IOPS values ​​sometimes collected too high

[wooyoungAhn]

Hello

Sometimes, Harvest reports a total of 1 billion IOPS. However, this value is not actually used by the ONTAP Cluster.

Although I cannot provide a screenshot from the client at the moment, the reported value appears unreasonably high for a very short period of time, similar to the graph below.

Currently, the customer has over 120 cluster connected to a single Harvest system.
Could having too many pollers connected cause Harvest to report incorrect values?

[rahulguptajss]

@wooyoungAhn Could you share Harvest version?
We have fixed similiar issues in past.
#2544
#2769
#2605

[rahulguptajss]

@wooyoungAhn Any updates on this? Thanks.

[wooyoungAhn]

sorry i'm late
I see similar symptoms in the qtree_nfs_ops metric
Is this also subject to modification?

[rahulguptajss]

@wooyoungAhn This is not expected for qtree_nfs_ops.
Could you please upload the relevant poller logs to the following link?
https://upload.nabox.org/tuny-byto-repo

[wooyoungAhn]

I will upload it after the version update. Thank you

[rahulguptajss]

Please upload logs once you notice the issue with new version.