I am trying to pull the CPU, Memory averages and maximums for the previous day then display the data for our management team. The numbers I am getting through power CLI are not adding up to what we are seeing in vCenter.
vCenter Version: 6.5.0 Update 3
Powershell Core Version: 6.2.3
Powershell Module Versions:
- VIM: 6.7.0.13964812
- VimAutomation.Cis.Core: 11.3.0.13964830
- VimAutomation.Common: 11.3.0.13964816
- VimAutomation.Core: 11.3.0.13964826
- VimAutomation.Sdk: 11.3.0.13964823
Data Snippets
CPU Maximum
Query
`get-stat -Entity $cluster -Stat cpu.usage.maximum -Start $yesterday | select MetricId,Timestamp,Value`
Returned Value
MetricId Timestamp Value vCenter
-------- --------- ----- -------
cpu.usage.maximum 10/31/19 9:00:00 AM 239.31 55.89
cpu.usage.maximum 10/31/19 8:30:00 AM 224.28 54.43
cpu.usage.maximum 10/31/19 8:00:00 AM 226.17 49.01
cpu.usage.maximum 10/31/19 7:30:00 AM 228.15 50.84
cpu.usage.maximum 10/31/19 7:00:00 AM 212.67 49.56
CPU Average
Query
`get-stat -Entity $cluster -Stat cpu.usage.average -Start $yesterday -IntervalMins 30 | select MetricId,Timestamp,Value`
Maximum
MetricId Timestamp Value vCenter
-------- --------- ----- -------
cpu.usage.average 10/31/19 9:00:00 AM 110.01 27.45
cpu.usage.average 10/31/19 8:30:00 AM 120.09 24.77
cpu.usage.average 10/31/19 8:00:00 AM 109.24 26.12
cpu.usage.average 10/31/19 7:30:00 AM 110.2 24.80
cpu.usage.average 10/31/19 7:00:00 AM 114.12 24.15
We have statics level 4 turned on for 5 and 30 minute data.
Only other way I can think to get the data that might be closer to reality would be to get the average of both sockets for each host in the cluster, then take those numbers and get the average of them, and report that at the CPU average and maximum for the entire cluster
