Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.1(EOL), 10.2(EOL), 10.3(EOL), 10.4(EOL)
-
None
Description
--source include/master-slave.inc
|
|
SET @@GLOBAL.rpl_semi_sync_master_enabled = 1; |
|
--connection slave
|
source include/stop_slave.inc;
|
SET @@GLOBAL. rpl_semi_sync_slave_enabled = 1; |
source include/start_slave.inc;
|
|
--connection master
|
SET @@GLOBAL.debug_dbug= "d,dbug_master_binlog_over_2GB"; |
CREATE TABLE t1 (a INT); |
INSERT INTO t1 SET a=0; |
|
--sync_slave_with_master |
The test case is derived from rpl.rpl_semi_sync_skip_repl . I would think that maybe the debug injection causes it intentionally, but since the failure only happens with semisync enabled, I suspect there might be a problem.
Not reproducible on 10.0.
Attachments
Issue Links
- is duplicated by
-
MDEV-24120 Assertion `thd_killed(current_thd) || !active_tranxs_ || !active_tranxs_->is_tranx_end_pos(trx_wait_binlog_name, trx_wait_binlog_pos)' failed in ReplSemiSyncMaster::commitTrx
-
- Confirmed
-
- relates to
-
MDEV-19376 Repl_semi_sync_master::commit_trx assertion failure: (thd_kill_level(thd) == THD_ABORT_ASAP) || !m_active_tranxs || !m_active_tranxs->is_tranx_end_pos(trx_wait_binlog_name, trx_wait_binlog_pos)
-
- Closed
-
julien.fritsch I crossed a test failure while was working on
MDEV-7974and got to dwell into it (naturally with the idea of clearing off my MDEV) to latter find outthat it's the test's fault and has nothing to do with my MDEV, and that there's an open bug mentioning the test.
The total work therefore as it often happens with test failures consists of 2/3 of analysis and 1/3 of the test refinement. I could've left out the fixing part, but that about 1 hr.
I hope this clarifies and justifies.
Cheers.