ST1901
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
STOR<br />
MAGAZINE<br />
STORAGE<br />
The UK’s number one in IT Storage<br />
January/February 2019<br />
Vol 19, Issue 1<br />
TRADING UP:<br />
European exchange moves to the cloud<br />
DISASTER RECOVERY:<br />
'Grappling with data loss'<br />
DIGITAL TRANSFORMATION:<br />
Preparing for Industry 4.0<br />
VDI BOTTLENECKS:<br />
Is all-flash the answer?<br />
COMMENT - NEWS - NEWS ANALYSIS - CASE STUDIES - OPINION - PRODUCT REVIEWS
TRADING UP:<br />
<br />
January/February 2019<br />
Vol 19, Issue 1<br />
CONTENTS<br />
STOR<br />
MAGAZINE<br />
STORAGE<br />
CONTENTS<br />
DISASTER RECOVERY:<br />
<br />
DIGITAL TRANSFORMATION:<br />
<br />
VDI BOTTLENECKS:<br />
<br />
<br />
Comment.......................................4<br />
GREATER REGULATION, ANYONE?<br />
06<br />
TECHNOLOGY FOCUS:<br />
OPTICAL DISK…...............................…6<br />
Hideharu Takeshima, CTO of Mitsubishi Chemical Media examines<br />
the ongoing role of archival optical disk an an increasingly data<br />
driven era<br />
INTERVIEW: STORAGECRAFT...….12<br />
Matt Medeiros, CEO of StorageCraft, speaks to Storage magazine<br />
editor David Tyler about how current rates of data storage growth have<br />
created the need for a new approach to intelligent data management<br />
14<br />
ANALYSIS: STORAGE TRENDS.….14<br />
Rainer Kaese, Senior Manager Business Development Storage<br />
Products at Toshiba Electronics Europe, examines likely storage<br />
trends for 2019<br />
20<br />
RESEARCH:<br />
DISASTER RECOVERY.....................20<br />
New research from Unitrends suggests that although DR testing is<br />
growing in popularity, there is still much to be done<br />
TECHNOLOGY FOCUS: VDI……..…24<br />
Deploying VDI can be demanding and expensive, argue Juan Mulford<br />
and Stefan Ferrari of AccelStor, but industry-validated, cost-effective,<br />
all-flash storage can help<br />
24<br />
COURT ON CAMERA………...........…28<br />
Safeguard your surveillance data or potentially fall foul of the law,<br />
warns Gary Watson, CTO of StorCentric and founder of Nexsan<br />
TECHNOLOGY FOCUS: HDD…..….30<br />
Jason Feist, Managing Technologist at Seagate Technology, describes<br />
how a laser no bigger than a grain of salt is just one crucial element<br />
of the next generation of hard drive technologies<br />
30<br />
STRATEGY:<br />
DIGITAL TRANSFORMATION.….…34<br />
Mark Jow, EMEA Vice President, Technical Services at Commvault, talks<br />
about preparing for an effective digital transformation journey<br />
www.storagemagazine.co.uk @STMagAndAwards Jan/Feb 2019<br />
^<br />
STORAGE<br />
MAGAZINE<br />
3
COMMENT<br />
EDITOR: David Tyler<br />
david.tyler@btc.co.uk<br />
SUB EDITOR: Mark Lyward<br />
mark.lyward@btc.co.uk<br />
REVIEWS: Dave Mitchell<br />
PRODUCTION MANAGER: Abby Penn<br />
abby.penn@btc.co.uk<br />
PUBLISHER: John Jageurs<br />
john.jageurs@btc.co.uk<br />
LAYOUT/DESIGN: Ian Collis<br />
ian.collis@btc.co.uk<br />
SALES/COMMERCIAL ENQUIRIES:<br />
Lyndsey Camplin<br />
lyndsey.camplin@btc.co.uk<br />
Stuart Leigh<br />
stuart.leigh@btc.co.uk<br />
MANAGING DIRECTOR: John Jageurs<br />
john.jageurs@btc.co.uk<br />
DISTRIBUTION/SUBSCRIPTIONS:<br />
Christina Willis<br />
christina.willis@btc.co.uk<br />
PUBLISHED BY: Barrow & Thompkins<br />
Connexions Ltd. (BTC)<br />
35 Station Square, Petts Wood<br />
Kent BR5 1LZ, UK<br />
Tel: +44 (0)1689 616 000<br />
Fax: +44 (0)1689 82 66 22<br />
SUBSCRIPTIONS:<br />
UK £35/year, £60/two years,<br />
£80/three years;<br />
Europe: £48/year, £85 two years,<br />
£127/three years;<br />
Rest of World: £62/year<br />
£115/two years, £168/three years.<br />
Single copies can be bought for £8.50<br />
(includes postage & packaging).<br />
Published 6 times a year.<br />
No part of this magazine may be<br />
reproduced without prior consent, in<br />
writing, from the publisher.<br />
©Copyright 2019<br />
Barrow & Thompkins Connexions Ltd<br />
Articles published reflect the opinions<br />
of the authors and are not necessarily those<br />
of the publisher or of BTC employees. While<br />
every reasonable effort is made to ensure<br />
that the contents of articles, editorial and<br />
advertising are accurate no responsibility<br />
can be accepted by the publisher or BTC for<br />
errors, misrepresentations or any<br />
resulting effects<br />
GREATER REGULATION, ANYONE?<br />
BY DAVID TYLER<br />
EDITOR<br />
As we start a new year, it seems that data protection is still on the minds of<br />
many and isn't going away any time soon. In fact, 80% of IT professionals<br />
believe more stringent data protection regulations will benefit their<br />
organisations, but only 20% are confident that they comply with the latest regulatory<br />
requirements. This is according to a survey conducted late last year at the Data<br />
Protection World Forum by Commvault.<br />
Many of the findings of the survey make for worrying reading: almost two thirds of<br />
respondents (64%) felt the changing business processes and cultural requirements<br />
necessary for the uptake of genuinely compliant data-centric practices across<br />
organisations were greater barriers than the technology issues. This is despite the<br />
fact that technology challenges such as cloud migration, integrating with legacy<br />
systems and managing unstructured data were still areas of concern.<br />
Despite this, almost 40% of respondents actually stated that they would welcome<br />
more data regulation. The thinking presumably here is that more regulation and<br />
enforcement could result in better data management practices being implemented in<br />
their organisations in the long term.<br />
Meanwhile, nearly one in five respondents (17%) said that the penalties for falling<br />
foul of current regulatory policy were still not strict enough. This means that<br />
regulations might need to be stricter in order to drive the real organisational changes<br />
required to recognise the potential business value<br />
"With many businesses still struggling to navigate the challenges of data protection<br />
regulations like GDPR, it's never been more important than it is today, to focus on<br />
the core principles of effective data management, beyond simply a tech perspective,<br />
but to also include cultural and process elements too," comments Jo Blazey, Global<br />
Data Governance Officer, Commvault."Every benefit you can gain from data tracks<br />
back to four simple premises: being able to see what data you have, where it's<br />
stored, how it's protected and what it's being used for. Compliance with current<br />
regulations requires changes in business and working cultural behaviours to achieve<br />
this but let's be clear, better data management is as much a business or legal priority,<br />
as it is a technology one."<br />
What do you think? Would you welcome even more regulation around data<br />
protection? I'd be interested to hear your views, via my email (see top left) as ever.<br />
^<br />
04 STORAGE<br />
MAGAZINE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk
TECHNOLOGY FOCUS: OPTICAL DISC FOCUS:<br />
DATA - THE OIL OF OUR TIME<br />
HIDEHARU TAKESHIMA, CTO OF MITSUBISHI CHEMICAL MEDIA EXAMINES THE ONGOING ROLE OF<br />
ARCHIVAL OPTICAL DISC IN AN INCREASINGLY DATA DRIVEN ERA<br />
With the popularisation of Social<br />
Networking Services (SNSs), the<br />
Internet of Things (IoT) and higher<br />
camera resolutions, the amount of data<br />
being generated will continue to increase<br />
dramatically, improving productivity and<br />
creating new business opportunities. We are<br />
in the Data Driven Era.<br />
Artificial Intelligence (AI) is considered to<br />
be the key technology in this era with<br />
Deep Learning the mainstream of AI,<br />
creating the neural network that learns<br />
from known input and output data and<br />
predicting optimum answers. The accuracy<br />
of the analysis is determined by how much<br />
learning data is given to the neural<br />
network. In addition, as new analysis<br />
algorithms are created, new knowledge<br />
can be obtained by analysing the past<br />
data with the new algorithms. Therefore,<br />
captured data that had previously not<br />
been considered useful may now prove<br />
highly valuable.<br />
Oil has been the world's most important<br />
substance since the mid-1950s - the<br />
lifeblood of the industrialised nations. Its<br />
products underpin modern society, mainly<br />
supplying energy to power industry, heating<br />
homes and providing fuel for vehicles and<br />
aeroplanes to carry goods and people all<br />
over the world.<br />
We are now living in a data-driven society<br />
and it can be argued that data is now more<br />
important and valuable than oil. Data can<br />
be used, but unlike oil, it does not have to<br />
be consumed. It can be kept for future<br />
analysis, so we have a duty to store data<br />
with the view that it may prove useful in the<br />
future. The challenge is how we should<br />
store these increasingly large volumes of<br />
data in a way that is cost effective and<br />
accessible for the future.<br />
06 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
TECHNOLOGY FOCUS: OPTICAL DISC<br />
the reliability of this storage method.<br />
DATA EXPLOSION<br />
As shown in Figure 1 (below right), due to<br />
the popularisation of IoT, the explosive<br />
growth of data is expected to exceed the<br />
storage capacity available to the world. The<br />
result is that valuable data is being<br />
discarded and lost. This dramatic growth in<br />
demand for data storage combined with the<br />
nature of this data - often referred to as<br />
Cold Data as it has low access frequency<br />
requirements and is stored for a long period<br />
of time - necessitates a change in how we<br />
store this data from now on.<br />
ARCHIVAL GRADE<br />
Optical discs are still one of the most<br />
popular forms of storage media in the<br />
world; for consumer applications, but also<br />
for business purposes, administrative<br />
materials and research materials in<br />
museums, libraries, and archives.<br />
For optical discs used for archiving, the<br />
reliability of the storage method and media<br />
is critical. Long-term preservation of data<br />
can be realised by combining longer<br />
lifetime optical discs, called Archival Grade<br />
Optical Media, which has proven higher<br />
reliability than regular optical media,<br />
together with a dedicated optical drive.<br />
addition, a dedicated recording drive is<br />
used and a specific recording method is<br />
adopted that draws out the best<br />
performance. There is also a drive<br />
employed that measures the quality of the<br />
recorded signal and periodically checks the<br />
quality of the recording.<br />
The optical media industry has benefited<br />
hugely from the adoption of international<br />
standards developed by ISO and JIS to<br />
ensure wide compatibility. Table 1 (above)<br />
shows the three standards that are used:<br />
Format Standard, which ensures<br />
compatibility with the drive used for<br />
recording or reproducing the data; Lifetime<br />
Measurement Standard, a test to assess the<br />
likely longevity of the media, and<br />
Operational Standard, which guarantees<br />
It has been reported that optical media<br />
using a stable inorganic material recording<br />
film can last for several hundred years - or<br />
even several thousands of years - when kept<br />
at a temperature of 25°C and a relative<br />
humidity of 50%. Also, using the<br />
parameters obtained by the lifetime<br />
specification and the Eyring formula, the<br />
estimated lifetime at different temperatures<br />
and humidity can be obtained.<br />
For example, even at a temperature of<br />
35°C and a relative humidity of 50%, the<br />
estimated lifetime is over a hundred years,<br />
and it can be presumed that data can be<br />
stored and managed for a long period of<br />
time even in a rough storage environment.<br />
Based on all of this evidence, it is possible<br />
to significantly extend the data migration<br />
interval which is carried out every three to<br />
five years on a normal hard disk, by using<br />
archival optical discs instead.<br />
In optical disc library systems, power is<br />
not needed except during recording and<br />
reproduction, so the amount of energy<br />
required and generation of heat is very<br />
small. Therefore, the use of archival optical<br />
discs make it possible to suppress energy<br />
For archival optical discs, Mitsubishi<br />
Chemical Media, which also owns the<br />
popular Verbatim brand, has improved<br />
resilience by using more durable materials<br />
compared to consumer optical discs, and<br />
strict quality control standards with tighter<br />
specifications to remove variation. In<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
07
TECHNOLOGY FOCUS: OPTICAL DISC FOCUS:<br />
can reduce the data migration cost and the<br />
power consumption cost when compared to<br />
RAID (hard disk) and tape, as can be seen<br />
here in Figure 2.<br />
and CO2 emissions associated with air<br />
conditioning.<br />
From a cost perspective, an optical disc<br />
library system can be considered to be an<br />
ideal long-term data storage method that<br />
In addition, it has been proven that<br />
archival optical disc can be durable in<br />
extreme environmental conditions. In a<br />
test where dual-layered archival BD-R was<br />
submerged at 1.5 m in the sea for a<br />
week, then taken out, washed with water,<br />
dried, and the Random Signal Error Rate<br />
(RSER) measured, the result confirmed<br />
that the data stored on the media could<br />
still be read.<br />
Furthermore, and very importantly, due to<br />
the industry standards employed, backwards<br />
compatibility is very good across<br />
generations of media, with old discs being<br />
able to be read on the latest generation of<br />
DiskStation DS1019+<br />
The perfect blend of performance & storage<br />
Up to 70 TB raw capacity in one box<br />
• Intel quad-core CPU, 1.5GHz burst up to 2.3GHz, 8GB RAM<br />
• 2x M.2 SSD slots supporting NVMe SSD cache<br />
• Scalable up to 10 drives, 140 TB raw capacity<br />
• AES-NI encryption engine & dual-channel H.264/H.265 4K transcoding<br />
• Backed by Synology’s 3-year limited warranty. Upgrade to 5-year warranty<br />
Available now!
TECHNOLOGY FOCUS: OPTICAL DISC<br />
optical drives. This helps with migration<br />
time and reduces costs. And the<br />
development is continuing with Mitsubishi<br />
Chemical Media working on new Archival<br />
Grade Optical Media with higher capacity<br />
in combination with optical drive<br />
manufacturers to increase the total capacity<br />
of optical library systems.<br />
A MULTITUDE OF OPTIONS<br />
Everyone is familiar with the small-scale<br />
systems that manually record data on<br />
optical discs using a PC and an optical<br />
disc drive and store them on shelves, but<br />
there are numerous options for larger<br />
scale systems.<br />
There are online systems in which a robot<br />
carries an optical disc, fills it in a drive, and<br />
records and reproduces it. There are also<br />
systems that incorporate different recording<br />
media (flash memory, hard disks, tapes,<br />
etc.) and provide a rational system that<br />
automatically selects optimal data storage<br />
locations by hierarchical technology.<br />
We are living in a time where data is one<br />
of the most valuable things that we create,<br />
and where the care of which could have a<br />
significant influence on the way our society<br />
will learn and develop. Optical media<br />
clearly has its part to play in this, providing<br />
a very reliable and efficient method of longterm<br />
data storage so that we can benefit.<br />
The industry standards that have been<br />
crucial in its success so far, ensuring<br />
compatibility across generations of optical<br />
media technology, will be maintained to<br />
continue to direct its development as higher<br />
capacity optical media and library systems<br />
are launched over the coming years.<br />
More info: www.verbatim.com<br />
with EW201<br />
www.synology.com
CASE STUDY: SENSIRION STUDY: SENSIRION<br />
SENSOR SENSIBILITY<br />
ADVANCED SENSOR MANUFACTURER SENSIRION HAS SHIFTED FROM BACKUP TO A SECONDARY<br />
STORAGE MANAGEMENT ENVIRONMENT WITH THE HELP OF COHESITY<br />
Headquartered in Stäfa,<br />
Switzerland, Sensirion is one of the<br />
leading manufacturers of digital<br />
microsensors and sensor systems.<br />
Established in 1998, the company<br />
employs a workforce of just under 750<br />
with operations in the U.S., South Korea,<br />
Japan, China, Taiwan, and Germany.<br />
Their product range includes gas and<br />
liquid sensors as well as differential<br />
pressure and environmental sensors for<br />
measuring temperature and humidity,<br />
volatile organic compounds (VOC), CO2,<br />
and particulate matter.<br />
Sensirion supports their customers with<br />
both standard and tailor-made sensor<br />
system solutions for a wide range of<br />
applications. Sensirion sensors are widely<br />
used in healthcare, industrial, and<br />
automotive applications as well as in<br />
analytical instruments, in the consumer<br />
goods industry, and in heating,<br />
ventilation, and air conditioning<br />
equipment.<br />
CHALLENGES OF EXPANSION<br />
With backup storage in need of expansion<br />
and operating costs on the rise,<br />
Sensirion's executives were on the lookout<br />
for a new storage solution.<br />
Previously, Sensirion had worked under a<br />
traditional data centre design. The<br />
10<br />
STORAGE<br />
MAGAZINE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk
CASE STUDY: CASE STUDY: SENSIRION<br />
"Cohesity is more than a simple backup solution; it also holds its own in other<br />
use cases. Auditing is clearly better. We do not have to set up our own<br />
reports and the solution is very transparent for the customer. It's easy to<br />
understand and future-proof."<br />
applications serving their workforce of just<br />
below 750 ran in two data centres, and<br />
around 350 VMs were deployed using<br />
Pure Storage as tier one storage, with a<br />
full flash solution for applications with<br />
high performance requirements.<br />
With backup storage needing expansion,<br />
rising technological demands, and<br />
growing licensing costs, the executives at<br />
Sensirion began to evaluate new<br />
solutions. Andreas Kuehne, Global IT<br />
Infrastructure Services Lead at Sensirion<br />
AG, puts it this way: "We took the<br />
optimisations and renewals we had in<br />
mind as a good opportunity for us to<br />
question our concept and find a<br />
replacement for our existing solution. We<br />
looked for a way to underpin the<br />
Snapshot for Pure Storage functionality.<br />
Furthermore, a new solution was to<br />
enable archiving in the cloud."<br />
Sensirion needed to address a number<br />
of challenges in a new storage<br />
architecture:<br />
Backup storage: Support nearing its<br />
end set the scene for a new storage<br />
solution.<br />
Pure Storage: Flash memory was at its<br />
limit and needed expansion.<br />
Cloud backup: Very costly due to<br />
missing or at best limited data<br />
deduplication and compression.<br />
Sensirion's executives entered into a<br />
multiple stage selection process based on<br />
a number of key criteria. The key<br />
requirement was to allow both Pure<br />
Snapshots and SMB Targets. Furthermore,<br />
archiving into the cloud was to be an<br />
option, allowing the company to shrink its<br />
tape infrastructure to a minimum or<br />
reduce its role to that of a static archive.<br />
A WHOLE NEW APPROACH<br />
The project officials decided to prepare a<br />
proof-of-concept (PoC) with expertise from<br />
information management and backup<br />
specialist Infoniqa. Sven Meyer, Senior<br />
System Engineer with Sensirion, explains<br />
the idea: "To begin with, we planned to<br />
keep the existing backup environment and<br />
use Cohesity only as a backup target for<br />
the time being. In addition, Pure<br />
Snapshots were to be replicated to<br />
Cohesity. We did not manage to do that<br />
altogether, however, and needed to find a<br />
different approach for the PoC."<br />
But although Cohesity did require<br />
customisation, it still won Sensirion's IT<br />
professionals over with its rapid and easy<br />
installation. One of the reasons tipping<br />
the balance was that Cohesity had more<br />
to offer than just a simple backup<br />
solution. Their portfolio encompassed<br />
other application options as well - going<br />
as far as a management environment for<br />
secondary storage, with archiving,<br />
analytics, file and object storage, and<br />
testing facilities all included. In addition,<br />
Cohesity provided a native cloud solution.<br />
MULTIPLE BENEFITS<br />
Sven Meyer cites strong points for<br />
choosing Cohesity: "Cohesity is more than<br />
a simple backup solution; it also holds its<br />
own in other use cases. Auditing is clearly<br />
better. We do not have to set up our own<br />
reports and the solution is very<br />
transparent for the customer. It's easy to<br />
understand and future-proof. Handling it<br />
is cool and the Sensirion engineers are<br />
quite happy using it. It meets our<br />
requirements."<br />
Cohesity DataProtect completed backups<br />
significantly faster than the existing<br />
solution had ever done. In addition, it<br />
showed a convincingly good performance<br />
in restore operations.<br />
Other key benefits with Cohesity include:<br />
Rapid and simple deployment<br />
Reliable operations<br />
Pleasant experience working with<br />
Cohesity's engineers<br />
Intuitive user interface<br />
Convenient, easy-to-manage file<br />
system<br />
First class support<br />
Additional potential, future-proof<br />
From the initial success of the Cohesity<br />
solution for backup and recovery,<br />
Sensirion will soon consider using other<br />
features of Cohesity. The team is looking<br />
at including archiving and data analysis,<br />
tape export, as well as cloud integration<br />
and disaster recovery from the cloud.<br />
"For us, Cohesity's ease of deployment,<br />
its operational simplicity, and its<br />
professional support were the deciding<br />
factors," concluded Andreas Kühne. "We<br />
can see additional potential to develop<br />
this backup solution into a secondary<br />
storage management environment -<br />
across sites and with cloud integration."<br />
More info: www.cohesity.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
11
INTERVIEW: STORAGECRAFT STORAGECRAFT<br />
GREATER THAN THE SUM OF<br />
ITS PARTS<br />
MATT MEDEIROS, CEO OF STORAGECRAFT, SPEAKS TO STORAGE<br />
MAGAZINE EDITOR DAVID TYLER ABOUT HOW CURRENT RATES<br />
OF DATA STORAGE GROWTH HAVE CREATED THE NEED FOR A<br />
NEW APPROACH TO INTELLIGENT DATA MANAGEMENT<br />
David Tyler: It's a couple of years<br />
now since StorageCraft saw a<br />
significant investment in its<br />
business from TA Associates, which was<br />
followed by the acquisitions of Gillware<br />
and then Exablox. To what extent has this<br />
helped you to deliver your proclaimed aim<br />
of a platform that combines primary and<br />
secondary storage?<br />
Matt Medeiros: We have always had a<br />
vision: StorageCraft had a legacy software<br />
offering for backup and recovery, but it<br />
wasn't a complete solution. We needed to<br />
move things on quickly around product<br />
development, and acquisition was one<br />
way for us to get products and solutions to<br />
suit our customers' needs, and also to<br />
have better control of our own destiny. At<br />
the same time we had to revamp our<br />
approach to the channel - not because<br />
what we were doing was wrong, but more<br />
because the solutions we are selling today<br />
are very different to what we were selling<br />
prior to the TA investment.<br />
Looking at our business model today we<br />
sell an awful lot of subscription services to<br />
the MSP market, around backup and<br />
recovery for their customers. Customers<br />
had been saying to us "We like your<br />
technology, but we don't want to have to<br />
do the heavy lifting of pairing your<br />
technology with the server, we don't want<br />
to have to optimise the solution - we<br />
would prefer you to do that."<br />
DT: So the combination of StorageCraft<br />
and Exablox technologies is more than the<br />
sum of its parts?<br />
MM: This was an opportunity to really<br />
disrupt the BDR (backup and disaster<br />
recovery) market; when you optimise the<br />
StorageCraft and Exablox technology you<br />
no longer have just a BDR offering, you<br />
have a scale-out intelligent data<br />
management solution. And that is<br />
precisely what we had defined as our<br />
vision.<br />
We acquired Exablox over a year ago,<br />
and we've been working on unifying the<br />
products, making sure they not only<br />
12 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
INTERVIEW: INTERVIEW: STORAGECRAFT<br />
"We're in danger of losing sight of the dilemma that if data is doubling every<br />
year, how does my IT budget keep up with that? We've simply got to get<br />
smarter."<br />
complement each other, but work really<br />
well with each other. Now we provide our<br />
ShadowXafe product, the next generation<br />
of our ShadowProtect software, and it had<br />
to be optimised and unified onto our<br />
platform. That's what we've been spending<br />
our R&D efforts on since the Exablox<br />
acquisition. Now those products are<br />
integrated, the companies integrated, and<br />
we are able to serve our customers in a<br />
much more seamless way.<br />
DT: How has this shift toward a broader<br />
data management offering affected<br />
business?<br />
MM: We officially started shipping product<br />
in October 2018, but well before that our<br />
Exablox business in Europe was growing at<br />
well over 100% CAGR. That's not just<br />
down to the product, but also to the sales<br />
model we deploy here, as well as the<br />
people we employ and the leadership of<br />
the management team.<br />
Our cloud offering is also enjoying well<br />
over 100% CAGR here in Europe. There is<br />
again a great connection unifying those<br />
products: we can seamlessly move data<br />
from a OneXafe solution to the cloud and<br />
back. Customers are definitely looking for<br />
that easy 'transportability', and the<br />
knowledge that their data is always going<br />
to be available to be recalled on demand.<br />
DT: Does this growth in cloud business<br />
mean a shift in focus for you toward more<br />
enterprise level prospects rather than your<br />
traditional SME market?<br />
MM: Some of our sales now would<br />
certainly suggest that they are well<br />
beyond what we'd typically think of as an<br />
SME type of investment, so yes I think so.<br />
Six-digit sales by customer are not<br />
uncommon for us these days. Equally it's<br />
not uncommon for a customer to be<br />
talking to us about not a terabyte of data<br />
in our cloud but a petabyte. Those are<br />
the kind of conversations we're having<br />
increasingly now: complexity, yes. Huge<br />
scale, yes as well.<br />
Customers are also respecting the level<br />
of agility that we've brought into the<br />
business - unlike a lot of the status quo<br />
storage providers, our users have an<br />
expectation on us to be more nimble. The<br />
definition there of agility is not about us<br />
'jumping through hoops', it's more about<br />
data being transportable, while still<br />
maintaining data integrity. Customers<br />
value that they don't have to go through a<br />
validation process to ensure that what<br />
they wanted transferred, did get<br />
transferred, whether that be on-premise or<br />
in the cloud.<br />
DT: How important is your partner network<br />
to the shift to cloud?<br />
MM: We do all of our business through<br />
partners, and in the UK a huge<br />
percentage of our business is specifically<br />
through MSP partners, and they are<br />
having that conversation internally:<br />
whether to carry on doing things in their<br />
own data centres, or to come to<br />
StorageCraft and exploit the IP that we've<br />
built up. We're seeing a large transition of<br />
people moving to our cloud over the last<br />
twelve months or so, because it is so<br />
robust and so simple as a solution - and<br />
we anticipate that accelerating this year.<br />
It comes down to this: whether you're an<br />
SME or a large enterprise, we have a<br />
foundational belief that not all data is<br />
equal. Certain data, you want to make<br />
sure is always accessible. And there is<br />
other data that, frankly, you just want to<br />
move it off to the cloud and "I'll get it when<br />
I get it."<br />
DT: It sounds like a large element of your<br />
sales proposition is based on the idea that<br />
you can help businesses to regain control<br />
of a storage infrastructure that might be at<br />
risk of collapsing under the weight of<br />
exponential data growth.<br />
MM: We're in danger of losing sight of the<br />
dilemma that if data is doubling every<br />
year, how does my IT budget keep up with<br />
that? We've simply got to get smarter.<br />
StorageCraft can give you intelligent<br />
tiering of your data, then you can make<br />
practical decisions on it. That's why we<br />
love our partner model - they are the<br />
professional services providers, who can<br />
go into a customer and analyse their<br />
business agenda while also helping them<br />
organise their data pools in the best way.<br />
That is what we mean by intelligent data<br />
management: you can go in, perform a<br />
targetted analysis and prescribe a simple<br />
approach toward how the customer<br />
should look at their data and recovery.<br />
But this requires work - it's not just a<br />
matter of flicking a switch. Someone has<br />
to profile the data: what's important to<br />
your business? The answer will be<br />
different depending on who you ask, of<br />
course. That's why our partners have to<br />
have a close working relationship with the<br />
business.<br />
It takes time to get there, but people are<br />
doing it - why are they willing to invest<br />
time and money to get this right? Because<br />
they just can't keep up with the exponential<br />
rate of growth of data. I've heard so many<br />
times: "My IT budget in 5 years is going to<br />
be 90% storage…" That's just wrong.<br />
More info: www.storagecraft.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
13
ANALYSIS:<br />
ANALYSIS: STORAGE TRENDS<br />
IT'S NOT RAINING DATA, IT'S POURING<br />
RAINER KAESE, SENIOR MANAGER BUSINESS DEVELOPMENT STORAGE PRODUCTS AT TOSHIBA<br />
ELECTRONICS EUROPE, EXAMINES LIKELY STORAGE TRENDS FOR 2019<br />
Where on earth are we going to put<br />
all this data? Thanks to engineers<br />
and programmers, disk drives are<br />
becoming more voluminous and combining<br />
them into efficient storage systems is getting<br />
easier. Taken together, we can see that the<br />
challenges we are facing today will be<br />
simpler to resolve.<br />
But with ever more data predicted to be<br />
generated by machines, such as<br />
autonomous vehicles and smart factories,<br />
coupled with the gigantic quantity of<br />
material already being stored and<br />
backed-up by humans, will we be able<br />
create enough storage for the coming<br />
decade's needs? Or will we have to<br />
contemplate a more ruthless approach<br />
and start to contemplate what warrants<br />
being stored at all?<br />
BALANCING HDD AGAINST SSD<br />
Not only does the amount of data that we<br />
store continue to grow unabated, its growth<br />
is faster than predicted. The expectation had<br />
been that, while the proportion of data<br />
stored on flash and SSD increased, there<br />
would be a drop in the quantity of data<br />
stored on hard drives and magnetic tape.<br />
However, it is clear today that all three<br />
technologies continue to grow simply<br />
because there is so much data to be stored.<br />
In 2019 it can be assumed that 90% of the<br />
capacity for typical cloud computing<br />
applications will be realised with hard disks,<br />
with some possibly on magnetic tape, and<br />
only 10% will be implemented with SSD. But,<br />
since Enterprise SSDs cost up to ten times as<br />
much as HDDs per unit capacity, the<br />
financial investment will be balanced with<br />
around 50% spent on HDDs and the same<br />
invested in SSDs. These storage systems<br />
cover the entire spectrum of applications,<br />
from "all-flash" appliances, to hybrid models<br />
with flash for cache or hot data and HDD<br />
for cold/warm data, through to pure hard<br />
disk-based storage servers.<br />
HELIUM HDDS OFFER 20TB<br />
All three major manufacturers are now<br />
shipping HDD models filled with helium,<br />
with 14TB capacities currently available.<br />
Over the coming years capacity can be<br />
14 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
ANALYSIS:<br />
ANALYSIS: STORAGE TRENDS<br />
"Not only does the amount of data that we store continue to<br />
grow unabated, its growth is faster than predicted. The<br />
expectation had been that, while the proportion of data stored<br />
on flash and SSD increased, there would be a drop in the<br />
quantity of data stored on hard drives and magnetic tape.<br />
However, it is clear today that all three technologies continue to<br />
grow simply because there is so much data to be stored."<br />
expected to increase at a rate of around<br />
2TB per year, meaning 20TB HDDs should<br />
be available at the beginning of the next<br />
decade. These hard drives are likely to be<br />
optimised for high capacity at a low price,<br />
but notable improvements in other<br />
technical parameters are not expected.<br />
One exception is power consumption,<br />
which will reduce as a result of the<br />
introduction of helium in HDDs.<br />
While air-filled 3.5" 7200rpm HDDs<br />
consumed a relatively constant 11W of<br />
power under load, regardless of capacity,<br />
the power consumption of helium-filled<br />
HDDs lie at around 6 - 7W. This is as a<br />
result of the lower friction of the lighter<br />
helium gas. Thus, the introduction of<br />
helium-filled hard drives will help to tackle<br />
the challenge of increasing energy<br />
consumption of data centres. Every watt of<br />
power saved by such drives results in less<br />
energy required by a data centre as well as<br />
less dissipated heat, resulting in more<br />
economical cooling.<br />
A knock-on effect of the reduced<br />
temperature is that helium-filled drives also<br />
have an increased reliability compared to<br />
air-filled drives in continuous operation.<br />
This results in far fewer failures and a<br />
longer life. Further increases in storage<br />
density are also in the pipeline, with<br />
technologies such as microwave assisted<br />
magnetic recording (MAMR) to be<br />
integrated into hard drive write heads.<br />
STORAGE ARCHITECTURES<br />
We can expect a continuing growth in<br />
top-load rackmount storage solutions due<br />
to capacity demands. While 60 bays in a<br />
4U format is standard today, there are<br />
already enclosures supporting 78 to<br />
around 110 bays for 3.5" hard drives.<br />
Instead of opting for hardware RAID, such<br />
quantities of drives are configured using<br />
software solutions.<br />
Modern software-defined storage systems<br />
will continue to dominate, along with<br />
scale-out designs such as Ceph clusters,<br />
with several storage servers being<br />
combined into larger units. Here data<br />
protection is no longer ensured through the<br />
redundancy of hard disks in the server.<br />
Instead, redundancy is implemented<br />
through the storage servers nodes<br />
available on the server network.<br />
DATA EXPLOSION?<br />
Today there is already an enormous<br />
amount of data being generated by<br />
people. When we also consider that this<br />
data is then backed up in data centres and<br />
the cloud, this only serves to multiply the<br />
amount of storage needed. To date, the<br />
quantity of machine-generated data has<br />
been, by comparison, rather low.<br />
However, this will change from 2019<br />
onwards as solutions and technologies<br />
such as autonomous driving, smart<br />
factories, Internet of Things (IoT) and home<br />
automation generate further data streams<br />
that need to be stored. The expected<br />
amount of data is so large that the current<br />
philosophy of data storage is under<br />
scrutiny. The harsh reality is that we will<br />
need to analyse data before it is stored to<br />
determine which data is really important<br />
and needs to be retained.<br />
AI, DEEP LEARNING AND<br />
BLOCKCHAIN<br />
New computing applications, such as<br />
Artificial Intelligence (AI), Deep Learning<br />
and Blockchain have increased the<br />
demands on processing performance<br />
dramatically. We can expect these<br />
technologies to generate much more data<br />
and demand access to storage solutions.<br />
Currently it is unclear precisely what<br />
impact they will have on storage<br />
requirements, as not enough is known<br />
about the applications and how they will<br />
be implemented. We should, however, start<br />
to acquire more clarity as we move through<br />
2019 and into the next decade. What is<br />
clear today is that these technologies will<br />
even more increase in the amount of data<br />
to be stored.<br />
More info: toshiba.semicon-storage.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
15
PRODUCT REVIEW REVIEW<br />
ALTARO VM BACKUP 8.2<br />
Businesses worried about protecting<br />
their virtualised environments can<br />
rest easy with Altaro VM Backup<br />
on their side. This slick software product<br />
can secure both VMware and Hyper-V<br />
VMs and is so easy to use it can be up<br />
and running inside 15 minutes.<br />
It's super value, with prices starting at<br />
only £445 per host for the Standard<br />
edition which protects five VMs and has<br />
no limitations of the number of host<br />
sockets or CPU cores. The Unlimited<br />
edition starts at £545 per host and adds<br />
deduplication, Exchange item-level<br />
restore plus cluster support while the<br />
Unlimited Plus edition adds a cloud<br />
management console for MSPs, cloud<br />
backup to Azure, WAN-optimised<br />
replication and CDP (continuous data<br />
protection).<br />
Deployment is swift: we loaded VM<br />
Backup on a Windows Server 2016 host<br />
in five minutes. The console presents a<br />
wizard where you declare Microsoft<br />
Hyper-V, VMware ESXi Hypervisor and<br />
VMware vCenter Server hosts plus<br />
backup locations - and you are ready<br />
to go.<br />
For testing, we used the lab's Hyper-V<br />
and VMware ESXi hosts and entered<br />
their IP addresses and credentials, after<br />
which VM Backup confirmed<br />
connections with them. For our backup<br />
locations, we declared an SMB share on<br />
a high-capacity Qnap NAS appliance<br />
for local backup and another share on a<br />
Synology appliance for off-site copies.<br />
Now comes the fun part, as most<br />
backup operations are drag and drop.<br />
From the Hosts view in the main console,<br />
we selected VMs from both our hosts and<br />
simply dragged them onto our backup<br />
locations.<br />
Backups can be run manually with one<br />
click but if you drag VMs onto predefined<br />
backup schedules, these will be<br />
automatically applied. Schedules are easy<br />
to create as you choose weekly or<br />
monthly recurrences, the days and times<br />
you want them to run and whether<br />
backups are followed by off-site copies.<br />
Predefined data retention policies are<br />
included but you can quickly create new<br />
ones for on-site and off-site backup<br />
locations. Just decide how many versions<br />
should be kept and for on-site backups,<br />
choose whether older ones are deleted or<br />
archived using GFS (grandfather, father,<br />
son) rotation.<br />
Enable CDP on selected VMs and you<br />
can schedule backups for as often as<br />
every five minutes. Application consistent<br />
backups can be applied to VMs running<br />
VSS-aware apps such as Exchange and<br />
SQL Server where VM Backup<br />
communicates directly with them to ensure<br />
transaction consistency.<br />
The Unlimited edition performs data<br />
deduplication by default on all VM<br />
backups, but you can disable it on<br />
selected VMs if you wish. To use off-site<br />
storage you set a master encryption key,<br />
and once configured it can also be used<br />
to encrypt local VM backups as well.<br />
Recovery features are outstanding as you<br />
can clone a VM, restore its virtual hard<br />
disk (VHD) to the same host or another<br />
one and perform granular restores of files<br />
or Exchange items. We cloned one of our<br />
Hyper-V Windows Server 2016 VM<br />
backups and the job automatically<br />
created a new VM for us on our host<br />
ready to be powered up.<br />
Businesses should test their data<br />
protection solution to ensure it'll work<br />
when they need it most and VM Backup's<br />
sandbox feature does just that. It verifies<br />
the integrity of selected backups but,<br />
more importantly, performs full test<br />
restores by cloning a VM backup to the<br />
host to ensure it'll boot - and all without<br />
impacting on the live source VM.<br />
Product: VM Backup 8.2<br />
Supplier: Altaro Software<br />
Web site: www.altaro.com<br />
Tel: +44 (0) 203 397 6280<br />
Sales: sales@altaro.com<br />
Price: Starts from £445 ex VAT<br />
VERDICT: With its intuitive console and slick drag and drop operations, Altaro VM Backup takes the pain out of protecting Hyper-V<br />
and VMware environments. It's remarkably easy to deploy, offers a wealth of valuable features and simply won't be beaten for value.<br />
16 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
CASE STUDY: EURONEXT STUDY: EURONEXT<br />
TRADING UP<br />
OVER THE PAST 2 YEARS TALEND HAS SUCCESSFULLY WORKED ON THE MIGRATION OF TRADING<br />
EXCHANGE EURONEXT'S DATABASE - ONE OF THE LARGEST IN EUROPE - TO THE CLOUD, HELPING<br />
THE COMPANY TO POSITION ITSELF FOR A FUTURE AS A 'DATA TRADER'<br />
Following its split from the New York<br />
Stock Exchange in 2014, Euronext<br />
became the first pan-European<br />
exchange in the eurozone, fusing together<br />
the stock markets of Amsterdam, Brussels,<br />
Dublin, Lisbon, and Paris. Euronext<br />
comprises close to 1,300 issuers, reporting<br />
a total market capitalisation of 3,700 billion<br />
euros at the end of March 2018.<br />
In 2016, Euronext began the typical<br />
process of migrating its data to the cloud.<br />
Except that this migration had nothing<br />
typical about it at all. First, the Euronext<br />
database contained 100 TB of data - one of<br />
the biggest in Europe. Then there was the<br />
fact that this was not just a simple transfer of<br />
a database to a hosted platform. The idea<br />
was to create a 'governed data lake' with<br />
self-service access for business units and<br />
clients in an effort to monetise new services<br />
and generate additional revenues.<br />
MIGRATING TO A GOVERNED<br />
CLOUD<br />
"We use Optiq, a trading platform with<br />
systems that work practically in<br />
nanoseconds," explains Abderrahmane<br />
Belarfaoui, Chief Data Officer (CDO) at<br />
Euronext. The huge Euronext database is the<br />
active memory of transactions handled<br />
directly by the stock exchange operator (1.5<br />
billion messages per day). "The database is<br />
compressed (at a rate of 400%) but some<br />
information was not being archived for lack<br />
of space," says Belarfaoui about the<br />
problems of the old system.<br />
Before 2016, Euronext stored its data on<br />
site, on hardware from one of the big<br />
names in the industry. But Euronext's storage<br />
needs continued to grow, especially<br />
following several acquisitions, such as the<br />
Dublin Stock Exchange and Fast Match in<br />
the US.<br />
"Our IT infrastructure had reached the end<br />
of its lifecycle in our European operations,<br />
where regulators were expecting that<br />
Euronext store more and more data,"<br />
Belarfaoui recalls. "Moreover, sometimes we<br />
had to wait six to twelve hours after market<br />
close on days with important events, such as<br />
the UK Brexit vote, before we could send the<br />
data to business units and clients."<br />
The situation prompted Euronext to look at<br />
moving to a hybrid cloud model. "We still<br />
keep trading platform information on an onsite<br />
server because lag times are not yet<br />
available on the cloud," explains Belarfaoui.<br />
"We also use AWS Managed Services in<br />
serverless mode together with Amazon S3 to<br />
have access to a data warehouse with<br />
unlimited storage capacity. For analysis, we<br />
use Amazon Redshift. And taking advantage<br />
of the cloud's great scalability, we can run<br />
the whole system while anticipating events<br />
18 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
CASE STUDY: CASE STUDY: EURONEXT<br />
"We can set up an environment for a data scientist in less than one day, compared to<br />
the 40 days it used to take, and we have moved from D+1 analytics to real-time<br />
analytics. This is fundamental to understanding markets, clients, competitors, and how<br />
they interact."<br />
that cause high volumes on the markets."<br />
Still, the transition to a Platform as a Service<br />
(PaaS) does require one key condition:<br />
remaining independent of the cloud provider.<br />
"The core of the data lake is managed by<br />
Talend. It was very important for us to keep<br />
this 'independence' compared to the layers<br />
below Talend. So, if tomorrow Euronext wants<br />
to change clouds, they can," says the CDO,<br />
happy about the greater flexibility. Euronext<br />
chose Talend Big Data to absorb real-time<br />
data in the data lake, including internal data<br />
from its own trading platform; and external<br />
data, such as from Reuters and Bloomberg.<br />
In an ultra-regulated world, Talend has also<br />
proven to be highly adept at meeting the<br />
challenges of data lake governance and<br />
regulatory compliance. Being able to safely<br />
open data involves knowing it inside out,<br />
keeping track of changes and the history of<br />
data feeds, and knowing how to classify them<br />
in a granular structure.<br />
This governance strategy is applied in very<br />
specific tools, such as the Talend Data<br />
Catalog. A dictionary is created together with<br />
each technical project for each individual<br />
market. These dictionaries are used to find the<br />
history of end-to-end data, from the sources<br />
to the reporting.<br />
"Now I can see when S3 is the data source, I<br />
can add value to the data, combine it with<br />
other data, and convert it into other data in<br />
Redshift," says Belarfaoui, who is very satisfied<br />
with the new process. "I can also add tags.<br />
Typically, we add the storage duration. For<br />
example, whether data has to be kept for ten<br />
years, or five years (per MIFID II), or if it<br />
should be archived."<br />
At the same time, data lineage with Talend<br />
drastically reduces impact analysis costs. "One<br />
simple example comes to mind: we plan to<br />
change the value of an index on the British<br />
stock market. Once we integrate it into our<br />
systems, it propagates itself pretty much<br />
everywhere. Currently we have to figure 200<br />
person-days just to find the index in our<br />
different systems. But with the dictionary, we are<br />
able to run this data lineage with just one click."<br />
MONETIZING STOCK MARKET DATA<br />
Two years after its launch, the governed lake<br />
project with Talend and AWS is a success.<br />
"The initial returns are more than positive,"<br />
says Belarfaoui. "On the technical side, we<br />
can manage ten times more iso-budget data."<br />
Beyond the improved architecture, the<br />
migration is also positioning Euronext to<br />
become a "data trader." The stock market<br />
operator wanted to be able to refine and add<br />
to its wealth of data in order to monetize it. In<br />
fact, the sale of data already brings in 20% of<br />
Euronext's revenues.<br />
"Traders actually sell, buy, and make their<br />
investment decisions in milliseconds. They<br />
have a huge appetite for aggregated data in<br />
real time. Who sells which stock, to whom, at<br />
what price and when. We are in the best<br />
position to track performance of the CAC 40<br />
or other indexes and sell that information to<br />
investors through our Datashop platform,"<br />
says Belarfaoui.<br />
In addition to clients, this project also<br />
involves giving data scientists and business<br />
units self-service access to this data, which<br />
they can analyse in data sandboxes for tasks<br />
such as market monitoring. Belarfaoui<br />
explains: "We can set up an environment for a<br />
data scientist in less than one day, compared<br />
to the 40 days it used to take, and we have<br />
moved from D+1 analytics to real-time<br />
analytics. This is fundamental to<br />
understanding markets, clients, competitors,<br />
and how they interact."<br />
This is a real turning point for Euronext. "In<br />
2016, we identified the need, but we didn't<br />
have the capacity to do it. At the time, we<br />
could only relay the volumes of market activity<br />
to market regulators (Mifid II). Today, we can<br />
dig deeper. Under the General Data Protection<br />
Regulation (GDPR), I have to know where<br />
personal data is stored. If I receive requests for<br />
modification or deletion, I can find the data,<br />
thanks to the dictionary," elaborates Belarfaoui.<br />
"Similarly, a user who searches a transaction<br />
can instantly see if it is confidential. Once data<br />
is identified as being critical, the data steward<br />
can deny user access."<br />
Euronext is just at the beginning of its digital<br />
transformation. A study is currently underway<br />
on the deployment of Talend's Master Data<br />
Management (MDM) solution. "We are<br />
working on 'golden sources' within all of our<br />
systems (CRM, trading, billing, finance,<br />
various departments, subsidiaries, etc.). The<br />
goal is to make all Euronext data even cleaner<br />
and of higher quality, such as by being sure<br />
that a client is consistently represented across<br />
all systems. Such standards will make our<br />
data even more usable," predicts Belarfaoui.<br />
More info: www.talend.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
19
RESEARCH:<br />
RESEARCH: DISASTER RECOVERY<br />
ENTERPRISES 'GRAPPLING WITH DATA LOSS'<br />
NEW RESEARCH FROM UNITRENDS SUGGESTS THAT ALTHOUGH DR TESTING IS GROWING IN<br />
POPULARITY, THERE IS STILL MUCH TO BE DONE<br />
Unitrends has shared survey<br />
findings from its fourth annual<br />
Cloud and Disaster Recovery<br />
Survey. The survey sheds light on the<br />
current state of data protection and<br />
disaster recovery (DR), as well as IT<br />
attitudes towards cloud usage and<br />
adoption. Incorporating input from over<br />
800 surveyed IT professionals across the<br />
world, the report also offers advice for<br />
implementing cloud into disaster<br />
recovery practices.<br />
need to protect more than 100 TB of<br />
data, more than double the figure<br />
compared to 2016. However, 30 percent<br />
of respondents report losing at least<br />
some of their data - a figure that has<br />
remained consistent since 2016. Data<br />
loss clearly continues to be a problem for<br />
the enterprise.<br />
The survey results show that many<br />
organisations are not following baseline<br />
best practices for data protection and<br />
DR. At the opposite end of the spectrum,<br />
leaders in DR are increasingly using the<br />
cloud to play a critical role in business<br />
continuity. Key survey findings include:<br />
Data growth continues to be<br />
exponential, however data loss continues<br />
at an unacceptably high rate. Twenty<br />
seven percent of respondents say they<br />
More organisations conduct regular DR<br />
testing and have a secondary recovery<br />
site. Compared to 2016, 46 percent<br />
more companies today say they conduct<br />
20 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
RESEARCH:<br />
RESEARCH: DISASTER RECOVERY<br />
DR testing every month. A full 75<br />
percent of IT professionals conduct DR<br />
testing at least annually (64 percent in<br />
2016). Additionally, the survey shows a<br />
16 percent drop in the number of<br />
organisations who lack a secondary<br />
recovery site to store data copies or host<br />
recovery operations, compared to 2016.<br />
However, there is a 24 percent increase<br />
in companies that use their own site or a<br />
co-location facility as their secondary<br />
DR site.<br />
Cloud continues to see acceptance, as<br />
its role in backup and data protection<br />
grows. The majority of survey<br />
respondents trust the cloud enough to<br />
use it for data protection and business<br />
continuity. Twenty two percent more<br />
companies use the cloud for backup and<br />
disaster recovery compared to 2016 -<br />
considerable growth in just two years'<br />
time. Indeed, cloud is replacing legacy<br />
media options to get backup data offsite,<br />
as more report storing backups in the<br />
cloud (36 percent) than using physical<br />
media (disk to tape, removable, tape)<br />
combined (31 percent).<br />
Cloud acceptance grows with resistance<br />
now settling around cost - not technical<br />
concerns. Among respondents who do<br />
not currently use the cloud, more say<br />
they plan to do so much sooner than<br />
compared to 2016 (when 55 percent<br />
said they had no plans). Cost is the most<br />
frequently cited reason today for nonadoption,<br />
compared to functional<br />
concerns in previous years.<br />
However mid-sized companies lag in<br />
cloud adoption, and cloud usage varies<br />
greatly by industry. Cloud adoption rates<br />
are not equal across companies of<br />
different sizes, as findings show that midsized<br />
corporate cloud adoption is 12 to<br />
18 percent lower compared to smaller<br />
and larger organisations, respectively.<br />
Not surprisingly, technology companies<br />
lead cloud adoption with 68 percent of<br />
respondents saying they use the cloud for<br />
business continuity and disaster recovery<br />
(BCDR) purposes.<br />
"Findings from this year's Unitrends<br />
Cloud and Disaster Recovery Survey<br />
unearthed some interesting trends, as we<br />
were able to analyse survey data across<br />
three consecutive years," comments Joe<br />
Noonan, vice president of product<br />
management and marketing, Unitrends.<br />
"Remarkably, 30 percent of organisations<br />
say they still experience data loss - a<br />
figure that has surprisingly remained<br />
consistent since 2016. It's clear from this<br />
data that there's still much work to be<br />
done to help organisations successfully<br />
implement and leverage backup and<br />
disaster recovery solutions and best<br />
practices. As a leader in the BCDR space,<br />
it's our responsibility to educate and<br />
empower enterprises with the knowledge<br />
and technologies they need to protect<br />
and recover their data."<br />
A full copy of the survey results is<br />
available from the website below.<br />
More info:<br />
www.unitrends.com/cloud_surveyresults_a<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
21
PRODUCT REVIEW REVIEW<br />
STORAGECRAFT ONEXAFE 4417 CONVERGED STORAGE<br />
StorageCraft's OneXafe appliances target<br />
mid-sized businesses that want to get<br />
their data backup and storage houses in<br />
order. They go way beyond the capabilities of<br />
typical legacy backup appliances by offering<br />
a converged scale-out storage and data<br />
protection platform for both physical and<br />
virtual environments.<br />
In this hands-on review, we look at the<br />
OneXafe 4417, a classy 2U appliance built<br />
around a top-quality Dell PowerEdge R740xd<br />
server platform. Powered by Xeon Scalable<br />
CPUs, it supports up to 17 internal LFF hard<br />
disks and StorageCraft allows you to bring<br />
your own drives or have it delivered<br />
preconfigured.<br />
The OneXafe appliances present their<br />
storage as SMB or NFS shares and a key<br />
feature is StorageCraft's patented distributed<br />
object-based file system, which delivers<br />
integral encryption, deduplication and<br />
compression. Vastly superior to classic RAID<br />
arrays, this technology allows up to seven<br />
OneXafe appliances to be placed in a ring, or<br />
cluster, where data blocks are automatically<br />
distributed across all members and drives for<br />
faster rebuilds and on-demand storage<br />
expansion.<br />
Management and monitoring is cloud-based<br />
and we had no problems registering our<br />
appliance with the StorageCraft OneSystem<br />
web portal. This presents an intuitive console<br />
where we could enable cluster encryption,<br />
create users and groups and provision secure<br />
access to SMB and NFS shares.<br />
This is a fairly simple use case for the<br />
OneXafe which really comes into its own<br />
when partnered with the Private OneSystem<br />
solution. Hosted on-site as a VMware VM,<br />
Private OneSystem joins seamlessly with<br />
StorageCraft's ShadowXafe backup software<br />
to deliver converged data protection and full<br />
storage management.<br />
The Private OneSystem console<br />
amalgamates everything you need for your<br />
data protection strategies into a single pane<br />
of glass. Its dashboard opens with an<br />
overview of system and host protection status<br />
along with data usage, throughput and rates<br />
of change.<br />
Storage locations are easily defined and you<br />
can declare OneXafe repositories along with<br />
SMB and NFS shares on third-party devices<br />
and iSCSI targets. A smart feature of<br />
ShadowXafe is that it can vastly improve<br />
backup speeds by communicating directly<br />
with the OneXafe repository using the gRPC<br />
(Google RPC) protocol and sending data to<br />
its object-based storage.<br />
ShadowXafe communicates with the VMware<br />
hypervisor for host-based VM backups with<br />
Hyper-V support coming soon. For physical<br />
system protection, you need to install a smallfootprint<br />
agent on each one.<br />
Policies determine backup job frequency<br />
plus data retention periods and control<br />
replication to secondary storage locations<br />
such as a remote OneXafe repository or the<br />
StorageCraft cloud. Context-based policy<br />
management allows backup requirements to<br />
be precisely matched to SLAs, RTOs and<br />
RPOs and once a policy is assigned to a<br />
virtualisation host, it will automatically protect<br />
new VMs as they created.<br />
System recovery couldn't be any easier as<br />
you select a machine or VM from the<br />
console, view and choose a recovery point<br />
and tell it to go ahead. Restoration options<br />
abound as, along with direct VM recovery,<br />
you can restore a physical machine as a VM<br />
and visa-versa.<br />
Bare-metal recovery of physical systems<br />
requires a recovery ISO agent to be created<br />
on bootable media. This connects with<br />
OneSystem, presents a list of recovery points<br />
and performs a full system restore.<br />
The VirtualBoot feature offers instant<br />
recovery in milliseconds, allowing<br />
StorageCraft to claim industry-leading RTOs<br />
and RPOs. It's extremely easy to use as it<br />
simply creates a new VM directly from the<br />
backup data in the OneXafe repository.<br />
Product: OneXafe 4417 Converged Storage<br />
Supplier: StorageCraft<br />
Web site: www.storagecraft.com<br />
Tel: +44 203 481 1240<br />
Price: Storage only from £19,913,<br />
Converged from £41,804 (all ex VAT)<br />
VERDICT: We found the appliance simple to manage and clearly capable of bringing order to data protection chaos. Businesses<br />
that can't afford any data loss downtime will find the StorageCraft OneXafe is an affordable converged protection solution offering<br />
lightning-fast recovery and easy scalability.<br />
22 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
Upgrade your storage to the next level!<br />
ORTIAL PLUS SATA III 2.5<br />
SOLID-STATE-DRIVE<br />
Increase the speed, durability and efficiency of your system with the<br />
Ortial+ solid-state-drive.<br />
Featuring the Silicon Motion SM2246EN controller and enabling read/write data<br />
transfer speeds of up to 560MB/s, the Ortial+ is available in multiple capacities<br />
upto 960GB.<br />
A solid-state-drive you can rely on!<br />
Ortial’s globally supported SSD’s are 100% compatible for use with all the leading<br />
brands of laptops and desktops with each upgrade designed and produced to achieve<br />
the exact specification of the OEM to ensure compatibility.<br />
HIGH<br />
PERFORMANCE<br />
TRIM<br />
GC<br />
IMPACT<br />
RESISTANT<br />
OPTIMAL<br />
DATA FLOW<br />
GUARANTEED<br />
COMPATIBILITY<br />
10 YEAR<br />
WARRANTY<br />
Buy now at:<br />
+44 (0) 1423 70 4700<br />
ortial.com
TECHNOLOGY FOCUS: VDI FOCUS: VDI<br />
ELIMINATING THE VDI PRICE/PERFORMANCE<br />
BOTTLENECK<br />
DEPLOYING A VDI SOLUTION CAN BE DEMANDING AND EXPENSIVE, WITH STORAGE IN<br />
PARTICULAR A MAJOR POTENTIAL HURDLE, ARGUE JUAN MULFORD AND STEFAN FERRARI OF<br />
ACCELSTOR. INDUSTRY-VALIDATED, COST-EFFECTIVE, ALL-FLASH STORAGE HOWEVER, CAN ENABLE<br />
BUSINESSES ACROSS ALL VERTICALS TO ADOPT VDI<br />
Virtual Desktop Infrastructure (VDI) allows<br />
IT organisations to simplify and automate<br />
the management of thousands of users,<br />
securely delivering desktop-as-a-service from a<br />
central, secure location. Adoption of VDI is<br />
highest in industries where there are both high<br />
levels of regulation and a distributed workforce<br />
or many remote offices (healthcare, banking,<br />
education, retail etc.).<br />
When effectively deployed, VDI delivers high<br />
returns on investment; reducing operating costs<br />
and closing security gaps through infrastructure<br />
consolidation, while also adding flexibility to the<br />
workforce which is now empowered to work<br />
from anywhere.<br />
THE VDI TRAP<br />
Because VDI touches so many users, it directly<br />
affects the productivity and bottom line of the<br />
organisation deploying it. This high level of<br />
visibility puts immense pressure on IT to deliver<br />
performance to ensure a positive user<br />
experience, and this added pressure is the<br />
single biggest roadblock for VDI adoption.<br />
Storage is widely regarded as the biggest<br />
hurdle to overcome from both a cost and<br />
performance perspective.<br />
In order to deliver a user experience on-par<br />
with users having their own desktop, the VDI<br />
infrastructure stack must service bursts of very<br />
high read/write I/O while maintaining ultra-low<br />
latency for all users at once. Failure to do so<br />
leads to the death blow of many VDI projects -<br />
users asking for their desktops back.<br />
According to IDC, most failed VDI<br />
deployments are blamed on storage, as<br />
bursts of high I/O request impact user latency<br />
across the VDI platform. It is unsurprising<br />
then that storage is often the most expensive<br />
and carefully selected hardware in successful<br />
VDI deployments.<br />
Traditional spindle (spinning disk) based and<br />
even hybrid (spindle mixed with flash) storage<br />
have failed to meet the demands of VDI<br />
because these systems depend on a read/write<br />
cache to service low-latency I/O requests -<br />
however once the cache is overwhelmed,<br />
latency increases, the users complain, and the<br />
project dies.<br />
THE ACCELSTOR DIFFERENCE<br />
The storage demands of VDI are so high that<br />
even all-flash storage systems that are<br />
dependent on NVMe or NVRAM caching for<br />
low-latency I/O can fail to deliver a user<br />
experience that leads to a successful VDI<br />
implementation.<br />
With AccelStor there is no cache. Only<br />
metadata is temporarily stored in global<br />
memory, so all write blocks are direct to disk -<br />
meaning all reads come directly from disk.<br />
This cache-less design means all I/O goes<br />
straight to/from the SSD media at near-cache<br />
speeds, significantly increasing performance,<br />
lowering latency, and providing better data<br />
resiliency than traditional cache-based systems.<br />
FLASH-OPTIMISED<br />
RAID was designed in the late 1980s to<br />
provide resiliency and increase performance for<br />
spinning HDD based storage systems. It was<br />
widely successful at the time, so successful in<br />
fact that it is now still the foundation used by<br />
many so-called "state-of-the-art" all-flash<br />
storage systems. These 'RAID optimised for<br />
24 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
TECHNOLOGY FOCUS: TECHNOLOGY FOCUS: VDI<br />
flash' based systems do two things that are<br />
actually far from optimised for flash:<br />
Write data to the SSDs randomly<br />
Write data more than once<br />
The AccelStor approach is very different. Born<br />
from an in-depth research project into the<br />
adverse effects of RAID on SSD media,<br />
AccelStor took a step back and uncovered a<br />
better way to use SSDs effectively.<br />
At the heart of every AccelStor system is a<br />
cache-less, RAID-less technology called<br />
FlexiRemap. The culmination of 10 years of<br />
research, 2 PhDs, and over 45 patents and<br />
counting; FlexiRemap harnesses the true power<br />
of NAND memory with gamechanging write<br />
I/O performance. With major funding from<br />
tech giants like Toshiba Memory and wide<br />
recognition like best-in-show at FMS 2016, this<br />
technology really is revolutionary.<br />
WHAT IS FLEXIREMAP?<br />
FlexiRemap got its name because it remaps<br />
incoming write I/Os into a single sequential<br />
write across every SSD in the system. To do<br />
this, it breaks the data down into 4K pages,<br />
which is, mathematically speaking, what fits<br />
best in most SSDs.<br />
Because the writes are sequential and<br />
writing in a way that is designed for NAND<br />
memory, they now occur much faster than in<br />
a RAID-based system (700,000 sustained<br />
write IOPS in a single appliance); there is no<br />
need to stage data to cache first, enabling a<br />
cacheless design.<br />
Additionally, because the writes are spread<br />
across every SSD in the system, read I/O gets<br />
serviced faster as well (1.1M 4K mixed<br />
read/write IOPS).<br />
Many RAID based appliances will use<br />
expensive SAS SSDs to support a shared<br />
storage architecture with dual controllers,<br />
however RAID is unable to utilise the potential<br />
of SAS SSDs and will on average only achieve<br />
10% of the potential IOPS from each SSD.<br />
FlexiRemap combined with a 'shared nothing'<br />
cluster architecture is able to out-perform<br />
these systems using SATA SSDs utilising 90%<br />
of their potential. This means a lower TCO<br />
and higher ROI. Essentially, FlexiRemap<br />
means you get what you pay for.<br />
Finally, unlike spinning hard drives, SSDs fail<br />
at a predictable rate based on the number of<br />
writes. So, unlike RAID-based systems,<br />
FlexiRemap technology writes data once and<br />
only once, increasing the longevity of each SSD<br />
by over 300%, keeping operating and support<br />
costs low as well.<br />
FlexiRemap technology therefore unlocks the<br />
true potential of flash, allowing AccelStor<br />
customers to do more with less across all types<br />
of high I/O, low-latency applications like VDI.<br />
A SIMPLE SOLUTION<br />
AccelStor NeoSapphire AFAs are preconfigured<br />
and simple to manage. The web-based GUI<br />
allows users to create volumes easily and<br />
present them to hosts. There is no cache to<br />
manage, the system just works.<br />
The NeoSapphire provides full integration as<br />
well with VMware environments, with support<br />
for VAAI block, SRM, Horizon View, and the<br />
VMware vSphere plugin integration tools<br />
people have come to expect are all standard.<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
25
TECHNOLOGY FOCUS: VDI FOCUS: VDI<br />
BETTER ROI FOR STORAGE<br />
AccelStor runs a "zero licence" operation,<br />
meaning that all features such as Clone,<br />
Snapshot, Replication, Deduplication and more<br />
are included out of the box, free of charge. As<br />
well as these features, AccelStor offers helpdesk<br />
support for the lifetime of the system, at no<br />
additional cost, minimising TCO and ensuring<br />
ease of budgetary administration.<br />
performance capabilities of VDI's heaviest and<br />
most resource-intensive workloads were tested.<br />
The results showed that no matter the<br />
workload, the user experience, or desktop<br />
application response-time remained<br />
unaffected, with average latencies well below<br />
the coveted 0.6 ms across the board (see<br />
chart below).<br />
The AccelStor NeoSapphire AFAs come<br />
equipped with FlexiDedupe, a deduplication<br />
algorithm that provides up to 10:1 data<br />
reduction with a minimal impact to<br />
performance in VDI environments. With the use<br />
of linked clones and FlexiDedupe, storage<br />
capacity efficiency has been shown to be as<br />
high as 98% (see chart).<br />
At full desktop density on a NeoSapphire<br />
AFA, AccelStor's FlexiRemap technology<br />
combined with thin provisioning and<br />
deduplication, reduce the storage cost as low<br />
as US$30 per power-desktop across the entire<br />
NeoSapphire line.<br />
TESTED AND VALIDATED<br />
The NeoSapphire product range has been<br />
tested and validated with the industry leading<br />
tool from Login VSI, the industry standard loadtesting<br />
tool for performance and scalability of<br />
VDI environments.<br />
In a test using 500 linked-clones on VMware<br />
vSphere and Horizon View, the storage<br />
CONCLUSION<br />
When storage performance really matters, and<br />
all-flash storage is a consideration for VDI, a<br />
system built on FlexiRemap technology will<br />
deliver the best performance for the price, bar<br />
none. With the ability to use enterprise SATA<br />
SSDs that last 3X longer and provide<br />
consistently low latency, these are systems<br />
tuned and tested to serve the needs of VDI.<br />
Please visit the website below for a free<br />
consultation and try a no obligation POC<br />
system to see how AccelStor's FlexiRemap<br />
powered NeoSapphire products can eliminate<br />
your VDI challenges today.<br />
More info: www.accelstor.com<br />
^<br />
26 STORAGE Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
CASE STUDY: CASE STUDY: DAIKIN<br />
A BREATH OF FRESH AIR<br />
AIR CONDITIONING SPECIALIST DAIKIN INDUSTRIES HAS ADOPTED A NEW STORAGE APPROACH FOR<br />
ITS MISSION CRITICAL BUSINESS APPLICATIONS<br />
as four SSDs per controller can be used,<br />
including SATA disks. In addition to faster read<br />
speeds, SSD cache reduces performance<br />
impact compared to spin-drive caches with<br />
rapid warm up and less drive wear.<br />
Configuring and managing SSD Cache<br />
pooling is easy via the intuitive SANWatch UI.<br />
Automated storage tiering optimises storage<br />
performance and increases ROI by leveraging<br />
high speed SSD for expedited access to hot<br />
data while using affordable drives such as NL-<br />
SAS for archiving purposes. To protect SSDs<br />
and the vital data they store, EonStor DS<br />
3000 systems offer SSD wear level indication<br />
that provides users with accurate real time<br />
SSD health info.<br />
Daikin Industries Ltd., a global leading<br />
air conditioning manufacturer with<br />
headquarters in Osaka, Japan, has<br />
adopted the Infortrend EonStor DS storage<br />
family in order to implement the entire suite of<br />
mission-critical business applications such as<br />
ERP, database, VDI, virtualisation, and backup<br />
in their Vietnamese operation.<br />
Daikin is the developer of numerous<br />
refrigerant volume systems and an innovator<br />
in split system air conditioners. It has<br />
operations in Japan, China, Australia, India,<br />
countries in Southeast Asia such as the<br />
Philippines and Vietnam, Europe, North<br />
America, and South America.<br />
Due to continuous business expansion,<br />
Daikin's Vietnam operations started to<br />
experience both impacted performance of<br />
their business applications and insufficient<br />
capacity in their storage system. After a<br />
rigorous selection process, Daikin Vietnam<br />
decided to replace its previous system with<br />
multiple high-performance EonStor DS 3024<br />
units. These storage devices feature highspeed<br />
Fibre Channel (FC) interfaces and<br />
Hybrid Drive technology to combine<br />
SSD/HDD, providing better performance and<br />
high capacity utilisation.<br />
The new setup has three subsystems. The first<br />
subsystem powers up Microsoft Hyper-V<br />
technology for server and desktop<br />
virtualisation, with more than 200 VMs (virtual<br />
machines). The second subsystem runs the<br />
ERP database system that uses Microsoft<br />
Dynamic AX 2012. The third subsystem<br />
supports backup of the company's missioncritical<br />
business application data.<br />
EonStor DS 3000 systems are designed for<br />
all-SSD storage and support large high speed<br />
SSD cache pools. SSD Cache speeds up read<br />
performance for priority data, and boosts<br />
cache pool capacities up to 6.4TB. As many<br />
"EonStor DS 3024 offers us a great pile<br />
of network storage at a very affordable<br />
price. It also supports a choice selection of<br />
high-performance data ports," said Mr.<br />
Nguyen Nhat Linh, system administrator at<br />
Daikin Vietnam.<br />
All EonStor DS 3000 systems are excellent<br />
platforms for database and VDI applications<br />
thanks to ample computing power, and their<br />
massive bandwidth serves to enhance media<br />
editing and surveying jobs, freeing users from<br />
bottlenecks. When looking to expand storage,<br />
up to an extra 444 drives can be connected<br />
easily using JB 3000 enclosures.<br />
"EonStor DS storage family is the industry<br />
price-performance leader with top rankings<br />
in both SPC-1 and SPC-2. It delivers<br />
superior IOPs, great capacity expansion, and<br />
reliable performance to power missioncritical<br />
business applications," commented<br />
Thomas Kao, Senior Director of Product<br />
Planning at Infortrend.<br />
More info: www.infortrend.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
27
STRATEGY:<br />
STRATEGY: SURVEILLANCE DATA<br />
COURT ON CAMERA<br />
SAFEGUARD YOUR SURVEILLANCE DATA OR POTENTIALLY FALL<br />
FOUL OF THE LAW, WARNS GARY WATSON, CTO OF<br />
STORCENTRIC AND FOUNDER OF NEXSAN<br />
Video surveillance has advanced<br />
rapidly since CCTV was invented in<br />
1942. In particular, the move to<br />
digital technology means its quality is<br />
now good enough to be routinely used by<br />
law enforcement agencies across the<br />
world, whether it's in court as evidence or<br />
to monitor and track individuals.<br />
Despite concerns about privacy, CCTV<br />
plays a fundamental role in daily life. In<br />
the UK alone, there are between 4 million<br />
and 5.9 million CCTV surveillance<br />
cameras protecting the public, their<br />
"Beyond the greater storage capacity that video<br />
surveillance footage demands, police and other law<br />
enforcement agencies are equally challenged by the<br />
manner in which electronic evidence must be stored. If<br />
the data collected is to be used to support legal cases, it<br />
is essential that it follows the legal, compliance and<br />
regulatory legislation - otherwise critical evidence could<br />
be deemed inadmissible."<br />
property and helping to prevent crime<br />
and bring criminals to justice. Whether it's<br />
recording the most serious crimes and<br />
terrorist incidents as they happen or the<br />
many examples of CCTV being used to<br />
help find missing people, it has become<br />
part of the fabric of modern society.<br />
However, this ongoing growth in the use<br />
of video surveillance brings with it new<br />
challenges for data storage, and as the<br />
role of CCTV grows ever more important,<br />
it is critical that it is managed and<br />
protected in the correct way.<br />
In particular, it's vital to consider data<br />
storage and security, ensuring it abides by<br />
the General Data Protection Regulation<br />
(GDPR) and The Regulation of<br />
Investigatory Power Act. For digital<br />
evidence, it's important to be able to offer<br />
seamless scalability to accommodate<br />
rapid growth, along with a<br />
comprehensive suite of data security<br />
features that meet strict requirements for<br />
file integrity, privacy, chain of custody,<br />
and compliance.<br />
CAPACITY PLANNING<br />
In terms of capacity, video and digital<br />
images are exceptionally heavy users of<br />
28 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
STRATEGY:<br />
STRATEGY: SURVEILLANCE DATA<br />
deleted or altered in any way. A chain of<br />
custody involves the chronological<br />
documentation of the custody, analysis,<br />
and configuration of physical or electronic<br />
evidence, to help prove in court that any<br />
evidence being used is correct and valid.<br />
storage space and the guidelines around<br />
managing evidentiary data are key to it<br />
being recognised as a valid source by the<br />
criminal justice system. Not only are more<br />
videos being created due to the everincreasing<br />
number of surveillance<br />
cameras, but also the resolution and<br />
specifications are constantly increasing in<br />
quality and therefore creating larger files.<br />
Many organisations require higherresolution<br />
video to ensure compliance<br />
with insurance providers, maximise loss<br />
recovery and strengthen criminal cases in<br />
court, for example. It is therefore essential<br />
that data is stored in a solution that is<br />
scalable and cost-effective, allowing<br />
customers to add capacity as and when<br />
needed, as running out of storage space<br />
is not an option.<br />
RETENTION RULES<br />
In addition to the daunting amount of<br />
storage that evidentiary videos and<br />
images consume, police departments and<br />
other law enforcement agencies must also<br />
address the retention requirements that<br />
govern how long video surveillance and<br />
images must be kept.<br />
Storage capacity needs (and costs)<br />
obviously increase the longer surveillance<br />
videos are retained; ideally, video should<br />
be stored no longer than required by law.<br />
Storage solutions that incorporate policybased<br />
rules for data retention can help<br />
police departments to, for example,<br />
significantly boost their storage efficiency<br />
by automatically keeping all surveillance<br />
videos only for their legally mandated<br />
retention period.<br />
Beyond the greater storage capacity that<br />
video surveillance footage demands,<br />
police and other law enforcement<br />
agencies are equally challenged by the<br />
manner in which electronic evidence must<br />
be stored. If the data collected is to be<br />
used to support legal cases, it is essential<br />
that it follows the legal, compliance and<br />
regulatory legislation - otherwise critical<br />
evidence could be deemed inadmissible.<br />
SECURITY AND CHAIN OF<br />
CUSTODY<br />
Security and chain of custody are perhaps<br />
two of the most important aspects to<br />
consider when it comes to ensuring digital<br />
video data is permissible in court. A<br />
digital chain of custody, including an<br />
audit trail, proves that any digital<br />
evidence is correct and has not been<br />
In the UK, digital evidence comes under<br />
the same laws as any other form of<br />
evidence, but new guidelines were<br />
published in 2012 in the ACPO Good<br />
Practice Guide for Digital Evidence to<br />
advise examiners on the authentication<br />
and integrity of evidence. Four principles<br />
were created to help decide whether<br />
digital evidence can be used in court,<br />
including:<br />
No action should change data that is<br />
relied on in court<br />
Anyone accessing the original data<br />
should be competent and provide<br />
evidence of the relevance of this<br />
An audit trail should be preserved and<br />
able to be checked by a third party<br />
The person in charge of the<br />
investigation must ensure the law and<br />
these principles are upheld<br />
Part of this process therefore relies on<br />
the security of the surveillance data - as<br />
no actions should cause the data to be<br />
changed, it is vital that it is stored<br />
securely enough to prevent anyone<br />
without authority having access to it at<br />
any point, thereby guaranteeing the<br />
integrity of the files. With adequate<br />
security measures in place within the<br />
storage solution, CCTV data is far more<br />
likely to be accepted in a court of law as<br />
evidence.<br />
By implementing high quality storage<br />
solutions, the owner of the surveillance<br />
data - whether that's local government or<br />
a business with its own CCTV - can be<br />
sure that it can be utilised as evidence to<br />
help win the case in question. And when it<br />
comes to catching criminals, it's better to<br />
be safe than sorry.<br />
More info: www.nexsan.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
29
TECHNOLOGY FOCUS: HDD FOCUS:<br />
HAMR: UNLOCKING THE NEXT STAGE OF<br />
DATA GROWTH<br />
JASON FEIST, MANAGING TECHNOLOGIST AT SEAGATE TECHNOLOGY, DESCRIBES HOW A LASER<br />
NO BIGGER THAN A GRAIN OF SALT IS JUST ONE CRUCIAL ELEMENT OF THE NEXT GENERATION<br />
OF HARD DRIVE TECHNOLOGIES<br />
One thing is for sure: the<br />
conventional way we store data<br />
is reaching its limit. For the past<br />
few decades we've grown accustomed to<br />
hard drive capacity increasing as surely as<br />
the passing of the seasons. But this will<br />
soon change.<br />
The world is producing more and more<br />
data. With the development of<br />
technologies like machine learning, 5G<br />
and AI this statement might not come as a<br />
surprise, but the sheer scale of the<br />
challenge ahead is worth remembering. By<br />
2025 the 'global datasphere' is expected to<br />
reach an astounding 175ZB, according to<br />
research from IDC sponsored by Seagate.<br />
Nearly half (49%) of that data will reside in<br />
the public cloud, further highlighting the<br />
urgent need for new technology to help<br />
deal with this massive increase in the<br />
production and storage of data at scale.<br />
Seagate is a company with a 40-year<br />
heritage in data storage and management.<br />
We understand these global industry<br />
trends, which is why we have invested our<br />
considerable research & development<br />
resources behind a technology called heatassisted<br />
magnetic recording, or HAMR for<br />
short. With HAMR, hard drive capacities of<br />
20TB and more will become commercially<br />
available from 2020, and the technology<br />
has the potential to make capacities of<br />
100TB or more a possibility within a few<br />
decades. The first HAMR drives are already<br />
out with Seagate customers for final testing<br />
and have been receiving excellent<br />
feedback. Here's the story of why we need<br />
a rethink of how we store data, and how<br />
we're charting a way forward with HAMR<br />
technology.<br />
A PROBLEM OF GRAINS<br />
While easily integrated and totally<br />
transparent to the end-user, HAMR is a<br />
highly complex technology sorely needed<br />
in the data storage industry in order to<br />
maximise capacity offerings. In order to<br />
appreciate why HAMR is so important, we<br />
first need to understand the current state<br />
of data storage technology and why we<br />
need innovative technologies in order to<br />
be sustainable. The vast majority of hard<br />
drives in the world today use technology<br />
called conventional magnetic recording<br />
(CMR). These hard drives are made up of<br />
disk-shaped platters, covered in a thin<br />
film of recording medium. This recording<br />
medium is made up of tightly-packed<br />
grains, which are polarized by magnetic<br />
fields produced by a write head in such a<br />
way that they can later be read as either<br />
'0' or '1'.<br />
Around a dozen of these grains make up<br />
a single bit of data, and the key to<br />
boosting the capacity of our hard drives<br />
has been developing new ways to make the<br />
grains smaller and pack as many as<br />
possible in a given space. This strategy has<br />
served the industry well for a number of<br />
years, but we're reaching the end of its<br />
usefulness. At a certain point, the grains on<br />
a storage medium are so small that they<br />
become unstable, potentially flipping<br />
between 0 and 1 when they're not<br />
supposed to. This can lead to data<br />
30 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
TECHNOLOGY TECHNOLOGY FOCUS: HDD<br />
"The theory behind HAMR technology is deceptively simple. In order to record<br />
data to the medium, it must first be heated up, to over 400°C. A tiny portion of<br />
the recording medium is brought to this temperature for a very short time -<br />
around one nanosecond - which allows the data to be recorded reliably and then<br />
'frozen' in place once the temperature has decreased."<br />
corruption and is clearly not acceptable;<br />
we need another solution if we are going<br />
to maintain a steady increase in hard drive<br />
capacity and keep up with the rapid<br />
growth in data storage needs projected<br />
over the coming years.<br />
INTRODUCING HAMR<br />
HAMR is a technology developed by<br />
Seagate that enables a significant increase<br />
in the amount of data that can be stored<br />
in a disk. Put simply, HAMR allows us to<br />
continue to shrink the size of the grains<br />
which makes it possible to increase areal<br />
density more than in previous types of<br />
hard drive.<br />
At the outset we knew that in order to<br />
reduce the grain size , we would need to<br />
experiment with the materials used to<br />
create the recording media. We tested a<br />
few different options, primarily highanisotropy<br />
(or "hard") magnetic materials<br />
such as iron-platinum alloys (FePt). The<br />
challenge with these materials is that, while<br />
they allow more bits to be recorded within<br />
a smaller space, these bits cannot be<br />
flipped under the normal temperatures that<br />
conventional hard disks operate at. In<br />
order to write data to drives with FePt<br />
recording media, the media needs to be<br />
hotter at the time of recording.<br />
PUTTING THEORY INTO PRACTICE<br />
The theory behind HAMR technology is<br />
deceptively simple. In order to record<br />
data to the medium, it must first be<br />
heated up, to over 400°C. A tiny portion<br />
of the recording medium is brought to this<br />
temperature for a very short time - around<br />
one nanosecond - which allows the data<br />
to be recorded reliably and then 'frozen' in<br />
place once the temperature has<br />
decreased.<br />
Seagate has tested approximately 40,000<br />
HAMR drives so far, and the technology<br />
will be commercially ready very soon.<br />
Crucially, the new technology in HAMR<br />
drives has no detrimental impact on<br />
overall temperature of the hard drive or<br />
wider system, or the overall durability of<br />
the drive. For example, one of our drives<br />
has clocked up more than 4000 TB and is<br />
still going strong after an incredible 16<br />
months of continuous HAMR writing.<br />
Reducing the grain size in our hard drives<br />
in this way required us to completely<br />
rethink how we build our products. The<br />
structure of existing hard drives using<br />
conventional perpendicular magnetic<br />
recording heads designed to operate at<br />
room temperature would simply not work.<br />
HAMR is an amazing combination of<br />
technologies incubated by Seagate. We<br />
have designed an entirely new recording<br />
head which incorporates a tiny laser, no<br />
bigger than a grain of salt, which is<br />
capable of heating the recording media to<br />
the right temperature. We needed to<br />
introduce a host of other new components<br />
to HAMR drives to get them operating<br />
effectively, such as optical waveguides and<br />
near-field transducers, as well as tackling<br />
the thousands of small issues that<br />
engineers encounter whenever developing<br />
a new product.<br />
THE FUTURE OF STORAGE<br />
Seagate is committed to making the most<br />
out of emerging technologies and their<br />
potential for improving customers' ability to<br />
store and use their data effectively, and<br />
HAMR technology is just the latest example<br />
of this. Back in 2014 we introduced<br />
shingled magnetic recording, which<br />
increased areal density by 10%, and the<br />
following year we developed heliumsealed<br />
drive technology, enabling drives to<br />
run with more disks while using less power.<br />
With today's HAMR technology we're<br />
currently on track to deliver five terabits<br />
per square inch, an unimaginable<br />
milestone just a few years ago. Keeping up<br />
with the constantly growing amount of<br />
data generated by the public and by<br />
businesses around the world is a challenge<br />
we work to solve every day.<br />
HAMR is going to deliver tangible storage<br />
improvements years into the future.<br />
According to our estimates, HAMR<br />
technology will enable us to achieve a<br />
30% compound annual growth rate<br />
(CAGR) in disk drive capacity over the next<br />
10 years. HAMR drives are just the<br />
beginning: the world's storage needs are<br />
increasing rapidly, and HAMR is the latest<br />
step in a long heritage of innovative<br />
systems and ideas to store the world's data.<br />
More info: www.seagate.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
31
CASE STUDY:<br />
CASE STUDY: MERCEDES-AMG PETRONAS<br />
DRIVING AUTOMATION<br />
THE MERCEDES-AMG PETRONAS MOTORSPORT TEAM HAS DEPLOYED RUBRIK TECHNOLOGIES TO<br />
HELP THEM BETTER LEVERAGE THE RACE DATA WHICH PLAYS A VITAL ROLE IN ACHIEVING THE<br />
INCREMENTAL GAINS THAT HELP THEM WIN TROPHIES<br />
Mercedes-AMG Petronas Motorsport,<br />
headquartered in Brackley, UK, is<br />
one of the most successful teams in<br />
the history of Formula One. The team<br />
includes five-time FIA Formula One Drivers'<br />
World Champion, Lewis Hamilton, widely<br />
regarded as one of the greatest drivers in the<br />
history of the sport. The team's 16 wins in both<br />
2014 and 2015 broke a new record for most<br />
wins in a single season, before bettering that<br />
in 2016 with 19 victories. The following year,<br />
in 2017, the team became the first to defend<br />
both Championships across a major<br />
regulation change.<br />
Matt Harris, Head of IT, manages a 30-<br />
person IT team that supports over 800 heavy<br />
technology users: "Our team is focused on<br />
winning both the Drivers' and Constructors'<br />
Championships and the right IT environment is<br />
essential to that pursuit. The difference between<br />
winning and losing in this sport comes down to<br />
fractions of a second. As a result, technology is<br />
weaved into every facet of our team's<br />
performance, from design to trackside."<br />
"Fully understanding and leveraging data and<br />
IT solutions is now so important to win races,"<br />
added Harris. "The team is constantly<br />
developing our cars in the quest for faster lap<br />
times, and analysis of race data plays a vital<br />
role in achieving the incremental gains that<br />
have put us at the top of the sport. We<br />
generate over 500 GB of data every race<br />
weekend from one car alone, so there is a<br />
massive amount of valuable information that<br />
needs to be protected."<br />
"The main currency in our organisation is<br />
time," said Harris. "Just as time is critical to<br />
the performance of our race cars, it is also<br />
vital to our team's ability to deliver the<br />
technical infrastructure that holds our<br />
organisation together. We are constantly<br />
looking for ways to improve our IT team's<br />
performance just as we strive to achieve faster<br />
lap times. As a result, we adopt technologies<br />
that allow us to save time or simplify our<br />
32 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
CASE STUDY:<br />
CASE STUDY: MERCEDES-AMG PETRONAS<br />
"Every year, the data that we produce in Formula One increases significantly.<br />
With Rubrik, we're confident we can not only protect that data, but scale to<br />
meet that growth. Rubrik is so much more than a business continuity solution -<br />
it's the future of data management."<br />
processes - and that's where Rubrik comes in."<br />
LEGACY HEADACHES<br />
"Our tape-centric legacy data management<br />
system had become too complex and<br />
unreliable to adequately protect our critical<br />
data," said Chris Green, IT Operations<br />
Manager. "We had a full-time employee<br />
dedicated to managing backups, yet we were<br />
missing 50% of our data protection goals.<br />
There were times where weeks would go by<br />
without some systems being successfully<br />
protected. We spent a year trying to fix the<br />
issues with no success."<br />
"We had no confidence in our previous<br />
solution and actually experienced data loss<br />
from the complexities of tape management,"<br />
said Green. "We had a data tape<br />
inadvertently returned to the scratch pool,<br />
making it impossible for us to recover key<br />
data several years ago. The team was<br />
sceptical of our legacy solution's ability to<br />
adequately restore our data, and we needed<br />
to assign another IT person to assist our<br />
backup expert with the restore. This is the<br />
kind of resource expenditure that we just<br />
could not continue with."<br />
"Backup is not the most glamorous part of IT,<br />
but when it goes wrong, it can be very tricky.<br />
Using the right solution is absolutely critical,"<br />
added Harris.<br />
UNRIVALLED SIMPLICITY<br />
"Some of the best decisions we've made as a<br />
team are when we prioritise simplicity," said<br />
Green. "We needed a data management<br />
solution that would lessen our team's<br />
workload while fully protecting the data that is<br />
crucial to our success. Rubrik offered a truly<br />
simple and elegant solution for data<br />
management. We also loved their API-first<br />
architecture, which would allow us to further<br />
drive value for the business with workflow<br />
automation. Our final decision was influenced<br />
by Rubrik's forward-looking data management<br />
philosophy, which will enable us to move<br />
faster as our needs grow and evolve."<br />
EMPOWERING BUSINESS<br />
"We are currently utilising Rubrik's role-based<br />
access control (RBAC) and multi-tenancy<br />
functionality," said Green. "One of the most<br />
appealing aspects of Rubrik is its ability to be<br />
rolled out to other teams. For example, we<br />
are in the process of upgrading our CAD<br />
platform and have allowed a number of key<br />
users to interact with Rubrik directly. Without<br />
having to open tickets with the IT team, they<br />
can do on-demand backups and self-service<br />
restores. This has led to cross-departmental<br />
gains in efficiency."<br />
"Traditionally, only the IT team would have<br />
the expertise necessary to perform data<br />
management functions for our engineering<br />
departments, but with Rubrik, the teams have<br />
direct control. Because Rubrik essentially<br />
manages itself, our team has more time to<br />
devote to strategic projects and finding ways<br />
to assist other departments," Green added.<br />
API-FIRST ARCHITECTURE<br />
"At the beginning of a race, we sign off new<br />
software code in a development lab and then<br />
ship that out to the race track," said Green.<br />
"Previously, we used virtual machine<br />
snapshots to seal the environment to a<br />
recovery point which involved shipping tapes<br />
and performing lengthy recoveries. We now<br />
use Rubrik's Live Mount to get up and running<br />
quickly and to ship new code faster. We are<br />
also looking into using Rubrik's REST APIs to<br />
create a lab orchestration package that<br />
speaks directly to Rubrik to further automate<br />
pre-race preparation."<br />
"The ability to use Rubrik's APIs with our<br />
software developers also eliminates the need<br />
for training," said Green. "They also no longer<br />
need to reach out to the IT team to do a<br />
backup. We supply them with a set of<br />
commands built with the APIs and all the<br />
complexity is hidden behind the scenes,<br />
making everyone's lives easier."<br />
BETTER VALUE BACKUP<br />
Mercedes-AMG Petronas Motorsport is using<br />
Rubrik to protect its 97% virtualised<br />
environment, including MS SQL, SharePoint,<br />
Exchange, Oracle, and NAS. Benefits include:<br />
90% management time savings<br />
Near-zero RTOs<br />
Significantly increased backup performance<br />
Simplified restore process<br />
Automated policy engine<br />
Ease of use<br />
Fast restores with Google-like search<br />
High performance with Pure Storage and<br />
Rubrik<br />
Data archival to private cloud<br />
"Every year, the data that we produce in<br />
Formula One increases significantly,"<br />
concludes Matt Harris, head of IT. "With<br />
Rubrik, we're confident we can not only<br />
protect that data, but scale to meet that<br />
growth. Rubrik is so much more than a<br />
business continuity solution-it's the future of<br />
data management."<br />
More info: www.rubrik.com<br />
www.storagemagazine.co.uk<br />
@STMagAndAwards<br />
Jan/Feb 2019<br />
STORAGE<br />
MAGAZINE<br />
33
STRATEGY:<br />
STRATEGY: DIGITAL TRANSFORMATION<br />
ARE YOU READY FOR INDUSTRY 4.0?<br />
MARK JOW, EMEA VICE PRESIDENT, TECHNICAL SERVICES AT COMMVAULT TALKS ABOUT<br />
PREPARING FOR AN EFFECTIVE DIGITAL TRANSFORMATION JOURNEY<br />
Are you suffering digital transformation<br />
fatigue? Believe me you are not alone.<br />
'Digital transformation' has been<br />
adopted as a general buzzword and is now<br />
used indiscriminately by vendors whose<br />
offerings have often only had the loosest<br />
connection to the pure meaning of the term.<br />
This is a shame because behind the forced<br />
associations and antipathy lies a key point that<br />
we should all be paying close attention to. As<br />
The Economist put it back in 2014, in the<br />
report The Third Great Wave: "This wave, like<br />
its predecessors, is likely to bring vast<br />
improvements in living standards and human<br />
welfare, but history suggests that society's<br />
adjustment to it will be slow and painful."<br />
Comparing the shift to digital business to the<br />
industrial revolution might sound like an<br />
overstatement, but the comparison is defensible<br />
- especially from the perspective that both were<br />
primarily economic phenomena, driven by<br />
technology innovation. Economic and societal<br />
drivers remain the same today, but rewritten<br />
with new technologies and the shifting<br />
expectations of 21st century societies.<br />
International analyst house Forrester talk<br />
about the 'shifting expectation' imperative in<br />
The Digital Business Imperative (2017): "Digital<br />
fundamentally changes your relationship with<br />
your customers. To remain competitive, you<br />
must re-engineer how your business creates<br />
value for your customers in the digital age."<br />
Re-engineering value is a good way of<br />
thinking about digital transformation because<br />
that takes us beyond the boundaries of<br />
technology to a place where the people,<br />
process, and cultural aspects of digital<br />
business are required to provide the 'big<br />
picture' business outlook. The challenge for<br />
many more established organisations<br />
operating today, is that they are competing<br />
against digital-native companies who have<br />
already reimagined and re-engineered how<br />
they create value for customers.<br />
As of May this year, the five most valuable<br />
Fortune 500 companies in the world were<br />
Apple, Alphabet, Amazon, Facebook and<br />
Microsoft. These companies have driven the<br />
development of the digital business world of<br />
today, and competing with these companies is<br />
tremendously difficult. The common element to<br />
all of these companies is the digital 'DNA' at the<br />
very core of their businesses and because of<br />
that gain a fundamental market advantage<br />
now and going forward.<br />
Recognition of the requirements and benefits<br />
of digital transformation for any business is<br />
clear. The path to achieving this transition<br />
successfully is much less obvious.<br />
Only recently IBM was quoted as saying that<br />
digital transformation projects on average take<br />
four years to complete, during which 85% of<br />
these undertakings actually fail. With these<br />
figures front of mind, it's crucial that<br />
organisations of all shapes and sizes should<br />
look at not just the technical questions about<br />
how they transform their operations, cultures<br />
and processes, but why they are undertaking<br />
these projects, and what success truly looks like<br />
at the end of the journey.<br />
As we start a new year, for many it will be<br />
the time to take stock of where they have<br />
reached along the path to digital<br />
transformation and consider what progress<br />
needs to made in 2019.<br />
More info: www.commvault.com<br />
34 STORAGE<br />
Jan/Feb 2019<br />
@STMagAndAwards<br />
www.storagemagazine.co.uk<br />
MAGAZINE
Join the Stress-Free Data<br />
Protection Revolution<br />
Introducing<br />
Where your data is always safe, always accessible, always optimized<br />
No need to lose sleep over storing and protecting data. StorageCraft makes it easy with its newly<br />
launched OneXafe converged data platform. It provides a unified solution for data protection and scaleout<br />
storage for both virtual and physical environments. By providing a comprehensive solution, OneXafe<br />
eliminates the need for siloed point solutions and minimizes costs incurred from standalone hardware<br />
and software solutions.<br />
Get a handle on explosive data growth now with StorageCraft OneXafe!<br />
www.StorageCraft.com/OneXafe<br />
StorageCraft is a registered trademark of StorageCraft Technology Corporation.<br />
All other brands and product names are trademarks or registered trademarks of their respective owners.