Doc: Kibana upgrade failures resulting in alias not pointing to the latest .kibana* (and related) indices #80283
Description
We have received quite a few reports from the field related to saved objects migration failures (since 6.5) during Kibana upgrades, resulting in Kibana failures to start up:
Another Kibana instance appears to be migrating the index. Waiting for that migration to complete.
If no other Kibana instance is attempting migrations, you can get past this message by
deleting index .kibana_<N> and restarting Kibana.
While we are working on improving saved objects migration, it can be helpful to document tips that can help avoid such failures in the first place, as well as clarifying the steps for remediating the situation once the Kibana upgrade has failed.
Opening this doc issue to track per chat with @joshdover @rudolf . Thanks!
Edit:
Docs are now available at https://www.elastic.co/guide/en/kibana/current/upgrade-migrations.html