Does anyone know why

spicedb-operator

would loop on this? I've previously seen the migration job fail due to github rate limiting or a bad DB uri string, but in those cases, I've always seen the migrate jobs and/or spicedbcluster pods come up and crash, where I could read failures in their logs explicitly. In this case, no spicedbclusters or migrate jobs ever materialize.

wait_for_migrations.go:34] migration job failed: Job has reached the specified backoff limit

pause.go:153] Operation cannot be fulfilled on spicedbclusters.authzed.com "spicedb-cluster": the object has been modified; please apply your changes to the latest version and try again
handler.go:31] "requeueing after api error" err="Operation cannot be fulfilled on 

spicedbclusters.authzed.com \"spicedb-cluster\": the object has been modified; please apply your changes to the latest version and try again" syncID="7rVSD" controller="spicedbclusters" obj={"name":"spicedb-cluster","namespace":"spicedb"}

Can you share the status and the labels on the SpiceDBCluster object? If the migration completely fails for some reason, the operator will "pause" that cluster. You should be able to unpause by removing a label on that object

Hm that's good to know, thanks! I just resolved the issue, though idk what led to it. There were no SpiceDBClusters at all, nor was there a migrate job even happening, which I guess is what this log was trying to tell me. I'm not sure what would lead

kubectl apply

to just not create the clusters? Luckily I put them in a

spicedb

namespace, which did get created. Deleting the empty namespace and reapplying everything did fix it.