5.6-beta -- init stop script (centos) does not wait for the server to be fully stopped

[Jay Janssen]

This breaks restart because stop exits early and the start action fails because the daemon is still there.

9225 pts/0 S 0:00 /bin/sh /usr/bin/mysqld_safe --datadir=/var/lib/mysql --pid-file=/var/lib/mysql/node
 9651 pts/0 Sl 0:01 \_ /usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --plugin-dir=/usr/lib64

[root@node1 mysql]# service mysql stop && service mysql start
Shutting down MySQL (Percona XtraDB Cluster)....... SUCCESS!
 ERROR! MySQL (Percona XtraDB Cluster) is running but PID file could not be found

 9225 pts/0 S 0:00 /bin/sh /usr/bin/mysqld_safe --datadir=/var/lib/mysql --pid-file=/var/lib/mysql/node
 9651 pts/0 S 0:02 \_ /usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --plugin-dir=/usr/lib64

[Raghavendra D Prabhu (raghavendra-prabhu)]

This looks like a race condition where the pid file has been
removed by mysqld while shutting down but process is still there
(but would exit in a second or so). This is the only possibility
from the error messages. So, this is more an issue with mysqld
itself than init script (which is similar for stop in PXC
5.5/PS 5.5/MySQL 5.5).

Only way around this would be add to a if kill -0 .... sleep 1; fi

Did the process finally exit (or was another service mysql stop
required?)

Also, can you share my.cnf as well.

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PXC-1526