Article is so full of it pitching DB2 that it's not even funny. Oracle, as an active MySQL contributor (InnoDB) would of course maintain and support it for the lower end niche market, probably adding an easier migration path up to the full Oracle installation but nevertheless, it makes no sense for them to kill the product with millions of installations which is not competing with any of their products. Nobody gets Oracle for a php startup.
Unlike Virtual Iron, which I think is in a direct competition with VirtualBox so why keeping the worse one?
Agreed that it's not the same sort of situation. MySQL is obviously well known and has significant market share, and although Oracle is a competitor to some extent, Postgres is probably much more similar in terms of who the typical customer is.
And anyway, how many people here have heard of Virtual Iron, much less used it? There's good reason for that -- it's really, really bad, or at least it was 6 months ago. We had it for a while because our MIS group was too cheap to buy VMware. Its Java-based frontend was clunky and unstable. VMs would frequently hang when powered down or rebooted through the console, stuck in a "shutting down" state that could be resolved only by rebooting the physical box. On a number of occasions, the server inexplicably got in a state where none of the VMs would power on until the box was rebooted. Snapshotting was based on LVM, totally different (in a bad way) from how every other virtualization product I've ever seen works. The "console" feature was based on VNC, and as you might guess, extremely slow.
After some time, we finally shelled out for VMware ESX/VirtualCenter, which isn't perfect, but still a much better product. I don't know why Oracle bought VI to begin with, but I'm not exactly sad to see it go.
Unlike Virtual Iron, which I think is in a direct competition with VirtualBox so why keeping the worse one?