Exclude VM's from efficiency (reclaimable waste) statistics

[tomtomv]

Hi all,  I have several VM's which I am contractually obliged to give a certain amount of vCPU and vRAM to.  As a result, these oversized VM's skew my reclaimable waste figures.  Is there a way of excluding VM's from efficiency monitoring, whilst still monitoring the usual health and risks?  How do others deal with this scenario where you cannot reclaim the waste?

Appreciate your thoughts

[mark_j]

In the policy that you're apply to these resources, set this in 4b. VMs are oversized when: , then set the value to more than "100" %. Done.. now those resources won't be considered in the oversized VM analysis nor will they be considered waste from a CPU/Mem perspective..

[tomtomv]

Hi Mark, Thanks for the recommendation.  I see your logic and have created/applied the policy however I am not getting the expected results (I still see lots of reclaimable waste for the VMs in question).

Im using VCOPS 5.8, not sure if it has changed much since prior versions, but the policy settings I have are:

Amount of CPU demand below (5-95%): As I interpret this, then oversized means its using less than % of its resource, so I have set this to the lowest value of 5%

Amount of memory demand below (5-95%): As above

is more than (0-100%) for the entire range:  I interpret this as the amount of time over 30 days that the above conditions are present, therefore I have set this to the highest value of 100%

If someone could clarify/correct my assumptions that would be great

thanks

[tomtomv]

It seems that the reclaimable waste takes a few days to generate so its all looking good now.  Mark thanks for pointing me in the right direction.  I guess my assumptions above must be correct.  Thanks again.