|
IBM’s AIX 5L V5.2 Improves UNIX Server
Utilization
By Charles King
IBM introduced AIX 5L Version 5.2 for its pSeries
UNIX servers this week. According to IBM, the new OS includes features and
enhancements that improve server utilization over competing UNIX systems from
vendors such as Sun Microsystems. AIX 5L Version 5.2 offers administrators
the means to create virtual servers on “dynamic partitions” as small as a
single processor and 250MB of memory. By contrast, the smallest partition
allowed on comparable Sun systems is two processors and 2GB of memory. Other
enhancements include dynamic Logical Partitioning (LPAR), which allows system
resources including processors, memory, and other components to be assigned
to independent partitions without rebooting the system; and Capacity Upgrade
on Demand (CUoD), which can be employed to implement new processors. CUoD and
dynamic LPAR support automatically bring a new processor online with no
interruption in service or performance. Logical Partitioning also allows
administrators to divide pSeries systems into smaller virtual servers concurrently
running either AIX 5L v5.2, v5.1, and/or Linux, so many systems may be
consolidated on to a single IBM eServer. IBM also announced Clustered Systems
Management (CSM) Version 1.3, which provides a single point-of-control for
installing, configuring, maintaining, and updating IBM xSeries servers running
Linux and IBM pSeries servers — or their logical partitions — running AIX.
AIX 5L V5.2 is currently available and CSM for IBM eServer 1350 and 1600
clusters is planned for October 25, with further product enhancements planned
for December 2002.
OS enhancements can be slippery little devils, and
UNIX upgrades are often particularly tough to get a handle on, but we see
IBM’s addition of LPAR and CUoD capabilities as a continuation of the
company’s long time server and business infrastructure strategies. On the
server side, dynamic logical partitions are yet another example of how IBM is
migrating mainframe-derived technologies to its other server product lines.
The company’s eLiza initiative may be the highest profile of these efforts,
but this new delivery of LPAR and CUoD capabilities to the company’s pSeries
products is another step along the same path of creating increasingly
powerful and flexible computers that have implications for virtually any and
every business process. The same strategic approach applies to the broadening
of CSM support for mixed UNIX/Linux (and Intel/Power) cluster environments.
Does that mean these enhancements to AIX 5L constitute a new digital business
revolution? Hardly, but they are positioned to notably ease the computing
headaches of businesses and their IT support personnel.
That may be the customer side, but what does AIX 5L
Version 5.2 mean to competitors such as Sun and HP? Both companies’ UNIX
servers utilize hardware-based partitioning which is less flexible than IBM’s
software approach. In practical terms, a 32-way IBM pSeries “Regatta” can be
sliced and diced into thirty-two partitions, while a 72-way Sun 15k “StarCat”
can support eighteen domains and a 64-way HP “Superdome” can support sixteen
partitions (nPars). This difference could make IBM solutions especially
attractive to customers who want to consolidate server workloads in order to
simplify infrastructure/management complexities and gain potential savings.
To our way of thinking, IBM’s enhancements to AIX 5L Version 5.2 suggest the
company clearly recognizes that long term strategic planning can offer
significant market results.
|
|
HP and Dell in Each Other’s Back Yards
By Jim Balderston
This week HP unveiled its new direct sales Web site,
where consumers and businesses can go to configure and buy PCs. The site
offers a wide range of options, including the adding of memory, enlarging or
enhancing graphics, storage, connectivity hardware, and the like. In short,
HP is offering both pre-built computers starting at under $500 to the custom-build
options than can run the price substantially higher. When doing a custom
build, the site calculates prices as the customer adds or subtracts items.
Meanwhile, Dell announced that it would be selling ink cartridges for its
recently announced printer line in conjunction with Lexmark, a printer
manufacturer. Dell will sell ink cartridges for its own products but will no
longer carry those for HP printers. Dell officials predicted they would drive
down the price of both printers and printing supplies like ink cartridges as
a result of entering this market.
Clearly this is a sign of the tightening spending by
both consumers and enterprises as both Dell and HP are entering markets or
channels that they not only have ignored recently but have even ridiculed.
This is a departure for both companies and one where each is eyeing the
possible revenues from the other’s once core business or business model.
Impacts of these moves — resulting in the companies becoming less and less
distinguishable at least on the consumer and SMB front — should be fairly
notable. Prospective customers will be able to not only do a certain amount
of head-to-head, apples-to-apples price comparisons, but also to compare
service and support options and costs. These factors, at least in the short
term, we believe will determine the relative success of each of these
companies’ forays into its competitor’s turf. But in the long run, it is
likely that price alone will diminish as the determinant factor in buying
decisions as matters such as product stability and support for retired
products may in the end become real differentiators that determine who wins
the battle. These companies are both selling commodities whose prices will be
largely dictated by supplies of the hardware components; their value add will
increasingly come in the form of the quality, depth, and consistency of the
service and support for the systems sold. The winner of that contest will be
the winner of this back yard brawl.
|
|
Security Extra? Not So, Says
Microsoft
By Jim Balderston
Published reports this week indicated that Microsoft
CTO Craig Mundie caused at least a minor fuss when he implied that Microsoft
might begin charging for additional levels of security within its product
base. Mundie made his comments at the RSA Conference in Paris, and those
comments drew a quick clarification from Microsoft HQ. Instead of describing
new charges for higher levels of security in Microsoft products, the
clarification went, Microsoft was exploring the idea of offering distinct
security products at some point in the future.
We have always been intrigued with the apparent
long-standing decision by the Redmond crew to steer clear of any sort of even
the simplest of security products. Yes, the company has been forced to
repeatedly offer security patches to its core product offerings, but the
company has never shown interest in getting into the leading edge security
product market. Instead, it has let others set the bar, and has followed a
“trailing edge” strategy in incorporating the more basic security features
into its core OS and application offerings.
It is entirely possible – and perhaps probable –
that Mundie’s comments were grossly misinterpreted. But even hinting at the
idea that security is an optional, extra cost add-on smells like an
over-reaching effort to squeeze more money out of a customer base that has
few options. Microsoft must understand that security is an essential part of
the operating system and its affiliated applications. Any other assumption
risks alienating customers in large numbers. Would the auto makers charge
extra for seat belts? Turn signals? A horn? Functional brakes? Of course not.
To date, gradually integrating security technologies into the OS has been a
relatively successful strategy for the Redmondites, even if regular security
patches must be distributed on a weekly basis. At least that steady stream of
responses to a constant and ever more threatening list of hacks and exploits
demonstrates some level of commitment to the user base. Software as a service
is a long-term commitment. Charging extra for next year’s installment of that
commitment seems not only short sighted, but notably customer abusive.
|
|
Duding Up the Enterprise?
By Charles King
In an interview this week, Dell Computer CEO Michael
Dell reiterated his company’s intention to push beyond the desktop market by
strengthening its enterprise server, storage, and service offerings.
According to Dell, most of the company’s recent profits came from non-desktop
products. In particular, Dell noted that in the past quarter the company had
shipped more than 250TB of storage per day. Additionally, the company’s
services business now contributes 10% of total revenues, and had 2,000
professional service engagements during the past year. Dell states that the
company will continue to expand its products and services globally, and cited
the networking and printer markets as areas with particular opportunities.
To begin, we admit that Michael Dell is a bright and
able guy whose company has been astute in its approach to the market. While
Dell Computer has never been a technical innovator (and is unlikely to ever
be one) the company is adept in identifying mature, relatively risk-free IT
sectors dependent on industry standard technologies, then seizing market
share by undercutting competitors with inexpensive alternatives, a strategy
that has been particularly resonant during the current recession. Dell makes
its profits by focusing on products that require minimal development effort
and by employing legendarily efficient supply chain processes. Dell’s success
will likely continue as it pushes further into the enterprise, an effort that
should be enhanced by the company’s long time strategic relationships with
Intel and Microsoft, and its more recent partnership with EMC.
However, we are less sanguine about Dell’s claims
concerning the company’s expanding services initiatives. While Dell Computer
has a seemingly limitless golden touch in any and everything requiring
creative manufacturing and distribution, it has never expressed much interest
in or capacity for the technical expertise required for successful
professional service offerings. Dell’s desire to press into the service arena
is smart enough, though it may be a touch belated. Every major IT vendor on
the planet has been moving in a similar direction for the past two years or
more, since recognizing that softening profits have become a natural feature
of the industry standard hardware landscape. With IBM as the poster child for
enterprise services success, and HP and other business vendors following
close in its wake, how will Dell to fit into this dynamic? Over the short
term, we believe the company is unlikely to succeed among enterprises with support-intensive
data environments that, understandably enough, will be unwilling to turn over
business critical processes to a service vendor with a limited track record.
Dell could partner its way around this roadblock, as the company has done
successfully in other areas where it lacked technical expertise or products.
While that strategy would cut into any short term profits Dell hopes to
realize from its services efforts, it could provide the time and cushion
necessary to develop adequate, reliable and believable enterprise service
offerings. Without such an effort, we expect Dell will find the IT services
sector slow going, at best.
|