We recently upgraded to OCUM 9.5 in preparation for an ONTAP upgrade to 9.5 , in Grafana since then we have noticed and only pretty much once a day that we see an astronomical spike to the 10's of thousands on CPU and Kahuna. Feels like bug in Grafana, we use Harvest 1.4.2 and NMSDK 9.5. Currently running ONTAP 9.3 which I was told was compatible with the Harvest version.
Just wondered if anyone had see this or identified as a bug?
Under the Node Dashboard, the System Utilisation panel shows Kahuna and CPU going into tens of thousands of percent. This is also shown under the CPU Layer Drilldown (Top CPU domains) under the same dashboard where wafl_exempt, nwk_exempt, exempt, raid_exempt and hostsOS all shows tens of thousands of percent utilisation. We dont see any spike in OCUM which matches this so feels like a bug in the stack elsewhere, maybe Harvest or NMSDK??
We use NABOX with the latest Harvest, NMSDK, and OCUM versions.