Is there any additional information I can provide that will help decipher the cause of this issue?

I'd really appreciate some feedback.

We got a similar problem. One of our 3-node MariaDB cluster node crashed.

CentOS 6.6 64Bit
MariaDB 10.0.16-1 galera 25.3.5-1

150303 12:06:02 [Warning] WSREP: SQL statement was ineffective, THD: 938807, buf: 390

QUERY: COMMIT

 => Skipping replication

150303 12:06:02 [ERROR] WSREP: FSM: no such a transition ROLLED_BACK -> ROLLED_BACK

150303 12:06:02 [ERROR] 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.

To report this bug, see http://kb.askmonty.org/en/reporting-bugs

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.

Server version: 10.0.16-MariaDB-wsrep-log

key_buffer_size=16777216

read_buffer_size=262144

max_used_connections=44

max_threads=252

thread_count=8

It is possible that mysqld could use up to

key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 1118402 K  bytes of memory

Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x0x7ff5507f5008

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 = 0x7ffe0f063d00 thread_stack 0x48000

(my_addr_resolve failure: fork)

/usr/sbin/mysqld(my_print_stacktrace+0x2b) [0xb995bb]

/usr/sbin/mysqld(handle_fatal_signal+0x398) [0x743ed8]

/lib64/libpthread.so.0() [0x32bb60f710]

/lib64/libc.so.6(gsignal+0x35) [0x32bb232625]

/lib64/libc.so.6(abort+0x175) [0x32bb233e05]

/usr/lib64/galera/libgalera_smm.so(galera::FSM<galera::TrxHandle::State, galera::TrxHandle::Transition, galera::EmptyGuard, galera::EmptyAction>::shift_to(galera::TrxHandle::State)+0x2d9) [0x7ffe17f5d4d9]

/usr/lib64/galera/libgalera_smm.so(galera::ReplicatorSMM::post_rollback(galera::TrxHandle*)+0x2e) [0x7ffe17f7828e]

/usr/lib64/galera/libgalera_smm.so(galera_post_rollback+0x68) [0x7ffe17f91d88]

/usr/sbin/mysqld() [0x6e4265]

/usr/sbin/mysqld() [0x6e42d8]

/usr/sbin/mysqld(ha_rollback_trans(THD*, bool)+0xde) [0x746dee]

/usr/sbin/mysqld(ha_commit_trans(THD*, bool)+0x1e2) [0x747142]

/usr/sbin/mysqld(trans_commit(THD*)+0x4c) [0x6b11cc]

/usr/sbin/mysqld(mysql_execute_command(THD*)+0x1020) [0x5dc110]

/usr/sbin/mysqld() [0x5e3452]

/usr/sbin/mysqld() [0x5e3d4b]

/usr/sbin/mysqld(dispatch_command(enum_server_command, THD*, char*, unsigned int)+0x17f0) [0x5e5a80]

What can we do to avoid the error? Is there a fix or something misconfigured?

Fos joergw : Can you share the transaction that triggered to this failure? Can you try upgrading to the latest mariadb/galera version to see if this still occurs?

Hi,

i am sorry that i cannot provide the requested query/transaction.
We just upgrade our cluster 4 weeks ago and the problem has not occurred again since upgrade.

I do not have the query run and I've had similar crashes occur since then but never on the same node.

Upgrading is currently not an option.

I'll paste the queries on the cluster across all 5 nodes if I get another crash.

joergw Thanks for confirming that.
Fos Since the bug does not seem to affect the latest versions, I am closing it.
But do post the queries/transaction in case you hit a crash, so that I can confirm
if its really fixed.

@Nirbhay I've upgraded all nodes to 10.0.19 and I just got the exact same error on a node in preproduction.

150701 23:10:45 [ERROR] WSREP: FSM: no such a transition ROLLED_BACK -> ROLLED_BACK
150701 23:10:45 [ERROR] 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.

To report this bug, see http://kb.askmonty.org/en/reporting-bugs

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.

Server version: 10.0.19-MariaDB-wsrep-log
key_buffer_size=134217728
read_buffer_size=31457280
max_used_connections=139
max_threads=501
thread_count=24
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 16552907 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x0x7f3a844ba008
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 = 0x7f3a6affec78 thread_stack 0x48000
/usr/sbin/mysqld(my_print_stacktrace+0x2b)[0xba1e4b]
/usr/sbin/mysqld(handle_fatal_signal+0x398)[0x748948]
/lib64/libpthread.so.0(+0xf710)[0x7f3b6e594710]
/lib64/libc.so.6(gsignal+0x35)[0x7f3b6cbf3925]
/lib64/libc.so.6(abort+0x175)[0x7f3b6cbf5105]
/usr/lib64/galera/libgalera_smm.so(ZN6galera3FSMINS_9TrxHandle5StateENS1_10TransitionENS_10EmptyGuardENS_11EmptyActionEE8shift_toES2+0x2d9)[0x7f3b6656b4d9]
/usr/lib64/galera/libgalera_smm.so(_ZN6galera13ReplicatorSMM13post_rollbackEPNS_9TrxHandleE+0x2e)[0x7f3b6658628e]
/usr/lib64/galera/libgalera_smm.so(galera_post_rollback+0x68)[0x7f3b6659fd88]
/usr/sbin/mysqld[0x6e9e75]
/usr/sbin/mysqld[0x6e9ee8]
/usr/sbin/mysqld(_Z17ha_rollback_transP3THDb+0xde)[0x74b76e]
/usr/sbin/mysqld(_Z15ha_commit_transP3THDb+0x200)[0x74bae0]
/usr/sbin/mysqld(_Z12trans_commitP3THD+0x4c)[0x6b4f4c]
/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x1020)[0x5de1e0]
/usr/sbin/mysqld[0x5e54f7]
/usr/sbin/mysqld[0x5e5ebb]
/usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x19fb)[0x5e7dfb]
/usr/sbin/mysqld(_Z10do_commandP3THD+0x1e2)[0x5e86c2]
/usr/sbin/mysqld(_Z26threadpool_process_requestP3THD+0xa7)[0x6dd3b7]
/usr/sbin/mysqld[0x71d99d]
/lib64/libpthread.so.0(+0x79d1)[0x7f3b6e58c9d1]
/lib64/libc.so.6(clone+0x6d)[0x7f3b6cca9b5d]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (0x7f3a6a9a5020): is an invalid pointer
Connection ID (thread ID): 2363
Status: NOT_KILLED*strong text*

Optimizer switch: index_merge=on,index_merge_union=on,index_merge_sort_union=on,index_merge_intersection=on,index_merge_sort_intersection=off,engine_condition_pushdown=off,index_condition_pushdown=on,derived_merge=on,derived_with_keys=on,firstmatch=on,loosescan=on,materialization=on,in_to_exists=on,semijoin=on,partial_match_rowid_merge=on,partial_match_table_scan=on,subquery_cache=on,mrr=off,mrr_cost_based=off,mrr_sort_keys=off,outer_join_with_cache=on,semijoin_with_cache=on,join_cache_incremental=on,join_cache_hashed=on,join_cache_bka=on,optimize_join_buffer_size=off,table_elimination=on,extended_keys=on,exists_to_in=on

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.
150701 23:10:48 mysqld_safe Number of processes running now: 0
150701 23:10:48 mysqld_safe WSREP: not restarting wsrep node automatically
150701 23:10:48 mysqld_safe mysqld from pid file /var/lib/mysql/mysql.pid ended

Fos Ok thanks, reopened this issue for further investigation.

@Nirbhay thanks. I've enabled wsrep_debug and general logging so I can try get additional information.

Fos Thanks! Please update when you find something of interest.

Same issue MariaDB Galera 10.0.22 crash_rollback_rollback.txt

In my case, I worked around the issue by using a procedure and checked if the session was in transaction before attempting a rollback:

  IF @@IN_TRANSACTION THEN  

    ROLLBACK;

  END IF;

stepan.patryshev, I will remove the unsupported FixVersion from this task. Please check if the task is still valid for 10.1+ and add the appropriate FixVersion

Both versions are EOL.