debian mysql restart and kill -9 two instances

[cfreak]

if you kill -9 the mysqld process while running /etc/init.d/mysql restart you end up with two instances of mysqld:

root 25066 0.0 0.0 70508 3328 ? Ss 16:51 0:00 \_ sshd: root@pts/1
root 25068 0.0 0.0 19344 2116 pts/1 Ss 16:51 0:00 | \_ -bash
root 25100 0.0 0.0 10796 1524 pts/1 S+ 16:52 0:00 | \_ /bin/bash /etc/init.d/mysql restart
root 25105 0.0 0.0 10812 1600 pts/1 S+ 16:52 0:00 | \_ /bin/bash /etc/init.d/mysql stop
root 25121 0.0 0.0 10812 756 pts/1 S+ 16:52 0:00 | \_ /bin/bash /etc/init.d/mysql stop
root 25122 0.0 0.0 23864 1676 pts/1 S+ 16:52 0:00 | \_ /usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf shutdown

root 15660 0.0 0.0 4004 640 ? S Jan23 0:00 /bin/sh /usr/bin/mysqld_safe
mysql 15818 141 86.8 116390336 114874020 ? Sl Jan23 16966:37 \_ /usr/sbin/mysqld --basedir=/usr --datadir=/mnt/mysqllv/data/ --user=mysql --log-error=/mnt/mysqllv/data//xxx.local.err --open-files-limit=65535 --pid-file=/var/run/mysqld/mysqld.pid --socket=/var/run/mysqld/mysqld.sock

# kill -9 15818

root 15660 0.0 0.0 4004 672 ? S Jan23 0:00 /bin/sh /usr/bin/mysqld_safe
mysql 25941 85.8 6.1 114221068 8126264 ? Sl 17:09 0:18 \_ /usr/sbin/mysqld --basedir=/usr --datadir=/mnt/mysqllv/data/ --user=mysql --log-error=/mnt/mysqllv/data//xxxlocal.err --open-files-limit=65535 --pid-file=/var/run/mysqld/mysqld.pid --socket=/var/run/mysqld/mysqld.sock
root 25981 0.0 0.0 4004 644 pts/1 S 17:09 0:00 /bin/sh /usr/bin/mysqld_safe
mysql 26139 27.2 5.8 113671044 7687392 pts/1 Sl 17:09 0:06 \_ /usr/sbin/mysqld --basedir=/usr --datadir=/mnt/mysqllv/data/ --user=mysql --log-error=/mnt/mysqllv/data//xxx.local.err --open-files-limit=65535 --pid-file=/var/run/mysqld/mysqld.pid --socket=/var/run/mysqld/mysqld.sock

i didn't try to reproduce it. version is percona-server-server-5.1 5.1.66-rel14.1-495.squeeze

[Valerii Kravchuk (valerii-kravchuk)]

I was able to reproduce this with PS 5.5.31-30.3 on Ubuntu 12.04 with the following command:

mysql -uroot -p... -e "insert into t1 select * from t1"; sudo service mysql restart & sudo kill -9 <PID>

where <PID> is the PID of mysqld process running and t1 is big enough InnoDB table (millions of rows) for INSERT ... SELECT to produce enough data changes so that shutdown is not instant.

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-1966