Volume / Volume Deep Dive Dashboard Latency Panel Description

[tsohst]

A note for the community

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please let us know in a comment

Problem

All latency panels in the Volume / Volume Deep Dive Grafana Dashboard are having the same description text.
Is there any ressource to get more info whats the meaning of these charts?

Configuration

-

Poller

Version

Poller logs

-

OS and platform

ONTAP or StorageGRID version

Additional Context

References

[rahulguptajss]

@tsohst Similar panels at the workload level are available in the Workload dashboard under Row Latency Breakdown and have descriptions. Could you check if those look good? If they do, we will update similar descriptions in these panels as well.

[tsohst]

Hello @rahulguptajss
these descriptions look better to me and provide more context to it.

I'm just confused why are they in % not in ms?
What is the % referred to?

[rahulguptajss]

Hello @rahulguptajss these descriptions look better to me and provide more context to it.

I'm just confused why are they in % not in ms? What is the % referred to?

The workload dashboard shows these as percentages. It indicates how much % of the total workload latency is consumed by a relevant subsystem. Do you think it would be better to have the percentage in the volume dashboard as well?

[tsohst]

Do you think it would be better to have the percentage in the volume dashboard as well?

Hm.. I'm looking for a quick and easy dashboard to have latency / performance issues in one place.
Like you know when someone comes around asking if there are any performance issues currenly I can just refer to that.
Like an flowchart or drilldown Cluster/Network/Disk [...] Volume/LUN maybe to see where the issues is or if there is a bottleneck somewhere in the cluster.

Yesterday I was also asked if we can see latency based on client connections?
This was for a SAN environment with multiple VMware ESXI Hosts.

[rahulguptajss]

Do you think it would be better to have the percentage in the volume dashboard as well?

Hm.. I'm looking for a quick and easy dashboard to have latency / performance issues in one place. Like you know when someone comes around asking if there are any performance issues currenly I can just refer to that. Like an flowchart or drilldown Cluster/Network/Disk [...] Volume/LUN maybe to see where the issues is or if there is a bottleneck somewhere in the cluster.

Yesterday I was also asked if we can see latency based on client connections? This was for a SAN environment with multiple VMware ESXI Hosts.

We do have performance panels for each object in their respective dashboards. We have also tried to create some high-level overviews through the cDoT and Datacenter dashboards in Harvest. We could potentially add more object performance panels there or create a new dashboard that consolidates all object-relevant performance panels in one place. Would that help?

We have top client-related issues opened here: Issue #2591 and Issue #2118. Could you please +1 these issues if they match your requirements, or open a new one if they do not?

[tsohst]

Thank you for your help!

[rahulguptajss]

@tsohst Descriptions have been added to the dashboards. You can either install the nightly build or simply import the dashboard from here.

Descriptions have also been added to the metric documentation, which you can find here.

[rahulguptajss]

verified in 24.08