Add default requests and limits for ocs-operator pods #3339
Conversation
|
Skipping CI for Draft Pull Request. |
openshift-ci
bot
added
the
do-not-merge/work-in-progress
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: mrudraia1 The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
mrudraia1
force-pushed
the
req-lim
branch
3 times, most recently
from
b49ab89 to
1a61568
Compare
openshift-ci
bot
removed
the
do-not-merge/work-in-progress
Signed-off-by: Mrudraia1 <[email protected]>
| }, | ||
| Limits: corev1.ResourceList{ | ||
| "memory": resource.MustParse("1.5Gi"), | ||
| "cpu": resource.MustParse("1m"), |
| "cpu": resource.MustParse("250m"), | ||
| }, | ||
| Limits: corev1.ResourceList{ | ||
| "memory": resource.MustParse("1.5Gi"), |
There was a problem hiding this comment.
Why does the exporter go till 1.5Gi, seems a lot.
There was a problem hiding this comment.
We have seen memory consumption reaching till 1.22Gi for performance, scale tests ran during testing, therefore setting a safe limit of 1.5Gi
There was a problem hiding this comment.
Isn't the exporter stateless? In your tests what made the mem increase and maybe we have mem leaks?
There was a problem hiding this comment.
I think it is stateless 🤔
The tests comprised of IO to fill up the cluster, PVC creation, deletion, PVC expansion, small file IOs and such.
Though, this was the only set of tests where we observed such a hike. In the other cases, the maximum consumption was ~ 700Mi.
There was a problem hiding this comment.
Need a bit more investigation, I understand even w/o this PR the exporter has no bounds but we better take this moment to re-look at the exporter and come to a conclusion for bumping the values.
There was a problem hiding this comment.
There was a problem hiding this comment.
Considering we don't have any ideas as to why we are seeing a spike, I would suggest to take this PR in and work on parallel to identify any memory leaks or such issues that could lead to a memory spike in exporter.
This PR adds the default requests and limits for the ocs-operator pods.
The default metrics are set for the following pods