Kubernetes namespace set in the UI isn't respected
We had our clusters created prior to 11.11 and https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/26565/diffs, meaning that the cluster is "GitLab-managed", but we're using a single namespace which is set in the UI, e.g. (we have a similar configuration for gitlab-ce
and gitlab-ee
).
Following a recent deploy, probably the one from yesterday:
deployerAPP [12:00]
deployer: Marin Jankovski is starting a deploy pipeline of 12.0.201906051128-30e31e4afb1.bd6aadb8c50 on gprd
Our review-deploy
jobs started trying to create their own namespaces, e.g. https://gitlab.com/gitlab-org/gitlab-ee/-/jobs/228719825:
** Checking that Kube domain exists... **
** Ensuring the gitlab-ee-278964 namespace exists... **
Name: gitlab-ee-278964
Labels: <none>
Annotations: <none>
Status: Active
No resource quota.
No resource limits.
when it was looking like the following before (https://gitlab.com/gitlab-org/gitlab-ce/-/jobs/213301203):
** Ensuring the review-apps-ce namespace exists... **
Name: review-apps-ce
Labels: <none>
Annotations: <none>
Status: Active
No resource quota.
No resource limits.
I believe this is wrong since we've explicitly declared the namespace that we want to use in the cluster settings?
I worked around that by creating the KUBE_NAMESPACE
CI/CD variable for both projects, according to
If you are using Auto DevOps, you will need to explicitly provide the KUBE_NAMESPACE deployment variable that will be used by your deployment jobs, otherwise a namespace will be created for you.
– https://docs.gitlab.com/ee/user/project/clusters/#gitlab-managed-clusters
That being said, this looks like a regression to me as Review Apps deployment were broken without anything changing on our end-user side.
/cc @mayra-cabrera