[fluss-server] Fix replicasOnOffline deadlock in addFetcherForReplicas

[platinumhamburg]

Search before asking

• I searched in the issues and found nothing similar.

Fluss version

0.9.0 (latest release)

Please describe the bug 🐞

When a tablet server receives a NotifyLeaderAndIsr request for a bucket with no elected leader (leaderId == null), the original code threw a STORAGE_EXCEPTION. The coordinator then marked that replica as offline via replicasOnOffline. Since isReplicaOnline() excludes replicas in the offline set, subsequent elections could never select these replicas as leader — causing a permanent no-leader state.

Solution

Replace the error branch with a guard condition (leaderId != null && leaderId >= 0). When no valid leader exists, skip fetcher setup silently and let the next LeaderAndIsr notification recover naturally.

Are you willing to submit a PR?

• I'm willing to submit a PR!