I'd say that cpu & memory reports for the Hosts, and at the Cluster level, would always be interesting...
E.G. memoryPressure is an nworks custom metric, that shows memory overcommittment (how much memory is allocated to VMs versus how much physical RAM in the Host systems).
cpuReady and cpuWait times in the VM CPU report can also point to resource problems in the hypervisor cpu scheduler.
And the Disk Performance History has not just I/O but also Latency counters, which can also show up bottlenecks in backend storage performance.
The two "Top N" Reports are also a really good way to quickly see which machines are hogging all the resources - both in terms of ESX Host, and VM. You can see the top users for any resource - cpu, memory, disk i/o and net i/o.
Below is the list of reports in the current 4.5 MP.
Hope this helps! Any questions let me know,
• CPU & Memory Performance History - ESX Clusters
• CPU Performance History - ESX Hosts
• CPU Performance History - Virtual Machines
• Disk Performance History - ESX Hosts
• Disk Performance History - Virtual Machines
• ESX Datastore Usage History
• ESX Host Configuration Report
• Virtual Machine Configuration Report
• Hardware Power Usage - ESX Hosts
• Hardware Temperature and Cooling - ESX Hosts
• Memory Performance History - ESX Hosts
• Memory Performance History - Virtual Machines
• Network Performance History - ESX Hosts
• Network Performance History - Virtual Machines
• Top N Resource Users by ESX Host
• Top N Resource Users by VM
• VM Loading - ESX Clusters
• VM Loading - ESX Hosts
Vice President, Product Management