iSCSI take-up in Europe 2004-2005Editor's intro:- A
year ago (2003), in an article on our news page called -
Running with iSCSI
, Alexandre Delcayre of FalconStor Software, Inc. considered
the prospects for the growth of iSCSI take-up in Europe during 2004. Back then,
he suggested that, after some initial hurdles, 2004 looked like a year for rapid
growth, especially among early adopters. Here, 12 months on, he looks at where
iSCSI is now, and offers a new opinion. | |
. |
 |
. |
|
About a year ago, I was
asked to consider the prospects for iSCSI in Europe in 2004. At that time and
after due thought, I offered the following opinions, saying that:
- IT purchasers should look to work with vendors and channel partners that
have established expertise in SANs, including IP-based SANs, because IP-based
SANs are not a new concept and have existed for a number of years
- iSCSI would need to go through a normal evolution cycle before becoming
mainstream in IT architecture. Before the standard ratification, iSCSI was a
concept for forward-thinkers to evaluate and incorporate into their medium-range
plans. In 2002 many such early adopters were already moving forward
aggressively, while the next round of technologists were starting [in 2003] to
design iSCSI into their future architecture o based on end-user and channel
feedback, I felt that there would be a period of rapid growth and that, while
2003 had been the year of standardisation and early-adoption, the sky could be
the limit in 2004.
Now, a year on, we can look at how iSCSI has really
done in the last twelve months: did it really come of age? At first
sight, there's no easy answer to that question, and there have certainly been
plenty of mixed signals.
At one end of the scale, I've been asked by a journalist whether
anyone is buying iSCSI, and whether it's going to take a price war to kick-start
buyer interest in the technology. That's a fairly controversial starting point,
and isn't one that I can justify in any way. But - journalistic interviewing
techniques aside - it suggested there are people out there who, even now, remain
sceptical of iSCSI.
At the other end of the scale is what I've
witnessed myself at FalconStor during the last twelve months. First, the launch
of Microsoft's Windows Storage Server 2003 brought the concept of storage
solutions within the attention span, if not yet the reach, of smaller
organizations. However, the need for a SAN is still there and an FC SAN will
clearly keep the cost of a storage solution up.
Then, in April of this year (2004) we formally announced our own iSCSI
Storage Server, designed specifically to transform Microsoft Windows Storage
Server 2003 into a unified storage platform, providing both file and block level
services, over an existing local area network. Organisations such as SMEs and
department-level customers that were unable to move to a Fibre Channel SAN due
to cost or complexity issues are now able to consolidate storage for Microsoft
Exchange Server, Microsoft SQL Server, Oracle, and other application servers.
Our announcement clearly struck a chord as, throughout Europe, we've
since been signing OEM deals with integrators and resellers to join our partner
programme for iSCSI Storage Server, in order to add iSCSI functionality to their
existing Microsoft Windows Storage Server 2003 devices. Our new partners
include companies like Acer (worldwide), Evesham Technologies (UK), and eSeSIX
and ARTEC (Germany).
If you consider not my opinion, but their opinions, to be the other
end of the scale, then we have two contrasting viewpoints to consider: some
negativity at one end, and a significant gearing up for iSCSI at the other.
Sitting somewhere in between these opinions, we'll doubtless find the
purchasers: organisations and businesses that know they need a storage solution
and are looking at their options.
For those considering iSCSI as part
of that storage solution, the key questions that need to be asked of your
reseller or integrator - and answered by them - are:
- When do I use the existing network instead of developing a separate IP SAN?
- Should I use standard NICs or should iSCSI HBAs be used?
- Should I use GbE for the backbone, or can I get by with 100 Mbit for
distributed servers?
- What are the critical tuning parameters? (GbE jumbo frames are just the
beginning.)
- What's the impact of network performance on application performance?
- What provisioning, monitoring and management tools are needed for the
network? For virtualising and managing storage capacity?
- What other storage services are needed (snapshot, mirroring, replication,
backup, etc.)?
- What are the security requirements?
With storage on the agenda of more companies than ever before through
the launch of Microsoft's Windows Storage Server 2003, and data no less critical
or vulnerable than it ever was, many companies will look at these questions and
their answers and conclude that iSCSI, together with storage management software
to provide business continuity and disaster recovery, will be the best way to
get the most out of their investment in storage.
The one thing we can
be certain of is that our industry is not prone to missing a trick by being too
late with a technology; if anything, we are sometimes a little early. With the
benefit of hindsight, it looks now as though I may have been over-optimistic
as to the levels of take-up of iSCSI in 2004 - but, ironically, because I was
right: iSCSI has, after all, had to go through a normal evolutionary cycle, and
that cycle has simply taken longer than the twelve months I predicted.
However, the rash of OEM signings for our iSCSI Storage Server
undoubtedly signifies that many influential people at the really sharp end of IT
- sales - now believe that iSCSI is, or is about to become, mainstream, and that
iSCSI is set for yet bigger and better things during the next twelve months.
Perhaps the sky will be the limit in 2005, instead.
| |
. |
...Later:-
iSCSI Market in 2004 was 10 times smaller than Hyped Analyst Predictions |
FRAMINGHAM, Mass. - March 4, 2005 - IDC reports that
Worldwide external disk storage systems factory revenues grew slightly in the
fourth quarter of 2004, posting just 1% growth over the same quarter one year
ago to $3.8 billion.
"In 2004, the market for iSCSI SAN
grew beyond the $100 million barrier, led mostly by the adoption of
midrange systems priced between $15,000 and $149,999," said Natalya
Yezhkova, senior research analyst, IDC Storage Systems.
Editor's comments:- these new figures explain the deep
disappointment which has been reported privately to me by many iSCSI vendors.
A few years ago IDC itself had predicted that the iSCSI market in 2004
would be worth over $1 billion.
Instead - the market barely reach a
tenth of that size. It shows that you can get burned if you base your business
plans on analyst projections. This has also affected several specialist iSCSI
publishers who were hoping to make a killing in this market - and have instead
resorted to running low cost google ads.
see also:-
iSCSI SSDs,
storage market research,
SSD market analysts | | |
|

| |
|