Description of problem: Critical daemonsets don't have a tolerate everything toleration which looks like this: ~~~ tolerations: - operator: "Exists" ~~~ This includes the sdn, ovs daemonsets in the openshift-sdn project, the sync daemonset in the openshift-node project and the fluentd daemonset in the logging project. Version-Release number of the following components: All 3.10 branch How reproducible: Always Steps to Reproduce: 1. oc taint node <node name> ataint=tainted:NoSchedule Actual results: Node is tainted, critical pods won't run Expected results: All critical daemonsets run on every node Additional info: https://bugzilla.redhat.com/show_bug.cgi?id=1635462 - openshift-sdn fixed in 3.11 https://bugzilla.redhat.com/show_bug.cgi?id=1635804 - openshift-sync pod fixed in 3.11
3.11 bug for fluentd ds https://bugzilla.redhat.com/show_bug.cgi?id=1685970
Thanks, I know we added this toleration to control plane pods and I think we discussed adding them to SDN and others. I'll see if I can find the outcome of that discussion.
https://github.com/openshift/openshift-ansible/pull/11309
*** This bug has been marked as a duplicate of bug 1685952 ***