Details
-
Bug
-
Status: Open (View Workflow)
-
Major
-
Resolution: Unresolved
-
10.4.12
-
None
-
None
-
debian
Description
Coming from: https://bugs.mysql.com/bug.php?id=99157 and https://jira.percona.com/browse/PS-6961
I have been done the same tests as reported on https://jira.percona.com/browse/PS-6961 with a host that has no activity apart from these events.
After 3 days the memory increased from:
First number: VIRT column reported on top
Second number: RES column reported on top
407.2g 68.4g
|
to
410.6g 71.7g
|
After stopping the event scheduler on that host and after 4 days the memory never grew anymore.
I have only tested this with 10.4.12 but we've had lots of memory issues on a host running 10.1 that has around 2.5k events which after a while gets OOM so it is likely possible that it is the same thing.
Unfortunately, MariaDB doesn't provide memory_global_by_current_bytes tables on P_S/sys so such a detailed profile isn't possible.
Attachments
Issue Links
- relates to
-
MDEV-25176 SHUTDOWN-caused SIGSEGV's, SIGABRTS's, error 22 and heap-use-after-free when using event_scheduler
- Confirmed