• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

esxi 5.1 Guest/Host CPU oddness

XOR != OR

[H]F Junkie
Joined
Jun 17, 2003
Messages
11,547
Morning folks,

I'm working on a rather strange issue and would appreciate some suggestions on where else to look.

ESXi host: 1 CPU, 8 cores, 24gigs.

Symptoms
Guest OS ( linux ) load avg hovering ~3-4. Services are impacted, although not to a horrible degree. Host reports show the guest using < 500Mhz of the processor, never going above that ( processor is 2.1ghz ).

Guest has a single vCPU assigned to it. Memory isn't over committed. Disk stats are normal to low ( I've seen spikes as high at ~4 CMDS/s ), latency overall is minimal ( < 1ms ).

Currently all other guests are shut down.

I've looked at various CPU/Memory/disk subsystem reports in the vSphere client. Nothing is jumping out at me.

Other ideas on where else to look would be great.
 
I should probably know what metric load avg 3-4 is referring to in linux but...i dont.

If you power on another VM with similar allocations (Perhaps a windows VM) does it behave similarly?

It doesn't have a CPU limit does or belong to a resource pool with a limit?
 
I should probably know what metric load avg 3-4 is referring to in linux but...i dont.
It's simply the number of processes that are waiting on the CPU.
If you power on another VM with similar allocations (Perhaps a windows VM) does it behave similarly?
Yes, windows guest similarly spec'd presents with the same symptoms.
It doesn't have a CPU limit does or belong to a resource pool with a limit?
Nope, checked about a dozen times.
 
Edit, reading 101, you have no other VMs running so that rules out what I said. Anyways, I'm curious how the VM stats read in esxtop. yellow-bricks.com has an esxtop guide and breakdown of thresholds that are accepted norms.
 
What kernel is the guest using (SMP, bigmem, normal?) and what distro?

64bit or 32?
 
Any power saving features enabled on the host? Set it for high performance and test. The board may think the CPU is idle since only one core is being used.


http://kb.vmware.com/kb/1018206
Quoting because this ended up being the problem. When I setup the server initially, I had set both the bios and the OS power saving to "performance". I got to thinking about the issues I was seeing, and check esxi's powersaving setting, which was fine. When we finally did a reboot we found that the bios power settings had been set to "balanced". Set it back to performance, everything cleared up.

Thanks for the input folks!
 
Back
Top