[HBASE-27988] NPE in AddPeerProcedure recovery

• Type: Bug
• Status: Resolved
• Priority: Trivial
• Resolution: Fixed
• Affects Version/s: 4.0.0-alpha-1
• Fix Version/s: 3.0.0-beta-1
• Component/s: Replication
• Hadoop Flags: Reviewed
• Assignee: ruanhui
• Reporter: ruanhui
• Created: 23/Jul/23 13:26
• Updated: 29/Jul/23 19:40
• Resolved: 28/Jul/23 15:20

Description

AddPeerProcedure will restore syncReplicationPeerLock when replayed in master recovery, however the replicationPeerManager has not been initialized when replay procedure, which will cause a nullPointerException and master to abort.

@Override
protected void afterReplay(MasterProcedureEnv env) {
  // ......
  if (peerConfig.isSyncReplication()) {
    if (!env.getReplicationPeerManager().tryAcquireSyncReplicationPeerLock()) {
      throw new IllegalStateException(
        "Can not acquire sync replication peer lock for peer " + peerId);
    }
  }
}  

Issue Links

• relates to: HBASE-27984 NPE in MigrateReplicationQueueFromZkToTableProcedure recovery
• links to: GitHub Pull Request #5331