slowness while getting /protocols/cifs/sessions

[nainfo]

Hello!

We are seeing slow response from multiple clusters while trying to GET /protocols/cifs/sessions.

At the same time, same clusters respond fine (quick) to other GET calls (ex volumes, aggregates, etc).

Is there a known performance issue with /protocols/cifs/sessions and are there any known workarounds?

Thank you,

Mark

[Ontapforrum]

Could you tell us - Ontap version you are running ? There are few cases related to slow RestAPI but I cannot locate anything specific to cifs sessions. Also, could you tell us how slow (in seconds) is querying cifs sessions compared to vol/aggr. This resource type (CIFS sessions) was introduced with the ONTAP 9.8 REST API and enhanced with ONTAP 9.9.

[nainfo]

Hello and thank you. We have tried it on 9.8p9, 9.10.1p10 and 9.10.1p12. The delay varies and on some clusters it could take hours to complete. On deeper look, it appears that REST API is somewhat equivalent to ‘cifs session show -active-volumes‘ and it is that -active-volumes part that makes it return with some (significant) delay. Even when directly running it via CLI, that command comes back slow, as it looks like it is doing some heavy lookups to get the active volumes.