Disable validation-webhook daemonset for managed clusters #361
Conversation
smalleni
force-pushed
the
managed_validation_webhook
branch
from
3d25801 to
abae979
Compare
smalleni
force-pushed
the
managed_validation_webhook
branch
3 times, most recently
from
c6da1e9 to
c6dc386
Compare
smalleni
force-pushed
the
managed_validation_webhook
branch
3 times, most recently
from
49d5f6d to
e18cd8f
Compare
smalleni
force-pushed
the
managed_validation_webhook
branch
from
e18cd8f to
8bfb31f
Compare
When running workloads on managed clusters with cluster-admin user, we are seeing an error like "admission webhook "regular-user-validation.managed.openshift.io" denied the request" when trying add a label to a node directly by using oc label node. The recommended way to add labels to nodes on managed ROSA clusters is by editing the machinepool. However, the Default machinepool cannot be edited to add labels and we see an error like "Labels cannot be updated on the Default machine pool". The only way to add a label is by disabling the validation-webhook daemonset and thereby admission control in the openshift-validation-webhook project that only exists on managed services clusters. We disable the daemonset by adding a fake nodeSelector before labeling the nodes and remove the nodeSelector after unlabeling the nodes. Adding a nodeSelector on top of the existing nodeAffinity means that both the conditions needs to be met for a pod to be scheduled. Also by adding the nodeSelector, the spec is not overwritten during reconcillation whereas changes to nodeAffinity are being overwritten. This change4 is important for managed clusters as we don't always have access to kubeconfig (when running in prow for example). Signed-off-by: Sai Sindhur Malleni <[email protected]>
smalleni
force-pushed
the
managed_validation_webhook
branch
from
8bfb31f to
5280bc7
Compare
curious, why do we need to label nodes? I assume this is used for the old node-density implementation available in e2e-benchmarking, the implementation based on the kube-burner's OCP wrapper doesn't need to label nodes anymore. Rather than keep fixing and updating the old implementation we should encourage users to use the new implementation |
| download_binary(){ | ||
| KUBE_BURNER_URL=https://github.com/cloud-bulldozer/kube-burner/releases/download/v${KUBE_BURNER_VERSION}/kube-burner-${KUBE_BURNER_VERSION}-Linux-x86_64.tar.gz | ||
| curl -sS -L ${KUBE_BURNER_URL} | tar -xzC ${KUBE_DIR}/ kube-burner | ||
| } | ||
|
|
||
| check_managed_cluster |
There was a problem hiding this comment.
Changes on this file are not required as I stated in my comment
Oh, I didn't realize that was the case, thanks for surfacing that. |
When running the node density workloads on managed clusters, starting
with 4.10.5 on ROSA, we are seeing an error like "admission webhook
"regular-user-validation.managed.openshift.io" denied the request" when trying add a label to a
node directly by using oc label node. The recommended way to add labels
to nodes on managed ROSA clusters is by editing the machinepool.
However, the Default machinepool cannot be edited to add labels and we
see an error like "Labels cannot be updated on the Default machine pool".
The only way to add a label is by disabling the
validation-webhook daemonset and thereby admission cotnrol in the
openshift-validation-webhook project that only exists on managed services
clusters. We disable the daemonset by adding a fake nodeSelector before
labeling the nodes and remove thenodeSelector after unlabeling the nodes.
Adding a nodeSelector on top of the existing nodeAffinity means that both
the conditions needs to be met for a pod to be scheduled. Also by adding
the nodeSelector, the spec is not overwritten during reconcillation whereas
changes to nodeAffinity are being overwritten.
Signed-off-by: Sai Sindhur Malleni [email protected]
Description
Fixes