applier fails with lock wait timeout exceeded on rsync SST donor

Test: Standard seesaw test with rsync SST.

During one full SST donor got stuck in " Flushing tables for SST...". Normally this stage is resolved when some client transaction lock wait times out and FTWRL can proceed. However, this time applier (or replayer) could not apply events. The question is, why would the BF transaction ever had to wait for lock until it times out.

Log snippet:

2013-11-27 02:08:26 32076 [Note] WSREP: Shifting SYNCED -> DONOR/DESYNCED (TO: 16875835)
2013-11-27 02:08:26 32076 [Note] WSREP: wsrep_notify_cmd is not defined, skipping notification.
2013-11-27 02:08:26 32076 [Note] WSREP: Running: 'wsrep_sst_rsync --role 'donor' --address 'gw:10023/rsync_sst' --auth 'root:rootpass' --socket '/run/shm/galera/local3/mysql/var/mysqld.sock' --datadir '/run/shm/galera/local3/mysql/var/' --defaults-file '/run/shm/galera/local3/mysql/etc/my.cnf' '' --gtid '8fc7b0b3-56a5-11e3-b5fe-1aa9415117a6:16875835''
2013-11-27 02:08:26 32076 [Note] WSREP: sst_donor_thread signaled with 0
2013-11-27 02:08:26 32076 [Note] WSREP: Flushing tables for SST...
2013-11-27 02:09:17 32076 [Warning] Slave SQL: Could not execute Update_rows event on table test.comm01; Lock wait timeout exceeded; try restarting transaction, Error_code: 1205; handler error HA_ERR_LOCK_WAIT_TIMEOUT; the event's master log FIRST, end_log_pos 202, Error_code: 1205
2013-11-27 02:09:17 32076 [Warning] WSREP: RBR event 3 Update_rows apply warning: 146, 16875911
2013-11-27 02:09:17 32076 [Warning] WSREP: Failed to apply app buffer: seqno: 16875911, status: 1
  at galera/src/trx_handle.cpp:apply():340
Retrying 2th time
2013-11-27 02:10:08 32076 [Warning] Slave SQL: Could not execute Update_rows event on table test.comm01; Lock wait timeout exceeded; try restarting transaction, Error_code: 1205; handler error HA_ERR_LOCK_WAIT_TIMEOUT; the event's master log FIRST, end_log_pos 202, Error_code: 1205
2013-11-27 02:10:08 32076 [Warning] WSREP: RBR event 3 Update_rows apply warning: 146, 16875911
2013-11-27 02:10:08 32076 [Warning] WSREP: Failed to apply app buffer: seqno: 16875911, status: 1
  at galera/src/trx_handle.cpp:apply():340
Retrying 3th time
2013-11-27 02:10:59 32076 [Warning] Slave SQL: Could not execute Update_rows event on table test.comm01; Lock wait timeout exceeded; try restarting transaction, Error_code: 1205; handler error HA_ERR_LOCK_WAIT_TIMEOUT; the event's master log FIRST, end_log_pos 202, Error_code: 1205
2013-11-27 02:10:59 32076 [Warning] WSREP: RBR event 3 Update_rows apply warning: 146, 16875911
2013-11-27 02:10:59 32076 [Warning] WSREP: Failed to apply app buffer: seqno: 16875911, status: 1
  at galera/src/trx_handle.cpp:apply():340
Retrying 4th time
2013-11-27 02:11:50 32076 [Warning] Slave SQL: Could not execute Update_rows event on table test.comm01; Lock wait timeout exceeded; try restarting transaction, Error_code: 1205; handler error HA_ERR_LOCK_WAIT_TIMEOUT; the event's master log FIRST, end_log_pos 202, Error_code: 1205
2013-11-27 02:11:50 32076 [Warning] WSREP: RBR event 3 Update_rows apply warning: 146, 16875911
2013-11-27 02:11:50 32076 [ERROR] WSREP: Failed to apply trx: source: fabb5d8b-56f7-11e3-97a2-bad682cc23e2 version: 3 local: 0 state: APPLYING flags: 129 conn_id: 15 trx_id: 56293974 seqnos (l: 19474, g: 16875911, s: 16875909, d: 16875910, ts: 630454850094192)
2013-11-27 02:11:50 32076 [ERROR] WSREP: Failed to apply trx 16875911 4 times
2013-11-27 02:11:50 32076 [ERROR] WSREP: Node consistency compromized, aborting...