Is there a way to track and collect disk and CPU queue numbers for ESXi 7?
Something similar to 'System\Processor Queue Length' and 'PhysicalDisk\Avg. Disk Queue Length' WMI performance counters in Windows - which can be collected in e.g. SolarWinds.
Why: arguably a very good way to track whether an ESXi is correctly sized for the workloads running on it, and whether a particular ESXi is struggling.
Why I want it? Because some of the Windows VMs running on our ESXis occasionally report high queue numbers indicating the VMs are struggling, and I'd like to see these metrics for the underlying hypervisor, not just the VMs.
Note: it's not vmstat or sar - those don't appear to expose the needed metrics. Additional context - in this awesome answer by Phil Frost on SO.
Thank you!
Hi,There was a similar topic before, you can check it out. Solved: Get CPU Memory Network Disk Usage - VMware Technology Network VMTN
It's similar (as in, in the same general area or topic) - but is not the same, and does not answer my question, possibly with the exception of cpu.ready.summation statistic discussed there.
Thanks!