vRop challenage

[djciaro]

Hi Guys,

We have been using vCenter operations manager since version 5.0.

Use case is for troubleshooting performance issues, dashboard for heatmaps and reports for an oversized and unused VMs (we have a month sprawl process) Our environment is very stable and we have more than enough resources available to cater for the number of VMs.

Management has asked for the following:

- establish usage thresholds based on market best practices applied to our architecture.

- build reports to report on usage vs thresholds.

- try to isolate usage of prod from the overall usage.  (dev/test and prod are on the same clusters - separation is at resource pool level)

For monthly reporting use a custom dashboard looking at CPU demand and memory usage (at cluster level).

I am looking for ideas / suggestions to meet the requirements of management.

Thanks

Ciarán

[mark_j]

I'll typically use Symptoms to reflect compliance with BPs. Usage vs thresholds, though.. you could use SMs to track hard thresholds, and create another SM to indicate the variance of usage vs threshold SM. That gets messy to report/visualize if the objects in your environment don't all align to the same BPs. At a certain point, though, you start drifting in to the capacity modeling that vR Ops does natively. I'd leverage the vR Ops cap modelling first (e.x. cap remaining), then 'maybe' use SMs to track the variance on items that aren't already part of each MP's capacity model.