We need to replace our old server and therefore requires a new Redhat server for our environment with "Synergy 18.104.22.168.001" (with Informix) and "Change 5.2".
Until now, I had a "HP Blade Server G1" with 8 cores, 32 GB RAM and direct attached SAN disks in use.
Our data center would mean that we are migrating to virtual "Redhat 6.1" "ESX Server." But I fear strong performance losses in the I / O area, because the ESX Server Shared SAN disk in use. What I want to avoid in any case is that there are reports "Waiting for CPU (I / O)".
I think not without reason, IBM recommends the use of local "Raw devices" for the DB file chunks.
For Oracle databases, there is even an article from Red Hat.
My question now is,
What should I ask for our data center requirements, with respect to the "Disk Performance"
Access time read / write
Throughput read / write.
And how can you measure or verify this.
thanks for the help