Handle scheduled task errors #3090
Conversation
|
As with #3077, I am making this PR directly into the |
sunaurus
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
2 times, most recently
from
12c666e to
103f744
Compare
|
Please change the target to main branch, we will publish 0.18 soon and wont make another 0.17 release. |
Got it, will rebase in a moment |
sunaurus
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
2 times, most recently
from
103f744 to
1c7bfd6
Compare
sunaurus
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
from
fc51418 to
28360de
Compare
|
Rebased onto main 👍 Note, my branch also included the changes from #3077 which was already merged but is currently missing in main |
|
These changes on the original 0.17 branch have been verified live in https://lemm.ee - deadlocks are no longer killing the scheduled tasks thread (but deadlocks still occur as expected) |
sunaurus
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
from
28360de to
0bb4a0e
Compare
| Err(e) => { | ||
| error!("Failed to update post_aggregates hot_ranks: {}", e) | ||
| } | ||
| } |
There was a problem hiding this comment.
Would be nice to use .inspect_err() here, but sadly thats only on nightly.
There was a problem hiding this comment.
Wouldn't .map_err() do the same thing here?
There was a problem hiding this comment.
Map functions need to return a value. It would probably work anyway, but might be confusing to read.
|
CI test failures seem unrelated to my changes. I think I can't re-run CI myself, is that correct? |
|
Not sure, anyway I restarted it now. |
Nutomic
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
from
0bb4a0e to
786c586
Compare
sunaurus
force-pushed
the
release/v0.17-handle-scheduled-thread-errors
branch
from
786c586 to
edd45a0
Compare
|
Thank you for submitting this PR. I've also seen similar errors in my logs. Looking forward to seeing this in a release version :) |
Currently, when an error occurs in a scheduled task, it panics and the thread dies while the main lemmy_server process continues working.
This PR removes the panics from scheduled tasks in order to allow the thread to keep working.
This is a partial solution to #3076