Hello I have a three node postgresql 9.4 cluster configured with one standby in sync mode and the other in asynch mode. If the master datbase is shutdown the pg promote happens as expected. However on the remaining standby node the RECOVERY.CONF file is misconfigured when issuing the command: Repmgr –f ./etc/repmgr/repmgr.conf —verbose standby follow The recovery.conf file get created like this: primary_conninfo = 'port=5432 host=10.999.99.999 user=repmgr application_name=postgres-1d-01.node.us-east-1.yyyy.xxxxx.com’ When it should look like this… primary_conninfo = 'host=postgres-1c-01.node.us-east-1.yyyy.xxx.com port=5432 user=replication password=<top secret password>’ Where is it picking up the user and application_name? This same this is happening when I use rampager to reclone my former master as a standby database. Regards John |
