|
Doing What You Gotta
Do: Microsoft Releases Windows Server 2003
By Clay Ryder
Microsoft has announced the general availability of
Windows Server 2003, Visual Studio .NET 2003, and SQL Server 2000 Enterprise
Edition (64-bit). The company indicated that this trio represents the
cornerstone of its enterprise strategy and lays the foundation for customers
to implement integrated, cost-effective solutions that connect information,
people, systems, and devices. “Doing more with less” was the mantra of the
day as the company demonstrated several new capabilities of WS 2003 targeted
at providing an interoperable enterprise infrastructure at lower cost, higher
performance, and lower IT operational cost (in part due to a new focus on
self-service IT) than its predecessor. Microsoft indicated that WS 2003
operates 30% more efficiently than Windows NT 4.0 and is the first Windows
Server release to support 64-bit Itanium systems. With this 64-bit support,
the company also announced that it is claiming the highest performance based
on the tpmC benchmark compared with any
non-clustered RISC-based UNIX solution. The WS 2003 family includes the
following configurations: WS 2003 Datacenter Edition (32-bit and 64-bit), WS
2003 Enterprise Edition (32-bit and 64-bit), WS 2003 Standard Edition, and WS
2003 Web Edition – all of which are now available – and the Windows Small
Business Server 2003, which will be available in Q3 2003.
Given the economic and IT doldrums that are
seemingly as difficult to escape as the proverbial death and taxes, it does
raise the question of whether or not this is an opportune time to launch a
major new release to Microsoft’s flagship server operating system. Although
WS 2003 offers several new capabilities, it is noteworthy that a substantial
portion of the product’s focus seems to be on enabling a more self-service
approach to corporate computing. Supporting multiple versions for
server-based documents (a ho-hum feat to those who used VMS, but a big deal
for Windows users) not only saves the user who accidentally trashed a
critical file, but more importantly eliminates frantic calls to the IT help
desk. In addition, the enhancements to SharePoint
Services have taken some of the mundane tasks IT faces, such as creating new
directories and file partitions, granting access to the appropriate
individuals, etc. when new project teams are established, and transfers this
activity to the users. Relieving IT of these burdensome tasks is reflective
of the doing-more-with-less theme, as Microsoft seeks to automate or create
self-service methods for fulfillment of mundane but essential IT processes.
It is also clear that WS 2003 is an operating system
that expects an n-tier environment in the enterprise to become the operating
norm. Features such as providing remote access to a user’s corporate desktop
illustrate that the server is more than a discrete resource, but can become a
conduit for enterprise computing that either facilitates information exchange
between machines or stands out of the way as warranted. Given Microsoft’s
continuing push for adoption of .NET based applications and services, it is
clear that the company is positioning itself as a provider of applied
infrastructure products, as opposed to a vendor of point products, be they
client/server operating systems or desktop applications. As standards-based
network services such as Web Services, XML engines, and other middleware
increasingly become the focal points of application development, thus
enabling J2EE and .NET as the effective operating environment for enterprise
applications, we are witnessing another layer of abstraction in both the
server itself and the resources under its control. Thus the nuts and bolts,
or more appropriately the hardware, firmware, and base level software
underpinnings of the enterprise IT infrastructure are no longer the design
criteria for application development as these give way to n-tier network
focused environments. But all this aside, WS 2003 represents a continuing
evolution of the server operating system, at least in Microsoft’s eyes. If
past history offers any indication of future behavior, we expect WS 2003 to
stand as another step in the Redmond Giant’s quest to advance its agenda and
crown itself as the king of the enterprise as it has in the consumer PC
market.
|
|
IBM Launches “Deep Computing” Initiative
By Charles King
IBM has announced its creation of the “Deep
Computing” strategic initiative, which will leverage IBM hardware, software,
emerging technologies, research initiatives, and industry expertise to
benefit high performance computing (HPC) customers. The company is appointing
a dedicated Deep Computing team that will take a comprehensive approach to
IBM’s HPC efforts, blending existing hardware and software with emerging open
standards such as Grid protocols into high performance solutions. Dave Turek was named vice president of the Deep Computing
group and Peter Ungaro will direct Deep Computing
sales.
We do not usually pay a great deal of attention to
the formation of new business groups, but a number of issues surrounding
IBM’s Deep Computing are well worth considering. First, IBM’s history and
strength in supercomputing and the HPC space makes the formation of a
dedicated team a natural step for the company to take. We assume that the
tram’s Deep moniker refers to the chess-playing Deep Blue system that beat grandmaster
Gerry Kasparov. In fact, the technologies developed for Deep Blue were
adapted for the IBM eServer p655, an 8-way
POWER4-based system developed specifically for clustered HPC applications.
The continuing migration of clustering technologies into the once monolithic
world of supercomputing is the second piece of the Deep Computing puzzle we
find particularly interesting. The time when high-performance computing meant
Big Iron is rapidly fading, replaced by highly clustered configurations of
scores or even hundreds of RISC and commodity servers. The penetration of
such systems can be seen clearly in the rankings organized by the Top500.org
Web site, which ranks supercomputers worldwide.
So what does all this have to do with IBM’s Deep
Computing initiative? While the sheer cost and complexity of HPC systems
tends to aim them primarily at research and government lab settings, the
continuing declines in hardware prices along with enhancements in performance
are making commercial applications of HPC increasingly viable and enticing.
IBM has found notable success with commercial HPC solutions for the
automotive, pharmaceutical, and oil/gas exploration industries, and we expect
those applications and sales wins to continue as hardware of every kind
becomes increasingly affordable. Hence the formation of Deep Computing, to
continue developing cutting edge solutions for traditional HPC customers, and
to drive commercial applications of those same technologies. Overall, we see
IBM’s Deep Computing initiative both as a simple continuation of one of the
company’s traditional strong points, and as a concerted effort to deliver
what were once considered highly esoteric computing technologies into broader
markets.
|
|
Who’s Next?
By Jim Balderston
Siebel Systems and PeopleSoft – two of the market’s
largest CRM vendors – released their earnings reports this week. Siebel
reported revenues of $332.8 million and a net income of $4.6 million for the
quarter. Revenues came largely from maintenance, consulting, and services, at
$220.7 million with license fees bringing in $112.1 million for the quarter.
Siebel’s profits were off 93% from the first quarter last year. PeopleSoft
announced that it had revenues of $460.3 million and a profit of $38 million.
Both figures were off from a year ago. License fees fell 39% compared to a
year ago, dropping to $80.8 million. PeopleSoft announced it was cutting 200
jobs.
We have not been shy about our criticisms of CRM
vendors and with these results in hand there is no reason to change course.
While many believe that all that is needed is a turnaround in the economy for
CRM vendors to once again be making financial waves in the marketplace, we
see signs that even a robust economy would not necessarily give these folks
the kind of boost that they enjoyed in the go-go days of CRM.
One would think that in a down market, CRM – or
customer care and feeding applications – would be in great demand as
companies try to hold onto the customers they have and maybe increase sales
to those customers without the high risk move of increasing staff. Yet these
numbers show that hypothesis is not being realized. Instead, these numbers
indicates that despite the need to hold onto customers in a down economy, CRM
is not the answer for many businesses. In fact, it would appear that CRM
spending was – and may always be — more of a discretionary outlay, one on a
par with advertising. And that the market – as the dozens of failed tech
publications can attest to – is very fickle indeed. While we believe that
offering applications and technology that will help companies keep their
customers close at hand is an intriguing opportunity and potentially valuable
offering, we are not sure that the last wave of CRM vendors has laid the
groundwork for the next generation of customer care applications. When one
considers the high cost and complexity of CRM offerings – as documented by
the revenues figures for Siebel Systems – one can’t help but think that
customer care was the not the prime motivation of product development at Siebel.
It would appear that long-term customer commitments and investment were the sought-after prime directive. As a result, the
idea of helping companies weather the down times by holding onto their
existing customer base may have fallen by the wayside. We suspect customer
care and feeding applications may well have to come from a different
direction, from different vendors, who actually might deliver on their
promises. Any takers?
|
|
SNIA-E Releases European Storage Survey
By Charles King
A market research report sponsored by the Storage
Networking Industry Association Europe (SNIA-E) suggests that enterprise
storage requirements are continuing to expand notably in the European market.
Omarketing Limited conducted the study, which
queried 100 respondents representing storage professionals from end-user
organizations throughout Europe. Responses were collected from September 2002
through January 2003 at a variety of European events and via the SNIA-E Web
site. According to the report’s executive summary, 30% of respondents claimed
that storage requirements had grown by 100% or more in the previous twelve
months, and 9% reported storage growth exceeding 200%. Additionally, the
group as a whole expects storage growth to continue at approximately the same
rate over the coming year. Key applications cited by respondents as driving
storage growth were email (63%) and databases (73%). eCommerce (22%) and CRM
(28%) represented the smallest growth categories trailing images, videos, and
MP3s (30%). Nearly four in ten (38%) survey participants said TCO is the key
driver in making storage purchasing decisions, though slightly smaller
numbers also cited storage management, downtime, integration, and scalability
as critical issues. Nearly a third of respondents expressed interest in
deploying virtualization solutions during the next six months.
The SNIA-sponsored study offers both items of
interest and cautionary notes regarding the value of market research. To
begin, we were troubled by how some data points were presented in the
executive summary. For example, the study suggests that 86% of respondents
claimed their storage needs increased by more than 50%, while 14% saw no
change or a decrease, leading to the claim that storage needs are
outstripping Moore’s Law. What we wondered was why no respondents claimed (or
perhaps were even asked) to indicate increases between 1-50%. A small thing,
perhaps, but one that niggles a bit. We were also disturbed by the study’s
suggestion that the respondents’ interest in virtualization indicated both
the timeliness of SNIA’s Bluefin-based
Storage Management Initiative and a maturing of the market for
virtualization. While users appear interested in virtualization and similar
centralized management solutions for heterogeneous storage environments, we
believe the SNIA’s conclusion has at least the
appearance of being a bit high handed, especially considering the
still-nascent state of virtualization technologies. We were also interested
to note that while virtualization was fourth on the list of storage areas of
interest for the next twelve months, the three technologies preceding it were
not mentioned at all in the summary.
Despite those shortcomings, the study did offer some
food for thought. Users’ ongoing interest in disaster recovery and data
backup were particularly intriguing when contrasted with their essentially
lackadaisical attitudes toward testing the systems and procedures involved in
these areas. Some may regard this as a simple sign of incompetence, but it may
instead indicate the level of difficulty end users have in establishing and
maintaining effective disaster recovery programs. If that is the case, it
could point to opportunities for storage service providers and storage
vendors focused on delivering utility-style storage services. We were also
intrigued by the 38% of users whose organizations deploy storage environments
that blend various mixes of DAS, SAN, and NAS. Clearly, European enterprises
have the same tendencies as their stateside counterparts in deploying storage
solutions however and wherever they please. This tendency seems to us to be
driving the interest in virtualization-style solutions, whether they come
from the SNIA or elsewhere. Overall, while we found the SNIA-E’s storage
study of some interest, we believe that if the group had been more balanced
in analyzing and presenting the data they collected, they would have brought
more credibility to themselves and their members.
|
|
Hot Fix? Not So Fast!
By Jim Balderston
Microsoft released its latest security patch for
Windows XP on April 16 and a number of users reported that the patch, which
was rated the second highest priority of “important” by Microsoft, affected
computer performance. Microsoft has said it is working to resolve the issues
that some users reported on a number of user group online discussion boards,
and is investigating the problem. The security flaw that the patch was
supposed to address could allow an attacker to alter their privilege level on
a susceptible system.
There may be a parallel between the Microsoft
“important” rating of this patch and the nation’s color coded security alert
system. One has to wonder to what degree Microsoft’s rating system is a
product of actual threats and one that is designed to fend off the hordes of
anti-Microsoft techies that regularly – and in many cases, accurately –
describe or reveal Microsoft operating system vulnerabilities.
Yet this incident raises a series of interesting
questions concerning the whole concept of software as a subscription, one
that is constantly updated, tweaked, and modified as new developments demand.
In such an environment, does the code base need to be fundamentally
re-engineered in a way that will prevent patches from causing havoc? Since we
are essentially in the early stages of such a software model, one can argue
with reasonable certainty that all the potential bugs in an ever-evolving
code base have not been worked out. While we believe that the
software-as-subscription model is one that is here to stay, we suspect that
its desirable properties – i.e., always up-to-date code – will require a
little more rework on the back end as a result of conflicts like this
example. Over time, the modularity of the code base that is the core of a subscription
model will mature, but to do so will require a substantial effort at both
increased stability and “healability” combined with
the need to maintain some level of backwards compatibility. Incremental
updates remain an improvement over wholesale OS upgrades or new releases, but
we suspect that the real work in perfecting the software-as-subscription
model lies out ahead of us, and as such remains a formidable challenge for
any and all that undertake it.
|
|
WebMethods and
Informatica Announce Business Activity Platform
Availability
By Myles Suer
WebMethods and Informatica this week announced the release of their
jointly developed Business Activity Platform, which combines WebMethods’ Integration Platform with Informatica's
PowerCenter data integration and PowerAnalyzer business intelligence software. The
companies claim their integrated offering creates real-time visibility in a
business intelligence dashboard, and say the new solution extends the typical
limits of business intelligence software by enabling real time data and
analysis, interpreting data to formulate and invoke optimal responses. Both
HP and Accenture have agreed to be systems
integrators for the Business Activity Platform. The product is now available.
Pricing was not announced.
Web Methods and Informatica
have performed an interesting balancing act with their new Business Activity
Platform, first in asserting that business intelligence cannot handle the
range of problems that the new system can, but also by claiming that the new
system offers users additional value by correlating real time data against
historical information. Since companies including Brio are making the same
argument, clearly Business Activity Monitoring has become the latest cool
enterprise software offering. But as with any new technology, product
definitions provide the basis of much of the attendant marketing hype. WebMethods and Informatica have
joined the party by asserting they can monitor any business process by, for
example, layering on top of data center management software from BMC Patrol
and HP Openview.
While we live in an age when software must
demonstrate a real ROI, to create such value systems need to take on unique
tasks or provide demonstrable solutions for specific problems. To our way of
thinking, bridging between analytics and business planning and budget offers
a unique value proposition. For this reason, we believe Cognos
and Hyperion have taken the right tack by focusing on the use of business
activity monitoring to create indicators of a company’s business health.
Although a personal dashboard such as that provided in the Business Activity
Platform can be a fun technology, having one that can monitor every
enterprise business process does not create unique, measurable ROI. In the end,
simply acquiring real time data is of limited usefulness without a larger
context.
|