 1287, 1301, 

1302, 1325, 1336, 1338, 1374, 

1380, 1383, 1385, 1399, 1400, 

1401, 1404, 1441, 1442, 1443, 

1445, 1446, 1447, 8787 

JBoss (only used for internal 

socket connections) 

1468 JBoss anonymous port 

1473 non-SSL port of ServerView’s 

directory service (OpenDS) 

1474 LDAPS, if OpenDS is configured 

as directory service. 

4444 OpenDS control port 

9212 PostgreSQL database server 

(only for Linux) 

9363 used by the Operations Manager 

to contact XEN daemons 

16509, 16514 used by the KVM service 

For detailed security information see the White Paper “Secure PRIMERGY 

Server Management“. 


To gather information, Operations Manager accesses the following ports of 

all network nodes of a subnet which is specified on ServerBrowser and managed 

servers. 

Port Used for 

80 Citrix 

135 Hyper-V 

161 SNMP 

443 VMware 

623 BMC (iRMC) 

3172 ServerView Remote Connector 

5988 VMware 

5989 VMware 

9363 XEN 

16509 KVM 

16514 KVM 

3.1.1.2 Role-based user administration 

JBoss also enables role-based user administration. For further information on 

managing certificates, see "User Management in ServerView" user guide. 

3.1.1.3 Managing certificates 


To communicate with the JBoss web server, web browsers always use an 

HTTPS connection (i.e. a secure SSL connection). Therefore, the JBoss 

web server needs a certificate (X.509 certificate) to authenticate itself at the 

web browser. The X.509 certificate contains all the information required to 

identify the JBoss web server plus the public key of the JBoss web server. 

For further information on managing certificates, see "User Management in 

ServerView" user guide. 



3.1.2 Installation of ServerView Installation Manager 

If Installation Manager as of version 7.804 is installed on the deployment 

server, you can use the Deployment Manager function to create installation 

groups for mass installation of PRIMERGY servers with Installation Manager 

configuration files. If Installation Manager as of version 6.711 is installed on 

the deployment server, you can select WinPE MDP as the deployment platform 

for cloning. 

You can also install Installation Manager from the ServerView Suite DVD 1 

as of version 7.810. 

l There are no dependencies between the Installation Manager 

package and the Deployment Manager package. These packages 

can be installed or uninstalled in any order. 

l Mass Installation functionality depends on ServerView Installation 

Manager. Therefore, you must also meet the software 

requirements of ServerView Installation Manager. For example, 

when you want to install a Linux system, you need to configure 

a FTP, HTTP or NFS server. For details, please refer to the 

ServerView Installation Manager user guide. 

3.1.3 Installation of ServerView Operations Manager 

Deployment Manager is integrated in the user interface of Operations Manager. 

The Operations Manager software package (as of version 5.00) must 

be installed locally or remotely before you start the Deployment Manager 

installation. 

This package is available on the ServerView Suite DVD 1, which is a component 

of the ServerView Suite. For more information on installing Operations 

Manager, see the Operations Manager installation guide on the Server- 

View Suite DVD 2. 

ServerView Deployment Manager reads the server list from the 

Operations Manager database. 


3.1.4 Installation of ServerView agents 

During the image creation process you can specify whether the server should 

be shut down before the image creation process is started, see section 

"Options step (Create Cloning Image wizard)" on page 170. 

If you want to do this, you need to install and run ServerView agents on the 

server: 

l On a Windows server platform, ServerView agents as of version 

3.10.05. 

l On a Linux server platform, ServerView agents as of version 3.10.06. 

During the installation of the ServerView agents you must allow SNMP Set 

operations. How to install and configure these agents is described in the 

Operations Manager installation documentation on the ServerView Suite 

DVD 2. 

You must also configure the SNMP service in the Properties window on the 

system. In the Security tab, add the community with the relevant SNMP 

rights. 

3.1.5 Deployment server 


Deployment Manager and Installation Manager support remote access to 

required resources over a LAN during the installation process. The deployment 

software is booted using the PXE boot service, allowing the unattended 

installation to be started on a remote server. 

In order to perform a remote installation using Installation Manager or image 

creation/cloning using Deployment Manager, the following software packages 

must be available on the deployment server: 



l Microsoft Windows Server 2008 or Windows Server 2012 

l You have to configure the Windows firewall on the 

reference system, see "Configuring the Windows 

firewall on the reference system" on page 160. 

l Please see the manual of "ServerView Installation 

Manager" for a list of operating systems that are supported 

by ServerView Installation Manager. 

l Installation Manager as of version 7.804 for PXE-based remote installation 

l Installation Manager as of version 6.711 and deployment service (cloning 

module) for PXE-based remote image creation and cloning 

l A DHCP service (if none is already present in the LAN segment) 

The deployment server and the target system must be in the same LAN 

segment (only with the deployment method Multicast). Only one deployment 

server can be in use but more than one can exist in one LAN segment. 

The IP address of a deployment server must be static. 

Installing the deployment server to PRIMEQUEST is not supported. 

PRIMEQUEST is supported only as a target server. 

PXE Service 

The Fujitsu PXE service must run on the deployment server. It comes with 

either Deployment Manager, Installation Manager (as of version 6.711) or 

Update Manager. These components use the same PXE service. 

Unless the PXE Service API is standardized, other PXE services are not supported. 

To avoid conflicts of PXE services during the installation, Deployment 

Manager automatically detects other PXE services on the 

LAN. This works only for aggressive services reacting to each 

PXE request independently of the client MAC address used. 

If Installation Manager and Deployment Manager and maybe Update Manager 

are installed on the same machine, the basic PXE service will be shared 


etween both modules. The installer of each product will enhance the basic 

PXE service by adding specific modules for their specific use of the basic 

service. With Installation Manager the PXE service will only be installed/enhanced 

if Classic + Remote Installation is selected in the first installation 

frame. 

You can only install one PXE service per system. This service uses port 

4011 (default) on the network. If more than one network interface card (NIC) 

is installed on the deployment server, the PXE service can be assigned to 

one of them. To reassign the NIC to the PXE service, edit the file localipaddress.txt 

in the installation folder .\\bin of the deployment service. 

DHCP Service 

If no DHCP service is installed on the deployment server, the PXE service 

uses the standard DHCP port. A PXE client must scan for a PXE service in 

the LAN segment by transmitting a broadcast on port 67. 

If a DHCP service is present on the deployment server, the PXE service is 

installed as a DHCP proxy service (activated DHCP option: 60). In this case 

the DHCP service is able to directly report the IP address of a PXE service 

(NBS = NetBootService) to a DHCP client, whereupon the client can directly 

access the PXE service by its IP address on port 4011. 

If the DHCP and PXE services are installed on different systems, the 060 

option must be disabled inside the DHCP service. 

For more details on the PXE protocol, see the detailed description of the PXE 

service. 

3.1.6 Network configuration 


Deploying operating systems requires a high-bandwidth network operating at 

least 100Mbps. In addition, special network services and techniques are 

used to control the transaction of transferred packages. 

When extensive cloning images are to be transferred to a large number of 

clients, the standard multicast protocol is used. Each data package is sent 

over the LAN only once. Recipients must be members of the relevant multicast 

session group in order to collect and read these packages. Otherwise 

the transferred packages are ignored. After completion of a data transfer, 



each recipient returns an acknowledgment of receipt to the multicast server. 

When all receipts have been submitted, the server sends out the next data 

package. 

If one of the session group members is using a 10Mbps LAN connection, the 

whole multicast session will run at 10Mbps. When it comes to multicast, 

therefore, it is highly advisable to set up a homogeneous LAN environment 

running at 100Mbps or faster. 

Managed switches and routers can control the transmission of multicast 

packages in order to prevent non-multicast session members from being bothered 

by them. This standard for IP multicasting in the Internet is called Internet 

Group Management Protocol (IGMP). With the IGMP protocol only 

members joining the current multicast session will see the sent out packages. 

3.1.6.1 LAN connections 

The LAN ports can be connected externally in one of the following ways (the 

following assumes that Deployment Manager and Operations Manager are 

installed on the same server): 

See "LAN connection 

topology and IGMP settings 

(Switch Blade)" on 

page 63, [02], [04]. 






page 63, [06], [08], [10], 

[12]. 






page 63, [05], [07], [09], 

[11]. 

For high-performance configurations a switch or router should be used 

instead of hubs. 

3.1.6.2 LAN connections (blade server) 

Each switch blade has ten downlink ports to the server blades and three 

uplink ports to level-3 switches. These LAN connections provide the only I/O 

functionality for the server blade. 

The two management blades have separate LAN connections for maintaining 

accessibility to the system even if the standard traffic network connections 

are down. 

In the Blade Server System manuals, you will find various examples of the 

blade system configuration and an overview of the LAN mapping. 

The management blade is fixed at 10Mbps only (no autosensing). The other 

end of the management blade LAN should be configured as “10Mbps/full 


duplex fix”. If autonegotiation is configured on the switch/router side, the final 

connection is negotiated to 10Mbps/half. Collisions will occur. 

3.1.6.3 LAN connection topology and IGMP settings (Switch Blade) 

The following table summarizes the LAN connection topology and IGMP settings 

of switch blades and external switches/routers: 

External 

switch 

Switch 

blade/ 

NIC 

External switch not used 

BMC/ Management 

blade LAN 

Performance of 

deployment 

[01] IGMP off not same segment not configurable 

[02] same segment governed by 10 Mbps 

*1 

[03] IGMP on not same segment not configurable 

[04] same segment IGMP effective 

External switch used 

IGMP 

off 

IGMP 

on 




*1 



*1 



*1 


[12] same segment IGMP effective 



*1: When a large number of broadcast packets is sent to a broadcast group 

which contains 10Mbps ports, the overall performance of the broadcast is 

governed by the slowest port, e.g. 10Mbps. 

IGMP enabling is only effective in cases [04] and [12]! 

3.1.6.4 Multiple segment cloning 

Deployment Manager uses unicast (image creation and cloning) and multicast 

(cloning only) data transfer. 

Multicast and unicast transfers both work with a data flow control based on 

acknowledgment packages. 

This means that for each data package an acknowledgment from the client is 

expected. Combined with certain timeouts, up to 2 retries are initiated before 

a connection is set to the status broken. 

A data package has bypassed the router and obviously reached the client 

(cause of the reaction on the client screen), but the acknowledgment has not 

reached the server. This fakes the mutual progress of preparing the first partition, 

which is not the case. 

To support cloning over multiple LAN segments via a router or VLAN border, 

you must set certain LAN ports to be passed through in the router configuration: 

The following ports of the router must be enabled in the direction from the 

deployment server to the target: 

Source: Deployment Server 

Destination: Target 

Destination ports: 4973 - 4989/udp 

The following ports of the router must be enabled in the direction from the target 

to the deployment server: 

Source: Target 

Destination: Deployment Server 

Destination ports: 

67/udp, 4011/udp (PXE) 


69/udp (TFTP) 

4972/udp 

4974- 4989/udp 

4974 - 4989/tcp 

If the Deployment Manager and Deployment Services packages are 

installed on two servers in different segments, port 4971 must be enabled in 

the router. Other multicast services in the LAN segment may use the same 

ports as well. 

The port selection by the current deployment server can be initialized in the 

registry settings of Deployment Manager at: 

HKEY_LOCAL_MACHINE\SOF- 

TWARE\FUJITSU\SystemcastWizard\CLONE 

Make the following settings: 

Portbase = 0x0000136e(4974) 

and 

PortRange = 0x00000010 (16) 

After a reboot, the new settings are active. These settings are of interest 

when multiple deployment servers or multiple multicast applications are used 

in one segment. 

If you want to transfer IP packages across multiple segment borders, the 

TTL counter in the header of an IP package must be set accordingly. With the 

TTL parameter, you can specify how many “hops” you allow multicast packets 

to make. Increasing the value means that multicast packets can be transferred 

over many routers. 

A configuration is mandatory if multiple segment cloning is required. There is 

one entry with which you can manipulate TTL in the deployment server operating 

system registry: 

HKEY_LOCAL_MACHINE\SOF- 

TWARE\FUJITSU\SystemcastWizard\CLONE\MRestore 

-> Value: TTL 

By default Deployment Manager uses the value 3. 




3.1.6.5 Multiple LAN ports of target systems 

It is possible to use the other LAN port of servers for the deployment process. 

For this purpose the other LAN port of each server must be configured 

using the Deployment Configuration wizard (see section "Remote Management 

Ports step (Deployment Configuration wizard)" on page 125 or 

"Remote Management Ports step (New Server wizard)" on page 139 (Lan 

Port for PXE Boot option.). 

When the other LAN port is to be used for deployment, both the deployment 

server and the other LAN port of the server must belong to the 

same LAN segment (only with the deployment method Multicast). 

The deployment server LAN port used for the PXE services is defined 

in the file localipaddress.txt. 

3.1.7 Creating an image repository 

l Create a shared network directory (UNC network path) on your deployment 

server (or on another server in the LAN segment). 

l Share this folder so that it can be accessed remotely. 

Both user profiles created in the course of the Deployment 

Manager installation must be allowed to fully access this 

network share, see section "Installation" on page 67. 

l Specify a repository name and the UNC network path in order to register 

the new share as an image repository in Deployment Manager. 

The images will be stored in this directory. 

For details refer to the section "Repositories" on page 106. 


3.2 Installation 

You have the following options for installing the Deployment Manager software: 

l Via the ServerView Suite DVD 1, see section Installation via the 

ServerView Suite DVD 1. 

l Download from the Fujitsu Technology Solutions web server: 

http://support.ts.fujitsu.com/com/support/downloads.html 

In Japan, it is also possible to download the software from 

the following web site: 

http://primeserver.fujitsu.com/primergy/ 

3.2.1 Installation via the ServerView Suite DVD 1 

The Deployment Manager software can be found in the directory 

SVSSoftware\Software\Deployment\DeploymentManager on the 

ServerView Suite DVD 1. 

You can start the installation via the ServerView Suite DVD 1. Proceed as follows: 

1. Insert the ServerView Suite DVD 1 in the DVD-ROM drive. If the 

DVD does not start automatically, click the setup.exe file in the root 

directory of the DVD-ROM. 

2. Select the option Select ServerView Software Products. 

3. Click Start. 

4. In the next window, select the required language. 

5. Select ServerView - Deployment Tools. 

6. Click RDSetup.exe. 

The installation wizard is launched with the start window. 




Figure 9: Start window 

Select the packages you wish to install. Click OK. For further information 

click Info. 

If you select both Deployment Manager packages, the Deployment 

Manager package is the first one to be installed. 

3.2.1.1 Installing the Deployment Manager package 

If you select the Deployment Manager package in the start window, the 

Fujitsu ServerView Deployment Manager Setup window opens. 

1. Click Next. 

2. Approve the license agreement by selecting the appropriate option button. 


4. Fill out your name and organization. 

5. Specify whether the Deployment Manager settings should only apply 

to the current user or to anyone who uses the target computer. Select 

the appropriate option button. 



7. Enter the license key. If you do not enter a valid license key, an evaluation 

key for 5 servers is created. After you start Deployment Manager 

a message window is displayed which shows when the evaluation 

key will expire. 

Valid license keys can be installed later, see section " Licenses 

Installed step (Settings wizard)" on page 332. 



The ServerView Operations Manager, ServerView Deployment 

Manager and ServerView Installation Manager packages must 

be installed on the same drive. The ServerView Installation Manager 

Data packages can be installed on any drive. 



9. Set up your user account. The user profile you specify must already 

be available on the target system or domain. 

You can also create a new user. Click Create New User in the window. 

This opens the Create New User window. Enter the new user 

account. 



If the Deployment Manager packages (Deployment Manager 

package and Deployment Services package) are installed on 

different computers, one of the following requirements must be 

met: 

l When using one or more domain accounts, both computers 

must be present in the same Windows network 

domain. In addition, the domain user account specified 

during the installation process must be a member of this 

network domain. 

l When using local user accounts, both computers may be 

present in different network domains or in no domain at 

all. However, both local user accounts must be allowed 

to fully access the shared image repository. This can be 

achieved by setting up identical user accounts for Deployment 

Manager and services on the computer containing 

the repository. User names and passwords must match 

those of the local user accounts. If only one user account 

is to be used, it must be reproduced identically on all computers 

containing the Deployment Manager package, 

the Deployment Services package and/or the image 

repository. 

The user account which is defined during the installation of Deployment 

Manager is assigned to the service ServerView Deployment 

Manager. The user account which is defined during the installation of 

the deployment service is assigned to the service Deployment Service. 

If the password of one of both users was changed by the administrator 

using the user management of the Control Panel, the appropriate 

password setting must also be changed in the server 

administration of the administration tools in the Control Panel. 

10. Select Start - Settings - Control Panel - Administrative Tools - 

Services. Double-click the service (ServerView Deployment Manager 

or Deployment Service) or select Properties from the context 

menu and change to the Log On tab. 



This account shows the user name whose password has been 

changed. Change the password of that user. To activate the new settings 

click OK and stop and restart the service. Afterwards you should 

be able to log on to Deployment Manager again. 



12. Enter the user account for the JBoss service. 

For this user name you must have set up a standard user account with 

no special rights. Enter the password and repeat it. When you have 

already installed the ServerView Operations Manager in the server, 

you cannot enter the user name and passwords. The username and 

passwords are already entered at the installation of the ServerView 

Operations Manager. 





14. Select the directory server. You can use the directory server OpenDS, 

which is automatically installed when you install ServerView Operations 

Manager. Or you want to use an existing directory server. If you 

select the Existing directory server option, you must make further 

settings for the directory service: 


15. Specify the following parameters for the directory service: 

Host (FQN) Fully-qualified name of the server on which the directory 

service is running. 

Port Port number used for access to the directory service. 

By default, port 636 is used. 

SSL This option is enabled by default to protect the data 

transfer. It is not advisable to use the directory service 

without SSL encryption. 

Base DN Highest level of the LDAP directory tree. 

User User ID for read access to the data. The user ID 

should only have basic read rights. 

Password Password for read access. 





17. If you select the No option, a standard installation is performed (recommended). 

Select Yes to support the high-availability function. For 

more information see section "High-Availability support" on page 339. 


19. Click Next to start the installation process. 

20. Click Finish to exit the wizard. 

21. After installation you have to specify the location where Operations 

Manager is installed. 


If Operations Manager is installed on the same system, select the 

Local on this server option and click Ok. Otherwise you must manually 

install a certificate at the remote server before you can continue 

the installation. The certificate is generated automatically by Server- 

View Deployment Manager installer and can be found in the following 

directory: 

%Program Files%\Fujitsu\ServerView Suite\svcommon\data\download\pki 

Copy the files in this directory to the following directory on the remote 

server: 

%Program Files%\Fujitsu\ServerView Suite\Remote Connector\pki 


ServerView Deployment Manager reads the server list from 

the Operations Manager database. 



If Operations Manager is installed on a remote server and 

Windows firewall is enabled on the remote server, you must 

configure the Windows firewall on the remote server. Select 

Start - Administrative Tools - Windows Firewall with 

Advanced Security, there select Inbound Rules in the left 

pane, and then select New Rule... from the context menu. 

Please add a rule of type Program for \Remote Connector\SVRemCon.exe. 

You can accept the default values 

offered in the wizard (choose SVRemCon as rule name). 

Once the installation has been successfully completed, you can start Deployment 

Manager, see section "Opening Deployment Manager" on page 87. 

In order to start Deployment Manager, a deployment server running 

the required deployment services must be installed in the LAN segment 

, see section "Installing the Deployment Services package" on 

page 78. 

3.2.1.2 Installing the Deployment Services package 

If you select the Deployment Services package in the start window, the 

DeploymentService - InstallShield Wizard window opens: 



2. If you select the No option, a standard installation is performed (recommended). 

Select Yes to support the high-availability function. For 

more information "High-Availability support" on page 339. 





4. Select the LAN port to be used for deployment and click Next. 

The PXE service requires a DHCP service to be present on the LAN. 


The selected LAN port (IP address) is stored in the localipaddress.txt 

file in the same directory as the PXE- 

Service.exe file. By default, this is: \DeploymentService\bin 

5. Specify whether the DHCP service is running on the PXE service target 

machine or another computer in the LAN segment. Select the 

appropriate option button. 


7. Fill out the user account under which the deployment service is to be 

run. 

You must enter this user account in the Settings wizard 

when adding a deployment server. 

The installation creates a (local) user group Deployment 

Admins on the deployment server. This user account is 

automatically added to the Deployment Admins group. 




8. The user account can either be a domain login [notation: 

\] or 

a local profile [notation: \]. 

In the latter case, details of the computer name can be left out. 


It is recommended that you enter the same values that 

were specified when installing the Deployment Manager 

package. 

10. Click Install to start the installation process. 

11. Click Finish to exit the wizard. 

12. For the Mass Cloning of Windows servers you need some files which 

are available on the ServerView Suite DVD 1 as of version 7.810. For 

more information, see the Hints section. Click OK to start the installation. 

Click Cancel if you want to install the files later. 

Configuring the Windows firewall 

After successful installation of the Deployment Services package, the following 

message is displayed: 


You can also change the firewall settings later at any time by calling the 

fw_setting.vbs script. 

The fw_setting.vbs script is copied to the \Deployment Manager directory. 

To execute the script proceed as follows: 

1. Start a command prompt. 

2. Change to the following directory: 

%Program Files%\Fujitsu\ServerView Suite\Deployment Manager 

3. Execute the following command: 

cscript.exe fw_setting.vbs 

3.2.2 Installation via the Fujitsu Technology Solutions web 

server 

You can download the Deployment Manager software from the Fujitsu Technology 

Solutions web server: 

http://support.ts.fujitsu.com/com/support/downloads.html 


You can enter the product name or you can be guided there by your choices 

in the successive selection lists offered. This method is described below by 

using an example: 

1. In the first selection list, choose the product line Industry standard 

server. 



2. Then select the product group PRIMERGY TXxxx. 

3. Then select the product family PRIMERGY TX150 S6. 

4. Then select the operating system Microsoft Windows Server 2008. 

5. Then klick Windows Server 2008 (x64). 

6. Select Server Management Software - ServerView - Deployment 


7. Download the relevant Deployment Manager version. 

3.3 Update installation 

You start an update installation as described in section "Installation via the 

ServerView Suite DVD 1" on page 67 

In the Start window select the packages you wish to update. The Deployment 

Manager and Deployment Services package are not selected if they 

are already installed. 

3.4 Uninstalling/Removing Deployment Manager 

l It is highly recommended that you uninstall Deployment Manager 

rather than deleting its folder. This is because various components 

of a program can be spread all over your system and in 

the registry. 

l The deployment services must be removed from the system 

before an update or reinstallation of Deployment Manager. If 

services remain on the system, the reinstallation may not be 

performed properly and error messages may be issued when 

the operating system is restarted. 


3.4 Uninstalling/Removing Deployment Manager 

l If you uninstall Deployment Manager (Deployment Manager 

and Deployment Services packages) and afterwards you 

install a new Deployment Manager version, you must clear the 

site cache from the corresponding browser. 

If you use Internet Explorer, select Internet Options ... in the 

Tools menu and click the Delete Files ... button on the General 

tab. 


and Deployment Services packages) and subsequently want 

to install an older version than before, you must delete the Java 

plug-in cache and remove the previous *.jar and *.gif files 

stored there. Otherwise a mixture of old and new modules will 

be running in the Deployment Manager front-end session and 

indeterminate errors may occur. 

You must also clear the browser cache for downloaded files. 

l You may see a dialog box during the uninstallation of the 

Deployment Services package showing a list of files which 

are being used. If you accept the default action Close and 

restart the applications automatically, there are some event 

log entries created by the Microsoft Windows RestartManager 

(e.g.: Application or service 'Deployment Service' could not be 

restarted.). If you want to prevent the creation of these event 

log entries, please select Do not close applications. (A 

Reboot may be required.) instead. 


and Deployment Services packages) you should execute the 

del_fw_setting.bat script. This script deletes the firewall settings 

that have been added with script fw_setting.vbs. This 

script is delivered with the ServerView Deployment Manager 

and ServerView Installation Manager packages. It is copied to 

the \Deployment 

Manager directory. 



There are different methods of uninstalling: 

1. Running uninstall from the Programs folder in the Windows Start 

menu. 

l Make sure the Deployment Manager front-end is not open. 

l Select Start - Programs - Fujitsu - ServerView Suite - 

Deployment Manager. 

l Select Uninstall Deployment Manager Completely. The 

packages which are installed are marked. 

l Confirm to uninstall the packages. 

2. Running Add/Remove Programs from the Control Panel. 

l Make sure the Deployment Manager is not running. 

l Click Start - Settings - Control Panel. 

l Double-click the Add/Remove Programs icon. 

l Find the entry for the packages (Deployment Service or Fujitsu 

ServerView Deployment Manager) and select the 

Add/Remove button to remove the package. 

These packages may have been installed on different 

computers. 

l A screen will prompt you for confirmation. Click yes to start the 

uninstallation process. 

l During the uninstallation of the Deployment Services package, 

you can select whether you want to delete the SCW database 

and the SCW log files. If you select to keep the files, the 

event log information is retained for future analysis. 

l There are no dependencies between Installation Manager 

package and Deployment Manager package. These packages 

can be installed or uninstalled in any order. 

l The Deployment Manager package must be uninstalled 

before the Operations Manager package is uninstalled. 


4 Working with Deployment Manager 

Web-based Deployment Manager is integrated in Operations Manager as a 

separate application. Deployment Manager handles all PRIMERGY servers 

that appear in the ServerView server list. 

Operations Manager is the first application to detect the PRIMERGY servers 

and their system information: 

l On request Operations Manager searches for the LAN to find all the 

existing servers in one segment. 

l Operations Manager requests a list of system information for each 

non-blade server and for the management blade of each server blade 

contained in a blade chassis. 

l Operations Manager creates a physical server list for each server and 

consolidates logical server groups based on the system information 

received. 

How to start Operations Manager is described in the “Operations Manager” 

manual. 

4.1 Opening Deployment Manager 

You can open Deployment Manager in the following ways: 

Opening via Operations Manager 

Deployment Manager is integrated in Operations Manager. In Operations 

Manager select the Deployment Manager item from the Deployment menu 

or select the Deployment Manager link in the Operations Manager main window. 

The Deployment menu will only appear in Operations Manager if 

Deployment Manager has been installed. 



Opening via the Windows start menu 

You can also start the Deployment Manager front-end via the Windows start 

menu: 

Start - Programs - Fujitsu - ServerView Suite - Deployment Manager - 

Start Deployment Manager 

Opening via a link 

If you specified during the installation that Operations Manager is located 

remotely on another server in the network, the Deployment menu is not 

added to the menu bar of Operations Manager. You can start Deployment 

Manager via the following link: 

http://:3169/svdm 

: .. 

4.2 Closing Deployment Manager 

You can close Deployment Manager by closing the main window: 

l To close the Deployment Manager main window, click the Close icon 

in the browser window. 

Deployment Manager waits until all currently running deployment jobs are finished 

before terminating. 

4.3 Actions after starting the Deployment Manager 

front-end 

If you are opening Deployment Manager via the Windows start menu or via a 

link, the login window of the Central Authentication Service is displayed. 


4.3 Actions after starting the Deployment Manager front-end 

In this window, enter the user name and the password of the ID under which 

you are authorized to use Deployment Manager. 

By default there are predefined user names with different roles: 

l Administrator for the role Administrator (default password: admin) 

l Operator for the role Operator (default password: admin) 

l Monitor for the role Monitor (default password: admin) 

l UserManager for the role UserAdministrator (default password: 

admin) 

For more information on role-based user management, see section "Rolebased 

permissions on accessing Deployment Manager" on page 90 and the 

"User management in ServerView” user guide. 

If you are opening Deployment Manager via Operations Manager, the 

login window of the Central Authentication Service is not displayed. 

You are already authorized to use Deployment Manager. 



If you start Deployment Manager for the first time after installation, you must 

specify global settings in the Settings wizard. You must work through all 

steps in the wizard, see section "Settings" on page 329. 

4.3.1 Role-based permissions on accessing Deployment Manager 

This chapter provides you with detailed information on the permissions granted 

by the user roles Administrator, Operator and Monitor. 

"Forbidden" components are deactivated in the Deployment Manager GUI 

Components and functions displayed in the Deployment Manager GUI are 

deactivated ("greyed out") if the role the user is assigned to does not have the 

required permissions. 

Description of authorization Permitted with user role 

Access the Workspaces „Mass 

Cloning“, „Mass Installation“ and 

“Crash Recovery“ 

Create a cloning image or snapshot 

image of a server 

Restore a cloning image or snapshot 

image to a server 

Install a server P 

Modify the global “Settings” P 

Administrator Operator Monitor 

P P P 

P P 

Create, modify and delete a server P P 

Perform power operations on a 

server 

Perform a "Generic Boot" or an 

"MDP Boot" of a server 

P 

P P 

90 Deployment Manager 

P

4.4 Deployment Manager main window 

After successful authentication the main window of Deployment Manager is 

displayed. 

The main window contains the following elements (from top to bottom): 

l the ServerView Suite header 

You can collapse the header with the icon on the right. 

l the Deployment Manager menu bars 

l the work area, which is divided into a left and right area. 

The work area on the left contains the Servers view and the Repositories 

view. The Servers view is displayed by default. 

On the right select one of the following work areas: 

Work Areas Description 

Mass Cloning Mass Cloning allows you to create an image of a server 

(the reference system) and to clone this image to any 

number of target servers in parallel. 

Mass Installation 

Crash Recovery 


Mass Installation allows you to install any number of target 

servers in parallel. 

Crash Recovery allows you to create and restore snapshot 

images of a server. 

The contents on the right change depending on your selection on the left. 



4.4.1 Servers view 

In the Servers view, the tree structure shows the servers contained in the 

ServerView server list. This was known as the "Physical View" in a previous 

Deployment Manager version. The Servers view and Mass Cloning workspace 

are selected by default after login. You can add additionally servers to 

the server list. 

The tree structure is divided into different groups, depending on your selection 

in the menu bar: 

Group Significance 

All Servers Predefined server group. The servers listed 

in this section are the servers which are 

managed in Operations Manager and/or the 

servers which you added by the New 

Server wizard. 

Servers by 

Cloning 

Image 

Servers by 

Installation 

List of all cloning groups. This is formally 

known as the "Deployment View". It contains 

the cloning groups which are created 

on the selected deployment server. The 

group has a cloning image assigned. If you 

select a cloning group, the servers in this 

group are displayed. By selecting Clone 

from the context menu of the group, all 

servers of the group can be cloned in parallel 

in one multicast job. 

Installation groups. Every server in the 

group can have a separate configuration file 

assigned. By selecting Install from the context 

menu of the group, all servers in the 

group can be installed in parallel. 

Selected 

menu item 

Mass Cloning 



Group Significance 

Servers by 

Boot Image 

Generic boot groups. By selecting Generic 

Boot/MDP Boot from the context menu of 

the group, all servers of the group are booted 

with the boot image that has been 

assigned to the group. 

Selected 

menu item 


The icon in front of each server shows the status of the PRIMERGY servers, 

see section "Icons" on page 102. 

The Deployment Manager functions are available via several context menus. 

4.4.2 Repositories view 

In the Repositories view, the tree structure shows all available image repositories 

and folders which are stored in image repository folders on the deployment 

server. Depending on the work area there is only a limited number of 

repository types displayed. 

The first time you start Deployment Manager after installation, you must 

specify at least one Disk Image or Installation Configuration repository. 

Later you can add additional folders and repositories. 

4.4.3 Registered Boot Images view 

4.4.4 Tabs 

In the Registered Boot Images view, there is a list of all registered boot 

images displayed. You can register and unregister generic boot images here. 

The right work area in the main window contains the following tabs: 

l Servers tab 

l Information tab 

l Tasks tab 




4.4.4.1 Servers (tab) 

The Servers tab displays server information of the selected servers in the 

Servers view. 

Figure 10: Servers tab 

The columns in the table have the following meanings: 

Column Meaning 

Process State The icon displays the process state. 

Action The icon displays the last deployment action issued for 

the server. 

Health State The icon displays the ServerView state. 

Server Name Name of the server. 

Slot Slot number. 



Server Type Type of the server. 

Cloning Image Path of the image file. 

Configuration 

File 

Path of the configuration file. 

Type PRIMERGY server type. 

For the meaning of the icons, see section "Icons" on page 102. 

4.4.4.2 Information (tab) 

The Information tab displays general information. 

Figure 11: Information tab 


Depending on the selected item in the Servers view, different information is 

displayed: 



Chassis Information 

Product Name PRIMERGY server type 

BIOS Version BIOS version 

IP Address (iRMC) IP address of the iRMC. 

Health State ServerView state 

Housing Name Name of the blade server 

Model Type Type of the server blade 

IP Address (MMB) IP address of the management blade. 

Remote Console http address. This address is used to 

access the iRMC (for server blades and 

PRIMERGY servers) or to access the management 

blade of a blade server. 

Network Information 

(displayed for non-server blades) 

The columns in the table below have the following meanings: 


MAC Address MAC address of the server. 

Current IP Current IP address. 

Config. IP Configured IP address. 

Config. Mode Static IP or DHCP. 

Slot Information 

(displayed for server blades) 

Slot Slot number 

Product Name PRIMERGY server name 

BIOS Version BIOS version 

Health State The icon displays the health state, see section 

"Icons" on page 102. 




MAC Address MAC address of the server blade. 

Current IP Current IP address. 

Config. IP Configured IP address. 

Config. Mode Static IP or DHCP. 

Assigned Image (only for a selected cloning group) 

Displays the assigned image name. 

Image Properties (only for a selected cloning group) 

Displays the image properties. 

4.4.4.3 Tasks (tab) 

You can watch the progress of a created task in the Tasks tab. 

Figure 12: Tasks tab 




Refresh 

Immediate update of the display. 

refresh every 

Specify a refresh interval. It is advisable to select refresh every to have 

the task status updated at regular intervals. 

Task List 



Scheduling The icon displays whether this is an immediate 

task or a scheduled task. 

State The icon displays the state of the task. 

Name Name of the task. 

Type Type of the task: Save Disk Image, Deploy 

Servers, Install Servers, Generic Boot, MDP 

Boot 

Scheduling Time Scheduling details of the task. 

Task Details 

The Task Details view displays details of the selected task in the Job Status 

tab and Event Log tab. 

Job Status tab 

The Job Status tab displays the following information about the current 

(latest) execution of the task: 

State The icon displays the state of the job. 

Start Time Time when the task begin. 

End Time Time when the task will finish. 

Phase Command Name of the running process. 

Phase Status Current state of the process. 




Job ID Job identification. 

State The icon displays the state of the job. 

Server Name Name of the server. 

Start Time Time when the task was started. 

End Time Time when the task was finished. 

Phase Command Current task phase. 

Detail Info A detailed status string. 

Detail Info displays detailed information about the process. 

Event Log tab 

The Event Log tab displays information about all past executions of the 

task. You can delete the event log via the Clear All Log button. 



State The icon displays the state of the task. 

Time Time when this event log entry was created. 

Result A detailed status string. 

Result displays detailed information about the execution of the task (probably 

in the past). 

You can perform different operations via the context menu, see section "Operations 

on the Tasks tab" on page 99. 

For the meaning of the icons, see section "Icons" on page 102. 

4.4.4.4 Operations on the Tasks tab 

You can perform the following operations on the Tasks tab: 

l Executing a task 

l Retrying a failed task 




l Editing and deleting task 

l Retrying a failed server 

l Canceling a task 

l Unregistering all clients at the PXE service (only for Generic Boot 

tasks) 

Executing a task 

1. Select the relevant task from the Task List. 

2. Select Execute Task from the context menu. 

Retrying a failed task 


2. Select Retry Failed Task from the context menu. 

Editing a task 

The job is not started for all servers, but only for the servers 

that failed during last execution of the task. 


2. Select Edit Task from the context menu. 

The corresponding wizard is displayed. 

Deleting a task 

1. Select the relevant task in the Task List. 

2. Select Delete Task from the context menu. 

Retrying a failed server 

1. Under Task Details select the relevant server from the table. 

2. Select Retry Failed Server from the context menu. 

Canceling a task 


2. Select Cancel Task from the context menu. 


When you specify iRMC Support, MMB SNMP Support or 

MMB Remote Manager Support in step Remote Management 

Ports in Deployment Configuration, the target 

server is forced to shutdown if you cancel the task. 

Unregistering all clients at the PXE service (only for Generic Boot 

tasks) 



2. Select Unregister All Clients from the context menu. 

This applies only to servers in Generic Boot tasks that have 

been started with selection Permanent for Registration at 

PXE Service (see "Settings step (Generic Boot wizard)" on 

page 288). The task must already have been completed. 


4.4.5 Icons 


The following list shows the Deployment Manager icons and their meanings. 

General icons: 

Icons Meaning 

Displays info text. 

Magnify result or show details in a larger dialog box. 

Server type icons: 

Icons Meaning 

Server 

Blade server 

Server blade 

BMC 

Power icons: 

Group (e.g. All Systems, Systems by Installation, Systems by 

Cloning Image) 

Icon Meaning 

Server is on. 

Server is off. 

Unknown power state. 


Deployment action icons: 

Icons Meaning 

Installation 

Cloning 

Not available 

Process state icons: 

Icons Meaning 

Running 

Okay 

Unknown 

Not available 

Health state icons: 

Icon Meaning 

Okay 

Unknown 

Warning 

Error 

Standard SNMP agent okay 

RSB mode 

Not manageable 

Ping 




Task scheduling icons: 

Icons Meaning 

Immediate task 

Scheduled task 

Task state icons: 

Icons Meaning 

Not started task 

Running task 

Retry task 

Task okay 

Task error 

Job state icons: 

Icons Meaning 

Scheduled job 

Initialized job 

Waiting job 

Running job 

Job okay 

Job error 

Canceling job 

Canceled job 


4.4.6 Wizards 

A wizard is an assistant that guides you through a task. 

A wizard usually consists of several steps. The number of steps used and 

their sequence are shown in a tree structure on the left. Steps that you have 

already completed are indicated in the tree. 

4.4.6.1 General buttons 

The buttons in the bottom right of each step allow you to progress through the 

wizard workflow. 

Next 

Opens the next step in the wizard. 

Previous 

Opens the previous step in the wizard. 

Cancel 

Cancels the wizard workflow without saving your changes. 

Finish 

Executes the wizard with your settings. 




4.5 Repositories 

The Repositories view displays on the left all images which are stored in 

image repository folders on the deployment server. 

Figure 13: Repositories view 

The Repositories tree view contains the following repository types: 

Disk Images 

These are the repositories of disk images where you save disk images 

which can then be used to define deployment groups for cloning servers. 

A repository of disk images is a directory with subdirectories where you 

store your disk images. 


Preparation Boot Images 

These are the repositories of preparation boot images. Preparation boot 

images are used to boot a server by using the PXE service. A repository 

of preparation boot images is a directory with subdirectories where you 

store the PXE boot images you have created in order to perform a 'generic' 

system preparation. 

Deployment Tables 

These are the repositories for deployment configurations. These repositories 

are used to export or import deployment configurations of selected 

servers. 

Installation Configuration 

These are the repositories of installation configuration files. These files 

must be created by Installation Manager. A repository of installation configuration 

files is a directory with subdirectories where you have stored 

the configuration files created with Installation Manager. Installation configuration 

files are needed when creating an installation group, which is 

only possible when Installation Manager is installed. 

Generic Boot Images 

These are the repositories of the generic boot images. These files must 

be created manually by the user. Generic boot images can be registered 

in the Registered Boot Images view. After registration, a generic boot 

image can be used to create a generic boot group. 

MDP Applications 

These are the repositories of the MDP applications. These files must be 

created manually by the user. MDP applications are needed to perform an 

MDP Boot of a generic boot group. 

For details on MDP (Multi Deployment Platform) please see the 

ServerView Installation Manager user guide. 

In the Repositories view you can perform the following functions: 


l Add a new repository, see section "Add Repository dialog box" on 

page 108. 



l Add a new repository folder, see section "Adding a new folder" on 

page 110. 

l Delete a repository or folder, see section "Deleting a repository or 

folder" on page 111. 

4.5.1 Add Repository dialog box 

In the Add Repository dialog box you can add repositories for all types of 

repositories: 

l Disk Images 

l Preparation Boot Images 

l Deployment Tables 

l Installation Configuration 

l Generic Boot Images 

l MDP Applications 

Figure 14: Add Repository dialog box 


Repository Name 

Name of the repository. 


Network Path 

In general the path name should be the network path to a shared directory 

(UNC notation). It can also be the absolute path to a local directory and 

not a network path, but this requires the Deployment Manager user interface 

and the deployment services to be installed on the same machine. If 

this is not the case, it must be the network path to a share in UNC notation: 

"\\server_name\share_name" 

It is important to set the security permissions correctly for the directory that 

is used as the repository. The user account(s) that you specified during the 

installation of Deployment Manager must have full control for this directory 

(and all subdirectories). This is also true for the sharing permissions. 



There are two services that access the image repository: ServerView 

Deployment Manager and Deployment Service. These two services 

may run under two different accounts that you specified during 

installation. These services need to access the image repository and 

need the right to read, write and list the directories. In general it is possible 

to install the Deployment Manager Web user interface to which 

the ServerView Deployment Manager belongs and the deployment 

services on different machines. And there can be a repository on a 

third machine. Therefore you must fulfil some requirements concerning 

the user account(s): 

4.5.2 Adding a new folder 

l The first possibility is that you specified a domain account during 

Deployment Manager installation. In this case the machine 

on which a repository is located needs to be in the same 

domain, and the domain account(s) that you specified for 

Deployment Manager must have full control of the image directory. 

l The second possibility is that you are using local user 

accounts. In this case the user accounts that you specified during 

installation must also be available on the machine on which 

a repository is located. Since you are using local account(s), 

you must create the same account(s) with the same password 

on the machine where the repository is to be located. 

In the Repository view you can add new folders. 

1. Select a repository or a subfolder of a repository on the left. 

2. Select New Folder from the context menu. 

New Folder is not supported for Generic Boot Images and MDP 

Applications repositories. 

The New Folder dialog box opens. 


Figure 15: New Folder dialog box 

Folder Name 

Name of the new folder. 

4.5.3 Deleting a repository or folder 

In the Repositories view you can delete a repository or a folder: 

1. Select the relevant repository or folder on the left. 

2. Select Delete Repository or Delete Folder from the context menu. 

Delete Folder is not supported for Generic Boot Images and 

MDP Applications repositories. 

A dialog box is displayed to confirm the deletion. 

4.6 Registered Boot Images 


The Registered Boot Images view displays a list of all registered boot 

images. 



Figure 16: Registered Boot Images view 

The built-in images SVIM32 and SVIM64 are 32 bit and 64 bit images of type 

SVIM Agent Image (MDP). 

If you want use MDP Boot to execute an MDP application, you can 

create a boot group from the built-in images SVIM32 or SVIM64. In 

most cases it is not necessary to add and register a boot image of 

type SVIM Agent Image (MDP). 

In the Registered Boot Images view you can perform the following functions: 

l Register a boot image, see "Register Boot Image wizard" on page 

113. 

l Unregister a boot image, see "Unregistering a boot image" on page 

117. 

l Unregister a client at the PXE service, see "Unregistering a client" on 

page 117. 


4.6.1 Register Boot Image wizard 

The Register Boot Image wizard allows you to register a boot image. To 

open the wizard, select Register Boot Image from the context menu of Registered 

Boot Images. 

Before you open the wizard, perform the following actions: 

1. Add a Generic Boot Images repository. Please use some shared network 

directory. Do not use a directory below the TFTP service directory. 

2. Manually copy the boot image files to a sub directory in the Generic 

Boot Images repository. 

During boot image registration, the boot image files will be copied 

from the repository to the TFTP service directory. 

4.6.1.1 Image ID step (Register Boot Image wizard) 


Image ID is the first step in the Register Boot Image wizard. In this step 

you must enter an Image ID. 



Figure 17: Image ID step 

Image ID 

An identifier for the boot image. Only alphanumeric characters are 

allowed. 

Description 

Optional: Specify any textual information that you want to be saved with 

the boot image. This information will be stored internally by ServerView 

Installation Manager. 

4.6.1.2 Boot Image step (Register Boot Image wizard) 

Boot Image is the next step in the Register Boot Image wizard. In this step 

you must select a boot image folder. 


Figure 18: Boot Image step 

Select a boot image folder in one of your Generic Boot Images repositories. 

4.6.1.3 Bootstrap Loader step (Register Boot Image wizard) 


Bootstrap Loader is the final step in the Register Boot Image wizard. In 

this step you must select a bootstrap loader and provide additional parameter. 



Figure 19: Bootstrap Loader step 

Boot Image Folder 

Select a bootstrap loader and select the image type. 

Image Type 

SVIM Agent Image (MDP) 

An image that is compatible with the built-in images SVIM32 and 

SVIM64 of ServerView Installation Manager. The image must be a 

WinPE based image with a ServerView Installation Manager agent 

included. If you create a boot group with such an image, you can use 

MDP Boot to execute an MDP application. 

Optionally, you can specify the following parameters: 


BCD File The filename of the BCD file. 

Bootstrap Loader for 

EFI Boot 

The filename of the bootstrap loader for 

EFI boot. 

BCD File for EFI Boot The filename of the BCD file for EFI 

boot. 

Path to TFTP Root Directory 

The path to the TFTP root directory. 

Generic Image 

Any other image. If you create a boot group with such an image, you 

can use Generic Boot to boot the generic image. 

4.6.2 Unregistering a boot image 

In the Registered Boot Images view you can unregister a boot image: 

1. Select the relevant boot image. 

2. Select Unregister Boot Image from the context menu. 

A dialog box opens to confirm the deletion. 

4.6.3 Unregistering a client 

In the Registered Boot Images view you can unregister a client at the PXE 

service: 

1. Select the relevant client in the Registered Clients list. 

2. Select Unregister Client from the context menu. 

A dialog box opens to confirm the deletion. 


This applies only to servers in Generic Boot tasks that have been 

started with selection Permanent for Registration at PXE Service 

(see "Settings step (Generic Boot wizard)" on page 288). The task 

must already have been completed. 



4.7 Configuration before cloning/installation tasks 

Before deploying the image to servers or before installing the servers you 

must configure the target system. 

After defining such a configuration you can export the deployment configuration 

and later import it. For more information, see section "Export 

Deployment Configuration dialog box" on page 130 and section "Import 

Deployment Configuration dialog box" on page 131. 

4.7.1 Deployment Configuration wizard 

The Deployment Configuration wizard allows you to define settings on the 

target system. To open the wizard, select Deployment Configuration from 

the context menu in the Servers view. 

4.7.1.1 General Settings step (Deployment Configuration wizard) 

General Settings is the first step in the Deployment Configuration wizard. 

In this step you can specify some basic parameters. 


Figure 20: General Settings step 

Name 

Server name. 

DNS Suffix 

If specified, displays the DNS suffix. 


The DNS Suffix is not applied to the target machine in case of 

"Mass Installation". 

GUID (UUID) 

Optional: GUID (globally unique identifier) of the server. 

One method for finding out the GUID: If you add a LAN port in the LAN 

Ports step, select the List button next to the MAC Address option. The 

Found MAC Addresses dialog box displays the GUIDs. 



Primergy Type 

When Installation Manager is installed, a list of supported PRIMERGY 

types appears, from which you should select one. 

Process State 

Selects the process state: 

unknown The process state is unknown. The server has not yet 

been deployed or installed by Deployment Manager. 

running A deployment job or installation is running. 

OK The server has been successfully deployed or installed by 


User Name 

Account name of the administrator. This is the account name used to 

select an administrator user account for a cloned system. After this 

server has been cloned, the password of this user account is set to the 

specified password. During installation of this server, only the specified 

password is used to set the password of the administrative account, 

which is Administrator on Windows systems and root on Linux systems. 

If you are installing a server, the account name has no meaning. If the 

administrator account is left empty but a password is specified, the password 

of the account Administrator on a Windows operating system and 

root on a Linux operating system will be set to the specified password 

after an image is cloned. 

Password/Repeat Password 

Installation: 

During installation of this server, the specified password is used to set 

the password of the administrative account, which is Administrator on 

Windows systems and root on Linux system. 

Deploy Image: 

The specified password will be set as the password of the selected 

account (see parameter "Administrator Account" above). The password 

change occurs during the post-preparation of the cloned system. 


4.7.1.2 LAN Ports step (Deployment Configuration wizard) 

LAN Ports is the next step in the Deployment Configuration wizard. In 

this step you can add, edit or delete LAN ports of the server. For each LAN 

port you must configure a MAC address. At least one network adapter must 

be specified. 

Click New to add a LAN port. Click Delete to remove the selected LAN port. 

You can only use these buttons for bare servers and servers on which the 

ServerView agents are not running. 

Figure 21: LAN Ports step 


General tab 

The General tab allows you to specify a MAC address. 



MAC Address 

MAC Address of the selected LAN port. Click List to open the Found 

MAC Addresses dialog box in which you can select a MAC address, see 

section "Found MAC Addresses dialog box" on page 144. 

Register Addresses in DNS 

Allows you to register the IP address automatically in DNS. The configuration 

of the value is only supported when cloning a Windows disk 

image. In addition, the image must be created with Remote Deploy as of 

version 4.0. 

DNS Suffix for this Connection 

Specify a DNS suffix for the selected network adapter. The configuration 

of the value is only supported when cloning a Windows disk image. In 

addition, the image must be created with Remote Deploy as of version 

4.0. 

Use Suffix in DNS Registration 

Enables a DNS dynamic update to register the IP addresses and the connection-specific 

domain name of this connection. The configuration of the 

value is only supported when cloning a Windows disk image. In addition, 

the image must be created with Remote Deploy as of version 4.0. 

IPv4 tab 

The IPv4 tab allows you to configure one or more IPv4 addresses. 

DHCP 

Assignment of a dynamic IP address. 

Static 

Assignment of static IP addresses. You must enter a fixed IP address 

and subnet mask. The default gateway is optional. 

DNS Servers 

You can assign a list of DNS servers to each LAN port. After an installation 

or cloning process the IP addresses are assigned. Click 

Add/Edit/Remove to add, edit or remove a DNS server from the list. 


IP Addresses 

For Mass Cloning, only the first two IP addresses are set. 

Gateways 

For Mass Cloning, only the first gateway address is set. For Mass 

Installation of Windows targets the gateway addresses are ignored. 

You can add, edit or delete IPv4 addresses, see "Add IPv4 Address dialog 

box " on page 146 and "Edit IPv4 Address dialog box " on page 147. 

IPv6 tab 


DHCP 



Static 

Assignment of one or more static IP addresses. You must enter a fixed IP 

address and subnet prefix length. The default gateway is optional. 

IP Addresses 


Gateways 



DNS Servers 


or cloning process the IP addresses are assigned to the corresponding 

LAN ports. Click Add/Edit/Remove to add, edit or remove a 

DNS server from the list. 


box" on page 148 and "Edit IPv6 Address dialog box" on page 149. 



For PRIMEQUEST partitions only: If a PRIMEQUEST Server Agent 

(PSA) is installed, you cannot set a static IPv4 address for the NIC of 

the PSA-to-MMB communication LAN. The PRIMEQUEST Server 

Agent must be setup manually after cloning or installation (see descriptions 

of configuring PSA in "PRIMEQUEST 1000 Series Installation 

Manual"). 

4.7.1.3 DNS Search Suffixes step (Deployment Configuration wizard) 

DNS Search Suffixes is the next step in the Deployment Configuration 

wizard. 

In this step you can optionally specify a list of DNS suffixes which are used 

to extend a server name to a fully qualified domain name (FQDN) when trying 

to resolve the IP address of a specified server name. When trying to find the 

IP address of a specified name, these suffixes are appended to the name and 

an address resolution is tried with the fully qualified domain name (FQDN) 

produced. This is tried for each specified suffix. The suffixes are tried in the 

order in which they are specified. 

The configuration of this value is only supported when cloning a disk image. 

It is not supported by a remote installation. In addition the image must be 

created with a Deployment Manager version 4.00 or higher. 

Click New..., Edit... or Delete to add, edit and remove suffixes. Use the 

arrow buttons to change the order. 


Figure 22: DNS Search Suffixes step 


4.7.1.4 Remote Management Ports step (Deployment Configuration wizard) 

Remote Management Ports is the next step in the Deployment Configuration 

wizard. In this step you can select a port for PXE boot and specify 

a method for remote management. 

If you select iRMC Support, MMB SNMP Support or MMB 

Remote Manager Support in this step, you need to turn off a target 

server before you start the deployment. You can perform the 

shutdown manually or by using the Force System Shutdown 

before Deployment option in the task you want to execute. 



Figure 23: Remote Management Ports step 

iRMC Support 

You can specify the following options: 

IP Address Specify the IP address that was assigned to the 

iRMC. This is needed for remote power-on and 

remote configuration of the boot sequence for PXE 

boot. 

User Name Preconfigured user name. 

Password / 

Repeat Password 

Preconfigured password. 

Timeout (s) Timeout value for access to iRMC. 

Retries Number of retries. 


For PRIMEQUEST, you can specify the following options: 

IP Address 

(MMB) 

Displays the IP address of the PRIMEQUEST 

MMB. 

User Name Preconfigured user name on the PRIMEQUEST 

MMB. The user name is valid for all partitions. 

Password / 





Click Test Connectivity to test whether the iRMC can be accessed via 

the specified parameters. 

Wake ON LAN Support 

In the case of Wake On LAN, Deployment Manager uses IP broadcast or 

Ethernet broadcast to send a magic packet as a UDP datagram to the 

subnet in which the target system is located. 

The following applies: 


l If the target system is in the same LAN segment as the deployment 

server, you do not need to specify an address under Broadcast 

Address. In this case, Deployment Manager automatically 

uses the limited broadcast address 255.255.255.255 and sends 

the magic packet to UDP port 9 as an Ethernet broadcast using 

the MAC Address of the target system. 

l If the target system is in a different LAN segment that is bridged 

by one or more gateways, enter either of the following under 

Broadcast Address: 

o Subnet broadcast address of the LAN segment in which the target 

system is located. The address must contain the value 

"255" in the device area (host area) (e.g. 192.168.2.255). In this 

case, the magic packet is sent to the gateway via one or several 

hops, and the gateway ultimately transmits the Ethernet 



broadcast to the subnet of the target system. 

o IP address of a BOOTP/DHCP server. In this case, you have 

to select the Broadcast to Bootstrap Protocol Server option 

and specify a valid IP address in the LAN segment of the target 

system under IP Address. 


Broadcast 

Address 

Broadcast 

to BootstrapProtocol 

Server 

IP 

Address 

Subnet broadcast address of the LAN segment in which 

the target system is located, or Unicast address of a 

BOOTP/DHCP server. If you specify the Unicast 

address of a BOOTP/DHCP server, you must select the 

Broadcast to Bootstrap Server option. 

The magic packet is sent to UDP port 67 (Bootstrap Protocol 

(BOOTP) Server); otherwise, it is sent to UDP port 

9. 

This option is required if you specify the Unicast address 

of a BOOTP/DHCP server under Broadcast Address. Furthermore, 

you should select this option if it is not guaranteed 

that all gateways included in a subnet broadcast 

are configured for "subnet broadcasting". 

Any Unicast address in the subnet of the target system. 

Using this IP address, the BOOTP/DHCP server determines 

the LAN interface (LAN port) via which it is to send 

the magic packet (in this case, a DHCP/BOOTP Reply 

Packet). 

Manual Management 

You must start or reboot the server manually. 

MMB SNMP Support (only for blades) 

The management blade of the blade server is accessed by SNMP. There 

must be an SNMP community with read-write access. 


MMB Remote Manager Support (only for blades) 

The management blade of the blade server is accessed via telnet. Readwrite 

access via SNMP is not necessary. Only blade servers BX600 

(with MMB S3) and BX900 are supported. 

This option can only be selected for Mass Cloning and Crash 

Recovery. For Mass Installation use the MMB SNMP Support 

option. 

4.7.1.5 Notes step (Deployment Configuration wizard) 

Notes is the final step in the Deployment Configuration wizard. In this step 

you can enter notes about this server. 

Figure 24: Notes step 




4.7.2 Export Deployment Configuration dialog box 

The Export Deployment Configuration dialog box allows you to export the 

deployment configuration for a list of selected servers in the server list to a 

file in XML format. This deployment configuration can then be imported. This 

enables you to change between different deployment configurations for 

servers whenever it is necessary. 

The exported files are always stored in the Deployment Tables repository in 

the Repositories view. 

To open the dialog box, select Export Deployment Configuration from the 

context menu in the Servers view. 

Figure 25: Export Deployment Configuration dialog box 


Deployment Tables 

Select the directory to which the deployment configuration is exported. 

File Name 

File name to which the deployment configuration is exported. 

Encrypt Passwords 

The exported deployment configuration contains several passwords. By 

default these passwords are exported in encrypted form. If you want to 

edit them, uncheck the option. 

4.7.3 Import Deployment Configuration dialog box 

The Import Deployment Configuration dialog box allows you to import the 

deployment configuration from a file that was created by a previous export 

run. 

To open the dialog box, proceed as follows: 


l Select the corresponding file in the Deployment Tables group in the 

Repositories view. 

l Select Import Deployment Configuration from the context menu. 

Figure 26: Import Deployment Configuration dialog box 



Add non existing Bare Servers 

The deployment configuration of bare servers which do not exist in the 

server list will be imported. 

Match Server ID 

The deployment configuration of servers whose server ID stored in the 

file also matches the server ID found in the server list will be imported. 

Match Slot ID (Only for blade servers) 

The deployment configuration of servers whose slot number stored in the 

file also matches the slot number found in the server list will be imported. 

4.7.4 Deleting an exported deployment configuration 

To delete an exported file, proceed as follows: 

1. Select the corresponding file in the Deployment Tables group in the 

Repositories view. 

2. Select Delete File from the context menu. 

4.8 Adding new servers 

You can add new servers to the ServerView server list. This list provides an 

overview of all the configured servers. 

To add a new server use the New Server wizard. 

4.8.1 New Server wizard 

The New Server wizard does not support blade server and PRI- 

MEQUEST. If you want to add new server blades or new PRI- 

MEQUEST partitions, add their MMB from ServerView Operations 

Manager. The server blades and PRIMEQUEST partitions are automatically 

added by adding the MMB. Do not directly add the server 

blades or PRIMEQUEST partitions to the server list of ServerView 

Operations Manager. 

The New Server wizard allows you to define settings for the new server. 


You open the wizard by selecting New Server from the context menu of All 

Servers in the Servers view. 

4.8.1.1 General Settings step (New Server wizard) 

General Settings is the first step in the New Server wizard. In this step you 

can specify some basic parameters. 

Figure 27: General Settings step 

Name 

Server name. 

DNS Suffix 

If specified, displays the DNS suffix. 


The DNS Suffix is not applied to the target machine in case of 

"Mass Installation". 



GUID (UUID) 

Optional: GUID (globally unique identifier) of the server. 

One method for finding out the GUID: If you add a LAN port in the LAN 

Ports step, select the List button next to the MAC Address option. The 

Found MAC Addresses dialog box displays the GUIDs. 

Primergy Type 

When Installation Manager is installed, a list of supported PRIMERGY 

types appears, from which you should select one. 

Process State 

Selects the process state: 

unknown The process state is unknown. The server has not yet 

been deployed or installed by Deployment Manager. 

running A deployment job or installation is running. 

OK The server has been successfully deployed or installed by 


User Name 

Account name of the administrator. This is the account name used to 

select an administrator user account for a cloned system. After this 

server has been cloned, the password of this user account is set to the 

specified password. During installation of this server, only the specified 

password is used to set the password of the administrative account, 

which is Administrator on Windows systems and root on Linux systems. 

If you are installing a server, the account name has no meaning. If the 

administrator account is left empty but a password is specified, the password 

of the account Administrator on a Windows operating system and 

root on a Linux operating system will be set to the specified password 

after an image is cloned. 

Password/Repeat Password 

Installation: 

During installation of this server, the specified password is used to set 


the password of the administrative account, which is Administrator on 

Windows systems and root on Linux system. 

Deploy Image: 

The specified password will be set as the password of the selected 

account (see parameter "Administrator Account" above). The password 

change occurs during the post-preparation of the cloned system. 

4.8.1.2 LAN Ports step (New Server wizard) 


LAN Ports is the next step in the New Server wizard. In this step you can 

add, edit or delete LAN ports of the server. For each LAN port you must configure 

a MAC address. One network adapter must be specified to be used for 

PXE boot. 

Click New to add a LAN port. Click the Delete button to remove the selected 

LAN port. You can only use these buttons for bare servers and servers on 

which the ServerView agents are not running. 

Bare server indicates a server where an operating system has not 

been installed yet. 



Figure 28: LAN Ports step 

General tab 

The General tab allows you to specify a MAC address. 

MAC Address 

MAC address of the selected LAN port. Click List to open the Found 

MAC Addresses dialog box in which you can select a MAC address, see 

section "Found MAC Addresses dialog box" on page 144. 

Register Addresses in DNS 

Allows you to register the IP address automatically in DNS. The configuration 

of the value is only supported when cloning a Windows disk 

image. In addition, the image must be created with RemoteDeploy as of 

version 4.0. 


DNS Suffix for this Connection 

Specify a DNS suffix for the selected network adapter. The configuration 

of the value is only supported when cloning a Windows disk image. In 

addition, the image must be created with RemoteDeploy as of version 

4.0. 

Use Suffix in DNS Registration 

Enables a DNS dynamic update to register the IP addresses and the connection-specific 

domain name of this connection. The configuration of the 

value is only supported when cloning a Windows disk image. In addition, 

the image must be created with RemoteDeploy as of version 4.0. 

IPv4 tab 


DHCP 


Static 

Assignment of static IP addresses. You must enter a fixed IP address 

and subnet mask. The default gateway is optional. 

DNS Servers 


or cloning process the IP addresses are assigned. Click 

Add/Edit/Remove to add, edit or remove a DNS server from the list. 

IP Addresses 



Gateways 




box " on page 146 and "Edit IPv4 Address dialog box " on page 147. 

IPv6 tab 




DHCP 


Static 

Assignment of one or more static IP addresses. You must enter a fixed IP 

address and subnet prefix length. The default gateway is optional. 

IP Addresses 


Gateways 

Displays a list of IPv6 gateway addresses. 

DNS Servers 


or cloning process the IP addresses are assigned to the corresponding 

LAN ports. Click Add/Edit/Remove to add, edit or remove a 

DNS server from the list. 


box" on page 148 and "Edit IPv6 Address dialog box" on page 149. 

4.8.1.3 DNS Search Suffixes step (New Server wizard) 

DNS Search Suffixes is the next step in the New Server wizard. 

In this step you can optionally specify a list of DNS suffixes which are used 

to extend a server name to a fully qualified domain name (FQDN) when trying 

to resolve the IP address of a specified server name. When trying to find the 

IP address of a specified name, these suffixes are appended to the name and 

an address resolution is tried with the fully qualified domain name (FQDN) 

produced. This is tried for each specified suffix. The suffixes are tried in the 

order in which they are specified. 

The configuration of this value is only supported when cloning a disk image. 

It is not supported by a remote installation. In addition the image must be 

created with a Deployment Manager version 4.00 or higher. 

Click New..., Edit... or Delete to add, edit and remove suffixes. Use the 

arrow buttons to change the order. 


Figure 29: DNS Search Suffixes step 

4.8.1.4 Remote Management Ports step (New Server wizard) 

Remote Management Ports is the next step in the New Server wizard. In 

this step you can select a port for PXE boot and specify one of the following 

methods for remote management: 

l iRMC Support 

l Wake on LAN Support 

l Manual Management 




Figure 30: Remote Management Ports step 

iRMC Support 


IP Address Specify the IP address that was assigned to the 

iRMC. This is needed for remote power-on and 

remote configuration of the boot sequence for PXE 

boot. 

User Name Preconfigured user name. 

Password / 






For PRIMEQUEST, you can specify the following options: 

IP Address 

(MMB) 

Displays the IP address of the PRIMEQUEST 

MMB. 

User Name Preconfigured user name on the PRIMEQUEST 

MMB. The user name is valid for all partitions. 

Password / 





Click Test Connectivity to test whether the iRMC can be accessed via 

the specified parameters. 

Wake ON LAN Support 

In the case of Wake On LAN, Deployment Manager uses IP broadcast or 

Ethernet broadcast to send a magic packet as a UDP datagram to the 

subnet in which the target system is located. 

The following applies: 


l If the target system is in the same LAN segment as the deployment 

server, you do not need to specify an address under Broadcast 

Address. In this case, Deployment Manager automatically 

uses the limited broadcast address 255.255.255.255 and sends 

the magic packet to UDP port 9 as an Ethernet broadcast using 

the MAC Address of the target system. 

l If the target system is in a different LAN segment that is bridged 

by one or more gateways, enter either of the following under 

Broadcast Address: 

o Subnet broadcast address of the LAN segment in which the target 

system is located. The address must contain the value 

"255" in the device area (host area) (e.g. 192.168.2.255). In this 

case, the magic packet is sent to the gateway via one or several 

hops, and the gateway ultimately transmits the Ethernet 



broadcast to the subnet of the target system. 

o IP address of a BOOTP/DHCP server. In this case, you have 

to select the Broadcast to Bootstrap Protocol Server option 

and specify a valid IP address in the LAN segment of the target 

system under IP Address. 


Broadcast 

Address 

Broadcast 

to BootstrapProtocol 

Server 

IP 

Address 

Subnet broadcast address of the LAN segment in which 

the target system is located, or Unicast address of a 

BOOTP/DHCP server. If you specify the Unicast 

address of a BOOTP/DHCP server, you must select the 

Broadcast to Bootstrap Server option. 

The magic packet is sent to UDP port 67 (Bootstrap Protocol 

(BOOTP) Server); otherwise, it is sent to UDP port 

9. 

This option is required if you specify the Unicast address 

of a BOOTP/DHCP server under Broadcast Address. Furthermore, 

you should select this option if it is not guaranteed 

that all gateways included in a subnet broadcast 

are configured for "subnet broadcasting". 

Any Unicast address in the subnet of the target system. 

Using this IP address, the BOOTP/DHCP server determines 

the LAN interface (LAN port) via which it is to send 

the magic packet (in this case, a DHCP/BOOTP Reply 

Packet). 

Manual Management 

You must start or reboot the server manually. 

MMB SNMP Support (only for blades) 

The management blade of the blade server is accessed by SNMP. There 

must be an SNMP community with read-write access. 


MMB Remote Manager Support (only for blades) 

The management blade of the blade server is accessed via telnet. Readwrite 

access via SNMP is not necessary. Only blade servers BX600 

(with MMB S3) and BX900 are supported. 

This option can only be selected for Mass Cloning and Crash 

Recovery. For Mass Installation use the MMB SNMP Support 

option. 

4.8.1.5 Notes step (New Server wizard) 

Notes is the final step in the New Server wizard. In this step you can enter 

notes about this server. 

Figure 31: Notes step 




4.8.1.6 Found MAC Addresses dialog box 

The Found MAC Addresses dialog box displays a list of MAC addresses 

found by the PXE service. This is a list of all MAC addresses that initiated a 

PXE request. By default all MAC addresses that belong to a server that is 

already in the server list are not shown in this list. Also only the last occurrence 

of a MAC address is listed by default. 

To open the dialog box, select List in the LAN Ports step in the New Server 

wizard. 

Figure 32: Found MAC Addresses dialog box 


Refresh 

Use this button to refresh the display. 

Exclude MAC Addresses Of Existing Servers 

When this option is not checked, the MAC addresses belonging to 

servers that are already in the server list are listed. 

Show Duplicates 

When this option is not checked, only the last PXE request of the server 

(MAC address) will be shown. 

The table in the dialog box lists the defined MAC addresses. The columns 

are explained below: 


MAC Address MAC address of the first LAN port of the server. 

GUID GUID (globally unique identifier) of the server. 

PXE Request Time Time of the last PXE request. 


After selecting a MAC address and clicking OK, the following window is displayed: 



Confirm with Yes to use the GUID for this server; otherwise confirm with No. 

4.9 IP addresses 

4.9.1 IPv4 IP addresses 

As of Deployment Manager version 5.40 you can specify more IPv4 addresses. 

Only two IPv4 addresses are set during Mass Cloning. 

You can add, edit or delete IPv4 addresses. 

4.9.1.1 Add IPv4 Address dialog box 

The Add IPv4 Address dialog box allows you to configure a static IPv4 

address. 

To open the dialog, select New in the IPv4 tab in the LAN Ports step of the 

New Server or Deployment Configuration wizard. 

Figure 33: Add IPv4 Address dialog box 


IP Address 

IPv4 uses 32-bit (four-byte) addresses, which limits the address space to 

4,294,967,296 (2 32 ) possible unique addresses. IPv4 addresses are 

usually written in dot-decimal notation, which consists of the four octets 

of the address expressed in decimal and separated by periods. 

Subnet Mask 

In IPv4 networks, the routing prefix is traditionally expressed as a subnet 

mask, which is the prefix bit mask expressed in quad-dotted decimal representation. 

4.9.1.2 Edit IPv4 Address dialog box 

The Edit IPv4 Address dialog box allows you to edit a static IPv64 address. 

To open the dialog, select Edit in the IPv4 tab in the LAN Ports step of the 


Figure 34: Edit IPv4 Address dialog box 


IP Address 

IPv4 uses 32-bit (four-byte) addresses, which limits the address space to 

4,294,967,296 (2 32 ) possible unique addresses. IPv4 addresses are 



usually written in dot-decimal notation, which consists of the four octets 

of the address expressed in decimal and separated by periods. 

Subnet Mask 

In IPv4 networks, the routing prefix is traditionally expressed as a subnet 

mask, which is the prefix bit mask expressed in quad-dotted decimal representation. 

4.9.2 IPv6 IP addresses 

As of Deployment Manager version 5.20 you can use the IPv6 address 

range. 

The previous globally available address space under IPv4 is quickly running 

out. This has been resolved by introducing the IPv6 address space. An IPv6 

address is 128 bits long and consists of eight 16-bit fields, each separated by 

colons. Each field must contain a hexadecimal number, as opposed to the 

dot-decimal notation used in IPv4 addresses. You can add, edit or delete 

IPv6 addresses. 

4.9.2.1 Add IPv6 Address dialog box 

The Add IPv6 Address dialog box allows you to configure a static IPv6 

address. 

To open the dialog, select New in the IPv6 tab in the LAN Ports step of the 

New Server or Deplyoment Configuration wizard. 

Figure 35: Add IPv6 Address dialog box 


IP Address 

IPv6 addresses are normally written as eight groups of four hexadecimal 

digits, where each group is separated by a colon (:). 

For example: 2001:0db8:85a3:0000:0000:8a2e:0370:7334 

Any leading zeros in a group may be omitted; thus, the given example 

becomes 

2001:db8:85a3:0:0:8a2e:370:7334 

One or any number of consecutive groups of 0 values may be replaced 

with two colons (::): 

2001:db8:85a3::8a2e:370:7334 

Subnet Prefix Length 

Length of the subnet prefix for the IPv6 address. For IPv6 unicast 

addresses this value should be 64 (the standard value). 

4.9.2.2 Edit IPv6 Address dialog box 

The Edit IPv6 Address dialog box allows you to edit a static IPv6 address. 

To open the dialog, select Edit in the IPv6 tab in the LAN Ports step of the 


Figure 36: Edit IPv6 Address dialog box 




IP Address 

IPv6 addresses are normally written as eight groups of four hexadecimal 

digits, where each group is separated by a colon (:). 

For example: 2001:0db8:85a3:0000:0000:8a2e:0370:7334 

Any leading zeros in a group may be omitted; thus, the given example 

becomes 

2001:db8:85a3:0:0:8a2e:370:7334 

One or any number of consecutive groups of 0 values may be replaced 

with two colons (::): 

2001:db8:85a3::8a2e:370:7334 

Subnet Prefix Length 

Length of the subnet prefix for the IPv6 address. For IPv6 unicast 

addresses this value should be 64 (the standard value). 

4.10 Managing the power state of a server 

You can manage the power state of selected servers via the Manage Power 

State dialog box. To open the dialog box, select one or more servers and 

choose Manage Power State from the context menu. 

4.10.1 Manage Power State dialog box 

The Manage Power State dialog box allows you to control the power state of 

one or more servers. 


Figure 37: Manage Power State dialog box 

Current Power State 

Displays the selected server. The icon next to it shows the current power 

state, see section "Icons" on page 102. 

Change Power State to 

New power state to be assigned to all servers in the list. 

On Server is on. Default value if the server is off. 

Force Off Server is off. Default value if the server is on. 

4.11 Collecting diagnostic information 

4.11 Collecting diagnostic information 

Sometimes it might be necessary to gather diagnostic information in order to 

send it to your Fujitsu support service. Via the Windows start menu please 

execute: 

Start - Programs - Fujitsu - ServerView Suite - Deployment Manager - 

Collect Diagnostic Information 



The diagnostic information will be saved to the \Deployment Manager\diag directory. 


5 Image creation 

Once a reference system is completely installed and finally configured for the 

end user, the image creation process can be started via Deployment Manager. 

Deployment Manager supports 

l snapshot image creation. A “snapshot image” of a server can be used 

to restore the server after a system crash. A snapshot image can be 

created for multiple selected servers at the same time. 

l file-system-dependent image creation and 

l file-system-independent image creation (raw image creation). 

Only when creating a non-snapshot image from a Windows system: 

If an error occurs during image creation, you should boot the Windows 

operating system normally. At the beginning of this boot there might be 

a "rollback" of changes already made. Do not interrupt this rollback. 

Before you start the image creation process, make sure that all connected 

network drives are disconnected. 

The created image is stored in an image repository. This is located in a 

shared folder on any file server in the network. 

An image consists of the following files: 

l .img 

The file contains the image. 

l .cfg 

The file contains information about the image. 

l .txt 

The file contains user-defined text. 



5.1 File-System-Dependent image creation 

The following file systems are supported: 

l FAT, FAT32 (Windows) 

l NTFS, NTFS5 (Windows), NTF5 

l EXT2, EXT3, EXT4 Reiser file system (V3.5, 3.6) (Linux) 

l Reiser file system with Deployment Manager version 4.0 

If the file-system-dependent image creation has been selected for a 

hard disk drive with unsupported file systems, the image is still created 

correctly. 

If a customer has an implementation of several partitions with different 

file systems, the deployment engine will check for each partition 

whether the file system is supported (and add partition to image in filesystem-dependent 

mode) or not (and add partition to image in raw 

mode) and will save all partition information in one image file. Otherwise 

(if file-system-independent or raw image creation has been selected), 

the file system type will not be checked and all data will be moved 

to the image file in raw mode (1:1). 

Partitions with unsupported file systems will be read out according to 

the file-system-independent image creation process. 

File-system-dependent image creation does determine the used clusters of 

each file in the file table of a file system, but only stores this file table and the 

used clusters in the image. With this method, the disk geometry is not limited 

because the used blocks are stored file-system-(partition)-oriented, not disksector-oriented. 

The image is compressed and encoded (for security reasons) and written to a 

location defined by a UNC network path (UNC = Universal Naming Convention). 

When using this kind of image, the cloning process first creates partitions 

with the appropriate size. Based on the information stored in the image header, 

these partitions are then formatted. Finally, the image is decoded and 


5.2 File-System-Independent image creation (raw image creation) 

decompressed on the fly and the extracted clusters are written directly to the 

correct location in the file system. With this method, only the user data of a 

file system must be copied to a target drive. 

If you have created an image of a 20 GB partition with 3 GB user data, only 3 

GB user data must be written on the hard drive. The rest of the 20 GB is 

defined by the formatting process (if verification is selected, this will take 

extra time). 

5.2 File-System-Independent image creation (raw 

image creation) 

The file-system-independent image is a 1:1 copy of a hard drive but is encoded 

and compressed in the same way as for the file-system-dependent method. 

The final image also has a similar size because of effective compression 

of the non-used areas. The hardcopy is read by sector containing information 

which depends on the drive geometry. This method requires that the hard 

drive of the target system is identical to the hard drive of the reference system. 

The big difference is the extraction of this image, which takes about three 

times as long but is independent of the file system and thus the operating system. 

On the target side, each byte of the raw copy of the image must be 

written to the target drive which takes the majority of the cloning process. 

If you have created an image of a 20 GB partition and 3 GB user data, you 

must write 20 GB to the hard drive again. On the other hand, you save the 

time required for formatting the drive (slow, especially with verification, but 

more flexible). 

For partitions which are converted as dynamic disks, the image creation can 

only be done in raw mode. 

Before you start the raw image creation you must clean the drive. 

This can be done via Installation Manager. Start the DiskManager 

tool and select CleanOut drive. 



5.3 Supported operating systems for image creation 

Deployment Manager supports the image creation for: 

l Microsoft Windows Server 2003 Standard/Enterprise (x86 and x64) 

l Microsoft Windows Server 2003 R2 Standard/Enterprise (x86 and 

x64) 


(x86 and x64) 

l Microsoft Windows Server 2008 R2 Standard/Enterprise/Datacenter/Foundation/Web 

(x64) 

Mass Cloning of Windows Domain Controllers (e.g. 

Windows Small Business Server) is not supported. 



l Resilient file system is not supported. 

l Storage Spaces and Virtual Disks created 

from Storage Spaces are not supported. 

l Before starting an image creation process for 

a Windows 2008 or Windows 2012 reference 

system, you must configure the Windows firewall 

on the reference system, see "Configuring 

the Windows firewall on the reference 

system" on page 160. 

l In cloning with personalization, only "full" Windows 

2008 or Windows 2012 installations are 

supported. Saving an image of a Windows 

2008 or Windows 2012 "core" installation is 

not supported. 

l Linux RedHat (AS/ES) 3.0 - not supported in Japan 

l Linux RedHat (AS/ES) 4.0 (x86 and EM64T) - not supported in Japan 




5.3 Supported operating systems for image creation 

Before image creation, a specific package must be 

installed on the reference system. This package is 

provided with the Deployment Manager software in 

the directory LINUX_support\RHEL6_Cloning_ 

Support. 

Before image creation, the NetworkManager should 

be disabled - otherwise the personalization may fail 

during cloning. Please set NM_CONTROLLED=no 

in the /etc/sysconfig/network-scripts/ifcfg-eth* 

files and issue the following commands: 

/sbin/service NetworkManager stop 

/sbin/chkconfig --del NetworkManager 

/sbin/chkconfig --add network 

/sbin/service network start 

l Linux SuSE Enterprise Server SLES 8 - not supported in Japan 

l Linux SuSE SLES 9 (x86 and EM64T) - not supported in Japan 

l Linux SuSE United Linux (with SLES 8) - not supported in Japan 


in Japan 


in Japan 

l Reiser file system versions 3.5 and 3.6 on SuSE SLES 9 and SuSE 

SLES 10 

Before image creation, a specific package must be installed on the reference 

system if Reiser is used as the root file system. This package 

is provided with the Deployment Manager software in the directory 

LINUX_support\ReiserFS_Cloning_Support. 



Hints: 

l All other operating system types can always be supported without 

personalization in raw snapshot mode. See also the Read- 

Me file. 

l Creating a cloning image of a disk with multiple operating systems 

is not supported. Creating a snapshot image is supported. 

l Image creation is only supported for Linux reference systems that 

have the Fstab option Mount in /etc/fstab set to Device name for all 

partitions. During a native installation of SuSE SLES 10 and SLES 11, 

this option is set to Device ID by default. You must manually change 

this option for all partitions. If you install SuSE SLES 10 and SLES 11 

unattended with an older Installation Manager version (up to 

V10.09.11), Device name is set by default for all partitions. 

If you install SuSE SLES 10 and SLES 11 unattended with Installation 

Manager as of version 10.09.12, you must change the Mount option 

from mount by id (the default) to mount by device in the Bootloader 

page of the Linux wizard. 

l If the Linux reference system (probably Red Hat 5) has LVM Volume 

Groups configured, the image creation is done in raw mode. Server- 

View Deployment Manager does not recognize LVM Volume Groups. 

l If the reference system has Native Multipath IO configured, the image 

creation is done in raw mode. ServerView Deployment Manager does 

not recognize Native Multipath IO configurations. 

5.4 Image creation of a Windows reference system 

During image creation the reference system is booted multiple times: 

l First the reference system is booted with Windows to do the system 

preparation: 

o The Client Agent package is temporarily installed on the reference 

system. After a successful installation the Deployment 

Agent service should be running. 


o The personalization files are copied to the reference system. 

o The LAN interfaces are disabled. 

l Then WinPE is booted and the image is saved to the deployment 

server. 

l Finally the reference system is booted again with Windows to rollback 

the changes that were done during system preparation. 

If an error occurs when saving the image to the deployment server, 

the reference system must first be booted once before you try again 

to create an image (the rollback must be done). 

If the system preparation fails, please have a look at the following log directory: 

\- 

DeploymentService\tftp\log\ID#.... 

The system preparation log file is named .log. 

If there is already a Client Agent package installed on the reference system, 

the image creation fails with result Serverscript returns error. The preparation 

log file contains the message ScwAgent has been already 

installed on the reference system. Please manually cleanup the reference 

system like described in the manual.. Please perform the steps 

described below before you try again to create an image. 

5.4.1 Manual rollback of changes done during system preparation 

If the automatic rollback fails, you can perform the following steps to undo 

the changes done during system preparation: 

1. Uninstall the Client Agent package: 

l Logon as local administrator. 

5.4 Image creation of a Windows reference system 

l Execute %Program Files%\SCWAgent\uninstall.bat. 

l Make sure that Windows has deleted the registry key 

HKLM\SYSTEM\CurrentControlSet\services\Deployment 

Agent. If this key still exists, please do not delete it manually. 



Please make sure that you are logged on as local administrator 

and try to call uninstall.bat again. 

l Delete %Program Files%\SCWAgent directory. 

l Remove the following registry key and its subkeys: 

o HKLM\Software\Fujitsu\SystemcastWizard\Agent 

2. Delete files on the reference system: 

l Logon as local administrator. 

l Execute %SystemDrive%\cleanup.bat. 

l Delete the following directories in %SystemDrive%: 

o Ipadj 

o Sysprep 

l Delete the following files in %SystemDrive%: 

o clcomp.dat 

o clcomp3.dat 

o cleanup.bat 

o Win.tag 

l Delete the following file: 

o %SystemDrive%\DeploymentInfo\DeploymentMode 

5.5 Windows 2008 and Windows 2012 reference systems 

When you use Windows 2008 or Windows 2012 as the reference system you 

must configure the Windows firewall. Note also the restrictions for Sysprep. 

5.5.1 Configuring the Windows firewall on the reference system 

Before starting an image creation process, you must configure Windows firewall 

with Advanced Security on the reference system. 



During system preparation, Deployment Manager tries to access the reference 

system with WMI and copies the Client Agent package (for temporary 

installation) to the reference system. If the reference system has Windows 

firewall switched on (which is the default after a Windows 2008 or 

Windows 2012 installation), perform the following steps on the reference system 

before starting the image creation process: 

1. Select Start - Administrative Tools - Windows Firewall with 

Advanced Security. 

2. Select Inbound Rules in the left pane. 

3. Enable the following Inbound Rules (set to Yes): 

File and Printer Sharing (SMB-In) 

Windows Management Instrumentation (DCOM-In) 

Windows Management Instrumentation (WMI-In) 

In a German installation the rules are called: 



Datei- und Druckerfreigabe (SMB eingehend) 

Windows-Verwaltungsinstrumentation (DCOM eingehend) 

Windows-Verwaltungsinstrumentation (WMI eingehend) 

Instead of performing these steps you can enter the following commands 

from a command prompt (replace the names of the rules with your language): 

netsh advfirewall firewall set rule name="File and Printer Sharing (SMB-In)" 

new enable=yes 

netsh advfirewall firewall set rule name="Windows Management Instrumentation 

(DCOM-In)" new enable=yes 

netsh advfirewall firewall set rule name="Windows Management Instrumentation 

(WMI-In)" new enable=yes 

The above commands enable the firewall rules for all profiles. If you 

want to enable the firewall rules for one profile only, please add the 

parameter profile to the netsh command (e.g. netsh advfirewall firewall 

set rule name="File and Printer Sharing (SMBIn)" new enable=yes 

profile=Domain). 

The System Access Data field in Disks step (Create Cloning Image wizard)" 

(see section "Disks step (Create Cloning Image wizard)" on page 172) 

can contain the IP address or the name of the server. If a server name is 

entered here, the following Inbound Rules must be enabled in addition: 

File and Printer Sharing (NB-Datagram-In) 

File and Printer Sharing (NB-Name-In) 

File and Printer Sharing (NB-Session-In) 

5.5.2 Sysprep restrictions 

Note the following restrictions for Sysprep: 

l Sysprep can only be called 3 times on Windows 2008 or Windows 

2012 

Make sure that Sysprep has been called no more than twice on the reference 

system otherwise Sysprep will crash during the deployment 


process. 

l Not all server roles support Sysprep. 

The following server roles will no longer function after deployment of 

Windows 2008 R2 or Windows 2012: 

o Active Directory Certificate Server (AD CS) 

o Active Directory Domain Services (AD DS) 

o Active Directory Federation Services (AD FS) 

o Active Directory Lightweight Directory Services (AD LDS) 

o Active Directory Rights Management Server (AD RMS) 

o DHCP Server 

o DNS Server 

o Fax Server 

o Network Policy and Access Services 

o UDDI Services 

o Windows Deployment Services 

The following server roles will no longer function after deployment of 

Windows 2008: 

o Active Directory Certificate Server (AD CS) 

o Active Directory Domain Services (AD DS) 

o Active Directory Federation Services (AD FS) 

o Active Directory Lightweight Directory Services (AD LDS) 

o Active Directory Rights Management Server (AD RMS) 

o DNS Server 

o Fax Server 

o File Services 

o Hyper-V 


o Network Policy and Access Services 



o Print Services 

o UDDI Services 

o Windows Deployment Services 

The following server roles will continue to function with restrictions 

after deployment of Windows 2008, Windows 2008 R2 or Windows 

2012: 

o Terminal Services/Remote Desktop Services 

Not supported in scenarios where the master Windows image 

is joined to a domain. 

o Web Server (Internet Information Services) 

Does not support Sysprep with encrypted credentials in applicationhost.config. 


6 Mass Cloning 

The work area Mass Cloning allows you to: 

l Create cloning images. 

l Create cloning groups (an image is assigned to the group). 

l Clone one or more servers with a cloning image. 

l Clone a group or members of a group. 

If the target server has an UEFI BIOS, each LAN port must be manually 

configured for either Legacy PXE or UEFI PXE boot. If the target 

server has a GPT boot disk, UEFI PXE boot must be 

configured. Otherwise Legacy PXE or UEFI PXE boot can be configured. 

If the target server has an UEFI BIOS, UEFI: PXE boot (or Legacy 

Boot - if Legacy PXE boot is configured) must be set before UEFI 

Shell in BIOS boot order. 

6.1 Creating a cloning image 

You can create cloning images. A cloning image can be cloned to a group of 

servers to be installed with the same configuration parameters. You can also 

clone an image without specifying a deployment group. To do this, select a 

server from the list and select Clone (Servers by Cloning Image group) or 

Clone with Image (All Servers group) from the context menu. 

After starting the image creation process (by clicking on Finish), the cloning 

module now boots the reference system referenced by its MAC address with 

a WinPE boot image via PXE and starts the image creation tool on WinPE. 



The automatic remote power-on and the remote PXE boot configuration 

of the BIOS require that the server can be managed by a 

management blade (blade system), an RSB or a BMC that supports 

IPMI 1.5 over LAN. For non-blade systems you must have specified 

the correct BMC settings. If the server cannot be managed in one of 

these ways, a dialog appears asking you to initiate a PXE boot manually. 

Scheduled tasks for non-manageable servers are not possible. 

6.1.1 Create Cloning Image wizard 

The Create Cloning Image wizard allows you to create cloning images. 

To open the wizard, select Create Cloning Image from the context menu of 

a selected server in the Servers view. 

6.1.1.1 Task Name step (Create Cloning Image wizard) 

Task Name is the first step in the Create Cloning Image wizard. In this 

step you must enter a task name. 


Figure 38: Task Name step 

Task Name 

Name of the task. Any string except the characters & and " is allowed. 

Image Information 


the image. This information will be stored in the file with the suffix .txt. 

6.1.1.2 Deployment Server step (Create Cloning Image wizard) 


Deployment Server is the next step in the Create Cloning Image wizard. 

In this step you can select the deployment server. 



Figure 39: Deployment Server step 

Deployment Server 

Name of the deployment server which is used for this task. 

6.1.1.3 Image Path and Name step (Create Cloning Image wizard) 

Image Path and Name is the next step in the Create Cloning Image wizard. 

In this step you can define the location (path) and the name of the image. 


Figure 40: Image Path and Name step 

Disk Images 

Displays existing image repositories in the tree view on the left. The table 

on the right displays the images in the repository. Select a folder in the 

repository. 

Image Filename 

Name of the image. You can use the following variables in the file name: 

%D This is replaced by the day (01, 02, ..., 31) of the month. 

%M This is replaced by the month (01, 02, ...,12). 

%Y This is replaced by the year (4 digits). 


%S This is replaced by the name of each server of which an image will 

be created. 



6.1.1.4 Options step (Create Cloning Image wizard) 

Options is the next step in the Create Cloning Image wizard. In this step 

you can specify server options before and after deployment. 

Figure 41: Options step 

Force System Shutdown before Deployment 

Force a system shutdown before creating the image. Operating-systemspecific 

ServerView agents must be installed on the relevant server. 

Shutdown Method 

Shutdown method: 


Graceful 

(via ServerView Agents) 

This shutdown method makes use of 

the ServerView agent functionality. 

ServerView agents running on the target 

systems are instructed to shut down the 

systems. This method requires Server- 

View agents to be installed and configured 

on all systems. It also requires 

specification of a shutdown user name 

and password. 

ACPI Deployment Manager will try to perform 

an ACPI shutdown of the system. If the 

running operating system supports 

ACPI, it should save unsaved data and 

clean up the file systems. 

Forced A hard power-off of all running target 

servers will be initiated if these servers 

support remote power-off by management 

blade or baseboard management 

controller. In this case 

unsaved data will be lost. Normally this 

is not a problem, since the disk will be 

overwritten by a new disk image. 

Shutdown Username 

Preconfigured user name. 

Shutdown Password/Repeat Shutdown Password 

Preconfigured user password. 


The shutdown user name and password are used for authentication 

of this shutdown request at the ServerView agent of the 

corresponding server. The ServerView Deployment Manager 

service remembers the last-used shutdown user name and password. 

These will be set as default shutdown user name and password. 

The last-used shutdown user name and password are lost 

when the ServerView Deployment Manager service is restarted. 



System Status after Deployment 

System status after creating the image: 

Shut down System The server will be shut down at the end 

of the process. 

Keep System Running The server will be kept running at the 

end of the process. 

GUID Check 

The GUID of a server as specified in the deployment configuration of that 

server is checked against the GUID that is sent as part of the PXE boot 

request sent by a server. The deployment of the server will only start if 

the GUID sent by the PXE boot request matches the configured GUID. 

6.1.1.5 Disks step (Create Cloning Image wizard) 

Disks is the next step in the Create Cloning Image wizard. In this step you 

can specify the disk from which the image will be saved. You can only select 

one disk. 


Figure 42: Disks step 


Logical Disk Number 

Select the logical disk number from which the image is to be created. By 

default the logical disk 0 is saved, which is the first logical device. 

If there are local disks and SAN disks connected to the system, 

the disk numbering depends on the driver load order (for deployment 

platform WinPE MDP). In most configurations with an Emulex 

board (e.g. in combination with an LSI SAS IME board) the 

SAN disks are listed before the local disks - if you accept the 

default value 0 for the logical disk number, the first SAN disk will 

be saved. As a workaround you could unplug the Fibre Channel 

cables before starting image creation. 



Raw Mode 

You can generate an image creation in raw disk mode. All sectors of the 

disks will be saved in the image. This image can only be deployed to 

servers which have disks with the same size and the same disk layout. 

File System Independent 

This method is used for non-supported file systems. When selected, the 


cloned to a partition of the same size. For more information see section 

"File-System-Independent image creation (raw image creation)" on page 

155. 

File System Dependent 

When selected, only the used blocks of a file system will be saved in the 

image. For more information see section "File-System-Dependent image 


This functionality is supported for the following file systems: 

l FAT, FAT32 

l NTFS, NTFS5 

l EXT2, EXT3, EXT4 

l Reiser FS version 3.5 and 3.6 on SuSE SLES 9 and 10. 

Other file systems and especially other versions of the Reiser file system 

are not supported by this functionality. 

If a disk contains several partitions for non-supported file system types, 

the whole partition including the non-used blocks will be saved. For the 

partitions with a supported file system, only the used blocks will be 

saved. This reduces the size of the image files and also increases the 

speed of the image generation. If you want to restore an image on a disk 

that has a different size than the source disk, this option should be specified. 

Verify File System 

The file system is checked before the image creation process is started. 


Raise Error for Unsupported File Systems 

Only use this option if the File System Dependent option is checked. If 

an unsupported file system is detected, the image creation process will 

raise an error and stop. 

Compress Image 

The image file is compressed during the image creation. 


Fast Image Creation 

When this option is checked, the image creation process uses an optimized 

method to access the target disk for the image creation. This optimization 

can be applied to EXT2, EXT3 and EXT4 file systems when the 

option File System Dependent is checked. For an image that was created 

with this option, the option Expand last partition to whole disk in 

the System Preparation step of the cloning is not available because 

applying the option Expand last partition to whole disk to such an 

image would raise an error. 

Operating System 

Specify the operating system installed on the reference system: Linux, 

Windows 2003, Windows 2008 or Windows 2012 

l Image creation of a Windows reference system only 

works if the interface used for the PXE boot is enabled on 

the reference system. 

l Before you can create an image of a Windows 2008 or 

Windows 2012 reference system, the following 3 

"Inbound Rules" must be enabled in Windows firewall on 

the reference system: File and Printer Sharing (SMB-In), 

Windows Management Instrumentation (DCOM-In) and 

Windows Management Instrumentation (WMI-In). 



l If an image of a Windows 2008 R2 or Windows 2012 reference 

system with static IP addresses was created, the 

following IP configurations on the reference system are 

lost and must be manually restored: Default Gateway, 

DNS Servers, WINS Server and NetBIOS over TcpIp. 

l Image Creation of a Windows 2008 R2 or Windows 2012 

reference system with static IP addresses fails if there is 

a gateway between the reference system and the deployment 

server. During rollback the reference system tries to 

connect back to the deployment server - this fails 

because the Default Gateway was not restored (see 

above). As a workaround please save the .img, .cfg and 

.txt files away while the rollback is still running - the image 

is okay and can be used for cloning. If you created an 

image of a GPT boot disk, please manually add an empty 

section NVRAM at the end of the .cfg file (this allows you 

to select the NVRAM Information Restoration during 

cloning). 

Note the following restriction for some PRIMERGY servers and 

server blades (e.g. TX150 S7, TX300 S5, TX200 S6, BX924 S3) 

and for PRIMEQUEST: 

If iRMC Support , MMB SNMP Support or MMB Remote 

Manager Support is selected in the Remote Management 

Ports step in the Deployment Configuration wizard, the local 

hard disk must be set before PXE in BIOS boot order. Otherwise 

the image creation of a Windows reference system fails because 

WinPE instead of the Windows operating system will be booted 

first. This restriction does not apply for Wake On LAN Support or 

Manual Management when PXE is required to be on first place in 

BIOS boot order. 

Additional Parameters for Linux: 


Location of /etc 

Specify the location of the /etc directory for the system from which you 

want to generate the image. By default this is /dev/sda2. 

Additional Parameters for Windows 2003: 

System Access Data 

Displays either the IP address or the host name from the selected server. 

This value is used to access this system during system preparation. 

Administrator Account 

Name of an account with administrator’s rights. This must be a local 

account (not a domain account) on the reference system. In some situations 

(e.g. when the reference system is a domain member) it is advisable 

to use the notation \Administrator instead of just 

Administrator, where is replaced by the host name of the reference 

system. 

Password / Repeat Password 

Administrator password of the specified administrator account on the reference 

system. 

Windows Product ID 

The product ID is mandatory for all Windows 2003 DVDs. 


Additional Parameters for Windows 2008 and Windows 2012: 

System Access Data 

Displays either the IP address or the host name from the selected server. 

This value is used to access this system during system preparation. 

Administrator Account 

Name of an account with administrator’s rights. This must be a local 

account (not a domain account) on the reference system. In some situations 

(e.g. when the reference system is a domain member) it is advisable 

to use the notation \Administrator instead of just 

Administrator, where is replaced by the host name of the reference 

system. 




Administrator password of the specified administrator account on the reference 

system. 

Installation Medium 

Installation medium that was used for the installation of the reference system: 

Fujitsu OEM Fujitsu OEM (SLP-2) multi-language DVD. 

Microsoft All other DVDs. 

Language Version 

Corresponding language version of the Fujitsu OEM DVD. Deployment 

Manager does not automatically recognize the installation language. 

Windows Product ID 

The product ID is mandatory for Windows 2008 and Windows 2012 Retail 

DVDs and Windows 2012 Volume DVDs. For Windows 2008 Volume 

DVDs please leave the Product ID empty. Do not enter a MAK (Multiple 

Activation Key) here. The MAK key can be entered during the image 

deployment. 

6.1.1.6 Bios Boot Type step (Create Cloning Image wizard) 

Bios Boot Type is the next step in the Create Cloning Image wizard. 

This step is only displayed when iRMC Support has been chosen as method 

for remote management in Deployment Configuration. The Bios Boot Type 

is needed to initiate the remote PXE boot of the target server. 

l The selection here is ignored if the target server does not 

have an UEFI BIOS. 

l Each LAN port must be manually configured for either Legacy 

PXE or UEFI PXE boot (in the BIOS of the target 

server). This configuration cannot be changed when initiating 

a PXE boot from the deployment server. The value that is 

specified here must match the manual configuration in the 

BIOS. 


Figure 43: Bios Boot Type step 

"PC compatible" Boot (legacy) 

Initiate a Legacy PXE boot at the target server. 

Extensible Firmware Interface Boot (EFI) 

Initiate an UEFI PXE boot at the target server. 

6.1.1.7 Scheduling step (Create Cloning Image wizard) 


Scheduling is the final step in the Create Cloning Image wizard. In this 

step you can create the image generation task as a scheduled task. 



Figure 44: Scheduling step 

Time Unit to perform this Task (only for later option) 

Specifies how often you want the task to be performed: Once, Daily, 

Weekly, Monthly. Additional settings are shown depending on your selection. 

Select Date and Time to start (only for later option) 

Specifies the start date (now or later) when you want the task to be performed. 

If you select later, you can also specify the start time for the 

task. 

Perform this Task (only for Daily) 

Selects whether the task should be performed Every Day, Weekdays or 

Every x days. 


Select days of the week you want this task to start (only for Weekly) 

Specifies that the task should be executed every week or every n week 

(n = 1 ..52). You can also select the days of the week. 

Select the day and the month you want this task to start (only for 

Monthly) 

Specifies for which months of the year the task should be executed. You 

can also specify on which day of the month a task should be executed. 

And you can specify a weekday and whether the task should be executed 

in the first, second, third, fourth or last week of a month. 

Retry Count 

Specifies the number of retries for this task if the job fails. A value 

between 0 and 10 can be specified. 

Retry Interval (minutes) (only for later option) 

The time (in minutes) before the next attempt if previous attempts at a 

task failed and the number of retries does not exceed the retry counter. A 

value between 1 and 360 minutes can be specified. 

Task Start Time Window (minutes) (only for later option) 

This value defines the maximum period (in minutes) for retries of a task. 

Outside this period no retries will be performed, even if all the retries specified 

by the retry counter have not been used. 

It can happen that a task is handed over to the deployment server but is 

only queued, because too many other tasks are already running. If the 

Task Start Time Window elapses while a task is in this state, the task 

will also be canceled. A running task will not be canceled when the Task 

Start Time Window elapses. 

Enable Segmentation 

Enable the option to specify the segmentation size. 


Segmentation Size 

Specifies the maximum number of targets that will be deployed in parallel. 

This might be useful when a job with a huge number of targets is created. 



6.1.1.8 Action after starting the "Create Cloning Image" task 

Basically, the target server behaves as follows after starting the Creating 

Cloning Image wizard. Please also refer to "Image creation of a Windows 

reference system " on page 158. 

Windows reference system 

When saving an image from a Windows Reference system, it is necessary 

to first perform a local boot of the reference system to install a “Client Agent” 

package. The following steps are done: 

1. After starting the task, power on the client. (Please see "Notes".) 

2. The client is sending a PXE request. However, this request is canceled. 

3. The client tries the next boot device (local hard drive) and boots the 

Windows operating system. 

4. Copy the Client Agent package from the deployment server to the 

Windows operating system which has just booted. 

5. After copying the Client Agent package, reboot the Windows operating 

system. 

6. The client is booted with WinPE. 

7. Start saving an image. 

8. After saving an image, reboot the client. 

9. The client boots the Windows operating system and deletes the 

Client Agent package. 

10. The client is shutdown. 

Linux reference system 

When saving an image from a Linux Reference system, the following steps 

are done: 

1. After starting the task, power on the client. (Please see “Notes”.) 

2. The client is sending a PXE request. 




5. After saving an image, shutdown the client. 


l When you select iRMC Support, MMB SNMP Support, 

MMB Remote Manager Support, or Wake on LAN Support 

in the Remote Management Ports step in the Deployment 

Configuration wizard, Deployment Manager turns on 

the client power automatically after starting the task. When 

you select Manual Management, you need to turn on the 

client manually. 

l When you select iRMC Support, MMB SNMP Support, or 

MMB Remote Manager Support in the Remote Management 

Ports step in the Deployment Configuration wizard, 

PXE is automatically set on first place in BIOS boot 

order when starting the client. 

l When you select Wake on LAN Support or Manual Management 

in the Remote Management Ports step in the 

Deployment Configuration wizard, you need to set PXE 

on first place in BIOS boot order of the client before you start 

the task. 

l If you select iRMC Support, MMB SNMP Support or 



you need to turn off a target server before you start the 

deployment. You can perform the shutdown manually or by 

using the Force System Shutdown before Deployment 

option in the task you want to execute. 






you need to set the user name and the password for 

iRMC or MMB in the Deployment Configuration wizard 

before starting the task. 

l When you select Keep System Running in the System 

Status after Deployment option, the client reboots after finishing 

the task. 

l When you execute a customer script, the client doesn't perform 

a shutdown after finishing the task. You need to perform 

a shutdown by your script or yourself. 


6.2 Cloning groups 

Before you start the cloning process you need to consider which servers are 

to get which image. Then you can build cloning groups consisting of one or 

more servers and one image file. 

Cloning groups with several servers help to reduce the workload of an administrator 

by organizing servers with the same configuration in one logical 

group. The servers getting the same image can then be cloned simultaneously 

by choosing only one deployment group. 

You can add, copy, edit or delete cloning groups. 

You can also clone an image without specifying a deployment group. 

To do this, select a server from the list and select Clone with Image 

from the context menu. 

6.2.1 Adding a cloning group 

The Add Cloning Group wizard allows you to create cloning groups. 

To open the wizard, select Add Cloning Group from the context menu in 

the Servers by Cloning Image group. 

6.2.1.1 Group Name step (Add Cloning Group wizard) 


Group Name is the first step in the Add Cloning Group wizard. In this step 

you can enter a group name. 



Figure 45: Group Name step 

Group Name 

Name of the group. 

Description 

Optional: Description of the group. 

6.2.1.2 Cloning Image step (Add Cloning Group wizard) 

Cloning Image is the next step in the Add Cloning Group wizard. In this 

step you can choose the image file for the deployment. 


Figure 46: Cloning Image step 

Image File 

Displays the selected image. 

NVRAM Information Restoration 

During image creation the NVRAM boot entries are saved if an UEFI 

PXE boot of a reference system with GPT boot disk is done. The .cfg file 

contains a section NVRAM if boot entries have been saved. You can 

select here whether the NVRAM boot entries should be restored on the 

target. 

Update boot 

entries (delete 

the original 

entries) 


The boot entries from the reference system are restored 

on the target. The original entries on the target are deleted. 

You should select this option when restoring a boot 

disk. 



Do not restore 

boot entries 

You should select this option when restoring a data 

disk. 

6.2.1.3 Group Members step (Add Cloning Group wizard) 

Group Members is the final step in the Add Cloning Group wizard. In this 

step you can select the servers for the group. 

Figure 47: Group Members step 

Show compatible servers only 

The Available Servers list displays all servers which are compatible. 

>> 

Copies all servers from the Available Servers list to the Selected 

Servers list. 




Group Name 


Description 


6.2.2.2 Cloning Image step (Copy Cloning Group wizard) 

Cloning Image is the next step in the Copy Cloning Group wizard. In this 




Image File 







target. 

Update boot 


the original 

entries) 





disk. 




boot entries 


disk. 

6.2.2.3 Group Members step (Copy Cloning Group wizard) 

Group Members is the final step in the Copy Cloning Group wizard. In 

this step you can select the servers for the group. 




>> 


Servers list. 




Group Name 


Description 


6.2.3.2 Cloning Image step (Edit Cloning Group wizard) 

Cloning Image is the next step in the Edit Cloning Group wizard. In this 




Image File 







target. 

Update boot 


the original 

entries) 





disk. 




boot entries 


disk. 

6.2.3.3 Group Members step (Edit Cloning Group wizard) 

Group Members is the final step in the Edit Cloning Group wizard. In this 





>> 


Servers list. 



6.3.1 Clone wizard 

The Clone wizard allows you to clone a group of servers with an image. 

To open the wizard, select Clone from the context menu in the Servers by 

Cloning Image group. 

6.3.1.1 Task Name step (Clone wizard) 

Task Name is the first step in the Clone wizard. In this step you can enter a 

task name. 


Task Name 



6.3.1.2 Deployment Server step (Clone wizard) 

Deployment Server is the next step in the Clone wizard. In this step you 

can select the deployment server. 




6.3.1.3 Disks step (Clone wizard) 

6.3 Cloning an image 

Disks is the next step in the Clone wizard. In this step you can select the 

disk number to which the image is to be restored. 




Restore to Logical Disk Number 

Optional: Select the disk on which the image is to be deployed. 







be overwritten. As a workaround you could unplug the Fibre 

Channel cables before starting the cloning. 

Partitions 

In the first column select whether the partition should be restored. 




Volume Label Volume label 

Attributes Attributes (and Active flag) 

Format File System 

Size Size of the partition 

Used Percentage of used space 

When the system partition is restored, all partitions before the 

system partition (for example the ESP and MSR partitions for a 

GPT disk) must be restored as well. 

6.3.1.4 System Preparation step (Clone wizard) 


System Preparation is the next step in the Clone wizard. In this step you 

can select the system preparation before image deployment. 



Figure 57: System Preparation step 

System Preparation Method 

Specify a system preparation method. Depending on the selection, further 

parameters are displayed: 

Unchanged 

When this method is selected, no further system preparation will be done 

before the disk image is cloned to the target servers. This requires that a system 

preparation (configuration of the RAID controller) was done manually for 

all target servers before the image cloning is started. 

All Primergy 

This method allows a RAID configuration for all RAID controllers supported 

by Installation Manager. 


The following parameters are displayed: 

Controller Vendor 

Select the controller vendor. 

Controller Family 

Select the controller family. 

Controller Model 

Select the controller model. 

Controller Number 

Select the controller number. 

Manual Configuration 

You can define additional parameters: 

Raid Level 

Select the RAID level. 

Number of Disks 

Select the number of disks. 

Use Hot Spare 

A standby hard disk can be used to replace a defective hard disk. 

l To create a RAID array on an SX940 storage blade with a 

Lynx controller, select LSI/Mylex RAID Controller as the 

controller vendor, LSI IME SAS as the controller family, 

Any as the controller model, and 1 as the controller 

number. 

l To create a RAID array on an SX940 storage blade with 

a Cougar controller, select LSI/Mylex RAID Controller as 

the controller vendor, LSI MegaRAID SAS as the controller 

family, RAID 5/6 SAS based on LSI MegaRAID as 

the controller model, and 0 as the controller number. 

Preparation Boot Image 





Boot Image Path 

Displays the path where the selected image is stored. 

Boot Image Type 

Select the boot image type: 

Floppy Disk Image The boot image is a "virtual" floppy image (1 

file of size 1.4 MB). 

Bootstrap Image The boot image consists of multiple parts, 

whereby the first part is PXE booted and 

requests the rest by tftp download. 

Expand last partition to whole disk 

The last partition of the restored disk image will be expanded to use the 

rest of the whole disk. 

6.3.1.5 Settings step (Clone wizard) 

Settings is the next step in the Clone wizard. This step allows you to specify 

the deployment method and the system shutdown method before cloning 

the image. 


Figure 58: Settings step 

Deployment Method 

Deployment method for the cloning: 


Multicast The disk image data is sent via multicast IP packets over 

the network. This means that a data packet is sent to all target 

servers involved once and not n times (where n is the 

number of servers involved). This allows you to deploy a 

large number of servers in a short time. You might experience 

problems with the multicast method if there is a router 

between the deployment server and a target server, since 

not all routers allow multicast transfer. 



Unicast The disk image data is sent via unicast IP packets over the 

network. This means that a data packet is sent to each 

server. When using the unicast deployment method, only 4 

servers can be deployed in parallel. If your deploy job 

involves more servers, the deployment of the other servers 

is queued. The other servers will be deployed as soon as the 

deployment of one server has finished. 


Deployment Manager will try to shut down the affected target systems 

before the deployment process is started: 

l When this option is checked, the target servers will be shut down 

before the deployment is started. 

l When this option is not checked, the target servers must be shut 

down, otherwise the deployment will fail. 

Depending on the selected shutdown method, a shutdown might 

be a hard power-off or a graceful shutdown. In the case of a hard 

power-off, unsaved data will be lost. This might not be a problem 

since the disk will be overwritten by a new disk image. But data 

on a second logical disk might be lost. 


Select the shutdown method: 

Forced A hard power-off of all running target servers will be initiated 

if these servers support remote power-off by management 

blade or baseboard management controller. In 

this case unsaved data will be lost. Normally this is not a 

problem, since the disk will be overwritten by a new disk 

image. 

ACPI Deployment Manager will try to perform an ACPI shutdown 

of the system. If the running operating system supports 

ACPI, it should save unsaved data and clean up the 

file systems. 


Graceful 

(via Server- 

View 

Agent) 

6.3.1.6 Post Deployment step (Clone wizard) 

This shutdown method makes use of the ServerView 

agent functionality. ServerView agents running on the target 

systems are instructed to shut down the systems. 

This method requires that ServerView agents are installed 

on all systems and are configured to allow this functionality. 

It also requires the specification of a shutdown 

user name and password. 

Post Deployment is the next step in the Clone wizard. This step allows you 

to specify options which are used after cloning. 

Figure 59: Post Deployment step 




Customer Script 

Specifies a customer script that is executed on the cloned system at the 

end of the deployment process. You can use this to modify and configure 

the cloned system for your requirements. In order to specify a customer 

script, you must specify its path name as follows: 

l For Windows target systems: 

The path name should be the network path name of the script in 

UNC notation. This means that you must use a network share 

where you store the script. You can also specify optional parameters 

for this script. 

Example: 

\\server\share\path\script_name [parameters] 

l For Linux target systems: 

The path name has the following format: 

\\server\path\script_name [parameters] 

//server/path/script_name [parameters] 

server FTP server name or the IP address of the FTP server. 

path Path to the script relative to the root of the FTP server. 

If you select the Customer Script option, the System Status after 

Deployment option is ignored. If the user wants to reboot after 

executing the customer script, you must add a reboot command at 

the end of the script. 

Windows Domain User 

If you deploy a Windows image (not a snapshot image), you can specify a 

Windows domain. 


In order to use the domain join functionality, the program netdom.exe 

from the ServerView Suite DVD 1 or a Windows installation 

CD (support directory 

SUPPORT\TOOLS\SUPPORT.CAB) must be copied to the following 

directories: 

Windows 2003 and Windows 2008: 

C:\Program Files\Fujitsu\DeploymentService\ 

PMKit\Win2003\boot\ipadj 


Windows 2003 X64, Windows 2008 X64 and Windows 2012: 


PMKit\Win2003x64\boot\ipadj 

The ServerView Suite DVD 1 contains the directory RD_FILES, 

which contains the corresponding files for Windows 

2003/2008/2012 (32-bit version only). During installation of Deployment 

Manager you can optionally copy the file to the deployment 

server. In this case you are asked to specify the location of the 

file. 

The domain join functionality can only work with images that were 

generated with RemoteDeploy as of version 3.30, and only if the 

file was copied to the correct directory before generating the 

image. The file must be part of the image in order to work correctly. 

If the file is available on the deployment server, it is automatically 

copied to the image during image generation. If the file is not part 

of the image, the deployment job will time out and you will see 

more information on the deployed target. 

Domain Name 

If this field is not empty, the post-preparation process running on the 

cloned Windows systems will try to join the specified Windows domain. 

User Name 

Specifies an account to be used for the domain join operation. This must 

be an account that has the right to perform a domain join on the domain 

server. By default this is a domain administrator account. 




Password/Repeat password for the account. 

Windows 2008 and Windows 2012 Volume Activation 

If you deploy a Windows 2008 or Windows 2012 image created from a volume 

installation, the cloned image can be automatically activated after 

the cloning process: 

Activation 

Method 

Internet 

Proxy 

KMS 

Server 

Select the activation method: MAK (Multiple Activation 

Key) or KMS (Key Management Service). 

Optional: Specify a proxy server (IP address or name) to 

connect to the Internet. If you want to specify a proxy 

server by using a hostname, you need to select the is a 

hostname option, too. You can also enter a port number. 

If you select the activation method KMS (Key Management 

Service), you must specify the IP address or the 

name of the KMS server. Optionally, you can enter the 

KMS port number. 

Key If you select the activation method MAK (Multiple Activation 

Key), enter the activation key. 

Set hostname in /etc/hosts 

If you deploy a Linux Red Hat image, the line with 127.0.0.1 in /etc/hosts 

is overwritten after cloning. 

For Do not set: 127.0.0.1 localhost.localdomain localhost 

For Set: 127.0.0.1 . 


System status of the target systems after the cloning: 

Shut down System The target servers involved will be shut 

down at the end of the process. 

Keep System Running The target servers involved will be kept 

running at the end of the process. 


6.3.1.7 Bios Boot Type step (Clone wizard) 

Bios Boot Type is the next step in the Clone wizard. 












BIOS. 








6.3.1.8 Scheduling step (Clone wizard) 

Scheduling is the final step in the Clone wizard. This step allows you to 

clone the image as a scheduled task. 










task. 



Every x days. 







Monthly) 





Retry Count 






















6.3.1.9 Action after starting the "Clone" task 

Basically, the target server behaves as follows after starting the Clone wizard. 


When restoring an image of Windows Reference system, the following steps 

are done: 




4. Start restoring an image. 

5. After restoring an image, reboot the client. 

6. The client boots the Windows operating system, starts the Client 

Agent, and executes Sysprep. 

7. After executing Sysprep, the client reboots. 

8. Start Windows setup 

9. The client reboots after Windows setup 

10. The client boots the Windows operating system and start Client 

Agent again. 



When restoring an image of Linux Reference system, the following steps are 

done: 





5. After restoring an image, shutdown the client. 




















the task. 
















the task. 


6.3.2 Clone with Image wizard 




The Clone with Image wizard allows you to clone a single server with an 

image. 

To open the wizard, select Clone with Image from the context menu in the 

All Servers group. 

6.3.2.1 Task Name step (Clone with Image wizard) 

Task Name is the first step in the Clone with Image wizard. In this step you 

can enter a task name. 





Task Name 


6.3.2.2 Deployment Server step (Clone with Image wizard) 

Deployment Server is the next step in the Clone with Image wizard. In this 

step you can select the deployment server. 




6.3.2.3 Disk Image step (Clone with Image wizard) 

Disk Image is the next step in the Clone with Image wizard. In this step 

you can select the disk image. 


Figure 64: Disk Image step 


Show compatible images only 

Displays all compatible images. Only valid for cloning of single servers 

without a group. 

Every server is identified by a D number (e.g. D2860 for PRIM- 

ERGY BX920). This D number is written to the .cfg file during 

image creation. An image is regarded as "compatible" with a server 

if the D number of the server is the same as the D number in the 

.cfg file of the image. Deployment Manager performs no additional 

compatibility checks. Servers with the same D number are handled 

as "image compatible". The D number of each server can be 

obtained from the serial number (Syyyy-Dxxxx) or from the BIOS 

welcome message on console at BIOS Boot phase. 



PRIMEQUEST cannot use this function because PRI- 

MEQUEST does not have a D number. 

Disk Images 


on the right displays the images in the repository. Select the image from 

the repository. 

Image File 







target. 

Update boot 


the original 

entries) 


boot entries 

6.3.2.4 Disks step (Clone with Image wizard) 




disk. 


disk. 

Disks is the next step in the Clone with Image wizard. In this step you can 

select the logical disk number. 






the disk numbering depends on the driver load order (for 

deployment platform WinPE MDP). In most configurations with 

an Emulex board (e.g. in combination with an LSI SAS IME 

board) the SAN disks are listed before the local disks - if you 

accept the default value 0 for the logical disk number, the first 

SAN disk will be overwritten. As a workaround you could 

unplug the Fibre Channel cables before starting the cloning. 

Partitions 















6.3.2.5 System Preparation step (Clone with Image wizard) 

System Preparation is the next step in the Clone with Image wizard. In 

this step you can select the system preparation before image deployment. 



Unchanged 





All Primergy 

















Raid Level 




Use Hot Spare 


l To create a RAID array on an SX940 storage blade with a 

Lynx controller, select LSI/Mylex RAID Controller as the 

controller vendor, LSI IME SAS as the controller family, 

Any as the controller model, and 1 as the controller 

number. 


a Cougar controller, select LSI/Mylex RAID Controller as 

the controller vendor, LSI MegaRAID SAS as the controller 

family, RAID 5/6 SAS based on LSI MegaRAID as 

the controller model, and 0 as the controller number. 







Select the boot image type: Floppy Disk Image or Bootstrap Image. 




6.3.2.6 Settings step (Clone with Image wizard) 

Settings is the next step in the Clone with Image wizard. This step allows 

you to specify the deployment method and the system shutdown method 

before cloning the image. 










































image. 




file systems. 

Graceful 

(via Server- 

View 

Agent) 






















6.3.2.7 Post Deployment step (Clone with Image wizard) 

Post Deployment is the next step in the Clone with Image wizard. This 

step allows you to specify options which are used after cloning. 


Customer Script 

Specifies a customer script that is executed on the cloned system at the 

end of the deployment process. You can use this to modify and configure 

the cloned system for your requirements. In order to specify a customer 

script, you must specify its path name as follows: 

l For Windows target systems: 

The path name should be the network path name of the script in 


UNC notation. This means that you must use a network share 

where you store the script. You can also specify optional parameters 

for this script. 

Example: 

\\server\share\path\script_name [parameters] 

l For Linux target systems: 

The path name has the following format: 

\\server\path\script_name [parameters] 

//server/path/script_name [parameters] 


server FTP server name or the IP address of the FTP server. 

path Path to the script relative to the root of the FTP server. 

If you select the Customer Script option, the System Status after 

Deployment option is ignored. If the user wants to reboot after 

executing the customer script, you must add a reboot command at 

the end of the script. 

Windows Domain User 

If you deploy a Windows image (not a snapshot image), you can specify a 

Windows domain. 



In order to use the domain join functionality, the program netdom.exe 

from the ServerView Suite DVD 1 or a Windows installation 

CD (support directory 

SUPPORT\TOOLS\SUPPORT.CAB) must be copied to the following 

directories: 

Windows 2003 and Windows 2008: 


PMKit\Win2003\boot\ipadj 

Windows 2003 X64, Windows 2008 X64 and Windows 2012: 


PMKit\Win2003x64\boot\ipadj 

The ServerView Suite DVD 1 contains the directory RD_FILES, 

which contains the corresponding files for Windows 

2003/2008/2012 (32-bit version only). During installation of Deployment 

Manager you can optionally copy the file to the deployment 

server. In this case you are asked to specify the location of the 

file. 

The domain join functionality can only work with images that were 

generated with RemoteDeploy as of version 3.30, and only if the 

file was copied to the correct directory before generating the 

image. The file must be part of the image in order to work correctly. 

If the file is available on the deployment server, it is automatically 

copied to the image during image generation. If the file is not part 

of the image, the deployment job will time out and you will see 

more information on the deployed target. 

Domain Name 

If this field is not empty, the post-preparation process running on the 

cloned Windows systems will try to join the specified Windows domain. 

User Name 

Specifies an account to be used for the domain join operation. This must 

be an account that has the right to perform a domain join on the domain 

server. By default this is a domain administrator account. 



Password/Repeat password for the account. 

Windows 2008 and Windows 2012 Volume Activation 

If you deploy a Windows 2008 or Windows 2012 image created from a volume 

installation, the cloned image can be automatically activated after 

the cloning process: 

Activation 

Method 

Internet 

Proxy 

KMS 

Server 

Select the activation method: MAK (Multiple Activation 

Key) or KMS (Key Management Service). 

Optional: Specify a proxy server (IP address or name) to 

connect to the Internet. If you want to specify a proxy 

server by using a hostname, you need to select the is a 

hostname option, too. You can also enter a port number. 

If you select the activation method KMS (Key Management 

Service), you must specify the IP address or the 

name of the KMS server. Optionally, you can enter the 

KMS port number. 

Key If you select the activation method MAK (Multiple Activation 

Key), enter the activation key. 

Set hostname in /etc/hosts 

If you deploy a Linux Red Hat image, the line with 127.0.0.1 in /etc/hosts 

is overwritten after cloning. 

For Do not set: 127.0.0.1 localhost.localdomain localhost 

For Set: 127.0.0.1 . 










6.3.2.8 Bios Boot Type step (Clone with Image wizard) 

Bios Boot Type is the next step in the Clone with Image wizard. 











BIOS. 







6.3.2.9 Scheduling step (Clone with Image wizard) 


Scheduling is the final step in the Clone with Image wizard. This step 

allows you to clone the image as a scheduled task. 










task. 



Every x days. 






Monthly) 





Retry Count 
























6.3.2.10 Action after starting the "Clone with Image" task 

Basically, the target server behaves as follows after starting the Clone with 

Image wizard. 


When restoring an image of Windows Reference system, the following steps 

are done: 





5. After restoring an image, reboot the client. 

6. The client boots the Windows operating system, starts the Client 

Agent, and executes Sysprep. 

7. After executing Sysprep, the client reboots. 

8. Start Windows setup 

9. The client reboots after Windows setup 

10. The client boots the Windows operating system and starts Client 

Agent again. 



When restoring an image of Linux Reference system, the following steps are 

done: 
























the task. 
















the task. 






6.4 Cloning/Installing Baseboard Management Controllers 

(BMCs) 

Baseboard Management Controllers (BMC) can be added to the ServerView 

server list via the server browser. A special BMC icon shows the BMC (only 

Fujitsu BMCs) in the ServerView server list. The icon of a BMC in the server 

list reflects the status of the Global Error LED. 

RemoteDeploy version 4.1 supports the cloning and installation of BMCs. 

To access a BMC, Deployment Manager requires a valid user ID with the following 

privileges: 

l Getting and setting the power status 

l Getting and setting the boot order 

The admin account on all current PRIMERGY servers has the required privileges. 

The BMC/iRMC settings cannot be set by using this function. The 

function can only clone/install the operating system to BMCs on the 

server list as a target servers. 

Cloning and installing older PRIMERGY servers 

The steps below are necessary if you want to clone or install one of the following 

older PRIMERGY servers. 

l RX200/TX200 

l RX220 

l RX300/TX300 

l RX100 S2 

l RX200 S2/TX200 S2 

l RX300 S2/TX300 S2 


l RX100 S3 

l TX150 S2 

l TX150 S3 

l TX150 S4 

In Japan, the PRIMERGY servers described above are not supported. 

1. Start Operations Manager, see the Operations Manager guide. 

2. Open the User Password Settings window by selecting Users/Passwords 

from the Administration menu. 

3. Delete the admin account. 

4. Add a new account with the required privileges. 

5. Add the admin account again. 

It is important that the privileged account is first in the search order. 

You can change the account for each server later in the Deployment Manager 

front-end via the Deployment Configuration wizard. 

6.4.1 Displaying BMCs 

6.4 Cloning/Installing Baseboard Management Controllers (BMCs) 

Displaying BMCs in Operations Manager 

The following window shows an example of BMC entries displayed in the 

Operations Manager ServerList window. 

Figure 71: BMC in Operations Manager 



Displaying BMCs in Deployment Manager 

The Deployment Manager database is periodically updated from the Server- 

View database. The BMC entries are displayed in the Servers view of the 

Deployment Manager front-end: 

Figure 72: BMC in Deployment Manager 

The host names generated for the BMCs are based on the last 6 bytes of the 

MAC address. 

BMCs can be added to deployment groups or installation groups. 

6.4.2 Changing the deployment configuration 

You can change the deployment configuration of a BMC in the Deployment 

Manager front-end. 

l Select a BMC in the All Servers view and choose Deployment Configuration 

from the context menu. The Deployment Configuration 

wizard opens. 

You can change the following settings: 

l In the General Settings step (first step) you should change the generated 

server name. 

l In the LAN Ports step (second step) DHCP is selected by default. 

You can enter a static IP address to replace the default deployment 

configuration. Click Add to add deployment configurations for add-on 

PCI boards. 

l The BMC Account Settings window (fifth window) displays the BMC 

configuration. You can change the user name and the password. 

Choose an account with the appropriate privilege. 


6.4 Cloning/Installing Baseboard Management Controllers (BMCs) 

A bare server entry will be added to the ServerView database. This entry 

takes the values from the BMC entry with the following changes: 

l The server name and system name are set to the generated server 

name or to the server name entered in the General Settings window. 

l The network address is set to "0.0.0.0". 

The GUID (UUID) is deleted in the ServerView database, but saved in the 

Deployment Manager database. After successful cloning or installation of the 

BMC, the GUID is set automatically in the ServerView database. 

In the ServerView database the BMC and the bare server entries exist in parallel. 

But in the ServerView server list only the BMC entry is displayed. 

In the Deployment Manager database, only the bare server entry exists. During 

the periodic update of the database, the ServerView Deployment Manager 

service matches the BMC entry from the ServerView database to the 

bare server entry. I.e. the Deployment Manager server list displays a bare 

server entry with the values of the BMC entry. 

Cloning/Installing a BMC without Changing the Deployment Configuration 

BMCs can be cloned or installed immediately without changing the deployment 

configuration. 

A default deployment configuration exists for every BMC: 

l Includes only the onboard interfaces (reported by the BMC). 

l DHCP is set for all interfaces. 

l The target system name is set to the generated server name. 

l The DNS suffix entry is empty. 

Once the cloning or installation process is started, a bare server with the generated 

server name is created. You cannot change the generated name later. 



6.4.3 Actions after cloning/installation 

The following actions are performed after successful cloning or installation of 

a BMC: 

l The ServerView Deployment Manager service changes the Bare- 

Server to type Server in the ServerView database. The following 

fields are also changed: 

o The UUID is set to the value from the corresponding BMC. 

o The FullQualifiedName is set. If a DNS suffix is specified in 

the Deployment Configuration wizard, the suffix is appended 

to the ServerName. Otherwise the FullQualifiedName is 

set to ServerName. 

l The Deployment Manager front-end immediately displays a server 

instead of a bare server in the Servers view. 

l Operations Manager initially displays a BMC entry and a server entry 

in parallel. 

l After cloning/installing, the server is booted with the installed Server- 

View agents. The ServerView Services service 

o contacts the DNS server with the given FullQualifiedName to 

get the network address of the server. 

o uses the network address to contact the ServerView agent on 

the server. 

o determines that the UUID delivered by the ServerView agent 

on the server is identical to the UUID of the BMC entry. The 

BMC entry is deleted from the ServerView database. 

l Finally, Operations Manager displays only the server entry. 


7 Mass Installation 

The Mass Installation work area allows you to install any number of target 

servers in parallel. 

If the target server has an UEFI BIOS, each LAN port must be manually 

configured for either Legacy PXE or UEFI PXE boot. If you 

configure UEFI PXE boot, the target server will be installed with a 

GPT boot disk, otherwise with an MBR boot disk. 

7.1 Installation Groups 

Before you create a new installation group you need to consider which 

servers should get which configuration file. You can build installation groups 

consisting of one or more servers (every server can be assigned a different 

configuration file - there is no configuration file assigned to the whole group). 

Installation groups with several servers help to reduce the workload of an 

administrator by organizing servers in one logical group. All servers in an 

installation group can be installed simultaneously. 



l If you assign configuration files to servers, make sure that 

these files are applicable for these servers. For example, an 

installation configuration file for a specific RAID controller 

requires that the installation target server contains this RAID 

controller. 

l The operating system which can be installed by Mass Installation 

is dependent on ServerView Installation Manager. Please 

see the manual of "ServerView Installation Manager" for 

details. 

l You need to create the installation configuration file from Server- 

View Installation Manager beforehand. For details, please refer 

to the manual for ServerView Installation Manager. 

l This functionality depends on ServerView Installation Manager. 

Therefore, you must meet the software requirements of Server- 

View Installation Manager. For example, when you want to 

install a Linux system, you need to configure a FTP, HTTP or 

NFS server. For details, please refer to the ServerView Installation 

Manager user guide. 

You can add, copy, edit or delete installation groups. 

7.1.1 Adding an installation group 

The Add Installation Group wizard allows you to create installation groups. 

To open the wizard, select Add Installation Group from the context menu 

in the Servers by Installation group. 

7.1.1.1 Group Name step (Add Installation Group wizard) 

Group Name is the first step in the Add Installation Group wizard. In this 

step you can enter a group name. 



Group Name 


Description 


7.1.1.2 Group Members step (Add Installation Group wizard) 


Group Members is the next step in the Add Installation Group wizard. In 





> 

< 

>> 

Copies the servers selected from the Available Servers list to the Selected 

Servers list. 

Copies the servers selected from the Selected Servers list back to the 

Available Servers list. 


Servers list. 



Assign >> 

Assigns the configuration file to the selected servers in the adjacent list. 

Assign to all 

Assigns the configuration file to all servers in the adjacent list. 

7.1.2 Copying an installation group 

The Copy Installation Group wizard allows you to copy an existing installation 

group which is to be used as a template. You can modify the setting to 

your own needs. 

To open the wizard, select Copy Installation Group from the context menu 

in the Servers by Installation group. 

7.1.2.1 Group Name step (Copy Installation Group wizard) 

Group Name is the first step in the Copy Installation Group wizard. In this 




Group Name 


Description 


7.1.2.2 Group Members step (Copy Installation Group wizard) 


Group Members is the next step in the Copy Installation Group wizard. In 







>> 

> 



7.1.2.3 Installation Configuration step (Copy Installation Group wizard) 

Installation Configuration is the final step in the Copy Installation Group 

wizard. This step allows you to select a configuration file from the repository 

which can later be supplied to one or more members of the group. 

Figure 78: Installation Configuration step 



Selects a configuration file from the Installation Configuration repository. 



Assign >> 


Assign to all 


7.1.3 Editing an installation group 

The Edit Installation Group wizard allows you to edit an existing installation 

group. 

To open the wizard, select Edit Installation Group from the context menu in 

the Servers by Installation group. 

7.1.3.1 Group Name step (Edit Installation Group wizard) 

Group Name is the first step in the Edit Installation Group wizard. In this 




Group Name 


Description 


7.1.3.2 Group Members step (Edit Installation Group wizard) 


Group Members is the next step in the Edit Installation Group wizard. In 







>> 

> 



7.1.3.3 Installation Configuration step (Edit Installation Group wizard) 

Installation Configuration is the final step in the Edit Installation Group 

wizard. This step allows you to select a configuration file from the repository 

which can later be supplied to all members of the group. 




Selects a configuration file from the Installation Configuration repository. 



Assign >> 


Assign to all 


7.1.4 Deleting an installation group 

You can delete installation groups: 

1. Select the relevant installation group in the Servers by Installation 

group. 

2. Select Delete Installation Group from the context menu. 

7.2 Installing servers 

Before you start the installation process you need to consider which servers 

should be installed. You can build installation groups consisting of one or 

more servers (every server can be assigned a different configuration file). It is 

also possible to install servers without specifying an installation group - you 

select a server from the list and then select Install from the context menu. 

l You need to create the installation configuration file from Server- 

View Installation Manager beforehand. For details, please refer 

to the manual for ServerView Installation Manager. 

l This functionality depends on ServerView Installation Manager. 

Therefore, you must meet the software requirements of Server- 

View Installation Manager. For example, when you want to 

install a Linux system, you need to configure a FTP, HTTP or 

NFS server. For details, please refer to the ServerView Installation 

Manager user guide. 


7.2.1 Install wizard 

The Install wizard allows you to install servers. 

To open the wizard, select Install from the context menu in the Servers 

view. 

7.2.1.1 Task Name step (Install wizard) 

Task Name is the first step in the Install wizard. This step allows you to 

enter a task name. 



Task Name 




7.2.1.2 Installation Configuration step (Install wizard) 

Installation Configuration is the next step in the Install wizard. This step is 

only available, if you start the Install wizard by selecting one or more 

servers.This step allows you to select a configuration file from the repository 

which can later be supplied to the selected servers. 



Select a configuration file from the Installation Configuration repository. 

7.2.1.3 Settings step (Install wizard) 

Settings is the next step in the Install wizard. It is the third step if you start 

the wizard by selecting an installation group. 



Assign to ServerStart/Installation Manager data 

Location of the Installation Manager data. Use the default settings or 

specify a new location and define a user name and user password for 

remote access. 

Content Tree (UNC) Shared network directory (UNC network 

path). 

Username Remote Access User name for remote access. 

Password / 


Password for remote access. 











Depending on the selected shutdown method, a shutdown 

might be a hard power-off or a graceful shutdown. In case of a 

hard power-off, unsaved data will be lost. This might not be a 

problem, since the disk will be overwritten by a new installation. 

But data on a second logical disk might be lost. 







problem, since the disk will be overwritten by a new installation. 




file systems. 

Graceful 

(via Server- 

View 

Agent) 











Shut down System The target servers involved will be shut down 

at the end of the process. 

Keep System Running The target servers involved will be kept running 

at the end of the process. 

7.2.1.4 Bios Boot Type step (Install wizard) 

Bios Boot Type is the next step in the Install wizard. 


This option is currently ignored in 

Mass Installation. After finishing 

Mass Installation, the target server 

will be kept running. 











BIOS. 








7.2.1.5 Scheduling step (Install wizard) 

Scheduling is the final step in the Install wizard. In this step you can create 

the install task as a scheduled task. 










task. 



Every x days. 







Monthly) 





Retry Count 






















7.2.1.6 Actions after starting the "Install" task 

This action is the same as in ServerView Installation Manager. 


l When installing a Windows operating system, ServerView 

Deployment Manager displays the installation status as Success 

after finishing the necessary process on WinPE. However, 

installation of target server continues for a while. 

















the task. 


















the task. 


7.3 Boot Groups 

7.3.1 Adding a boot group 

The Add Boot Group wizard allows you to create boot groups. 

To open the wizard, select Add Boot Group from the context menu in the 

Servers by Boot Image group. 

7.3.1.1 Group Name step (Add Boot Group wizard) 

Group Name is the first step in the Add Boot Group wizard. In this step 

you must enter a group name. 





Group Name 


Description 


7.3.1.2 Boot Image step (Add Boot Group wizard) 

Boot Image step is the next step in the Add Boot Group wizard. In this 

step you can choose an image from the list of registered boot images. 


Select an image from the list of registered boot images. 


7.3.1.3 Group Members step (Add Boot Group wizard) 

Group Members is the final step in the Add Boot Group wizard. In this 



> 

< 



Servers list. 





>> 


Group Name 


Description 


7.3.2.2 Boot Image step (Copy Boot Group wizard) 


Boot Image step is the next step in the Copy Boot Group wizard. In this 

step you can choose an image from the list of registered boot images. 





7.3.2.3 Group Members step (Copy Boot Group wizard) 

Group Members is the final step in the Copy Boot Group wizard. In this 




> 

< 

>> 



Servers list. 




Servers list. 



Group Name 


Description 


7.3.3.2 Boot Image step (Edit Boot Group wizard) 

Boot Image is the next step in the Edit Boot Group wizard. In this step you 

can choose the image file for the deployment. 






Unregister all clients for the previously selected image at the PXE service 

If you select another boot image, you can activate this option. This 

applies only to servers in Generic Boot tasks that have been started with 

selection Permanent for Registration at PXE Service (see "Settings 

step (Generic Boot wizard)" on page 288). 

7.3.3.3 Group Members step (Edit Boot Group wizard) 

Group Members is the final step in the Edit Boot Group wizard. In this 







Servers list. 

> 


7.4 MDP booting of servers 

Before you can perform an MDP boot for one or more servers, you must 

create a boot group that has an image of type SVIM Agent Image (MDP) 

assigned. With MDP boot, you can execute an MDP application on one or 

more servers of a group. 

The boot group can be created from the built-in images SVIM32 or 

SVIM64. In most cases it is not necessary to add and register a 

boot image of the type SVIM Agent Image (MDP). 

Before calling the MDP Boot wizard, you must add an MDP Applications 

repository and manually copy your MDP application to a 

sub directory in the MDP Applications repository. 

For details on MDP (Multi Deployment Platform) refer to the ServerView 

Installation Manager user guide. 

7.4.1 MDP Boot wizard 

The MDP Boot wizard allows you to execute an MDP application on one or 

more servers of a boot group. 

To open the wizard, select MDP Boot from the context menu in the Servers 

by Boot Image group. 

MDP Boot is offered in the context menu of a boot group if the 

group has an image of type SVIM Agent Image (MDP) assigned. 

7.4.1.1 Task Name step (MDP Boot wizard) 

Task Name is the first step in the MDP Boot wizard. In this step you must 

enter a task name. 



Task Name 


7.4.1.2 MDP Application step (MDP Boot wizard) 


MDP Application is the next step in the MDP Boot wizard. In this step you 

can select an MDP application. 



Figure 97: MDP Application step 

Select an MDP application in the tree view. 

7.4.1.3 Settings step (MDP Boot wizard) 

Settings is the next step in the MDP Boot wizard. This step allows you to 

specify the system shutdown method before starting the MDP boot and the 

SVIM agent status after deployment. 












power-off, unsaved data will be lost. 






Forced A hard power-off of all running target servers will be initiated if 

these servers support remote power-off by management 

blade or baseboard management controller. In this case 

unsaved data will be lost. 



ACPI, it should save unsaved data and clean up the file systems. 

Graceful 

(via 

Server- 

View 

Agent) 

This shutdown method makes use of the ServerView agent 

functionality. ServerView agents running on the target systems 

are instructed to shut down the systems. This method 

requires that ServerView agents are installed on all systems 

and are configured to allow this functionality. It also requires 

the specification of a shutdown user name and password. 









These will be set as default shutdown user name and 

password. The last-used shutdown user name and password 

are lost when the ServerView Deployment Manager service is 

restarted. 

SVIM Agent Status after Deployment 

SVIM Agent status of the target systems after the execution of the MDP 

application: 


Disabled The SVIM Agent will return the control back to 

WinPE. 

Agent PowerOff The SVIM Agent will shut down the target. 

Agent Reboot The SVIM Agent will reboot the target. 

Agent Idle The SVIM Agent will keep running in idle mode. 

7.4.1.4 Bios Boot Type step (MDP Boot wizard) 

Bios Boot Type is the next step in the MDP Boot wizard. 












BIOS. 



Figure 99: BIOS Boot Type step 





7.4.1.5 Scheduling step (MDP Boot wizard) 

Scheduling is the final step in the MDP Boot wizard. In this step you can 

create the MDP Boot task as a scheduled task. 










task. 



Every x days. 







Monthly) 





Retry Count 






















7.5 Generic booting of servers 

Before you can perform a generic boot for one or more servers, you must 

create a boot group that has an image of type Generic Image assigned. With 

generic boot, you can boot a generic image on one or more servers of a 

group. 

There is only limited task progress handling possible. As soon as a 

PXE request has been received from a client, the state of this client 

is set to Success. After the PXE requests from all clients in the 

boot group have been received, the task state is set to Success. 

7.5.1 Generic Boot wizard 

The Generic Boot wizard allows you to boot one or more servers of a boot 

group with a generic image. 

To open the wizard, select Generic Boot from the context menu in the 

Servers by Boot Image group. 

Generic Boot is offered in the context menu of a boot group if the 

group has an image of type Generic Image assigned. 

7.5.1.1 Task Name step (Generic Boot wizard) 


Task Name is the first step in the Generic Boot wizard. In this step you 

must enter a task name. 




Task Name 


7.5.1.2 Settings step (Generic Boot wizard) 

Settings is the next step in the Generic Boot wizard. This step allows you 

to specify the system shutdown method before starting the generic boot and 

the registration type at the PXE service. 












power-off, unsaved data will be lost. 









this case unsaved data will be lost. 




file systems. 

Graceful 

(via Server- 

View 

Agent) 








Registration at PXE service 

The registration type at the PXE service: 

OneShot 

The clients will PXE boot only once. After a PXE request has been 

received from a client, the client is automatically unregistered at the 

PXE service. 

Permanent 

The clients are registered permanently at the PXE service. If there is 

another PXE request received from a client, there is another generic 

boot of the client done. 

After the task has been completed, you must manually unregister 

the clients at the PXE service. See Unregistering all 

clients at the PXE service in section "Operations on the 

Tasks tab" on page 99 and "Unregistering a client" on page 

117. 


Delay (in seconds) to unregister the client after PXE request has been 

received 

If OneShot is selected, the client should not be unregistered immediately 

after the PXE request has been received. Specify the delay here. 

7.5.1.3 Bios Boot Type step (Generic Boot wizard) 

Bios Boot Type is the next step in the Generic Boot wizard. 












BIOS. 








7.5.1.4 Scheduling step (Generic Boot wizard) 

Scheduling is the final step in the Generic Boot wizard. In this step you can 

create the Generic Boot task as a scheduled task. 










task. 



Every x days. 







Monthly) 





Retry Count 






















8 Crash Recovery 

The Crash Recovery work area allows you to create and restore snapshot 

images. 

If the target server has an UEFI BIOS, each LAN port must be 

manually configured for either Legacy PXE or UEFI PXE boot. 

If the target server has a GPT boot disk, UEFI PXE boot must 

be configured. Otherwise Legacy PXE or UEFI PXE boot can 

be configured. 

8.1 Creating a snapshot image 

You can create a snapshot image, which can be used as a kind of backup of 

one particular server. A snapshot image can be created for selected servers 

at the same time. After restoring a snapshot image, the restored system will 

have the same host name and the same IP address as the source system. 

Therefore a snapshot image is not useful for cloning several servers. 

8.1.1 Create Snapshot Image Wizard 

The Create Snapshot Image wizard allows you to create a snapshot image. 

To open the wizard, select Create Snapshot Image from the context menu 

of a selected server in the Servers view. 

After starting the image creation process (by clicking Finish), the cloning 

module now boots the reference system referenced by its MAC address with 

a WinPE boot image via PXE and starts the image creation tool on WinPE. 

The automatic remote power-on and the remote PXE boot configuration 

of the BIOS require that the server can be managed by a 

management blade (blade system), an RSB or a BMC that supports 

IPMI 1.5 over LAN. For non-blade systems you must have specified 

the correct BMC settings. If the server cannot be managed in one of 

these ways, a dialog appears asking you to initiate a PXE boot manually. 

Scheduled tasks for non-manageable servers are not possible. 



8.1.1.1 Task Name step (Create Snapshot Image wizard) 

Task Name is the first step in the Create Snapshot Image wizard. In this 

step you must enter a task name. 


Task Name 


Image Information 


the image. This information will be stored in the file with the suffix .txt. 


8.1.1.2 Deployment Server step (Create Snapshot Image wizard) 

Deployment Server is the next step in the Create Snapshot Image wizard. 






8.1.1.3 Image Path and Name step (Create Snapshot Image wizard) 

Image Path and Name is the next step in the Create Snapshot Image wizard. 

In this step you can define the location (path) and the name of the image. 



Figure 107: Image Path and Name step 

Disk Images 



repository. 

Image Filename 

Name of the image. You can use the following variables in the file name: 

%D This is replaced by the day (01, 02, ..., 31) of the month. 

%M This is replaced by the month (01, 02, ...,12). 

%Y This is replaced by the year (4 digits). 

%S This is replaced by the name of each server of which an image will 

be created. 


8.1.1.4 Options step (Create Snapshot Image wizard) 

Options is the next step in the Create Snapshot Image wizard. In this step 

you can specify server options before and after deployment. 

Figure 108: Options step 


Force a system shutdown before creating the image. Operating-systemspecific 

ServerView agents must be installed on the relevant server. 


Shutdown method: 




Graceful 

(via ServerView Agents) 

This shutdown method makes use of 

the ServerView agent functionality. 

ServerView agents running on the target 

systems are instructed to shut down the 

systems. This method requires Server- 

View agents to be installed and configured 

on all systems. It also requires 

specification of a shutdown user name 

and password. 

ACPI Deployment Manager will try to perform 

an ACPI shutdown of the system. If the 

running operating system supports 

ACPI, it should save unsaved data and 

clean up the file systems. 

Forced A hard power-off of all running target 

servers will be initiated if these servers 

support remote power-off by management 

blade or baseboard management 

controller. In this case 

unsaved data will be lost. Normally this 

is not a problem, since the disk will be 

overwritten by a new disk image. 














System status after creating the image: 

Shut down System The server will be shut down at the end 

of the process. 

Keep System Running The server will be kept running at the 

end of the process. 

GUID Check 





8.1.1.5 Disks step (Create Snapshot Image wizard) 


Disks is the next step in the Create Snapshot Image wizard. In this step 

you can specify the disks from which the image will be saved. 




Click Add to add a new disk. 

Logical Disk Number 

Select the logical disk number from which the image is to be created. By 

default the logical disk 0 is saved, which is the first logical device. 







be saved. As a workaround you could unplug the Fibre Channel 

cables before starting image creation. 


Raw Mode 

You can generate an image creation in raw disk mode. All sectors of the 

disks will be saved in the image. This image can only be deployed to 

servers which have disks with the same size and the same disk layout. 

File System Independent 

This method is used for non-supported file systems. When selected, the 


cloned to a partition of the same size. For more information see section 

"File-System-Independent image creation (raw image creation)" on page 

155. 

File System Dependent 

When selected, only the used blocks of a file system will be saved in the 

image. For more information see section "File-System-Dependent image 


This functionality is supported for the following file systems: 

l FAT, FAT32 

l NTFS, NTFS5 

l EXT2, EXT3, EXT4 


l Reiser FS version 3.5 and 3.6 on SuSE SLES 9 and 10. 

Other file systems and especially other versions of the Reiser file system 

are not supported by this functionality. 

If a disk contains several partitions for non-supported file system types, 

the whole partition including the non-used blocks will be saved. For the 

partitions with a supported file system, only the used blocks will be 

saved. This reduces the size of the image files and also increases the 

speed of the image generation. If you want to restore an image on a disk 

that has a different size than the source disk, this option should be specified. 

Verify File System 

The file system is checked before the image creation process is started. 



Raise Error for Unsupported File Systems 

Only use this option if the File System Dependent option is checked. If 

an unsupported file system is detected, the image creation process will 

raise an error and stop. 

Compress Image 

The image file is compressed during the image creation. 

Fast Image Creation 

When this option is checked, the image creation process uses an optimized 

method to access the target disk for the image creation. This optimization 

can be applied to EXT2, EXT3 and EXT4 file systems when the 

option File System Dependent is checked. For an image that was created 

with this option, the option Expand last partition to whole disk in 

the System Preparation step of the cloning is not available because 

applying the option Expand last partition to whole disk to such an 

image would raise an error. 

This option must not be checked if there are NTFS partitions (for 

Windows). 

Click Remove to delete the settings for the disk. 

8.1.1.6 Bios Boot Type step (Create Snapshot Image wizard) 

Bios Boot Type is the next step in the Create Snapshot Image wizard. 











BIOS. 







8.1.1.7 Scheduling step (Create Snapshot Image wizard) 


Scheduling is the final step in the Create Snapshot Image wizard. In this 

step you can create the snapshot image generation task as a scheduled 

task. 










task. 



Every x days. 






Monthly) 





Retry Count 
























8.1.1.8 Action after starting the "Create Snapshot Image" task 

Basically, the target server behaves as follows after starting the Create 

Snapshot Image wizard. 

Windows / Linux reference system 

When saving an image from a Windows Reference system and Linux Reference 

system, the following steps are done: 





5. After saving an image, the client is shutdown. 

















the task. 

















the task. 

8.2 Restoring a snapshot image 

You can restore a snapshot image. After restoring a snapshot image, the 

restored system will have the same host name and the same IP address as 

the source system. 

8.2.1 Restore Snapshot Image Wizard 

The Restore Snapshot Image wizard allows you to restore a snapshot 

image. 

To open the wizard, select Restore Snapshot Image from the context menu 

of a selected server in the Servers view. 

8.2.1.1 Task Name step (Restore Snapshot Image wizard) 


Task Name is the first step in the Restore Snapshot Image wizard. In this 

step you can enter a task name. 




Task Name 


8.2.1.2 Deployment Server step (Restore Snapshot Image wizard) 

Deployment Server is the next step in the Restore Snapshot Image wizard. 






8.2.1.3 Disk Image step (Restore Snapshot Image wizard) 


Disk Image is the next step in the Restore Snapshot Image wizard. In this 

step you can select the disk image. 



Figure 114: Disk Image step 

Use Cloning Images for Crash Recovery 

Cloning images can be used for Crash Recovery. 

Disk Images 



repository. 

8.2.1.4 Disks step (Restore Snapshot Image wizard) 

Disks step is the next step in the Restore Snapshot Image wizard. In this 

step you can select the logical disk number. 




Restore Disk Image 

This option can only be selected if the image contains multiple disks (multiple 

drive snapshot cloning) and the user can select for each one whether 

they want to restore it. 











be overwritten. As a workaround you could unplug the Fibre Channel 

cables before starting the cloning. 

Partitions 












8.2.1.5 System Preparation step (Restore Snapshot Image wizard) 

System Preparation is the next step in the Restore Snapshot Image wizard. 

In this step you can select the system preparation before image deployment. 



Unchanged 





All Primergy 

















Raid Level 




Use Hot Spare 



a Lynx controller, select LSI/Mylex RAID Controller as 

the controller vendor, LSI IME SAS as the controller 

family, Any as the controller model, and 1 as the controller 

number. 


a Cougar controller, select LSI/Mylex RAID Controller 

as the controller vendor, LSI MegaRAID SAS as the controller 

family, RAID 5/6 SAS based on LSI MegaRAID 

as the controller model, and 0 as the controller number. 







Select the boot image type: Floppy Disk Image or Bootstrap Image. 




Convert MBR to GPT 

Initialize the target disk as GPT disk and convert MBR partitions to GPT 

partitions. This is only possible for data disks (not for system disks). 

8.2.1.6 Settings step (Restore Snapshot Image wizard) 


Settings is the next step in the Restore Snapshot Image wizard. This step 

allows you to specify the deployment method and the system shutdown 

method before cloning the image. 










































image. 




file systems. 



Graceful 

(via Server- 

View 

Agent) 



















GUID Check 





8.2.1.7 Post Deployment step (Restore Snapshot Image wizard) 

Post Deployment is the next step in the Restore Snapshot Image wizard. 

This step allows you to specify the system status after cloning. 









8.2.1.8 Bios Boot Type step (Restore Snapshot Image wizard) 


Bios Boot Type is the next step in the Restore Snapshot Image wizard. 













BIOS. 







8.2.1.9 Scheduling step (Restore Snapshot Image wizard) 

Scheduling is the final step in the Restore Snapshot Image wizard. This 

step allows you to clone the image as a scheduled task. 









task. 





Every x days. 





Monthly) 





Retry Count 






















8.2.1.10 Action after starting the "Restore Snapshot Image" task 

Basically, the target server behaves as follows after starting the Restore 

Snapshot Image wizard. 

Windows / Linux reference system 

When restoring an image of Windows reference system and Linux reference 

system, the following steps are done: 

























the task. 
















the task. 


9 Messages 

The GUI polls the Deployment Service service at regular intervals for error 

and event messages. You can specify the refresh interval in the Global 

Options step in the Settings wizard, see section "Global Options step (Settings 

wizard)" on page 336. 

These messages are displayed in the Messages List dialog box. To open the 

dialog box, select the item Messages from the menu bar. 

You will be informed of new incoming messages. This is done in a message 

window in the top right corner of the main window. 

The message window closes when you select Messages from the menu bar 

to display the Messages List dialog box. If you do not want to look at the 

messages immediately, you can close the message window by selecting 

Close. If you do not want the message window to appear again (in the current 

session), select Don't show again. 

9.1 Messages List dialog box 

The Messages List dialog box displays all messages received from the 

ServerView Deployment Manager service. 


9 Messages 

Figure 121: Messages List dialog box 

The columns in the table are explained below: 


Timestamp Time the message was read from 

the ServerView Deployment Manager 

service. 

Message Message text. 

Message 

Displays the complete message text from the selected message. 


10 Settings 

You can define global settings for Deployment Manager. This is done in the 

Settings wizard. 

The Settings wizard starts for the first time when you start Deployment Manager 

after installation. You must work through all steps of the wizard, to specify 

the settings for the first time. 

You can open this wizard again at any time to change the settings. To start 

the Wizard, select Settings from the menu bar. 

10.1 Deployment Server step (Settings wizard) 

Deployment Server is the first step in the Settings wizard. In this step you 

can select deployment servers for Mass Cloning and Crash Recovery. 

Deployment servers are responsible for preparing the servers and their environment 

from a central instance via the LAN. The preparation of a deployment 

server is described in the “Installation Manager” guide. 


10 Settings 


>> 

> 

Copies the servers selected from the Deployment Servers list back to 

the Server List list. 

Log On Properties 

This opens the Log On Properties dialog box for the selected Deployment 

Server, see section "Log On Properties dialog box" on page 331. 

10.1.1 Log On Properties dialog box 

Figure 123: Log On Properties dialog box 

10.1 Deployment Server step (Settings wizard) 

Username 

User name for logging on to the deployment server. This must be one of 

the user names specified during the Deployment Manager installation 

(Deployment Services package). 

Password 

Password for logging on to the deployment server. 


10 Settings 

10.2 Repositories step (Settings wizard) 

Repositories is the next step in the Settings Wizard. In this step you must 

specify at least one installation configuration repository or one image repository. 

You can add repositories and folders via the context menu. 

Figure 124: Repositories step 

10.3 Licenses Installed step (Settings wizard) 

Licenses Installed is the next step in the Settings wizard. In this step you 

can see the currently installed licenses for Deployment Manager. 


Figure 125: Licenses Installed step 

Installed Licenses 

Displays the number of installed licenses. 

10.3 Licenses Installed step (Settings wizard) 

Used Licenses 

Displays the number of occupied licenses. Whenever a server is 

deployed, it occupies a license. If a server is already using a target 

license, it does not occupy a new target license when it is deployed 

again. A server that occupies a license can be deployed as many times 

as necessary. 

Click Add to open the Add New License Key dialog box. 


10 Settings 

10.3.1 Add New License Key dialog box 

The Add New License Key dialog box allows you to add a new license 

string. 

You cannot use the Deployment Manager functionality without a license. As 

well as the information that defines the validity of a license, a license string 

contains information about the number of target systems that can be 

deployed with this license. Multiple Deployment Manager licenses can be 

installed. In this case the number of deployable targets of all valid Deployment 

Manager licenses are added. Deployment Manager allows to deploy as 

many servers as the sum of target licenses over all valid Deployment Manager 

licenses estimates. 

You can enter a license key during Deployment Manager installation. 

If you do not enter a valid license key, an evaluation key is created. 

To open the dialog box click Add in the Installed step of the Settings wizard. 

Figure 126: Add New License Key dialog box 

License Key 

Enter the new license string that you received from your provider. The 

license you specify will be added to the list of licenses. 


10.4 Licenses Used step (Settings wizard) 

Licenses Used is the next step in the Settings wizard. Each server in the 

list occupies one license. In this step you can release licenses from a selected 

server. 

Figure 127: Licenses Used step 

Installed Licenses 

Displays the number of installed licenses. 

10.4 Licenses Used step (Settings wizard) 

Used Licenses 

Displays the number of occupied licenses. Whenever a server is 

deployed, it occupies a license. If a server is already using a target 

license, it does not occupy a new target license when it is deployed 


10 Settings 

again. A server that occupies a license can be deployed as many times 

as necessary. 

To release an occupied target license, select a server in the list and click 

Release License. 

10.5 Global Options step (Settings wizard) 

Global Options is the final step in the Settings wizard. In this step you can 

specify global settings for Deployment Manager. 

Figure 128: Global Options step 


GUI Options 

Messages 

update 

interval 

(s) 

Trace 

Level 

(Frontend 

only) 

10.5 Global Options step (Settings wizard) 

Refresh interval in which messages are read from Server- 

View Deployment Manager. 

For diagnostic purposes you can specify one of the following 

trace levels: 

No Tracing: No trace output will be generated. 

Severe: Severe (not recoverable) errors only. 

Info: Recoverable errors and important information. 

Fine: Debug output of important information, most of it 

reflecting user actions. 

Calltrace: System-level debug output. 

Finer: Detailed system-level debug output, parameters sent 

to and data parsed by ServerView Deployment Manager 

service. 

Data: Original data as received from ServerView Deployment 

Manager service. 



11 High-Availability support 

As of version 4.0 you can use Deployment Manager in a high-availability configuration. 

This means that 

l The location of the databases (for Deployment Manager and Deployment 

Services packages) is configurable. You can use an external 

filer as storage for the databases. The two databases for the two packages 

can be on different filers. You cannot use a mapped network 

drive. 

l Deployment Manager can be installed on two servers which use the 

same databases. The Deployment Manager and Deployment Service 

services are not started on the servers - it is up to the cluster software 

to start the services on the server it chooses as the active one. 

Only one instance should run at a time. 

l The ServerView Deployment Manager service and the Deployment 

Service service can detect the abnormal shutdown of the service 

(crash) and write a message to the event log. 

l After a crash the ServerView Deployment Manager service and the 

Deployment Service service can recover by setting the status of the 

jobs that were interrupted by the crash to Error. 

l ACID support 

In databases, ACID stands for Atomicity, Consistency, Isolation, and 

Durability. These are considered to be the key transaction processing 

features of a database management system. Without them, the integrity 

of the database cannot be guaranteed. 

Requirements 

Both servers must have the same host name and IP address. If cluster software 

is installed, this requirement should be provided by the cluster software. 

Otherwise it can be achieved when the two servers are not up and running at 

the same time. 




11.1.1 ServerView Operations Manager installation 

Deployment Manager must have access to an Operations Manager installation 

on the local server or on a remote server. The Deployment Manager 

package installer only offers the ServerView Operations Manager Location 

window if a Full installation on first server has been selected in the 

High Availability window. This window is not displayed if a Limited installation 

on second server is installed. 

The parameters given in the ServerView Operations Manager Location 

window are added to the Deployment Manager database that is stored on a 

filer. In the failover scenario, the second server therefore uses the same Operations 

Manager location as the first server on which the Operations Manager 

location was entered. 

For the configuration of the Operations Manager location, see section "Installing 

the Deployment Manager package" on page 68. 

If both servers are members of a Windows cluster, there are 3 entries 

in Operations Manager database - one for the cluster (with the cluster 

"virtual" IP address) and two for the servers. You should select the 

cluster and not the currently active server when you add a new deployment 

server. You are allowed to add both servers to the list of deployment 

servers in addition, but this deployment server might not be 

accessible, depending on which host is the currently active one and 

has the ServerView Deployment Manager service started. 

Using two local Operations Manager installations 

If the Operations Manager installation is to be used, Operations Manager 

must be installed on both servers, and the ServerView databases on both 

servers should be kept identical. 

Using one remote Operations Manager installation 

In this case Operations Manager must be installed on one remote server 

only, and the servers must only be added to the ServerView database once. 


11.1.2 Deployment Manager installation 

During the installation of the Deployment Manager package the High Availability 

window is displayed: 

Figure 129: High Availability window - Support High Availability 

Support High Availability 

Select the option Yes. 




Installation Order 

Full installation 

on first 

server 

Limited 

installation 

on second 

server 

If you select this option, the installation is similar to a 

standard installation without high availability but with the 

following differences: 

l Specify a location of the Deployment Manager database 

that is accessible to the second server. The 

default value offered is only appropriate for a standard 

installation without high availability and must 

be changed to a directory in a mounted LUN (Logical 

Unit Number) on a filer. A mapped network 

drive cannot be used. 

l The startup type of the ServerView Deployment 

Manager service is set to Manual and the service 

is not started automatically after installation. It is 

up to the cluster software to start the service on the 

system it chooses as the active one. 

As a prerequisite a Full installation on first server must 

already have been performed. 

11.1.3 Deployment Services installation 

l Make sure that the location of the Deployment Manager 

database points to an already existing database 

on a filer. If the database is not found by the 

installer, an error message is displayed and the 

installation will fail. The database is not created if 

this option is selected. 

l The startup type of the ServerView Deployment 

Manager service is set to Manual. It is up to the 

cluster software to start the service on the system 

it chooses as the active one. 

During the installation of the Deployment Services package the Destination 

Folder window is displayed: 


Figure 130: Destination Folder window - Support High Availability 

Support High Availability 

Select the option Yes. 




Installation order 

Full installation 

on first 

server 

Limited 

installation 

on second 

server 

If you select this option, the installation is similar to a 

standard installation without high availability but with the 

following differences: 

l Specify a location of the Deployment Service database 

that is accessible to the second server. The 

default value offered is only appropriate for a standard 

installation without high availability and must be 

changed to a directory in a mounted LUN on a filer. 

A mapped network drive cannot be used. 

l The startup type of the Deployment Service service 

is set to Manual and the service is not started 

automatically after installation. It is up to the cluster 

software to start the service on the system it chooses 

as the active one. 

A Full installation on first server must already have 

been performed. 

l Make sure that the location of the Deployment Service 

database points to an already existing database 

on a filer. If the database is not found by the 

installer, an error message is displayed and the 

installation will fail. The database is not created if 

this option is selected. 

l The startup type of the Deployment Service service 

is set to Manual. It is up to the cluster software 

to start the service on the system it chooses as the 

active one. 


11.2 Hints 

11.2.1 User Accounts 

During installation of the Deployment Manager and Deployment Services 

packages, the user accounts for the ServerView Deployment Manager 

service and Deployment Service service must be given. These user 

accounts are used to access the databases on the filer. The following combinations 

are recommended: 

11.2.2 Repositories 

l During the installation the user specifies a local account with the 

same name and password on both servers. 

11.2 Hints 

l During the installation on both servers the user specifies the same 

domain account. 

Repositories should not be on one of the two servers, but in a directory in a 

mounted LUN on a filer or on a mapped network drive. When adding a repository, 

a UNC path name that is stored in the shared Deployment Manager 

database must be specified. Make sure that this UNC path is reachable from 

both servers. 

11.2.3 Actions Needed When No Cluster Software is Available 

Both servers do not have Deployment Manager and the Deployment Service 

started after installation. It is assumed that some cluster software starts the 

services on the server that it chooses as the active one. 

If there is no cluster software available, you must start the services on one of 

the servers (the one that you chooses as the active one). 



11.3 Failover scenarios 


are Installed on Same Server 

A typical failover scenario is: 

l The first server crashes (more precisely: the server that has been 

selected as the active one by the cluster software crashes - it might 

as well be the server installed as the second server), ServerView 

Deployment Manager service and Deployment Service service are 

not stopped correctly. 

l The cluster software starts the ServerView Deployment Manager 

service and Deployment Service service on the second server. 

l The ServerView Deployment Manager service and Deployment 

Service service on the second server detect that a crash has 

occurred. 

l The ServerView Deployment Manager service and Deployment 

Service service add event log entries. 

l The Deployment Service service cancels all running jobs and marks 

the jobs with the status Error in the Deployment Service database. 

l The ServerView Deployment Manager service tries to cancel all running 

jobs. This does not work (because the Deployment Service service 

already canceled them), so the ServerView Deployment 

Manager service marks the jobs with the status Error in the Deployment 

Manager database. 


are Installed on Different Servers 

In the following 

l first server Mgr/second server Mgr denote the servers where the 

Deployment Manager package is installed. 


l first server Depl/second server Dep denote the servers where the 

Deployment Services package is installed. 

The first server with the Deployment Manager package crashes 


l The first server Mgr crashes, the ServerView Deployment Manager 

service is not stopped correctly. 

l The cluster software starts the ServerView Deployment Manager 

service on the second server Mgr. 

l The ServerView Deployment Manager service on the second 

server Mgr detects that a crash has occurred. 

l The ServerView Deployment Manager service adds an event log 

entry. 

l The ServerView Deployment Manager service tries to cancel all running 

jobs. This should work (because the ServerView Deployment 

Manager service did not crash) and the ServerView Deployment 

Manager service finally sets the status Canceled in the Deployment 

Manager database. If canceling the jobs does not work, the Server- 

View Deployment Manager service marks the jobs with the status 

Error in the Deployment Manager database. 

The first server with the Deployment Services packages crashes 


11.3 Failover scenarios 

l The first server Depl crashes, the Deployment Service service is 

not stopped correctly. 

l The cluster software starts the Deployment Service service on the 

second server Depl. 

l The Deployment Service service on the second server Depl 

detects that a crash has occurred. 

l The Deployment Service service adds an event log entry. 

l The Deployment Service service cancels all running jobs and marks 

the jobs with the status Error in the database. 



l The ServerView Deployment Manager service running on the first 

server Mgr finally retrieves the status from the Deployment Service 

service (reads status Error from Deployment Service database) and 

sets the status Error in the Deployment Manager database. 


12 Cloning deployment process 

The following description of the cloning deployment process 

assumes that the deployment platform Caldera DOS is used. Most 

of the description is valid for deployment platform WinPE MDP as 

well. 

A cloning session consists of several phases and offers different alternatives 

to be used for each phase: 

1. Power control, i.e. remote PXE boot of target server 

o Initiated by RemCtrl.dll (default for ServerView Suite) 

o Initiated by SNMP commands accessing management blade 

or ServerView agents 

o Initiated by IPMI commands via Kalypso BMC 

o Initiated manually 

2. System preparation 

o Via DOS agent at cloning (used by Haribote for server blades) 

Concerns unattended RAID preparation of blade controller 

only. 

o Via Installation Manager based on WinPE PXE boot 

n SCU Server Management settings in SM-BIOS 

n Unattended RAID preparation for all PRIMERGY 

servers 

o Boot of system preparation images created by yourself (as 

used for MS ADS (Automated Deployment Services)) 

All actions applicable on a DOS or MiniLinux platform boot are 

possible. You will need to obtain appropriate tools yourself to 

make the system preparations on the target server. 

o Manually by the administator 

n via Installation Manager in expert mode locally 

n via BIOS extensions of PCI controller in server boot 

phase 

n via Remote Management remote console 



3. Clone of an image 

All physical clients of a deployment group are cloned with the 

assigned image. Image-related actions can only be activated on logical 

groups. 

If a server was removed from one logical group and is to be moved to 

another logical group, the Status of Cloning reflects whether the newly 

assigned image is already cloned or not. If not, and Deployment Manager 

is closed and restarted, it will not be reassigned to the old group 

because of the new image reference, but will be identified as a cloned 

client of that group. 

4. Post-preparation phase 

o Individualization of a clone via 

n LAN parameters 

n host name 

n Microsoft Windows system ID: SID 

o Start of customer scripts 

n ServicePack/Quickfix update 

n Driver update 

n Install applications/services 

n Configuration of operating system or application, e.g. 

Windows scripting 

These phases are described in more detail in the following sections. 

The figure below shows an overview of the phases during a cloning process. 


Figure 131: Deployment phases - overview 



As an example these are the basic deployment steps for a blade server system: 

1. The blade server chassis is powered on and the management blades 

start discovering the hardware environment (number of server blades, 

switch blades and status of redundant second management blade). 

2. On request, Operations Manager searches the LAN to find all present 

management blades in one segment. 

3. Deployment Manager retrieves the list of management blades found 

from Operations Manager and requests a list of system information 

directly from the management blade about each blade server installed 

in a blade chassis. 

4. Deployment Manager offers the administrator a physical server list 

based on the management blade information to create logical server 

groups based on the system information received. 

5. The administrator may change these logical groups by adding or 

changing logical-group-specific parameters, and finally he/she activates 

a deployment process on a logical group. 

6. Deployment Manager creates a list of related server blades and their 

MAC addresses and generates a cloning job for the cloning engine. 

7. The cloning engine identifies each physical server blade on the LAN if 

required. It prepares LAN switch configurations for PXE access and 

initiates a Unicast or Multicast cloning process on each client. 

8. The cloning engine contacts the management blade via RemCtrl.dll to 

initiate a PXE boot of each related server blade. 

9. The server blade PXE BIOS contacts the deployment PXE server to 

receive the first DOS boot image and starts that image containing the 

cloning agent. 

10. The cloning agent prepares the server blade hardware like a RAID 

array and downloads via TFTP a second master image directly onto 

the prepared storage volume, which is reachable via an Int13h BIOS 

call with the logical device ID: 0. 


11. The final clone will be patched for individual operating system parameters 

such as IP address, host name and system ID (Windows only). 

12. The cloning agent initiates a normal reboot of its current server blade. 

13. The final operating system boots: 

a. In the case of Linux, cloning is done by activating new LAN settings 

at run level 0. 

b. In the case of Windows, a RunOnce script is started to activate 

Microsoft system preparation and probably a customer 

script for additional configuration steps. 

With RemoteDeploy V3.0 this is done by controlling the post-preparation 

phase using a temporarily installed agent. 

12.1 Power control via remote control 

The remote control API (RemCtrl.dll) is used to initiate requests to the target 

server especially for power control and detection of servers. The service processors 

such as management blade (SNMP) and Kalypso (IPMI) access are 

supported via RemCtrl.dll. 

For PRIMERGY servers which support none of these service APIs, a corresponding 

dialog is displayed on the Deployment Manager front-end to make 

the user initiate tasks manually instead of remotely. The advantage of this is 

that the program using the RemCtrl.dll always uses the same API. 

By using the RemCtrl.dll, the ServerView suite provides a common API for 

power control of servers regardless of their managed server type. The different 

types of access are managed inside RemCtrl.dll, but the user of the 

API must offer the appropriate parameters used to enter each BMC-API. 

12.1.1 Management blade power control 

The management blade requires the following parameters to access the 

power control functions: 

l Management type = MMB 

12.1 Power control via remote control 

l IP address or host name of BMC (used by RemCtrl.dll) 



l SlotID 

l SNMP community string 

l Timeout of access BMC API 

l Number of retries to access BMC 

For accessing the BMC by IP address or host name, Deployment Manager 

must be installed. (The ServerView Deployment Manager service is used 

to determine the MAC address via its Deployment Manager database.) If you 

use the MAC address for the same function, the ServerView Deployment 

Manager service is not necessary. If you use the MMB control, PXE boot 

mode can be initialized and ACPI PowerOff is supported. 

12.2 System preparation phase 

The real cloning process is independent of the server configuration. It just 

requires a server to have the following attributes: 

l The server is in PXE boot mode or can be booted in PXE mode 

remotely. 

l A DOS image with a minimum size of 1.44 MB can be booted via 

PXE. 

l One or more BIOS devices are available based on Int13h (which of 

them is used for cloning is selected by its logical device ID; default is 

Log-ID: 0). 

l The deployment server is available over the whole cloning cycle via 

LAN. 

A system preparation phase occurs immediately before the cloning to set up, 

for example, a RAID array on the target server. 

Deployment Manager provides the following alternatives: 

l Manual preparation 

l Automatic preparation as part of a DOS session 

l Preparation based on WinPE 

l Customer system preparation image 


12.2.1 Manual preparation 

Within Deployment Manager nothing is done for system preparation. The 

Deployment Manager continues directly with the generic cloning job. The 

administrator must make sure that the attributes mentioned in section 

"Power control via remote control" on page 353 are implemented and a logical 

Int13h device is reachable via BIOS on DOS. 

A typical method of RAID configuration is to use the BIOS extensions 

offered by each RAID controller using a key combination in the BIOS boot 

phase. 

If remote PXE boot is not applicable, the BIOS boot device table must have 

the PXE boot device configured statically as the first boot device. In this 

mode, the PXE server can control whether a PXE boot will be performed or 

whether a PXE timeout will occur and the boot process will continue with the 

next boot device in the list (which should be the hard disk containing the final 

cloning image). 

Server Management BIOS settings or Server Management configurations 

usually done by the SCU of Installation Manager must also be made manually 

(if required for the operating system; this is not required for Deployment 

Manager). 

12.2.2 Automatic preparation as part of a DOS session 

This is the standard preparation used for blade servers only as provided by 

the Haribote cloning engine. 

The DOS session for cloning a blade server is usually split into two boot 

phases in the case of a RAID configuration: 

1. Booting the cloning agent on DOS via PXE. 


2. Starting a vendor DOS tool for detection of storage devices and initiating 

the defined RAID configuration. 

3. Initiating a second PXE reboot which boots the DOS cloning agent 

again. 

Now the system BIOS can support access to the new RAID array 



based on an Int13h device call. The cloning process will start on this 

prepared logical device. 

4. A final normal reboot finalizes the cloning session by entering the 

post-cloning phase. 

The RAID preparation is the only system preparation done by the Haribote 

cloning engine in Deployment Manager. This method is not used for nonblade 

servers. The ServerView Deployment Manager service differentiates 

between blade and non-blade servers. 

12.2.3 Preparation based on WinPE 

For system preparation of all PRIMERGY servers, the preparation modules 

of Installation Manager are used. 

The ServerView Deployment Manager service calls the Installation Manager 

Extension via an API as described in section"Software architecture" on 

page 46. 

The Installation Manager module stack for remote installation is extended by 

a separate module management controlled by the Installation Manager Extension 

and Installation Manager manager. These modules can be used as 

individual functions via the Installation Manager Extension API. 

Deployment Manager uses the following functions: 

l Boot server with WinPE by PXE 

l Get system information of server 

l This output contains all information on the server physics collected by 

the Installation Manager hardware detection module. 

l Set/Get RAID configuration 

This allows complete unattended RAID configuration for all RAID controllers 

used by PRIMERGY servers 

If a bare server is to be entered to the server list, the Installation Manager 

Extension API is used as follows: 

l The Add Server dialog offers to collect additional server information. 

l The ServerStart agent is started and WinPE is booted via PXE. 


l GetSystemInfo is called to get the server parameters. 

l The ServerStart agent is stopped and the server is powered off. 

l A new entry is created in the ServerView server list with the retrieved 

parameters. 

This allows the following cloning process: 

Deployment 

Manager 

Installation 




SCW (System 

Cast 

Wizard) 

l define RAID configuration 

l define cloning parameters and deployment group 

l start ServerStart agent and boot WinPE via PXE 

l configure RAID parameters 

l stop ServerStart agent and WinPE, release PXE service 

l enter cloning job for DB API 

l prepare cloning job image 

l initialize PXE service for DOS cloning agent boot 

image 

l set PXE mode via remote control 

l start PXE reboot via remote control 

Check if the ServerView agent is available: If yes, initiate 

a reboot via ServerView agent. If no, initiate a 

shutdown followed by a power on 

l boot PXE DOS image 

l perform cloning 

12.2.4 Customer system preparation image 


You can prepare your own system preparation phase if you want a particular 

server type to be supported in a particular way (must not be a PRIMERGY 

server). 



For that purpose a generic system-preparation image PXE boot method is provided 

before starting the generic cloning phase. 

The image must meet the following requirements: 

l The size of the image is less than 1.44 MB 

l The image is capable of being booted from and used as a local floppy 

disk copy. 

This means that all the contents of a floppy disk that can be booted 

from the target server can also be used as a PXE image. 

At the end of the system preparation process, a normal reboot is initiated. If 

the boot device table in the system BIOS has a PXE-capable LAN card as 

the first device, a PXE boot is automatically initiated instead. Any operating 

system can be used for the system-preparation image. The PXE boot works 

independently of the operating system. 


Finally, to continue with a cloning process automatically the following process 

is implemented within Deployment Manager: 


Manager frontend 



Installation 




SCW (System 

Cast Wizard) 

l define cloning parameters and target servers 

l request PXE boot image path 

l start ServerStart agent on server side without booting 

WinPE 

l boot given system preparation image via PXE 

l check bootstrap boot status and return if PXE 

request from selected server was received 

l stop ServerStart agent, release PXE service only 

l finalize cloning job definition: without system preparation, 

no PXE boot, just PXE server initialization 

l enter cloning job for DB API 

l prepare cloning job image 

l initialize PXE service for DOS cloning agent boot 

image 

l set PXE mode via remote control 

l set PXE mode via remote control, but do not initiate 

a PXE boot, reboot is initiated by preparation 

image process 

l boot cloning agent DOS image via PXE 

l perform cloning 


The critical path in this scenario is the timing. The time between booting the 

system preparation image via PXE and the next PXE request from this 

server initiated by a reboot after finalizing the system preparation must cover 

the minimum of 30 seconds. If not, a PXE boot timeout will occur. This can 

be prevented by simply rebooting this target server in PXE mode. This is 



especially the case if more than four sessions are activated in the Haribote 

cloning engine. 

Within a 1 Gbit environment a higher session value is possible by changing a 

special registry key. This value controls the number of parallel sessions for 

Unicast cloning sessions as well. 

You can start a Unicast cloning job of, for example, 10 servers in one job, but 

only four servers are cloned in parallel at a time. 

With Unicast mode each server is a session, with Multicast each job is a session. 

12.3 Supported storage devices 

12.3.1 SCSI/IDE drives 

SCSI and IDE drives are handled by the system BIOS by default as Int13h 

devices and do not require special tools for modification. 

If the SCSI/IDE drive is mapped as the logical Int13h device, cloning will 

work without any system preparation. Manual mode should be used as the 

system preparation method. 

12.3.2 RAID devices 

It is possible to create an image on a different RAID level than is used for the 

subsequent cloning task. The RAID controller chip set must support a BIOS 

configuration extension for the manual system preparation method. This 

extension can be used on a reference server with a connected local console 

or via Remote Management. 

For blade servers only, the RAID functionality is prepared as part of the cloning 

process on DOS via the Haribote cloning engine directly. 

For non-blade servers, unattended RAID configuration is used by the Installation 

Manager Extension which boots WinPE on the target system. With 

this method, all PRIMERGY servers and their RAID controllers are automatically 

supported as declared for the currently installed Installation Manager 

release. 


After creation of a RAID array, an additional reboot must usually be initiated 

to activate the RAID array and make the new volume accessible to the cloning 

agent via BIOS support. 

This second reboot is again a PXE boot and reloads the same DOS image as 

before. 

12.3.3 FC and iSCSI Devices 

Since RemoteDeploy V3.0, generic storage devices are supported which are 

visible as Int13h DOS devices and are detected by the system BIOS. 

Within the image creation or deployment configuration session of the Deployment 

Manager Web user interface, you can define the logical ID of such an 

Int13h DOS device. This is necessary if more than one storage adapter has 

enabled Int13h BIOS support, or one adapter provides more than one bootable 

volume. FC and iSCSI adapters basically behave in the same way as 

SCSI or RAID controllers in detecting and creating a boot device list of their 

attached storage devices. 

Once this storage device list is visible on DOS as Int13h device (the adapter 

BIOS must be enabled for the last boot), Deployment Manager can perform 

image creation and mass cloning or mass remote installation on/from such 

devices generically. 

Based on this method, it is possible to clone on nearly all storage devices 

found as DOS Int13h device. 

Especially for FC and ISCSI device usage, the default UNDI LAN adapter no 

longer works for Broadcom NICs and must be replaced by the NDIS DOS 

driver, which is also provided with the Deployment Manager software. How 

to do this is described in section "Adding NDIS Driver to Deployment Manager" 

on page 363. 

12.3.4 Partitioning and File System Formatting 


Partitioning is required to store the operating system master image in a partition. 

The sector algorithm used needs this partition orientation to decompress 

the used sectors to the right place. 



If an image is created in raw mode, all format information is stored inside the 

image as part of the data packages. No pre-formatting is required for cloning. 

If an image was created with file system optimization, the target volume will 

be pre-formatted to offer data block organization on the storage volume to put 

the used data sectors in the right place. This method requires a much smaller 

image size than the image created in raw mode. The cloning images are created 

on a partition basis even if this is offered on a disk basis in the Deployment 

Manager GUI. 

The following file system formats are supported depending on the used operating 

system: 

File system 

FAT 16 P 

FAT 32 P 

NTFS4 P 

NTFS5 P 

NTFS5+ P 

DOS Windows Linux ESX 

Ext2 P P 

Ext3 P P 

Ext4 P 

Reiser file system (v3.5/3.6 on SuSE SLES 

9/10) 

1 ) A specific package must be installed on the reference system before 

image creation if Reiser is used as the root file system. This package is provided 

with the Deployment Manager software in the directory LINUX_support\ReiserFS_Cloning_Support. 

Basically the file system analysis is done independently of the operating system 

used. But with regard to the operating system types supported for individualization, 

there are limits imposed by the operating system (e.g. W2k3 does 

not support FAT16). 

But unknown or unsupported file systems can always be cloned in raw mode. 

362 Deployment Manager 

P 1 )

12.3.5 Multi-boot Operating System Partitioning 

A hard disk with more partitions with different operating system instances 

can be used if you install a boot loader in the master boot record (MBR) of 

this hard disk. 

For non raw mode image cloning, the MBR is created by the cloning agent 

and therefore no multi-boot loader is installed and supported. 

This kind of installation can be cloned in raw mode only, but without individualization 

after cloning. 

12.3.6 Adding NDIS Driver to Deployment Manager 

With RemoteDeploy V3.0, NDIS and UNDI LAN protocol drivers are used by 

the DOS clone agent. 

The following types of NIC vendor are currently supported: 

l Intel pro1000 family by NDIS driver on all LAN ports (up to 4) 

l Other (such as Broadcom) 

o Broadcom NDIS family driver for all known PRIMERGY 

servers released at Deployment Manager release time. Supports 

all LAN ports (up to 4) 

o by generic Intel UNDI driver on all LAN ports (up to 4) 

o by customer NDIS driver (by default, only the first LAN port is 

supported) 

To add an additional NDIS driver or switch Broadcom to NDIS, the following 

steps are required to change the Deployment Manager installation folder: 

1. Copy all driver files (typically *.DOS only) to ...\Program Files\Fujitsu\ServerView 

Suite\DeploymentService\tftp\agent\dos\boot 

on the deployment server. 

2. Modify the existing DetNic.inf file by adding a new entry with PCI vendor 

and device ID with the same format as the existing one. 

3. Completely reboot the deployment server once. 




For Broadcom NICs, a sample is provided with the Deployment Manager software 

in the directory setup\tftp\agent\dos\boot\AdditionalDrivers. 

For Broadcom NICs, two NDIS drivers are already configured. These NDIS 

drivers are stored in the B57.dos and bxnd20x.dos files in the ...\Program 

Files\Fujitsu\ServerView Suite\DeploymentService\tftp\agent\dos\boot 

directory on the deployment server. Modifying these files is not recommended. 

Bear in mind that, typically, NDIS drivers support only the first LAN port by 

default. You can change this by creating a protocol.ini file with vendor-specific 

settings. 

Please check your NDIS driver documentation for details of LAN port selection, 

assignment and bindings. 

12.4 Image creation 

The reference image is generated on a reference server with certain static 

machine-related parameters: 

l IP address 

l Host name 

l Windows SecureID 

These parameters can be changed during the cloning process to adapt to 

each target client. 

The ServerView Deployment Manager service delivers the final values of 

each client inside the cloning job description based on the deployment table 

settings. The image must be modified directly after it is cloned to the target 

client storage volume. 

In figure "Deployment phases - overview" in section "Cloning deployment 

process " on page 349, the flow chart of a cloning process describes this procedure 

in Block "C" for a Windows cloning. The file clcomp.dat, created for 

the before image creation, is now patched with the final individualization 

parameters and the next RunOnce process to be started will use these 

parameters in a system preparation session. 


For Linux cloning, the parameters are stored directly in the original operating 

system configuration files and are directly valid at the next boot. Therefore no 

post-cloning phase is required as for Windows. 

The procedure is always the same, regardless of whether Unicast or Multicast 

is used. 

Additional dynamic parameters per client are 

l Administrator account 

l Administrator password 

These parameters are available in the deployment table. 

12.5 Tag file handling 

A tag file can be used to detect whether a running system is still being modified 

by the deployment process. The tag file is stored in a specific directory. 

This directory might also be used to store log files. The names and the handling 

will be part of the official specification and will also be communicated to 

customers. 

The tag file can be used by third-party software (application or services, for 

example) to detect whether a cloned system is still being modified by the 

deployment process. 

12.5.1 Directory and tag file 

Operating System 

Directory Tag File 

12.5 Tag file handling 

Windows %Systemdrive%\DeploymentInfo\ DeploymentMode 

Linux /etc .DeploymentMode 

Deployment Manager log files on the target Linux system are stored in the 

/var/log/remotedeploy directory. 



12.5.2 Creating a tag file on Windows systems 

The tag file will be created during image creation at the beginning of the preparation 

of the reference system. If the %Systemdrive%\DeploymentInfo 

directory does not exist, it will be created. If the directory exists, any existing 

log files that are known as Haribote log files will be removed; other log files 

should not be removed. If the DeploymentMode file already exists, it is 

simply retained; this does not cause an error. 

12.5.3 Creating a tag file on LINUX systems 

The tag file will be created at the end of the image cloning while the DOS cloning 

agent is running. The DOS cloning agent creates the tag file in the directory 

/etc. If the tag file already exists, this does not cause a problem. 

12.5.4 Removing the tag file 

The tag file will be removed at the end of the Haribote post-preparation. If a 

customer script was specified for a cloning job, the tag file will be removed 

just before the customer script is executed. If an error occurs during the postpreparation, 

the tag file will not be removed. 

12.6 PXE protocol 

The following description is based on the Intel PXE specification from 1999 

release 2.1. 

Basically a DHCP server is a must for using PXE boot functionality in the current 

LAN segment! (A "DHCP proxy helper" functionality can be enabled in 

the router and works as well.) 

In principle the PXE protocol works as follows: 

1. The deployment server initiates a power-on of a particular server. 

2. Assuming the BIOS parameters are set to PXE boot, after power-on 

the BIOS ignores all alternative existing boot devices and starts the 

PXE LAN boot extension. 


12.6 PXE protocol 

3. The PXE BIOS issues a DHCP request to a DHCP server via broadcast. 

4. The DHCP server offers an IP address and additional LAN parameters 

such as the boot (deployment) server IP address (if it is configured 

by the administrator). 

5. A boot service broadcast is initiated: 

l If the boot server IP address is provided, the PXE BIOS can 

contact the boot server directly anywhere in the LAN (performing 

a check on port 4011). 

l If the boot server IP address is not provided, a broadcast on 

port 67 is sent which has to bypass a switch (e.g. with virtual 

LAN software) or hub depending on where the server is connected 

to the LAN. 

6. The boot server offers a packet with the name of the boot image. 

7. The PXE BIOS initiates an MTFTP or TFTP to receive the boot image 

file. 

8. The boot image is copied to address 07c0h and started in floppy emulation 

mode. 

This first boot image (normally based on DOS) with the maximum size of 

2.88 MB (Deployment Manager uses only 1.44 MB) is used to prepare the 

storage devices so that the second operating system image can be stored 

from the deployment server onto the target drive. 

A RAID array may have to be configured beforehand, and an additional 

reboot may have to be initiated again using PXE to activate the created RAID 

array (depending on the RAID controller type). 

A second PXE boot can be avoided if an already initialized RAID array with 

the right RAID level is detected. 

For security purposes, the cloning image must be decompressed and decrypted. 

Decompressing also stores the image in such a way that only used hard 

disk sectors are transferred and must be placed on the right position on the 

drive. Decrypting is used to ensure that this image will not be sent by an 

intruder. 



12.7 LAN traffic and deployment methods 

Because of very large operating system images of up to several GB, different 

deployment methods are used to keep bandwidth on the network for other 

processes. 

12.7.1 Unicast and Multicast 

Transfer with a Unicast IP protocol, only one target system can receive a 

packet at any one time. The packet header includes the MAC address of only 

one recipient. If individual data per recipient is to be distributed, this is the preferred 

method. 

The first DOS boot image containing the dynamic parameters of each client 

(IP address, host name) will be distributed in this way. 

The disadvantage is that if, for example, 100 clients are to be cloned, this 

DOS image must be created 100 times, each with its individual parameters 

for cloning its target client. 

If a RAID1 is created, and a second reboot with the same DOS image is 

required to activate the RAID array, these 100 DOS images must be transferred 

again, each with a size of up to 1.44 MB. This takes a few minutes to 

be prepared. 

Multicast is the typical method for cloning many servers over the LAN using 

the same image each time. 

The server sends only one image packet-by-packet over the LAN to the 

clients using a Multicast IP packet. 

The clients are assigned via an IP Multicast address. Each client assigned 

via such a temporary address catches this kind of broadcast and stores the 

operating system image in its memory or disk, concatenating the user data. 

Since each Multicast client receives the same image, dynamic parameters 

such as IP address, host name and SecureID are modified on the client side. 

The multicast protocol used by Deployment Manager uses an acknowledgement 

package for each transferred data package from each listening 


client. The slowest one controls the speed of the process. If more than two 

retries to a certain client are necessary for a data package, this client will be 

phased out and the protocol continues without it. That means, if one of the target 

servers is connected by a 10 Mbit connection, the whole Multicast protocol 

runs at 10 Mbit/s only, irrespective of which speed is possible with the 

other members of that Multicast group. With Unicast it is different on account 

of a peer-to-peer connection between each server and client. 

With Unicast mode each server is a session; with Multicast each job is a session. 

12.7.2 Switch, Hub and Bridge Configuration 

12.7 LAN traffic and deployment methods 

The PXE protocol searches for the PXE server and the DHCP server via 

broadcast on port 67. 

If these servers are placed behind bridges, hubs or switches with activated 

virtual LAN software, these devices must be programmed port-by-port to 

bypass these broadcasts. 

These external devices usually provide configuration APIs, accessible via 

vendor-specific configuration tools or via a programming interface. 



13 Appendix - network techniques 

13.1 MAC address handling 

13.2 PXE 

The MAC (Media Access Control) address is a hardware address that 

uniquely identifies each node of a network. If a server is already being managed 

by Operations Manager, the MAC address is automatically read from 

the ServerView Operations Manager database. 

There are several ways to find out the MAC address of a server: 

l You can unplug the server and read off the imprinted MAC address 

from the MAC/iSCSI address label on the rear side of the server. 

l You can start the server and read off the MAC address on the information 

screen of the BIOS. 

There are several ways to find out the MAC address of a server blade: 

l If you have already configured the access via the management blade, 

you can use the management blade’s Web interface (http:\\; default user: root; default 

password: root). 

l You can also unplug the server blade and read off the imprinted MAC 

address from the MAC/iSCSI address label of the server blade. 

l You can start the server blade and read off the MAC address on the 

information screen of the BIOS. 

Short for Pre-Boot Execution Environment. 

13.1 MAC address handling 

PXE is a boot mode of the LAN adapter. It does not become active until the 

system BIOS activates the LAN adapter as boot device during system boot 

and jumps to it. For this to happen, the LAN device must be set to the highest 

priority in the system BIOS boot device table. No jumpers are required for 

this. 



PXE is a mandatory element of the WfM (Wired for Management) specification. 

To be considered compliant, the PXE must be supported by the 

computer's BIOS and its NIC. 

PXE Boot Algorithm 

2-3 The boot process on the client side starts with a PXE broadcast to 

the DHCP server to receive a temporary IP address (mandatory). 

4-5 A similar broadcast discovers the PXE boot server on port 67 or 

4011 (depending on the information issued by the DHCP server) 

requesting a boot image name. 

6-8 If the required information has been provided, a TFTP session is 

started to receive the boot image from the PXE server. The image 

size must not exceed 1.44 MB (floppy disk emulation mode). This 

boot image is copied to memory address 07C0h and started by the 

BIOS. 

9 As long as the operating system kernel is not started and neither are 

the kernel drivers, any LAN access is performed using the PXE 

BIOS for further TFTP sessions. 


13.3 DHCP 

DHCP (Dynamic Host Configuration Protocol) is a protocol for assigning 

dynamic IP addresses to devices in a network. With dynamic addressing, a 

device can have a different IP address every time it connects to the network. 

In some systems, the device's IP address can even change while it is still 

connected. DHCP also supports a mix of static (reserved) and dynamic IP 

addresses. 

Further information on DHCP can be obtained at the following internet 

addresses: 

l Red Hat Linux 9: Red Hat Linux Customization Guide 

http://www.redhat.com/docs/manuals/linux/RHL-9-Manual/customguide/ch-dhcp.html 

l DHCP.org - Resources for DHCP (contains further Links) 

http://www.dhcp.org/ 

l ntfaq.com - Frequently asked questions concerning DHCP 

http://www.ntfaq.com/Articles/Index.cfm?DepartmentID=774 

l DHCP-Handbook 

http://www.dhcp-handbook.com/dhcp_faq.html 

13.4 VLAN (Virtual Local Area Network) 

The following VLAN description is based on blade servers. 

13.3 DHCP 

A virtual LAN is a network of computers that behave as if they are connected 

to the same wire even though they may actually be physically located on different 

segments of a LAN. VLANs are configured through software rather 

than hardware, which makes them very flexible. One of the biggest advantages 

of VLANs is that when a computer is physically moved to another location, 

it can stay on the same VLAN without any hardware reconfiguration. 



VLAN Configuration 

The PXE client running on the system does not support VLANs and does not 

send untagged frames. 

VLAN requirements 

l The external switch in the LAN must support VLANs. 

l The driver for the server network connectors (NIC) must allow integration 

in several VLANs. 

l The image from the PXE server must already contain a driver which is 

pre-configured for VLANs. 

Example of a VLAN Configuration 

l VLAN IDs must be assigned for all segments on the segment switch, 

except for the deploy segment. 

o I.e. data traffic toward the deploy segment is transmitted 

untagged. 

o If the PXE client sends an untagged frame, it is allocated to all 

three source ports by the switch blade. The untagged frame is 

only transferred to the untagged port (connected to the deploy 

segment) by the segment switch. 

o After the operating system has been booted on the server 

blade, it identifies several separate segments due to the VLAN 

configuration. 

l All ports in the switch blade must recognize all VLAN IDs and must be 

configured for untagged frames at the same time. Even though the 

switch blade does not perform a segmentation, without VLAN configuration 

it would abandon all frames that carry a VLAN tag. 

Summary 

l PXE service data is only forwarded to the deploy segment. 

l The VLAN configuration provides the splitting of segments after the 

boot process. 


l Data exchange among the segments is only possible via a router. 

l Full redundancy can be configured for both network controllers. 

13.5 Example: Deployment over VLAN 


The following example shows the deployment configuration of a BX300 blade 

server (contains one switch blade and four server blades) over VLAN. 



Configuration / Preparation 

Configuration: 

Make sure that an image is installed on all server blades. 

Client 30.90.1.40 255.0.0.0 

Cisco switch 30.90.1.30 255.0.0.0 

Switch blade 1 30.90.1.21 255.0.0.0 

Switch blade 2 removed 255.0.0.0 

Server blade 1 DHCP: 10.0.0.0 255.0.0.0 




Management blade 30.90.1.20 255.0.0.0 

Connect the components, see figure above. 


Deploying images into different VLANS 

Requirements: 

l Two VLANs which are not connected to each other. An image should 

be deployed into these VLANs from a third VLAN which is able to communicate 

with the other two. 

l For the whole network you must have one deployment server, one 

DHCP server and one WINS server running. 

Perform the following steps: 

1. Creating VLANs on the switch blades 

a. Assign ports 1, 2 and 13 to VLAN 10 untagged and remove 

them from VLAN 1 or any other VLANs. 

b. Set PVID for ports 1 and 2 to 10 only. 

c. Assign ports 3, 4 and 13 to VLAN 20 untagged and remove 

them from VLAN 1 or any other VLANs. Leave port 13 in VLAN 

10 also. 

d. Set PVID for ports 3 and 4 to 20 only. 

e. Assign port 13 to VLAN 30 untagged. 


f. Create a .1Q Trunk on port 13 and connect it to Cisco port 2. 

(Set frame type to tagged in egress rules (VLAN port configuration).) 

g. Assign VLAN 30 the IP address: 30.90.1.21 for the switch management 

(with CLI). 

2. Creating VLANs on the Cisco switch 

a. Create a .1Q Trunk on port 2 which is connected to the switch 

blade. 

b. Assign port 1 to VLAN 30 untagged which is connected to the 

client. 

c. Assign VLAN 30 the IP address: 30.90.1.30 for the switch management 



3. Configuring the client 

a. Install a WINS_Server on the client and configure it. 

b. Under Control panel - Add/Remove Programs - MS Components 

- Network Services, add WINS. 

No configuration is necessary for WINS. 

c. Configure the DHCP server so that there is a range for each 

subnet: 

Range WINS (044) Gateway (003) 

VLAN 10 10.90.1.1 to .10 30.90.1.40 10.90.1.30 

VLAN 20 20.90.1.1 to .10 30.90.1.40 20.90.1.30 

VLAN 30 30.90.1.1 to .10 30.90.1.40 30.90.1.30 

d. Set up MMB with a new IP address (30.x.x.x) 


e. Set Cisco ports 3 and 4 to static access VLAN 30 and 10 Mb/s 

full or half duplex (depends on the MMB setting). 

f. Try to ping the server blades in VLAN 10 and VLAN 20 from the 

client. Also ping switch blade 1, management blade and Cisco 

switch. What IP address do they have? What is the result? 

Server blade 1 




Switch blade 

Management blade 

Cisco switch 

g. Is your result okay? Why? 


IP address Result of ping 

4. Configuring routing on the Cisco switch (part 1) 

a. Create Vlan 10 and Vlan 20 under Vlan - Vlan - Configure 

Vlans. 



b. Assign each VLAN an IP address. 

Under Administration - IP Addresses: 

c. Remove the IP address for Vlan1. 

d. Click Apply. 

e. Enable IP Routing under Device - IP-Routing - Protocols. 

Set the check mark to enable IP routing for 3550: 


f. Click Apply. For 3750 under Device – IP-Routing – Protocols 

select enable\disable. 

g. Try to ping the server blades in VLAN 10 and VLAN 20 from the 

client. Also ping switch blade 1, management blade and Cisco 

switch. 

What IP address do they have? What is the result? 





Switch blade 





Management blade 

Cisco switch 

h. Is your result okay? Why? 


i. Enable the IP-Helper function. 

You must configure a relay device when a switch sends broadcast 

packets that need to be responded to by a host on a different 

LAN. Examples of broadcast packets that the switch 

might send are DHCP, DNS and, in some cases, TFTP packets. 

You must configure this relay device to forward received 

broadcast packets on an interface to the destination host. If the 

relay device is a Cisco router, enable IP routing (ip routing global 

configuration command), and configure helper addresses by 

using the ip helper-address interface configuration command. 

Cisco CLI: 

en 

conf t 

int vlan 10 

ip helper-address 30.90.1.40 

Same for VLAN 20 with the same address. 

j. Enable the forwarding for UDP ports: 

Cisco CLI (Global parameter (config)) 

en 

conf t 

ip forward-protocol udp (portnumber) 

Do this for the ports Bootps 67, Bootpc 68 and TFTP 69. 

5. IP address for server blades 

a. Execute an ipconfig /release and ipconfig /renew in the 

DOS window for each server blade. 


. Then execute an ipconfig /all. Fill out the following lines. 






IP address gateway WINS ping 

c. Try to ping the server blades in VLAN 10 and VLAN 20 from the 

client. Fill out the column above. 

d. Try to ping from a server blade in VLAN 10 and the VLAN 20 

server blade. 

What is the result? ___________________________________ 

__ 

6. Configuring routing on the Cisco switch (part 2) 

In order to define a secure routing, you must create an access list: 

a. Create an access list: 

conf t 

access-list 100 deny ip 10.0.0.0 

0.255.255.255 20.0.0.0 0.255.255.255 

access-list 100 permit ip any any 

access-list 110 deny ip 20.0.0.0 

0.255.255.255 10.0.0.0 0.255.255.255 

access-list 110 permit ip any any 

interface VLAN10 

ip access-group 100 in 

interface VLAN20 

ip access-group 110 in 

b. Try to ping from a server blade in VLAN 10 and the VLAN 20 

server blade. 




__ 

c. Try to ping the server blades in VLAN 10 and VLAN 20 from the 

client. 


__ 

d. Enable multicast routing: 

conf en 

Password: xxx 

cisco1#conf t 

cisco1(config)#ip multicast-routing distributed 

(on Cisco 3750) 

cisco1(config)#ip multicast-routing (on 

Cisco 3550) 

Enable PIM Dense: 

cisco1(config)#int vlan 10 

cisco1(config-if)#ip pim dense-mode 

cisco1(config-if)#exit 







e. Check IGMP snooping (it should be enabled by default) 

cisco1#sh ip igmp snooping 

vlan 1 

---------- 

IGMP snooping is globally enabled 

IGMP snooping is enabled on this Vlan 

IGMP snooping immediate-leave is disabled on 

this Vlan 

IGMP snooping mrouter learn mode is pim- 



dvmrp on this Vlan 

IGMP snooping is running in IGMP_ONLY mode 

on this Vlan 

vlan 10 

---------- 




this Vlan 

IGMP snooping mrouter learn mode is pimdvmrp 

on this Vlan 


on this Vlan 

vlan 20 

---------- 




this Vlan 


on this Vlan 


on this Vlan 

vlan 30 

---------- 




this Vlan 


on this Vlan 


on this Vlan 



Deploying 

Now the entire configuration is ready for deploying: 

1. First try to save the image from server blade 1. 

2. Once the save process is complete, try to deploy this new image on 

server blades 2 - 4.

ServerView Deployment Manager V6.20 - Fujitsu Technology ...

Create successful ePaper yourself

Delete template?

Save as template?