[MDEV-26059] Query in the range optimizer doesn't react to KILL Created: 2021-06-30 Updated: 2022-06-20 Resolved: 2021-12-06 |
|
| Status: | Closed |
| Project: | MariaDB Server |
| Component/s: | Admin statements |
| Affects Version/s: | 10.2, 10.3, 10.4, 10.5, 10.6 |
| Fix Version/s: | 10.4.19, 10.5.10 |
| Type: | Bug | Priority: | Major |
| Reporter: | Kyle Joiner (Inactive) | Assignee: | Sergei Petrunia |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Issue Links: |
|
||||||||
| Description |
|
Customer has been affected by the inability to stop a run away query that was application generated without causing an outage to restart the server when KILL is ineffective. Even hard kill is not always reliable to stop a query/connection in all situations. An option that kills a query in any state that is not writing a DML or DDL would help immensely. |
| Comments |
| Comment by Sergei Golubchik [ 2021-07-01 ] |
|
KILL is cooperative, it cannot kill a query in any state. It is supposed to kill a query reasonably fast though, if it doesn't — it's a bug. If you have an example of a query that doesn't respond to kill — please, provide it and it'll be fixed. |
| Comment by Sergei Golubchik [ 2021-12-06 ] |
|
Closed, as fixed in |