At the moment I need to check every graph for each VM or datastore to analyze the IOPS.
To determine an average load for an array I have to sum up all numbers from all datastores or VMs and I also have to check if (peak) loads happened at the same time.
If I would simply add all numbers without checking if the load/peak happened at the same time, the data/numbers would be wrong.
What we would love to see is a kind of stacked graph for all VMs or datastores.
This would tell us that our datacenter needs avg. xxx IOPS and we got peak loads of round about xxxx IOPS.
This numbers would help to size the backend properly.
At the moment it looks like this:
One stacked graph would help to get a much better overview.