[pgpool-general: 5349] can not connect to slave after failover

[Dustin Kempter]

Hi all,

Im having an issue. So I have 3 vm servers running Centos 6 and postgres 
9.4. I have 1 set up as a pgpool server. The other 2 are in a master 
slave configuration via hot_standby. The set up all works. From the 
pgpool server I can connect to both db servers via psql -h and ssh keys 
are set up. I have a script for fail over. In my test I will shut down 
the master, then pgpool runs my script and it reaches out to the slave 
and promotes it. However theres an issue. Pgpool then can not connect to 
the promoted slave. If as postgres I run psql -h I can connect fine. I 
checked the log file for pgpool and im getting 2 lines looping over and 
over.


2017-02-27 (current time) pid 2765: LOG: find_primary_node: checking 
backend no 0
2017-02-27 (current time) pid 2765: LOG: find_primary_node: checking 
backend no 1

Those just repeat over and over, and any attempted connection just hangs.


Please advise. Thanks in advance!!