Closed gohilankit closed 2 months ago
/ok-to-test
/approve
[APPROVALNOTIFIER] This PR is APPROVED
This pull-request has been approved by: deepakkinni, divyenpatel, gohilankit
The full list of commands accepted by this bot can be found here.
The pull request process is described here
What this PR does / why we need it: This PR is increasing CSI attacher worker threads in Supervisor manifests to 100.
Which issue this PR fixes (optional, in
fixes #<issue number>(, fixes #<issue_number>, ...)
format, will close that issue when PR gets merged):CSI provisioner worker threads are set to 100 by default, but CSI attacher worker threads are set to only 25 for Supervisor. This results in volume provisioning at a faster rate but they can't be utilized in pods at the same rate. We should have same worker threads for create/delete and attach/detach operations.
We have seen issues during failure testing at scale that pods remained in ProviderFailed. Increasing the number of attacher threads have helped resolve such issues.
Testing done: Pre-checkin pipelines were run after changing the attacher threads. Most of the tests were successful except 2, which are not related to this change.
Release note: