[ERROR] Transaction not registered for MySQL 2PC, but transaction is active

Does anyone know what the following error message means

  • “[ERROR] Transaction not registered for MySQL 2PC, but transaction is active”

Environment:

  • OS: Debian Wheezy, up to date patched.
  • Percona version: percona-xtradb-cluster-56 5.6.20-25.7-886.wheezy amd64
  • XTRABackup: percona-xtrabackup 2.2.5-5027-1.wheezy amd64
  • Galera percona-xtradb-cluster-galera-3.x 3.7.3256.wheezy amd64
  • Database size : ~40GB

Detailed description:
I have the following problem that I’ve not been able to solve. When our application starts and modifies the database schema ( using Liquibase ), the following error message will be in the error logs followed by a crash dump

  • “Transaction not registered for MySQL 2PC, but transaction is active”
    The error will occure when Liquibase tries to add index to a table or alter the table structure like add new column to a table. The database is backup from Percona 5.5, and I have run mysql_upgrade before synchronizing the cluster with the new database.

To make things weird, here’s few of my notes

  • Everything works well with PXD 5.6 if I will use test database that is about ~400MB ( from percona 5.5 also )
  • Everything works well if I will use Percona XtraDB Cluster 5.5. No errors or warnings.
  • I can run single ‘CREATE INDEX’ or ‘ALTER TABLE’ clauses from mysql command line without problems when using Percona 5.6, but the same statements will fail when run throught Liquibase. ( This is why I’m asking about what the error means, as I’m wondering if it could be caused by Liquibase. Still dont understand why it works with small database )
2014-10-10 14:20:29 4876 [ERROR] Transaction not registered for MySQL 2PC, but transaction is active
2014-10-10 14:20:38 7f95fd985700 InnoDB: Assertion failure in thread 140282181474048 in file row0mysql.cc line 3990
InnoDB: Failing assertion: table->n_rec_locks == 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.
11:20:38 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=402653184
read_buffer_size=2097152
max_used_connections=60
max_threads=502
thread_count=62
connection_count=60
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 2457013 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x3045010
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 = 7f95fd984e50 thread_stack 0x30000
/usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x8c8e1e]
/usr/sbin/mysqld(handle_fatal_signal+0x36c)[0x680ffc]
/lib/x86_64-linux-gnu/libpthread.so.0(+0xf030)[0x7f962ef06030]
/lib/x86_64-linux-gnu/libc.so.6(gsignal+0x35)[0x7f962d1401a5]
/lib/x86_64-linux-gnu/libc.so.6(abort+0x180)[0x7f962d143420]
/usr/sbin/mysqld[0x98e654]
/usr/sbin/mysqld[0x8edf77]
/usr/sbin/mysqld(_Z15ha_delete_tableP3THDP10handlertonPKcS4_S4_b+0xb5)[0x5e1e65]
/usr/sbin/mysqld(_Z14quick_rm_tableP3THDP10handlertonPKcS4_j+0x12a)[0x73ff1a]
/usr/sbin/mysqld(_Z17mysql_alter_tableP3THDPcS1_P24st_ha_create_informationP10TABLE_LISTP10Alter_infojP8st_orderb+0x2681)[0x748fc1]
/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x37f5)[0x6f9695]
/usr/sbin/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x588)[0x6fbbb8]
/usr/sbin/mysqld[0x6fbcbd]
/usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0xc4e)[0x6fcdbe]
/usr/sbin/mysqld(_Z10do_commandP3THD+0x1f1)[0x6fde21]
/usr/sbin/mysqld(_Z24do_handle_one_connectionP3THD+0x27d)[0x6cf8ed]
/usr/sbin/mysqld(handle_one_connection+0x42)[0x6cf972]
/usr/sbin/mysqld(pfs_spawn_thread+0x140)[0xb13270]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x6b50)[0x7f962eefdb50]
/lib/x86_64-linux-gnu/libc.so.6(clone+0x6d)[0x7f962d1e9e6d]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (3051600): CREATE INDEX `xxxIndex` ON `xxxTable`(`xxxColumn`)
Connection ID (thread ID): 59
Status: NOT_KILLED

Hi I just check your issue and it seems like you hitting a bug. This bug was fixed in 5.6.21.

Link:

[url]https://bugs.launchpad.net/percona-xtradb-cluster/+bug/1282707[/url]

Cheers
Eddi