I'm building a script to query the vSphere API for various critical performance metrics for CPU, MEM, Network and Disk performance metrics from vCenter.
I'll then use this data to set various thresholds to generate alerts and also graph performance.
My question is - should I query the API using the PerformanceManager methods or simply grab the "quick stats" data objects?
Here is what I found in this vSphere documentation: vSphere Documentation Center
"You can obtain near real-time summary information about performance or utilization without using the PerformanceManager methods. vSphere servers maintain “quick stats” data objects for hosts (HostListSummaryQuickStats), virtual machines (VirtualMachineQuickStats), and resource pools (ResourcePoolQuickStats). For more information about these objects, see the vSphere API Reference."
It seems that the PerformanceManager method provides far more "Performance Providers" than the "quick stats" as summarized below:
PerformanceManager >> Performance Providers:
Cluster Services
CPU
Management Agent
Memory
Network
Resource Scheduler
Storage Capacity:
Datastore / Virtual Machine
Storage I/O:
Datastore
Disk
Virtual Disk
Storage Adapter
Storage Path
System
Virtual Machine Operations
Quick Stats
Hosts
VMs
Resource Pools
Anyone grappled with this question before?
Thank you!
Amir