07 January 2013 Karim Elatov

From “vSphere Monitoring and Performance”, here are the CPU counters:

esxtop cpu counters p1 VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Here are more of CPU counters:

esxtop cpu p2 VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

From the same pdf, here are the memory counters:

esxtop mem p1 VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Here are more counters from the same page:

esxtop mem p2 VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

And here is the last page for the memory counters:

esxtop mem p3 VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

From this Communities page:

perf charts cpu counters VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

and here are the memory counters:

perf charts mem counters VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

You can also check out the full list here.

Troubleshoot ESXi host and Virtual Machine CPU performance issues using appropriate metrics

From this pdf:

esxtop cpu performance VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Here is a good summary:

%RDY: Percentage of time a VM was waiting to be scheduled. Possible reasons: too many vCPUs, too many vSMP VMs or a CPU limit setting (Trouble when >10) %CSTP: Percentage of time a ready to run VM has spent in co-descheduling state (Trouble when > 3 to resolve, decrease number of vCPUs on the VM) $MLMTD: Percentage of time a ready to vCPU was no schedules because of a CPU limit setting (Trouble when > 1, to resolve; remove the CPU limit) $SWPWT: How Long a VM has to wait for swapped pages read from disk. (Trouble when > 5, possible reason; memory overcommitment

Also this is from a 4.x document but it applies to 5.0 as well. Here is the document: “Performance Troubleshooting for VMware vSphere 4.1”. From that document:

check for host cpu saturation VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Here is a similar diagram for Guest CPU saturation:

check for guest cpu saturation VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Troubleshoot ESXi host and Virtual Machine memory performance issues using appropriate metrics

From this pdf:

esxtop memory issues VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Here is a good summary:

MCTLSZ: Amount of guest physical memory (MB) the ESXi Host is reclaiming by balloon driver. (Trouble when > 1) ZIP/s: Values larger than 0 indicates that the host is actively compressing memory (Trouble when > 1) UNZIP/s: Values larger than 0 indicate that the host is accessing compressed memory (Reason for this behavior is memory overcommitment) SWCUR: Memory (in MB) that has been swapped by VMkernel (Possible cause is memory overcommitment, Trouble when > 1) CACHEUSD: Memory (in MB) compressed by ESXi Host (Trouble when > 1) SWR/s SWW/s: Rate at which the ESXi host is writing to or reading from swapped memory (Trouble when > 1)

And from “Performance Troubleshooting for VMware vSphere 4.1”:

check for vm swapping VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

and here is another flow chart for VM memory issues:

check vm fo compression VCAP5 DCA Objective 6.2 – Troubleshoot CPU and Memory Performance

Use Hot-Add functionality to resolve identified Virtual Machine CPU and memory performance issues

If you discover then the VM needs more resources, then follow the instruction laid out in VCAP5-DCA Objective 3.2 to hot add memory or CPU


blog comments powered by Disqus