Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.5, 10.6, 10.7(EOL), 10.8(EOL), 10.9(EOL), 10.10(EOL)
-
None
Description
We should get the change buffer merge process reported in the log In the same way as it was done in MDEV-22577 "innodb_fast_shutdown=0 fails to report purge progress"
Currently, with innodb_fast_shutdown=0, the change buffer merge can take some time, a user might think that the server hans on shutdown as there are no updates in the server log file.
Attachments
Issue Links
- relates to
-
-
- Closed
-
-
-
- Closed
-
-
-
- Closed
-
- mentioned in
-
Page Failed to load
Activity
The revised progress report will look like this:
|
10.6 ae6ebafd819d48c965d2615fc78f1f950e0fbf40 |
2022-11-14 15:44:37 0 [Note] InnoDB: Starting shutdown...
|
2022-11-14 15:44:37 0 [Note] InnoDB: Dumping buffer pool(s) to /dev/shm/10.6m/mysql-test/var/mysqld.1/data/ib_buffer_pool
|
2022-11-14 15:44:37 0 [Note] InnoDB: Buffer pool(s) dump completed at 221114 15:44:37
|
2022-11-14 15:44:52 0 [Note] Completing change buffer merge; 409 page reads initiated; 1 change buffer pages remain
|
2022-11-14 15:45:07 0 [Note] Completing change buffer merge; 409 page reads initiated; 1 change buffer pages remain
|
2022-11-14 15:45:22 0 [Note] Completing change buffer merge; 409 page reads initiated; 1 change buffer pages remain
|
This is from a modified test innodb.ibuf_not_empty, actually reproducing a shutdown hang that is basically the same as MDEV-20934. I filed MDEV-30009 for that.
There actually is some progress reporting for the change buffer merge, once per 60 seconds, implemented in the function srv_shutdown_print_master_pending(). But those messages were not present in the log extract that was posted in
MDEV-29905. This would suggest that the case involves a corrupted change buffer, similar toMDEV-20934.In any case, it does make sense to reduce the reporting interval from 60 to 15 seconds, to be consistent with
MDEV-22577.