With semi-sync replication, master server crash after slave server shutdown

[go-to-k]

・Version
Mysql Version: 5.7.16
mysql-audit Version: audit-plugin-mysql-5.7-1.1.1-660-linux-x86_64
Linux version: 3.10.0-327.18.2.el7.x86_64 ([email protected])(gcc version 4.8.3 20140911 (Red Hat 4.8.3-9) (GCC) )

・Events

1. With semi-sync replication (composing one master and one slave mysql server), both master and slave install plugin "semisync_master.so" and "semisync_slave.so".
   For after fail-over, i write in my.cnf in both servers like that
   rpl_semi_sync_master_enabled=1
   rpl_semi_sync_slave_enabled=1

2. I stop the "slave" mysql server during behaving replication correctly, then the "master" mysql server crashes and i couldn't connect to master.

[slave ~]$ service mysqld stop
Redirecting to /bin/systemctl stop mysqld.service

[master ~]$ mysql -u root -p
Enter password:
ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/lib/mysql/mysql.sock' (111)

3. After several seconds, master server starts crash recovery automatically, and i became able to connect to the master server.

[master ~]$ mysql -u root -p
Enter password:
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 2
Server version: 5.7.16-log MySQL Community Server (GPL)
...

・Error log (in master server):

09:39:04 UTC - mysqld got signal 11 ;
This could be because you hit a bug. It is also possible that this binary
or one of the libraries it was linked against is corrupt, improperly built,
or misconfigured. This error can also be caused by malfunctioning hardware.
Attempting to collect some information that could help diagnose the problem.
As this is a crash and something is definitely wrong, the information
collection process might fail.

key_buffer_size=8388608
read_buffer_size=131072
max_used_connections=4
max_threads=800
thread_count=2
connection_count=2
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 940467 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x0
Attempting backtrace. You can use the following information to find out
where mysqld died. If you see no messages after this, something went
terribly wrong...
stack_bottom = 0 thread_stack 0x40000
/usr/sbin/mysqld(my_print_stacktrace+0x3b)[0xef182b]
/usr/sbin/mysqld(handle_fatal_signal+0x461)[0x7ad8c1]
/lib64/libpthread.so.0(+0xf100)[0x7fb4f2dfe100]
/usr/lib64/mysql/plugin/libaudit_plugin.so(+0xfafd)[0x7fb1e5364afd]
/usr/lib64/mysql/plugin/libaudit_plugin.so(+0x11009)[0x7fb1e5366009]
/usr/lib64/mysql/plugin/libaudit_plugin.so(+0xeb35)[0x7fb1e5363b35]
/usr/sbin/mysqld[0x7ada1a]
/usr/sbin/mysqld(Z24plugin_foreach_with_maskP3THDPFcS0_P13st_plugin_intPvEijS3+0x1c7)[0xce5737]
/usr/sbin/mysqld[0x7adb0b]
/usr/sbin/mysqld(_Z18mysql_audit_notifyP3THD30mysql_event_general_subclass_tiPKcm+0x1f4)[0x7ae074]
/usr/sbin/mysqld(my_message_sql+0xbf)[0x79ee9f]
/usr/sbin/mysqld(my_error+0xe0)[0xeebd40]
/usr/sbin/mysqld[0xc1b42b]
/usr/sbin/mysqld[0xc1b7f0]
/usr/sbin/mysqld(my_net_read+0x334)[0xc1c574]
/usr/lib64/mysql/plugin/semisync_master.so(_ZN12Ack_receiver3runEv+0x3f2)[0x7fb1e4a8dab2]
/usr/lib64/mysql/plugin/semisync_master.so(ack_receive_handler+0x19)[0x7fb1e4a8dd79]
/usr/sbin/mysqld(pfs_spawn_thread+0x1b4)[0x12645b4]
/lib64/libpthread.so.0(+0x7dc5)[0x7fb4f2df6dc5]
/lib64/libc.so.6(clone+0x6d)[0x7fb4f18b2ced]
The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
information that should help you find out what is causing the crash.

Also, in the case when stop "master" (not slave) with semi-sync replication, same logs were written in master's error log.

[aharonrobbins]

Thanks for the report. We will investigate.

[karthikg001]

+1
This bug hampers usage in 5.7 semi-sync setup.

[aharonrobbins]

Hi. We have fixed this issue in the latest commit. Pre-compiled binaries have been uploaded to Bintray. Thanks!