Node is attached in state "catchup"

Skip to first unread message


Sep 16, 2021, 10:16:16 AMSep 16
to repmgr
I have a 4 node system and simulated several successive server outages of different masters. When then a node rejoin (or standby follow) is done, a standby is attached in state "catchup" (NODE REJOIN successful) to a new master.
After this repmgr service status shows a "not running" repmgd on the rejoined server, although the process itself runs.
Shortly afterwards there appears an "[ERROR] unable to execute get_primary_current_lsn()" and "[ERROR] get_primary_node_id(): unable to execute query" in the joined server log.
The state of the repmgrd seems only to change after a former master (which shall be also rejoined) has reappeared, and only then it is available as new master if necessary.
Any hints if this can be solved would be appreciated.
Reply all
Reply to author
0 new messages