Hide Forgot
Description of problem: During rpm upgrade, although there is no changes with etcd configure, the etcd service was still restarted, The upgrade should only restart service when there are packages or configuration changes. Version-Release number of selected component (if applicable): openshift-ansible-3.3.35-1.git How reproducible: always Steps to Reproduce: It is easy to reproduced by run upgrade playbook against the latest OCP. In Production Environment, the step is as below: 1. Install OSE 3.2 with external etcd. 2. Upgrade to OCP 3.3 ansible-playbook /usr/share/ansible/openshift-ansible/playbooks/byo/openshift-cluster/upgrades/v3_3/upgrade.yml 3. Check the etcd service status. Actual results: etcd service was restarted. Expected results: etcd service won't be restarted Additional info:
Now that we're upgrading etcd on each upgrade it's incredibly likely that we don't actually need to restart etcd. Marking UpcomingRelease
I have observed unnecessary restart during v3.5 and v3.5 testing. Could we say the bug is fixed?
Anping, you mean you have or have not observed unnecessary restart?
If not, then yes, lets CLOSED CURRENTRELEASE.
Scott, Yes