Originally from the User Slack
@Andrey_Kojevnikov: Hi.
I’ve got a node stuck in DJ state as it was destroyed during join.
There is no way to conveniently replace or remove it, as it is both “exists” and “bootstrapping”.
Is there a way to fix this?
Seems like the only relevant stuff in docs is under “Manual Recovery Procedure”, which I really would like to avoid.
The answer is “upgrade to 6.2.3”
https://github.com/scylladb/scylladb/issues/20082
GitHub: node could stay in gossip forever if node start gossiping but topology coordinator didn’t get request to bootstrap · Issue #20082 · scylladb/scylladb