We are running ESX 3.0.1 on an HP BL460 4way Intel dual core server with 16gb memory.
There are 14 guest servers on the host with a mix of RH Linux and Windows.
The Memory and CPU Stats for the ESX Server show the CPU avg about 25% and the mem avg util at 52% and the max at 60%.
However, we have server of the RH Linux servers begins stopping processes. Processes stopped appear to be random on the guest server.
Here is an example of the error messages:
messages:Jul 8 15:59:20 rstapp1 kernel: oom-killer: gfp_mask=0xd0
messages:Jul 8 15:59:21 rstapp1 kernel: Out of Memory: Killed process 11012 (java)
messages.1:Jul 3 20:03:22 rstapp1 kernel: oom-killer: gfp_mask=0xd0
messages.1:Jul 4 23:19:49 rstapp1 kernel: oom-killer: gfp_mask=0xd0
messages.1:Jul 4 23:19:49 rstapp1 kernel: Out of Memory: Killed process 7091 (ExecuteThread: )
The servers have plenty of memory with 1gb and 4gb per server.
Does anyone know the solution to this problem. Help!
What does sar say about the guest memory usage?
Also, try stopping the vmware tools and see if that improves it - the balloon driver could be pissing it off.