Failing assertion: foreign->referenced_table ->n_foreign_key_checks_running > 0

[Facundo Martinez]

I have a 2 nodes cluster in a Centos 6.5 insallation. My packages are the followings:
percona-release-0.0-1.x86_64
percona-xtrabackup-2.1.9-744.rhel6.x86_64
Percona-XtraDB-Cluster-shared-56-5.6.15-25.5.759.rhel6.x86_64
Percona-XtraDB-Cluster-client-56-5.6.15-25.5.759.rhel6.x86_64
Percona-Server-shared-51-5.1.73-rel14.11.603.rhel6.x86_64
Percona-XtraDB-Cluster-server-56-5.6.15-25.5.759.rhel6.x86_64
Percona-XtraDB-Cluster-galera-3-3.5-1.216.rhel6.x86_64

Cluster node crashes unexpectedly at random intervals of time. The only message that prompts in the error logs are the followings:

2014-05-16 09:10:44 7f3964676700 InnoDB: Assertion failure in thread 139884474361600 in file row0upd.cc line 442
InnoDB: Failing assertion: foreign->referenced_table ->n_foreign_key_checks_running > 0
InnoDB: We intentionally generate a memory trap.
InnoDB: Submit a detailed bug report to http://bugs.mysql.com.
InnoDB: If you get repeated assertion failures or crashes, even
InnoDB: immediately after the mysqld startup, there may be
InnoDB: corruption in the InnoDB tablespace. Please refer to
InnoDB: http://dev.mysql.com/doc/refman/5.6/en/forcing-innodb-recovery.html
InnoDB: about forcing recovery.
12:10:44 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.
Please help us make Percona XtraDB Cluster better by reporting any
bugs at https://bugs.launchpad.net/percona-xtradb-cluster

key_buffer_size=33554432
read_buffer_size=131072
max_used_connections=72
max_threads=502
thread_count=70
connection_count=70
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 233047 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x2f109cc0
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 = 7f3964675d38 thread_stack 0x40000
/usr/sbin/mysqld(my_print_stacktrace+0x35)[0x902645]
/usr/sbin/mysqld(handle_fatal_signal+0x4c4)[0x6801e4]
/lib64/libpthread.so.0(+0xf710)[0x7f39b111c710]
/lib64/libc.so.6(gsignal+0x35)[0x7f39af565925]
/lib64/libc.so.6(abort+0x175)[0x7f39af567105]
/usr/sbin/mysqld[0xa17900]
/usr/sbin/mysqld[0xa1947e]
/usr/sbin/mysqld[0xa1994a]
/usr/sbin/mysqld[0x9fd3bb]
/usr/sbin/mysqld[0x96e33f]
/usr/sbin/mysqld(_ZN7handler13ha_update_rowEPKhPh+0xb6)[0x5c9436]
/usr/sbin/mysqld(_Z12mysql_updateP3THDP10TABLE_LISTR4ListI4ItemES6_PS4_jP8st_ordery15enum_duplicatesbPySB_+0x177e)[0x77393e]
/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x239d)[0x7039ed]
/usr/sbin/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x608)[0x707908]
/usr/sbin/mysqld[0x707a31]
/usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x1ad4)[0x709cb4]
/usr/sbin/mysqld(_Z10do_commandP3THD+0x1e3)[0x70b083]
/usr/sbin/mysqld(_Z24do_handle_one_connectionP3THD+0x17f)[0x6d492f]
/usr/sbin/mysqld(handle_one_connection+0x47)[0x6d4b07]
/usr/sbin/mysqld(pfs_spawn_thread+0x12a)[0xb3c0da]
/lib64/libpthread.so.0(+0x79d1)[0x7f39b11149d1]
/lib64/libc.so.6(clone+0x6d)[0x7f39af61bb6d]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (7f312000f4e0): is an invalid pointer
Connection ID (thread ID): 2142
Status: NOT_KILLED

You may download the Percona XtraDB Cluster operations manual by visiting
http://www.percona.com/software/percona-xtradb-cluster/. You may find information
in the manual which will help you identify the cause of the crash.
140516 09:10:44 mysqld_safe Number of processes running now: 0
140516 09:10:44 mysqld_safe WSREP: not restarting wsrep node automatically
140516 09:10:44 mysqld_safe mysqld from pid file /srv/data/mysql.pid ended
140516 09:19:32 mysqld_safe Starting mysqld daemon with databases from /srv/data/
140516 09:19:32 mysqld_safe WSREP: Running position recovery with --log_error='/srv/data//wsrep_recovery.mdcQ8Y' --pid-file='/srv/data//srv3ar-recover.pid'

No logs before this except the ones from start-up.

I'm attaching al conerning logs from both nodes and configuration files.

In advance, we thank you very much for all the support.
Kind regards,
Facundo.

[Raghavendra D Prabhu (raghavendra-prabhu)]

@Facundo,

If you are able to reproduce this, can you provide 'bt', 'bt full' and 'thread
apply all bt full' from the core files generated (enable coredumps before this).

[Facundo Martinez (fmartinez-q)]

@Raghavendra

We had a new crash and I could make a core-dump out of it. I'm attaching what you've requested along with the logs.

We also changed the number of threads from 64 to 32 and that seems to help in delaying the problem, if that's a thing.

Thank you in advance.

Facundo

[Raghavendra D Prabhu (raghavendra-prabhu)]

@Facundo,

Do you still have the core file with you?

Before generating the bt etc. can you please install the debuginfo package
corresponding to the server package you have.

Many symbols are still missing, so it would help to fill the gaps.

[Facundo Martinez (fmartinez-q)]

@Raghavendra

Though I still have with me the dumps, we downgraded to Percona 5.5, which was more stable than 5.6 for us. And since this is in a productive server, I'm not able to install any other package at the moment.

If I install the symbols in another server, will it work?