wsrep_exec_mode remains in LOCAL_COMMIT after AC rollback

[Teemu Ollakka]

In 5.6 tree wsrep_rollback() sets thd->wsrep_exec_mode= LOCAL_COMMIT. Since thd->transaction.cleanup() does not get called for autocommit statements after rollback, missing call to wsrep_cleanup_transaction() makes thd->wsrep_exec_mode to stay in LOCAL_COMMIT even after rollback finishes and causes problems when the next transaction starts.

[Teemu Ollakka (teemu-ollakka)]

Fix committed in lp:codership-mysql/5.6 rev 3974.

[Raghavendra D Prabhu (raghavendra-prabhu)]

Seeing this crash from codership binaries:

========================================================
mysqld: /media/Tintin/Work/code/codership/codership-5.6/sql/wsrep_hton.cc:79: void wsrep_register_hton(THD*, bool): Assertion `thd->wsrep_exec_mode != LOCAL_COMMIT' failed.
01:06:34 UTC - mysqld got signal 6 ;
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.
We will try our best to scrape up some info that will hopefully help
diagnose the problem, but since we have already crashed,
something is definitely wrong and this may fail.

key_buffer_size=8388608
read_buffer_size=131072
max_used_connections=20
max_threads=151
thread_count=19
connection_count=19
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 68310 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x2386230
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 = 7f8d442e6e78 thread_stack 0x40000
/chip/bin/mysqld(my_print_stacktrace+0x2c)[0x9474da]
/chip/bin/mysqld(handle_fatal_signal+0x2f7)[0x67f887]
/usr/lib/libpthread.so.0(+0xf870)[0x7f8d575a3870]
/usr/lib/libc.so.6(gsignal+0x39)[0x7f8d55b393d9]
/usr/lib/libc.so.6(abort+0x148)[0x7f8d55b3a7d8]
/usr/lib/libc.so.6(+0x2e506)[0x7f8d55b32506]
/usr/lib/libc.so.6(+0x2e5b2)[0x7f8d55b325b2]
/chip/bin/mysqld(_Z19wsrep_register_htonP3THDb+0x3a)[0x7c6c39]
/chip/bin/mysqld(_Z17trans_commit_stmtP3THD+0x463)[0x7ad1b3]
/chip/bin/mysqld(_Z21mysql_execute_commandP3THD+0x76f0)[0x71222d]
/chip/bin/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x4e7)[0x71373e]
/chip/bin/mysqld[0x7140c1]
/chip/bin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x132d)[0x715965]
/chip/bin/mysqld(_Z10do_commandP3THD+0x715)[0x7180d3]
/chip/bin/mysqld(_Z24do_handle_one_connectionP3THD+0x143)[0x6db350]
/chip/bin/mysqld(handle_one_connection+0x40)[0x6db4db]
/usr/lib/libpthread.so.0(+0x80a2)[0x7f8d5759c0a2]
/usr/lib/libc.so.6(clone+0x6d)[0x7f8d55be943d]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (7f8ccc004c50): DELETE FROM `table100_innodb_key_pk_parts_2_int` WHERE `col_char_12_key` = 88014848 ORDER BY `col_char_12`,`col_char_12_key`,`col_int`,`col_int_key`,`pk` LIMIT 3
Connection ID (thread ID): 23
Status: NOT_KILLED

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.
Writing a core file

=================================================================

(gdb) bt
#0 0x00007f8d575a10b1 in pthread_kill () from /usr/lib/libpthread.so.0
#1 0x0000000000946cd8 in my_write_core (sig=6) at /media/Tintin/Work/code/codership/codership-5.6/mysys/stacktrace.c:422
#2 0x000000000067fa02 in handle_fatal_signal (sig=6) at /media/Tintin/Work/code/codership/codership-5.6/sql/signal_handler.cc:248
#3 <signal handler called>
#4 0x000...

[Raghavendra D Prabhu (raghavendra-prabhu)]

gdb

a) thread apply all bt
b) thread apply all bt full

[Raghavendra D Prabhu (raghavendra-prabhu)]

vardir.

[Teemu Ollakka (teemu-ollakka)]

Additional fix in http://bazaar.launchpad.net/~codership/codership-mysql/5.6/revision/3981

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PXC-1477