Migrate cluster policy storage is broken

[deads2k]

openshift/openshift-ansible#4476 exposed this, but merged anyway.

[mfojtik]

bumping to P0 as this is now merge/test queue blocker. @smarterclayton FYI

[mfojtik]

@smarterclayton is somebody working on this? i see that last change to migrate storage was you :-) This is failing for every single test :-(

@sdodson I thought the revert will fix this.

[mfojtik]

also the message changed slightly:

Output to stdout: summary: total=0 errors=0 ignored=0 unchanged=0 migrated=0 error: exited without processing all resources: Get https://ip-172-18-1-109.ec2.internal:8443/apis/batch/v1/jobs: dial tcp 172.18.1.109:8443: getsockopt: connection refused Output to stderr: error: exit directly

seems like the server died?

[smarterclayton]

@enj is working on this, yes

[smarterclayton]

Check the logs generated by your start - last time it was the PR, not this code. This just happens to be the first thing that runs after the master starts.

[mfojtik]

that one seems to be me:

Error starting "openshift.io/sdn" (User "system:serviceaccount:openshift-infra:sdn-controller" cannot get clusternetworks at the cluster scope)

fun.

[enj]

Fixed by #14764

[deads2k]

@sdodson the storage update should be fixed now.