Hi, should it be possible to use “allow_service_cpu_overcommit” in our self-managed Scaleway cluster? We did everything by documentation, also changed this field on cluster’s advanced settings, but when click “Update” to apply changes, get this error. We see in Kubernetes dashboard, that overcommit changes are not applied in our services.
@ce_gagnaire it seems all “Advanced settings” are unusable with self managed clusters. Should we manage the advanced settings via helm chart? We couldn’t find relevant info in the documentation.
@Qovery_Team Could you guys take a at this? This CPU overcommit feature is very important to our infrastructure and we can’t move further without it being solved.
Hello,
It is indeed a bug on our side, we don’t allow modifying cluster advanced settings for self-managed cluster, but it is needed to allow CPU overcommit.
We are going to fix that in the coming days
2 Likes
Hello,
The issue is fixed. You need to update your engine version running on your cluster to this one public.ecr.aws/r3m4q3r9/engine:12c88f5
Regarding the step to enable cpu over commit.
- Enable cpu/ram over commit in the cluster settings. Only saving is enough, no need to update your cluster (and you can’t as you are on a self-managed one).
-
Once enabled at cluster level, go into your application advanced settings, and replace the “null” by the value you want. (it needs to be above the requests)
-
Re-deploy your service, and the new limit should be set
Be sure to udpdate your engine version to public.ecr.aws/r3m4q3r9/engine:12c88f5
2 Likes
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.