Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.5
-
None
Description
The following intermittent result difference was observed immediately after MDEV-16678 was pushed to 10.5.
CURRENT_TEST: main.backup_interaction
|
--- /usr/share/mysql/mysql-test/main/backup_interaction.result 2019-12-10 14:25:41.000000000 -0500
|
+++ /dev/shm/var/4/log/backup_interaction.reject 2019-12-10 16:16:44.997159774 -0500
|
@@ -105,6 +105,7 @@
|
select lock_mode from information_schema.metadata_lock_info;
|
lock_mode
|
MDL_BACKUP_START
|
+MDL_SHARED
|
backup stage block_commit;
|
commit;
|
backup stage end; |
According to the buildbot cross-reference, this is the very first time the failure was observed. There were 2 failures on the very first push:
http://buildbot.askmonty.org/buildbot/builders/kvm-deb-bionic-ppc64le/builds/2948/steps/mtr/logs/stdio
http://buildbot.askmonty.org/buildbot/builders/kvm-deb-eoan-amd64/builds/234/steps/mtr/logs/stdio
I am able to repeat it locally:
|
10.5 adb117cf6901a0e0da7577267be1209359892207 |
main.backup_interaction 'innodb' w6 [ 7 fail ]
|
Test ended at 2019-12-11 09:15:35
|
|
|
CURRENT_TEST: main.backup_interaction
|
--- /mariadb/10.5/mysql-test/main/backup_interaction.result 2019-10-28 17:09:53.002690675 +0200
|
+++ /mariadb/10.5/mysql-test/main/backup_interaction.reject 2019-12-11 09:15:35.661386904 +0200
|
@@ -105,6 +105,7 @@
|
select lock_mode from information_schema.metadata_lock_info;
|
lock_mode
|
MDL_BACKUP_START
|
+MDL_SHARED
|
backup stage block_commit;
|
commit;
|
backup stage end;
|
This was an optimized ASAN debug build with
ASAN_OPTIONS=abort_on_error=1 ./mtr --repeat=100 --parallel=auto main.backup_interaction{,,,,,,,,,,,}
|
Attachments
Issue Links
- relates to
-
MDEV-16678 Use MDL for innodb background threads instead of dict_operation_lock
-
- Closed
-