Comment 7 for bug 1292842

a) Yes - in that it happens often. I have not tried to reduce it down to a simple example.

b) To be deterred. I don't think so at this time. I'll turn on slow logs with time set to zero and do a review.

c) +---------------------------+-------+
    | Variable_name | Value |
    +---------------------------+-------+
    | wsrep_load_data_splitting | ON |
    +---------------------------+-------+

d) all attached them to the report now
Mark Grennan | Database Administrator
Mobile: +1 (405) 343-9332 | <email address hidden>
Live 24/7 Support +1 (866) 429-5578 | <email address hidden>

Weather Decision Technologies, Inc - Office: +1 (405) 579-7675 260 | Fax: +1 (405) 579-7780
201 David L Boren Blvd, Suite 270 Norman OK 73072

READ CAREFULLY. The information in this electronic mail and any attachments may
be proprietary, private or otherwise confidential and are transmitted for the
exclusive and restricted use of the intended recipient. If the reader of this
message is not the intended recipient, you are notified that retention, use,
dissemination, distribution, or copying of this message or any attachments, or
the use of any information transmitted, is strictly prohibited. If you receive
this message in error, please notify us immediately by electronic mail or
telephone +1 (405) 579-7675 and delete this message and any attachments.

.

-----Original Message-----
From: <email address hidden> [mailto:<email address hidden>] On Behalf Of Raghavendra D Prabhu
Sent: Friday, June 20, 2014 9:54 AM
To: Mark Grennan
Subject: [Bug 1292842] Re: WSREP: FSM: no such a transition ROLLED_BACK -> ROLLED_BACK with LOAD DATA INFILE in 5.5
Importance: High

@Mark,

a)
Is this reproducible with

LOAD DATA LOCAL INFILE "tmp20131.txt.001" REPLACE INTO TABLE point_obs.metar_nc (date_time,stn,lat,lon,data_source,temp_F,dewpt_F,dir,vrb_wind,speed_mps,gust_mps,slp_mb,altim_mb,visib_m,cld_cover,cavok,precip_type,tmin6_F,tmax6_F,tmin24_F,tmax24_F,snow_mm,snow24_mm,precip_1hr_mm,precip_24hr_mm,weather,waveht_m,wavedir,waveper,sst_F,sensor_status,auto_indicator,raw_metar

in a repeatable way?

b) Are there more than one LOAD DATA INFILE running (the RQG does more than one
in concurrently).

c) Do you have wsrep-load-data-splitting ON or OFF?

d) Can you also share your my.cnf/error log etc.

--
You received this bug notification because you are subscribed to the bug
report.
https://bugs.launchpad.net/bugs/1292842

Title:
  WSREP: FSM: no such a transition ROLLED_BACK -> ROLLED_BACK with LOAD
  DATA INFILE in 5.5

Status in Percona XtraDB Cluster - HA scalable solution for MySQL:
  New
Status in Percona XtraDB Cluster 5.6 series:
  New
Status in Percona XtraDB Cluster trunk series:
  New

Bug description:
  This is reproduceable with load-data grammar in
  bzr+ssh://bazaar.launchpad.net/~raghavendra-prabhu/randgen/pxc/

  #0 0x00007f69e86b90c1 in pthread_kill () from /usr/lib/libpthread.so.0
  #1 0x00000000006a951f in handle_fatal_signal (sig=6) at /tmp/bld1/sql/signal_handler.cc:250
  #2 <signal handler called>
  #3 0x00007f69e6c53389 in raise () from /usr/lib/libc.so.6
  #4 0x00007f69e6c54788 in abort () from /usr/lib/libc.so.6
  #5 0x00007f69e61bc1bc in galera::FSM<galera::TrxHandle::State, galera::TrxHandle::Transition, galera::EmptyGuard, galera::EmptyAction>::shift_to (this=this@entry=0x7f69380d2cf0,
      state=state@entry=galera::TrxHandle::S_ROLLED_BACK) at galera/src/fsm.hpp:81
  #6 0x00007f69e61b36c6 in set_state (state=galera::TrxHandle::S_ROLLED_BACK, this=0x7f69380d2c60) at galera/src/trx_handle.hpp:229
  #7 galera::ReplicatorSMM::post_rollback (this=0x171b9d0, trx=0x7f69380d2c60) at galera/src/replicator_smm.cpp:964
  #8 0x00007f69e61c4589 in galera_post_rollback (gh=<optimized out>, trx_handle=0x29eb7b8) at galera/src/wsrep_provider.cpp:380
  #9 0x0000000000659cb8 in wsrep_rollback (thd=0x29e9b60, all=<optimized out>, hton=<optimized out>) at /tmp/bld1/sql/wsrep_hton.cc:208
  #10 0x0000000000659d88 in wsrep_rollback (hton=<optimized out>, thd=<optimized out>, all=<optimized out>) at /tmp/bld1/sql/wsrep_hton.cc:196
  #11 0x00000000006abe0e in ha_rollback_trans (thd=thd@entry=0x29e9b60, all=<optimized out>) at /tmp/bld1/sql/handler.cc:1615
  #12 0x00000000006ac222 in ha_commit_trans (thd=thd@entry=0x29e9b60, all=all@entry=false) at /tmp/bld1/sql/handler.cc:1459
  #13 0x0000000000650529 in trans_commit_stmt (thd=thd@entry=0x29e9b60) at /tmp/bld1/sql/transaction.cc:383
  #14 0x000000000059cfe1 in mysql_execute_command (thd=thd@entry=0x29e9b60) at /tmp/bld1/sql/sql_parse.cc:5157
  #15 0x00000000005a4cc1 in mysql_parse (thd=thd@entry=0x29e9b60, rawbuf=rawbuf@entry=0x7f6938004ba0 "LOAD DATA INFILE '/tmp/gentest18131.tmp' INTO TABLE `table1_innodb_compressed_key_pk_parts_2_int`",
      length=length@entry=97, parser_state=parser_state@entry=0x7f69a04f8620) at /tmp/bld1/sql/sql_parse.cc:6475
  #16 0x00000000005a563c in wsrep_mysql_parse (thd=thd@entry=0x29e9b60, rawbuf=0x7f6938004ba0 "LOAD DATA INFILE '/tmp/gentest18131.tmp' INTO TABLE `table1_innodb_compressed_key_pk_parts_2_int`", length=97,
      parser_state=parser_state@entry=0x7f69a04f8620) at /tmp/bld1/sql/sql_parse.cc:6279
  #17 0x00000000005a7c37 in dispatch_command (command=<optimized out>, command@entry=COM_QUERY, thd=thd@entry=0x29e9b60,
      packet=packet@entry=0x296e0c1 " LOAD DATA INFILE '/tmp/gentest18131.tmp' INTO TABLE `table1_innodb_compressed_key_pk_parts_2_int`", packet_length=packet_length@entry=98) at /tmp/bld1/sql/sql_parse.cc:1241
  #18 0x00000000005a8c52 in do_command (thd=0x29e9b60) at /tmp/bld1/sql/sql_parse.cc:870
  #19 0x0000000000642e98 in do_handle_one_connection (thd_arg=thd_arg@entry=0x29e9b60) at /tmp/bld1/sql/sql_connect.cc:1426
  #20 0x00000000006430aa in handle_one_connection (arg=0x29e9b60) at /tmp/bld1/sql/sql_connect.cc:1338
  #21 0x00007f69e86b40a2 in start_thread () from /usr/lib/libpthread.so.0
  #22 0x00007f69e6d03d1d in clone () from /usr/lib/libc.so.6

To manage notifications about this bug go to:
https://bugs.launchpad.net/percona-xtradb-cluster/+bug/1292842/+subscriptions