This one is stumping me. I opened a low priority ticket to get some official info, but i figured i would ask here to see if anyone else has had a similar experience:
Background: we received an alert from prism central that one of our cluster's was going to run out of CPU soon. I checked in the hypervisor (esxi) and in prism element and in central under analysis but i could not find an issue. Then, i checked planning and it shows us running out of CPU very soon.
At first i thought the initial load up was skewing the data, but when i looked closer, i noticed that the value planning was using for our current usage was at 50Ghz, when the hypervisor is reporting 12Ghz (cluster of 3).
I pulled all of the other cluster's planning data and all of them have the same problem, CPU usage is being greatly over reported on the planning chart.
Charts under analysis agree with the hypervisor, but these are looking at % usage, not hz.
What gives? is there a bug in how planning is calculating CPU usage? or perhaps my hypervisor is under reporting by a factor of 3? (seems unlikely)
Thanks for letting us know. Can you send me the support ticket? I will let our engineers look at it.
Can you tell us what is the version of your PC and ncc? We did discover a bug in ncc 3.1.2 and PC 5.1.3 that is very similar to what you observed. Just wanted to confirm.
case number 00254576. Low priority, we aren't at risk or anything. mostly, either I am reading the data wrong or planning is calculating wrong (or somwhere in the middle). either way, we aren't seeing any impacting issues at the moment, so take your time.
@penguindows, I think that is the same bug. All you need to do is to upgrade to ncc 3.5 in PC or you can upgrade your PC to 5.5. Both were released yesterday.