Virtual Irony for Oracle

Where’s Oracle VM 3, and why should you be waiting for it? Here lies a puzzling story.

On May 13, 2009, in the course of developing Oracle VM 3, Oracle acquired Virtual Iron Software Inc and with it, their rather good VM product Virtual Iron. In June 30th 2009, scarcely a month later, they had ditched the product, saying “Oracle will suspend development of existing Virtual Iron products and will suspend delivery of orders to new customers”. It was a move that puzzled everyone at the time. Virtual Iron had been a US start-up with a Xen/Java server virtualization product based, like Oracle VM, on the open source Xen hypervisor. The bulk of Virtual Iron’s 2,000 customers were small- and medium-sized companies who were using it as a cheaper and simpler alternative to VMware. The product was, evidently, well-received by its customers who then, it seems, shrugged and moved to VMWare. Oracle seemed to have acquired Virtual Iron merely to bolster their Oracle VM team and to enhance the rather lacklustre Oracle VM 2.2 with Virtual Iron technologies. These technologies that attracted Oracle were Capacity and power management, an Open scriptable API, and a scalable, modular management framework.

We all thought that the effect of this cull of a promising company would be to accelerate the development of Oracle VM 3, predicted originally, and unofficially, for release in 2009 by Wim Coekaerts, vice president of Linux and virtualization engineering.

Sadly, it didn’t really work out that way. 2009 Oracle OpenWorld came and went, as did 2010 Oracle OpenWorld: No Oracle VM 3. Finally, It went into Beta in December 2010. And in February was reported by Herbert van den Bergh to be in the final production phase. At the time of writing, the best guess for its belated appearance is 2011 Oracle OpenWorld.

So why do we need Oracle VM 3.0? Whereas Oracle VM 3.0 could, once it arrives, persuade Red Hat users to move to Oracle for their Linux, virtualization and application server support, it is less likely to induce VMware/Oracle customers to migrate their Oracle workloads to Oracle VM 3.0. The aging Oracle VM 2.2 didn’t compare well with competing products regarding load-balancing; it has no snapshots, suffers tricky backups and cloning, and doesn’t have memory consolidation, so Version 3 would need to leapfrog the current industry standards. Even though we are assured by insiders that Oracle VM 3 will be the best VM product to host Oracle Database, there is a another problem for anyone with Microsoft Windows servers. Neither Exchange 2007 nor Windows 2008 are currently supported to run on Oracle VM. It seems that Oracle has no VM product that is certified to run Microsoft software at present. If you are aiming to have one VM solution for all your servers, including Oracle Database, you’re currently out of luck if you prefer it to start with the letter O.