Oracle Exadata: User Experiences with Bugs and Patches
Lately, I’ve been hearing a lot about Oracle Exadata performance. It certainly performs quite well. Although it is not necessarily faster than an IBM system. For instance, I know of a recent customer bake-off where an IBM Power 780 system was 2.4 times faster than a half-rack Exadata system. And when you combined the performance difference with IBM’s price advantage, it made the decision a no-brainer for the customer. Naturally, each customer bake-off has so many variables, as to make it useful only for that customer. The key thing to remember is that you need to keep your vendors honest by performing these competitive bake-offs, and not simply comparing a new system’s performance to an old system.
But anyway, the real reason for this blog post is to remind you that product maturity is important and to remind you of the 15 years of product maturity that have gone into the IBM Smart Analytics System. A level of product maturity that is not yet present in Oracle Exadata. Oracle are relative newcomers when it comes to developing integrated hardware/software systems, or engineered systems as they like to call them. And this shows when it comes to the challenges currently facing Exadata users. In the following presentation, I have assembled some references to independent experiences with Oracle Exadata. In each case, I include information about the source of the information, whether it is a Web page or a session at an Oracle event.
Admittedly, every product has bugs and issues. My intention here is simply to highlight that Oracle Exadata may not necessarily be the “IT nirvana” that Larry Ellison may portray it to be. While we have heard Oracle touting the ease of managament of Oracle Exadata, the reality is that this is a very complex system with many of the issues you might expect with a complex system that is so early in its maturity. And, of course, remember that as your system grows, the system will only be as reliable as the underlying software (ie. Oracle RAC).