Resource monitoring (add CPU?)

[burdweiser]

So I'm having a great debate about assigning another CPU to a server that seems slow to some users. The server in question is running an app called "change gear". It is Win 2K8x64, 2GB of RAM and 1CPU (intel x7460 2.66Ghz from a Dell R900). Looking in vCenter, I see that the guest server is running at 49mhz and the CPU ready time in esxtop is around .03%. Seems to be running fine right? Well, in Spolight (quest software), we see that the server has a warning about processor queue lenghts and that there are about 800 threads running. Quest software is stating there is a bottleneck with the CPU and we should add a CPU to the server. Is this assumption correct?

I'm just not seeing the same results in vCenter.

[RParker]

Granted, this is not to bad, but we only have about 15 VMs on the host servers. We will be adding much more as time goes on.

Which means the RTO for ESX is 50% of your justification it was originally. I am sure they looked at the consolidation ration around at least 4 or 6 vCPU per core, right? If that's the case, 15 VM's means LESS VM's per host, which means MORE hosts, which means they aren't getting the MOST out of it. All to find out in the end, that 2 vCPU per VM is negligible impact vs 1 vCPU. To me the solution is clear, this wagon has 1 horse, either its going as fast as it can or its not. Its not going to be faster if they keep stopping every mile to change horses.....

[nonu]

I agree with RParker 2vCPU per VM is negligible impact vs 1 vCPU.

I still can't understand.. what's the issue here.. is the performance an issue... is the consolidation an issue.. or it's the Spotlight :D....

I would go ahead and add the additional VCPU.. provided you have license for SMP.. and I think your Queue length for the processor would come down.. and at the same time.. ready times.. won't be impacted as much as you think... number of CPU is kind of thing which does not have any specific answer.. 1 is good or 2 or 4... all depends on the kind of application you are running.... we had an issue here.. where we got better performance on the 2VCPU system rather than one with 4VCPU... so does that make 4VCPU.. scenario.. impractical.. no..

It's just the indication that. app isn't as efficient to make use of 4VCPU... so you go with 2.. and in your case.. if CPU usage is less.. and process queue still keep on growing.. it does indicate me that.. adding an additional CPU won't hurt..

[burdweiser]

The servers I am virtualizing are IIS. They only use 49mhz on the physical box with 2CPU's. I made a copy of the physical ones to test with. The one with 1CPU had the processor queue lenght, but the second with 2CPUs did not (as you can see from my earlier post). But when I observe the IIS server with 2CPUs, the %RDY time is twice as much as the one with 1CPU (granted we are talking about small numbers, but these will grown over time with the addition of new VM's). I have similar boxes in production. I'm being told that we are planning on putting around 60 production server on these hosts, but I'm trying to balance things for growth beyond that. I really think we are going to end up with twice as much VM's.

Our test and dev clusters just got out of control with 4CPU servers all over the place. I'm trying to draw a line in the sand so the same thing does not happen to production.

[nonu]

All in all I can %rdy isn't everything.. esxtop provides lot of other useful information abut how 2VCPU Vm's are behaving..