[pgpool-general: 4459] pgpool fails to follow new master after failover

[Travis Kirstine]

I've been testing pgpool with repmgr and have in streaming replicaiton / watchdog mode with two backend nodes and have run into a issue


When the current postgresql master node fails pgpool executes the failover script to promote the postgresql node to master successfully however pgpool fails to follow that node as the new master and rejects on db connections.


If I restart / switch pgpool node it will start working again.




Failover Script ###


#!/bin/sh

# Failover command for streaming replication.

# Arguments: $1: failed node id. $2: new master hostname.


failed_node=$1

new_master=$2


(

date

echo "Failed node: $failed_node"

set -x

# Promote standby/slave to be a new master (old master failed) using repmgr

/usr/bin/ssh -T -l postgres $new_master "/usr/pgsql-9.4/bin/repmgr -f /var/lib/pgsql/repmgr/repmgr.conf standby promote 2>/dev/null 1>/dev/null <&-"

exit 0;

) 2>&1 | tee -a /tmp/failover_stream.sh.log








-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20160216/f685ab06/attachment-0001.html>