There is a problem with the replication.
The slave between db2 and pxc stucks with the message “waiting for GTID to be written to binary log”. In this case there is no error message. Meters “Seconds_Behind_Master” grows.
In binary logs can be seen that replication is stopped after the “commit” command.
Replication restart does not help (stop slave, start slave, reset slave).
Only restarting mysql process helps.
db1 masked on pxc using replicate_ignore_db.
Versions and settings:
pxc
centos7
5.6.28-76.1-56-log
binlog_format = ROW
log-slave-updates = 1
gtid_mode = ON
enforce_gtid_consistency = 1
wsrep_restart_slave = 1
db1
centos 7
5.6.30-76.3-log
gtid_mode = ON
log_slave_updates = 1
enforce_gtid_consistency = ON
db2
centos 7
gtid_mode = ON
log_slave_updates = 1
enforce_gtid_consistency = ON
Between db1 and db2 - auto_position = 1
Between dv1 (db2) and pxc - auto_position = 0
