gellner
commented
5 years ago Looks like this was resolved in https://github.com/UKCloud/openshift-deployment-ansible/pull/133
Closed gellner closed 5 years ago
gellner
commented
5 years ago Looks like this was resolved in https://github.com/UKCloud/openshift-deployment-ansible/pull/133
The iptables rules added in allserverspostdeployment don't persist over reboot.
In particular, this has caused private-routers on worker-infra nodes to become inaccessible when it is rebooted. When only only worker-infra is rebooted, this makes the environment prone to losing private-router connectivity when one worker-infra dies.
At this time it looks like the only option is to run iptables-save.