#qos metrics, default.yaml - nabox 3.3

1 messages · Page 1 of 1 (latest)

terse robin
#

Hi, we are having issues getting workload metrics graphed since an upgrade to 23.11.01 and going from NABox 3.1 to 3.3.

#

We have default.yaml editted as follows:

#
  Workload:             workload.yaml
  WorkloadVolume:       workload_volume.yaml

# The following workload templates may slow down data collection due to a high number of metrics.
  WorkloadDetail:       workload_detail.yaml```
#

And custom.yaml shows

#
objects:
  Volume: custom_volume_blacklist.yaml
  Workload: workload.yaml
  WorkloadDetail: workload_detail.yaml
  WorkloadVolume: workload_volume.yaml
  WorkloadDetailVolume: workload_detail_volume.yaml```
#

I can see in the harvest-2 container that this is also set after restarting the container, logs show the metrics being collected fine but no metrics visible in grafana

#

2023-11-16T13:07:00Z INF collector/collector.go:510 > Collected Poller=***** apiMs=212 calcMs=0 collector=Zapi:QosWorkload instances=656 metrics=1968 parseMs=9 pluginMs=0

dense totem
#

@terse robin Do you see logs containing ZapiPerf:WorkloadDetailVolume or ZapiPerf:WorkloadDetail?

terse robin
#

yes

dense totem
#

Did you reset the dashboards after upgrading to 23.11?

terse robin
#

no we did not

dense totem
#

Could you please do that and recheck?

terse robin
#

Why do I want to reset the dashboards? I have lots of bespoke ones I don't want to loose

#

Will I loose them?

dense totem
#

Yes, with Reset you'll loose any changes done to Harvest shipped dashboards.

#

From which version of Harvest did you upgrade to 23.11?

terse robin
#

I'd rather not loose them :), it was 19.10

dense totem
#

We made changes to workload metrics in 23.08 which requires dashboard reset for NABox.

terse robin
#

I'm happy to snapshot the VM, reset the dashboards see if it fixes it and then revert

#

Will it just impact the dasborads in Harvest cDot?

#

We have other folders where we've move our editted dashboards to from previous iterations

dense totem
#

You can move existing dashboards to a different folder and then try Reset.

terse robin
#

yeah we've already done that. I'll test now

dense totem
#

Great!

terse robin
#

that fixed it!

#

I always like a quick fix. Many thanks

dense totem
#

Nice!

terse robin
#

Whilst I have you both do you know what the metric Latency from Suspend is actually capturing?