Oracle Storage Guy

VMware: Maritz out, Gelsinger in. What does it mean for Oracle?

As many of you have heard by now, Paul Mauritz is out as CEO of VMware, and Pat Gelsinger (formerly COO of EMC, reporting directly to Joe Tucci) is replacing him.

The relevant question for those of us who care passionately about Oracle virtualization using VMware vSphere is:

What does this mean for VMware's BCA (Business Critical Applications: stuff like Oracle) strategy (which under Maritz was somewhat lacking)?

In my view, this is very, very good news indeed. The reason is simple: Paul Maritz was former Microsoft, and thus very Microsoft-centric. Under Maritz, VMware very successfully penetrated the Microsoft space (especially Exchange). BCA, and especially Oracle were not that well supported by VMware's high level leadership in my view.

There is no question in my mind that Pat Gelinger gets it with respect to the importance of Oracle for VMware's future. (As I have said before, I believe that Oracle is the most interesting virtualization target environment for VMware, period.)

Gelsinger has shown consistent interest in Oracle, as his presentation at Oracle OpenWorld 2011 shows. (His co-presentation which included a vCenter demo with Chad Sakac at OOW 2011 is classic.)

We wish Pat Gelinger the very best in his new role and have high hopes for him as the leader of VMware.

Please Vote for My Session at VMworld 2012

I have once again proposed a session at VMworld 2012. This session is titled as follows:

Near Zero Downtime Migration of Oracle Database from Legacy RISC-based UNIX to x86-64 / Linux on VMware vSphere

The reason I believe this session is so important is simple: Every Oracle customer and his dog is desperately trying to get off of expensive, slow, proprietary RISC-based UNIX and onto the x86 environment. The reason for this desperation is simple: Cost. These environments are becoming hideously expensive platforms on which to run Oracle. Huge core counts. And performance is frequently terrible.

For example, EMC IT just replatformed off of their E25K and onto x86 / VBLOCK, in the process improving performance by 2 - 10 X and saving over $7 million per year. Yeah, that's real money.

Trouble is, EMC IT did their migration cold, and it took 16 hours to do so. Many, many Oracle customers have mission critical databases on RISC UNIX which have little or no downtime window. These customers need a migration path off of RISC UNIX which involves minimal downtime.

Another source of trouble: Oracle has such a process, which they refer to as O2O. They offer a service to get the customer off of RISC UNIX and onto x86. But where does the customer end up? Onto ExaData. Not good considering that EMC is the majority incumbent storage company for storing Oracle on RISC-based UNIX. We stand to lose all of these customers to Oracle in other words, if we simply ignore this problem.

The one clear differentiator we have against ExaData: VMware. By migrating the customer off of RISC-based UNIX directly onto VMware vSphere, we innoculate them against ExaData forever. That's because, as we all know, the value proposition for running Oracle under vSphere is very compelling, and quickly becomes addictive once the customer adopts it. vSphere has a high rate of stickiness in other words, especially in Oracle shops.

Bottom line: I think this is the most important subject that we can be talking about right now to Oracle customers.

If you would like to vote for my session, please go to this site (requires a login, but registration is free), and select:

Session Type: Breakout Session

Tracks: Applications

Subtracks: Business Critical Applications

Search: "oracle"

My session is session 1270: Near Zero Downtime Migration of Oracle Database from Legacy RISC-based UNIX to x86-64 / Linux on VMware vSphere