sunoraclefaster

I saw the press release on the stunning 27 node RAC, but somehow I missed how Oracle uses this achievement in advertising. The press release talks about throughput whereas the picture:

hides the main message behind images of children’s bicycle (Superdome) and what looks like an F1 car from 70s (for an IBM P7), whereas Oracle’s result is associated with a fighter. I think this is wrong to make such an analogy and makes one to believe that one system has an incredibly fast response times – which is obviously not true. They all have similar numbers (except for some strange figures for IBM) – I mean, they are not far away from each other (OK, Oracle’s result looks better to me because it is more consistent – or maybe I misinterpreted FDRs), and the result just shows the transactions throughput. I really can’t imagine who needs this 27 heads monster or why Oracle did that benchmark after all when there are smaller yet powerful alternatives.

WLS + Oracle benchmark – 2

Another press release from Oracle on the same topic: benchmark of WLS 10.3.3 (not released yet) + Oracle 11.2.0.1 @ Dell PowerEdge R910 Server with 4×8-core beasts X7560 (announced today by Intel). Benchmark was done by Oracle itself (previous were conducted by Cisco & HP). Results and FDA are here. DB & WLS configuration are pretty much the same as in previous run, but there are some variations in numbers, like ~400GB redo logs instead of ~140GB; with one big change: compatible=11.2.0.1.

WLS + Oracle benchmark

Today Oracle announced this press release. Cisco Systems tested WebLogic SE 10.3.3 (not available yet) + single-node Oracle 11gR2 @ Cisco UCS C250 M2 Rackmount Server with 2×6-core Intel Xeon X5680 (not available yet) – here are the results. I want to highlight some specifics about this test bed – after all, it’s interesting to know what’s behind the scene.