I’ve been getting a lot of questions about why one would choose Exadata Database Machine X2-8 over Exadata Database Machine X2-2. That’s actually a tough question, however, some topics do spring to mind. I’ll start a list:
- The Exadata Database Machine X2-8 only comes in full-rack configurations. No way to “start small.”
- The Exadata Database Machine X2-2 only (immediately) supports Oracle Linux. If Solaris is attractive to you then the X2-2 is not an option at the time of this blog entry. That is slated to change soon.
- Database Host RAM. The aggregate database grid RAM in a full-rack X2-2 system is 768 GB but 2 TB with the X2-8. The list is quite long for areas that benefit from the additional memory. Such topics as large user counts (consolidation or otherwise), join processing, and very large SGA come to mind. And, regarding large SGA, don’t forget, the Exadata Database Machine supports in-memory Parallel Query as well.
Not on the numbered list is the more sensitive topic of processor power. While these sorts of things are very workload-dependent, I’d go with 16 Intel Xeon 7500 (Nehalem EX) processors over 16 Intel Xeon 5600 (Westmere EP) for most any workload.
So, readers, what reasons would motivate you in one direction or the other?
Kevin Closson's Blog: Platforms, Databases and Storage 
Hi Kevin,
Would you be able to conduct some webcasts on Exadata in the interest of the community. Oracle RAC SIG does a lot of such presentations and they help us a lot. If we can do something similar, I think it would be great.
Regards
Vijay
I see one very big disadvantage to the X2-8 platform: stability. If you only have 2 servers running your database, any problem halves your power( also in case of online patching). So if you need good uptime, you either only use 50% of the power or you need a multirack installation.
For me this would rule out X2-8 for any environment that needs only one rack.
Ikke,
Ok, that’s why there are choices. It sounds as though you’d rather have more, smaller servers (X2-2).
I would tend to agree with ikke, if you lose one server in an 8 server config, you still have 87.5% of the capacity remaining. Lose one out of two, and you have lost half. According to Oracle, if you have a large OLTP workload (one big one or lots of little ones), you may be better off with the larger memory footprint of the two node X2-8. For Warehouse environments, the 8 node scale out work better, they say.
Seems a shame that we are only given a couple of server/ disk configs. We have a greater need for Server capacity than we do storage, so we could wind up buying addional Exadata racks just to get CPU power and wind up with way more storage than we need.
Hi ,
Guess with 2-node X2-8 we will see less Interconnect traffic and more capability per node. Perhaps that could be the reason for having lesser nodes.
Would having lesser nodes bring down the electricity bill making it greener 🙂
Thanks.
How does the x3 dynamic and the changed cpus tilt this question. Some napkin based calculations I did seem to put the x3 in a losing position cpu wise.