Library of Congress NAVCC Overview and Update - (lib.stanford ...
Library of Congress NAVCC Overview and Update - (lib.stanford ...
Library of Congress NAVCC Overview and Update - (lib.stanford ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>Library</strong> <strong>of</strong> <strong>Congress</strong> <br />
Archive Environments<br />
Presented by:<br />
Carl Watts<br />
Information Technology Specialist<br />
Enterprise Server Engineering
<strong>Overview</strong><br />
• Four Archive Environment<br />
– Packard Campus Audio <strong>and</strong> Visual Archive<br />
• Production Environment<br />
– <strong>Library</strong> <strong>of</strong> <strong>Congress</strong> Bit Preservation Program<br />
• LCBP Platform<br />
• GPFSLCBP Platform<br />
– Legacy Archive
PCAVC Architecture V2<br />
• 2GBps maximum throughput<br />
• Commissioned 2007<br />
• Currently Stores:<br />
– 1.3 PB<br />
– about 140K files<br />
– current growth: 95 TB / month<br />
– about 9K files per month
Packard Campus Production Archive<br />
POD<br />
1Gbps<br />
10Gbps<br />
POD<br />
Archive Data<br />
Network<br />
Workflow Server<br />
Production Data<br />
Network<br />
<strong>NAVCC</strong><br />
Network Core<br />
Archive<br />
Management<br />
Server<br />
<strong>NAVCC</strong> Culpeper<br />
POD<br />
Cisco<br />
MDS<br />
DWDM<br />
30TB<br />
HSM Cache<br />
HSM Cache<br />
Cisco<br />
MDS<br />
DWDM<br />
SL8500 Tape <strong>Library</strong><br />
T10000 Tape Dives<br />
10TB<br />
SL8500 Tape <strong>Library</strong><br />
T10000 Tape Dives<br />
Archive Management<br />
Server<br />
Remote Site<br />
Network<br />
<strong>Library</strong> <strong>of</strong> <strong>Congress</strong><br />
Remote Site<br />
Architecture Version 2
PCAVC Architecture V3 (in test)<br />
• Goal <strong>of</strong> 6GBps ingest rate<br />
– Oracle M9000<br />
• Increase HSM Cache Size<br />
– DDN SFA10000 to replace Sun FLX380<br />
• Change SMB data transfers<br />
– Signiant Content Distribution Management<br />
• Improve Tape Monitoring<br />
– Crossroads ReadVerify Appliance
LC Bit Preservation #1<br />
• 4Gbps maximum throughput<br />
• Commissioned 2008<br />
• Currently Stores:<br />
– 830 TB<br />
– about 32M files<br />
– current growth: 45 TB / month<br />
– more than 1 million files per month<br />
• Advance Tape Monitoring
LC Bit Preservation #1 <strong>and</strong> #2<br />
Internet / Internet 2<br />
NDIIPP Partners<br />
<strong>and</strong> Contractor<br />
1Gbps<br />
Primary Datacenter<br />
DC<br />
Ingest<br />
Workstations<br />
Archive<br />
Management<br />
Server<br />
GPFSLCBP<br />
External<br />
Production Network<br />
20TB<br />
HSM Cache<br />
GPFS/HPSS<br />
Archive<br />
Management<br />
Server<br />
LCBP<br />
20TB<br />
HSM Cache<br />
SAMFS<br />
Web Capture<br />
External Ingest<br />
Servers<br />
Cisco<br />
MDS<br />
SL8500 Tape <strong>Library</strong><br />
T10K(B) Tape Dives<br />
External Content<br />
Servers<br />
Content<br />
Processing<br />
Servers<br />
DWDM<br />
SL8500 Tape <strong>Library</strong><br />
T10K(B) Tape Dives<br />
TS3584 Tape <strong>Library</strong><br />
LTO4 Tape Dives<br />
Tape Monitoring<br />
<strong>Library</strong> <strong>of</strong> <strong>Congress</strong><br />
Remote Site<br />
TS3584 Tape <strong>Library</strong><br />
LTO3 Tape Dives<br />
Cisco<br />
MDS<br />
Remote Site<br />
Network<br />
DWDM<br />
Tape Monitoring<br />
Archive Management<br />
Server<br />
10TB<br />
HSM Cache
LCBP Environment Improvements<br />
• Second Archive for risk mitigation<br />
– 2x tape technology<br />
– 2x OS<br />
– 2x HSM<br />
– second vendor<br />
- adds 3 rd <strong>and</strong> 4 th copy<br />
• Increase network b<strong>and</strong>width to 10GigE
Migration <br />
Migration <br />
Migration
Near Future Directions<br />
• Moving toward digital submission<br />
• Adding Storage Resource Management<br />
• Reducing the cost <strong>of</strong> acquisition cost for storage<br />
– Segmenting Tier 2 storage into 3 sub-tiers<br />
– $1k or less per TB (10X or less consumer price per TB)<br />
• Heterogeneous Global Namespace<br />
• Geographical Dispersed Archive<br />
• Erasure Archive
Thank You<br />
Contact Information: Carl Watts<br />
Email: cwat@loc.gov<br />
LinkedIn: www.linkedin.com/in/carlwatts<br />
Join the PASIG Group on LinkedIn<br />
Some <strong>of</strong> the opinions expressed during this presentation are my own <strong>and</strong><br />
do not necessarily represent those <strong>of</strong> the <strong>Library</strong> <strong>of</strong> <strong>Congress</strong>