Details
-
Bug
-
Status: Closed (View Workflow)
-
Critical
-
Resolution: Fixed
-
None
Description
There's a race between FLUSH LOGS and binlog-group commit for Binlog-checkpoint notification which must be done *after* the committed group transactions' affected
data are flushed to disk. However the notification may happen before that fore the reason
FLUSH LOGS handler does to acquire necessary mutexes.
A test case follows with a patch to be committed soon.
Attachments
Issue Links
- relates to
-
MDEV-24530 binlog-optimize-thread-scheduling=0 case of binlog rotate via FLUSH LOGS may obsolate binlog file too eary
-
- Open
-
-
MDEV-24302 RESET MASTER hangs as Innodb does not report on binlog checkpoint
-
- Closed
-
-
MDEV-28948 FLUSH BINARY LOGS waits/hangs on mysql_mutex_unlock(&LOCK_index) forever
-
- Closed
-
Activity
Transition | Time In Source Status | Execution Times |
---|
|
1d 23h 47m | 1 |
|
2m 46s | 1 |
|
5d 21h 30m | 1 |
|
98d 2h 50m | 1 |