Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
-
None
Description
(Initially found by Elena)
I'll attach a testcase, which eventually gets three threads to be stuck in "Waiting for table level lock" state.
The testcase must be run like this: ./mysql-test-run t/t45.test --repeat=100 (can also add --testcase-timeout=600 to ease debugging). Execution will get stuck after a minute. On my machine (pylon-fedora15), only release build gets stuck, and only when the server is started by MTR.
Being stuck looks like this:
+----+------+-----------------+------+---------+------+------------------------------+------------------------------------------------+----------+
|
| Id | User | Host | db | Command | Time | State | Info | Progress |
|
+----+------+-----------------+------+---------+------+------------------------------+------------------------------------------------+----------+
|
| 23 | root | localhost | test | Sleep | 17 | | NULL | 0.000 |
|
| 24 | root | localhost | test | Query | 17 | Waiting for table level lock | ANALYZE TABLE E | 0.000 |
|
| 25 | root | localhost | test | Query | 17 | Waiting for table level lock | ANALYZE TABLE A | 0.000 |
|
| 26 | root | localhost | test | Query | 17 | Waiting for table level lock | SELECT * FROM mysql.index_stat, test.E LIMIT 1 | 0.000 |
|
| 27 | root | localhost:52086 | NULL | Query | 0 | NULL | show processlist | 0.000 |
|
+----+------+-----------------+------+---------+------+------------------------------+------------------------------------------------+----------+
|
Attachments
Issue Links
- is duplicated by
-
-
- Closed
-
- relates to
-
MDEV-3806 Engine independent statistics
-
- Closed
-