In-Depth
Analysis: The Oracle-Intel-HP Fallout Revisited
IT pros believe Oracle's move has less to do with cost savings and more to do with a strategy to negatively impact the hardware sales of partner/rival HP.
IBM Corp. has embraced Oracle Corp.'s decision to drop support for Intel Corp.'s Itanium microprocessor as both a market opportunity and as a way to settle a long-standing score. Big Blue isn't alone; most of Oracle's rivals issued competitive responses to try to attract shell-shocked HP customers.
For IBM, however, the net effect of Oracle's move is especially ironic.
First, the background. Hewlett-Packard Co. (HP) likes to position its Itanium-powered HP-UX platform as a mainframe replacement alternative. It's been boosted in this regard by Gartner Inc., which on a few occasions has recommended that customers consider migrating off of aging mainframe systems and on to "modern" platforms -- such as HP-UX.
To be sure, Gartner hasn't issued a blanket recommendation to migrate off of Big Iron; however, in certain scenarios (for example, aging mainframe systems running legacy non-relational database workloads such as IDMS or Datacom), Gartner analysts have counseled customers to consider ditching Big Iron altogether.
IBM, not surprisingly, hasn't liked this. After all, notes Timothy Sipples, an IBM Enterprise Software Architect who contributes regularly to Big Blue's official Mainframe blog, if customers had followed Gartner's advice and moved to a "modern" alternative -- say, a combination of HP-UX and Oracle middleware -- they'd be effectively high and dry.
"Shouldn't those same Gartner analysts apologize for giving such profoundly bad advice to their customers?" asked Sipples in a blog post earlier this month.
Sipples noted that IBM already markets several offerings that could be of some help to orphaned Itanium customers. These include its IBM Database Migration Toolkit (which can help customers shift from Oracle 11g running on HP-UX to DB2 running on AIX, z/OS, or Intel x64 hardware); its IBM Migration Assistant for Oracle Tuxedo (which can help customers shift Tuxedo-based applications from HP-UX to Big Blue's TXSeries or CICS platforms); its IBM WebSphere Application Server Migration Toolkit (a WebLogic migration tool); and its IBM Migration Factory (which includes a built-in migration scenario from Itanium/HP-UX to AIX-running-on-POWER).
In a separate post, Sipples argued that Oracle's move is tantamount to a death sentence for Itanium. Not that Itanium was ever in any sense a runaway success to begin with.
"Oracle's announcement is Itanium's death sentence. It's also [a] brutal but highly effective business strategy for Oracle in its competition with HP," he maintains.
This isn't exactly a controversial claim. In their own assessment of Oracle's move, Gartner analysts George Weiss, Andrew Butler, and Donald Feinberg concluded that "HP-UX users must now reevaluate their platform and life cycle support options."
IBM's Sipples says it's clear what will happen once customers perform this reevaluation: software dependency will (in most cases) trump hardware preference.
"Oracle will probably lose some HP customers who are ... concerned about doing business with a vendor that abandoned them, but Oracle is betting, correctly, that most of its customers, when forced to choose sides, will choose Oracle's software over HP's hardware," he writes.
Customer Doubts
Sipples could be overestimating the importance of software, however.
According to a new survey conducted by Gabriel Consulting Group (GCG), for example, HP customers aren't just angry with Oracle; they profoundly distrust the software giant, too.
Whether this mistrust, profound or otherwise, is to Oracle's detriment -- or whether it's strong enough to offset the considerable leverage Oracle enjoys from its combined database, middleware, and application stack -- remains to be seen.
GCG says its survey collects responses from 450 IT pros -- almost 70 percent of whom are employed by companies at least 1,000 employees (and almost half of which work for organizations with at least 10,000) -- about Oracle's move.
Respondents support a variety of operating environments -- including HP-UX (cited by 58 percent of shops, almost 20 percent of which are self-described "heavy" HP-UX users) and Solaris (cited by 64 percent of shops, 18 percent of which are self-described "heavy" Solaris users).
Users of HP server hardware were likewise heavily represented, with more than half of all respondents describing themselves as "Heavy" HP server shops. (All told, HP had a hardware presence of some kind in 87 percent of shops.)
By the same token, Oracle's database platform was used by a vast majority (88 percent) of respondents, with 60 percent describing themselves as "heavy" Oracle users.
The GCG survey, then, comprises input from a large percent of the customers likely to be most affected by Oracle's move.
GCG found that almost half (48 percent) of respondents don't accept Oracle's claim that the company decided to discontinue support for Itanium because Intel had disclosed plans to phase out Itanium. Just 29 percent of respondents accepted Oracle's account, with nearly a quarter (23 percent) reporting that they were "not sure" what to believe. One-half of respondents endorsed Intel's account, agreeing with the statement that Intel "know[s] their own roadmaps." Just 16 percent disagreed; just over a third (34 percent) said they weren't sure.
Respondents were more divided about Oracle's motivation, with a plurality (49 percent) either agreeing (34 percent) or "strongly agreeing" that Oracle wished to "reduce the number of platforms [it] support[s]." On the other hand, 39 percent disagreed with this explanation, and 12 percent said they weren't sure.
There was more considerably more consensus about the competitive fallout from Oracle's move: most respondents (64 percent) endorsed the statement that "Oracle will use as a justification to raise license/support costs for Itanium customers."
Just 14 percent said that they "disagree" or "strongly disagree" with this explanation. Slightly more than a fifth (22 percent) said they weren't sure.
Meanwhile, a clear majority of respondents believe Oracle's move will have a negative effect on HP's HP-UX and NonStop platforms: 77 percent of respondents judged Oracle's decision a "competitive move to kill HP's HP-UX and NonStop products." Just 10 percent disagree with this assessment; 13 percent report that they're unsure.
More worrisome, at least from Oracle's perspective, is that almost four-fifths (79 percent) interpret the decision as the "First step in Oracle's plan to put all competitors at disadvantage [versus] Oracle [hardware] products." Just 10 percent of respondents disagreed with this assessment; 11 percent said they were unsure.
"It's clear that customers believe Oracle dumping support for Itanium isn't just 'business as usual.' In their minds, Oracle has a much more ambitious agenda, and this move against HP is the just the first -- or latest – step," writes principal analyst Dan Olds, in GCG's What's Up with Oracle survey report.
If HP, Why Not Others?
Olds isn't the only industry watcher who's reached such a conclusion. Jonathan Eunice, founder and principal IT advisor with consultancy Illuminata, says Oracle's move begs a bevy of uncomfortable questions -- particularly for large Oracle customers running on non-Oracle hardware platforms.
"What about Oracle running on IBM's Power Systems? Oracle running on Red Hat Enterprise Linux? Oracle running on Dell or Fujitsu gear? Are those at risk next? It's certainly possible. Oracle CEO Larry Ellison has described the mainframe era and model in glowing terms; if he's hell-bent on going there, and on pulling customers into Oracle's 'full stack' fold, then Oracle's traditional commitment to heterogeneous multi-vendor support goes by the wayside," wrote Eunice, on his company blog.
"If I were an Oracle customer running on any non-Oracle hardware, I'd be asking for long-term support and software update assurances, in writing, with clear and enforceable penalties for non-performance. Verbal reassurances would no longer be sufficient."
The upshot, Eunice argued, is that the nature of IT competition is changing.
"You can very rationally argue that too much money is made on platform X or with partner Y, for a company like Oracle to stop doing business there. You could credibly argue that companies like Oracle have too many precious joint customers, accounting for too much revenue and profit, for such stoppages," he wrote.
"But you probably could have made very similar arguments about Oracle supporting HP and Integrity servers, too. 'That can never happen here' is like last month's assurances about the safety of the Fukushima Daiichi nuclear plants -- arguments begging to be disproved as facts on the ground change."