While upgrading cluster from 4.11 to 4.12 , selinux was disabled on RHCOS node which failed upgrade.

    4.12 and 4.13

    I have tried upgrading a different cluster but could not reproduce this issue. We are basically looking for a root cause here as this issue seems weird to me.

1.   As mentioned above , customer simply upgraded the cluster. Master nodes were updated fine.
2. During mc rollout process on worker nodes , some of them got effected with this issue.
3. We checked that selinux was enabled before mco touches the node.
4. When it starts putting new configuration on the node , suddenly selinux gets disabled and we get to see below mentioned error with degraded mcp.
5. When we reboot this node , seliux gets enabled again and upgrade starts normally. Only 3-4 nodes got effected with this issue.

E0329 17:03:04.881721 1182865 writer.go:200] Marking Degraded due to: changing selinux context on path /run/mco-extensions/os-extensions-content-1062585222: error running chcon -R -t var_run_t /run/mco-extensions/os-extensions-content-1062585222: chcon: can't apply partial context to unlabeled file 'yp'
chcon: can't apply partial context to unlabeled file 'tmp'
chcon: can't apply partial context to unlabeled file 'debug'
chcon: can't apply partial context to unlabeled file 'rhsm'
chcon: can't apply partial context to unlabeled file 'mail'
chcon: can't apply partial context to unlabeled file 'lpd'
chcon: can't apply partial context to unlabeled file 'spool'
chcon: can't apply partial context to unlabeled file 'run'
chcon: can't apply partial context to unlabeled file 'preserve'

    Upgrade failed because selinux was disabled on RHCOS node.

    Upgrade should not be failed and selinux issue should not be there