Details
-
Task
-
Status: Open (View Workflow)
-
Major
-
Resolution: Unresolved
-
None
-
None
Description
The configuration looks like this:
systemctl edit mariadb.service
add the following for example
|
systemctl edit mariadb.service
|
|
[Service]
|
MemoryMax=18000M
|
MemorySwapMax=750M
|
|
|
systemctl daemon-reload
|
|
or vi(m) / nano
|
|
nano /etc/systemd/system/mariadb.service.d/override.conf
|
If the memory limit is reached, the server crashes and restarts.
This memory limit behaves as if there was no more memory available on the system.
Could memory drain intelligence be improved?
Attachments
Issue Links
- relates to
-
MDEV-24670 avoid OOM by linux kernel co-operative memory management
-
- Closed
-
Yes, MariaDB needs to have better checks for out-of-memory conditions.
In case of bigger allocations, MariaDB will detect that and give an error to end user.
However if one cannot allocate any memory even for small things, then MariaDB may crash.
This is something we have to improve over time.