Hi all, I'm using PostgreSQL 9.1.7's replication function for my HA project and getting trouble when terminate the hangup sender process. My cluster configuration: Node1(MASTER) Node2(STANDBY) Execution sequence: --- step 1. execute pg_basebackup on Node2 step 2. While Node2's pg_basebackup is processing (wal sender was created in Node1), turn off the Node2 by power off buton. --- Problem: --- the sender process in Node1 for base_backup of Node2(SLAVE) was hanging up (in about 30 minutes). --- Question: --- Is there's a solution to terminate that [wal sender] in Node1 but not restart or kill (-9) the process? --- I've tried with replication_timeout=60s (default) best regards, Huong DM. |