crash-diagnostics
crash-diagnostics copied to clipboard
[Feature] Ability to curl etcd health endpoints to return the metric histograms
On CAPI clusters it is possible to use the certificates on the control plane to query etcd metrics when troubleshooting performance with etcd. This could perhaps also run Storage I/O benchmarking the nodes to retrieve Disk I/O performance. This would be a good additional example
We could enhance it to calculate 99% disk latency
started a spike on this https://github.com/vmware-tanzu/crash-diagnostics/pull/227/files