mysqld got signal 11, Galera 3.7

[Anna]

Hi,

I have 2+1 nodes Percona Cluster (2 worker nodes +garbd ).

mysql> show variables like "%version%"; +
| Variable_name | Value |
+
| innodb_version | 5.6.20-68.0 |
| protocol_version | 10 |
| slave_type_conversions | |
| version | 5.6.20-68.0-56 |
| version_comment | Percona XtraDB Cluster (GPL), Release rel68.0, Revision 888, WSREP version 25.7, wsrep_25.7.r4126 |
| version_compile_machine | x86_64 |
| version_compile_os | Linux |
+

wsrep_provider_name Galera
wsrep_provider_vendor Codership Oy <email address hidden>
wsrep_provider_version 3.7(r7f44a18)

Yesterday suddenly one node crashed with the following error:

13:45:01 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.
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=8388608
read_buffer_size=131072
max_used_connections=60
max_threads=153
thread_count=33
connection_count=31
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 69257 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x2db48d0
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 = 7fa721437d38 thread_stack 0x40000
/usr/sbin/mysqld(my_print_stacktrace+0x35)[0x8f5835]
/usr/sbin/mysqld(handle_fatal_signal+0x4b4)[0x664384]
/lib64/libpthread.so.0(+0xf710)[0x7fa781d67710]
[0x7fa6e40000c8]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (7fa6e4004c10): is an invalid pointer
Connection ID (thread ID): 3432
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.
141015 15:45:01 mysqld_safe Number of processes running now: 0
141015 15:45:01 mysqld_safe WSREP: not restarting wsrep node automatically
141015 15:45:01 mysqld_safe mysqld from pid file /var/lib/mysql/mysql02_cluster.pkw.de.pid ended

IST was successful only after deleting galera.cache and grastate.dat.

Please let me know if you need some other information.

[Raghavendra D Prabhu (raghavendra-prabhu)]

From s

tack_bottom = 7fa721437d38 thread_stack 0x40000
/usr/sbin/mysqld(my_print_stacktrace+0x35)[0x8f5835]
/usr/sbin/mysqld(handle_fatal_signal+0x4b4)[0x664384]
/lib64/libpthread.so.0(+0xf710)[0x7fa781d67710]
[0x7fa6e40000c8]

Did you try to replace with cp/mv libgalera_smm.so when PXC was running? In that case, for dlopened libraries, it can crash like this.

If you need to replace, use install.

[Anna (sonyque)]

Hi,

no, i didn't. It is being crashed at random time. For today i have:

- second node - crashed 3 times
- first node - crashed 1 time

Btw deleting galera.cache and grastate.dat do not help anymore. Now i have to remove all percona packages, clear /var/lib/mysql/ folder, install percona back and start sync.

[Raghavendra D Prabhu (raghavendra-prabhu)]

@Anna,

Can you enable coredump (and install debug info packages) and a provide backtrace from them? You can use https://gist.github.com/ronin13/8d403c2f88826fcc7aba as the gdb script.

Also, providing error log and my.cnf should be helpful.

[Anna (sonyque)]

Hi,

sorry for delayed reply. Please find all requested files in the attachment. But please take the error log from my first post. Unfortunately i remembered to late to keep the logs within that recovery process :(

[Anna (sonyque)]

Hi,

this log could be helpful. I used this example for gdb backtrace:

http://www.leonardoborda.com/blog/gdb-series-attach-gdb-to-a-program-that-is-already-running-part-1/

Within the test ((gdb) continue) mysql has crashed. The log in attach contains that information. The backtrace you can find in my previous post (gdb-mysql02_cluster.txt).

gdb2_mysql02_cluster.txt and gdb2_mysql01_cluster.txt was done using the script from URL you provided.

[Anna (sonyque)]

New update:

Oct 21 08:08:15 mysql01_cluster kernel: mysqld[31742]: segfault at 51 ip 00007f266e9c0867 sp 00007f2501634140 error 4 in libgcc_s-4.4.7-20120601.so.1[7f266e9b1000+16000]

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PXC-1755