I have come across an issue with on of the usage meters where it was unable to connect to the vCenter systems for 2 and half days in a month.
Question is, is the Avg Capped Billed vRAM (GB) that is reported in the monthly usage report valid or will this be short?
Yes, but only assuming everything is steady-state (same VM's running) - I will provide an example for you.
Let's say you missed 2 days of collections, of 48 hours in a 30 day period. For the remaining 28 days (or 672 hours), I had a monthly report of 1,000GB of Advanced SP Bundle.
I can then extrapolate this out on an hourly basis - 1,000GB over a 672 period is 1000/672 = 1.488GB vRAM per hour.
Therefore, over a 30-day full month, or 720 hours, I would take 1.488GB * 720 = 1,071.36, or 1,071 since we do not round up in Usage Meter.
Also, you can see how many collections you missed under the Report tab on the UM Portal.
Thanks,
-Daniel
Since collection is on an hourly basis and we average out the usage over 30 or 31 days (720 or 744 hours), it will be short.
Sure, makes sense.
Is there a way that I could manually workout what the shortfall is? Or would that be too complex?
Yes, but only assuming everything is steady-state (same VM's running) - I will provide an example for you.
Let's say you missed 2 days of collections, of 48 hours in a 30 day period. For the remaining 28 days (or 672 hours), I had a monthly report of 1,000GB of Advanced SP Bundle.
I can then extrapolate this out on an hourly basis - 1,000GB over a 672 period is 1000/672 = 1.488GB vRAM per hour.
Therefore, over a 30-day full month, or 720 hours, I would take 1.488GB * 720 = 1,071.36, or 1,071 since we do not round up in Usage Meter.
Also, you can see how many collections you missed under the Report tab on the UM Portal.
Thanks,
-Daniel
Thank you Daniel, that's exactly what I had done so pleased that I had this correct.
Luckily the environment has been static for a little while now.
Thanks, Ben.