VMware Cloud Community
jcossota_san
Contributor
Contributor
Jump to solution

vROps 6.6 - Total Capacity value in Capacity Remaining WRONG (Projects also busted?)

When I select Cluster > Analysis > Capacity Remaining the Total Capacity value is incorrect for both CPU and Memory. 

Under Workload the Capacity value is accurate.  For instance:

vrops_workload_tc.jpg

This is incorrect, which skews calculations (of course I'm under tight deadline for compute procurement):

vrops_capremain_tc.jpg

Additionally, the Projects page doesn't seem to be correct either.  It shows the most constrained resource for the above selected cluster is CPU at 96% demand!  As you can see above Memory is the most constrained resource.

I have an SR open just wanting to see if I'm looking at something wrong, my policy is busted, or if anyone else has come across this.

Reply
0 Kudos
1 Solution

Accepted Solutions
jcossota_san
Contributor
Contributor
Jump to solution

Turns out the Total Capacity metric under the Capacity Remaining tab is an average.  Some clusters experienced events where hosts were either disconnected or unavailable, and this affected the Total Capacity metric.  Capacity under the Workload tab represents the real-time value.  Not ideal when you are performing capacity planning using current data, but there's a way to even it back out.

VMware suggested that I modify the policy for the Cluster Compute Resource adapter by changing the Capacity Calculation setting from Current to Trend.  This balanced my Total Capacity metric back out.

vrops-policy.jpg

View solution in original post

1 Reply
jcossota_san
Contributor
Contributor
Jump to solution

Turns out the Total Capacity metric under the Capacity Remaining tab is an average.  Some clusters experienced events where hosts were either disconnected or unavailable, and this affected the Total Capacity metric.  Capacity under the Workload tab represents the real-time value.  Not ideal when you are performing capacity planning using current data, but there's a way to even it back out.

VMware suggested that I modify the policy for the Cluster Compute Resource adapter by changing the Capacity Calculation setting from Current to Trend.  This balanced my Total Capacity metric back out.

vrops-policy.jpg