some cluster nodes needed to be restarted, the scheduling did not work in all cases, so a few services were offline.

Now all services recovered.