I'm in the process of evaluating the fitness of vCops for a set of specific needs I have been given, and I'm struggling with one specific use case. I'm relatively new to vCops, so please, excuse my lack of deeper knowledge on the tool.
One of the requirements is analyzing performance data for a set of datastores as a group.
I have created a new "Resource Tag" which includes different values which identify each different group, for instance:
Resource Tag Name: StorageGroup
Possible values: Group1, Group2, Group3
I have tagged different datastores as belonging to either Group1, Group2, etc... and only to a specific group.
This way I can create dashboards based on the Group Tag as a whole, and for instance show Health score for a group with a "health status" widget on a custom dashboard. Or for instance, create heat maps, grouped by StorageGroup. So far so good.
BUT, now I need to be able to graph specific performance METRICS for the whole Group, and so far, I have been clueless as to how do it.
To clarify, for instance: as the resource tag groups only datastores, I want to be able to SUM "datastore->usage avegare" metric for all the datastores on a group, and graph that value as a group.
What is the best way to do this?
Thanks and kind regards.
Go in to the vSphere UI and create a custom group. Add the datastores to the group dynamic or statically (dynamic is preferable). Create the group type to be environment or whatever you prefer, it is mostly preference for organization.
Then go the Custom UI, you'll see the tag in the left. If you don't, go to manage resource kind tags and make that resourcekind visible (tag name).
You'll now have a container that contains all of your datastores that you can use as a tag filter on widgets and achieve what you're describing.
creating a dynamic group through the default "VCops vsphere-UI" interface, provides some additional aggregated metrics for each group, however it seems to miss some important ones, let me explain:
For a dynamic group of Datastores, vcops only shows a subset of grouped metrics under "All metrics->Datastore":
And some of the missing ones are key to what I intend to monitor, for instance: Disk Command Latency, Max. Observed Read/Write latency, Commands per second (IOPS), etc... btw, most of those metrics reside under "Devices" branch of the "all metrics" view for Datastore objects. That branch is missing from the dynamic datastore group "all metrics" view.
Is there a way to force vCops to include and aggregate those metrics in a datastore dynamic group as well?
Ideally, I would like to show total IOPS for a group of datastores, by adding up the individual IOPS of each datastore contained within.
You'd want to create a supermetric and apply it to this group. That would do the aggregation you're looking for.