Sudden loss of all Schedulers / Jobs #975
Replies: 7 comments
-
Sounds very curious. First I'd enable logging and see if there's something odd happening at the time of disappearance. Secondly if you have such dynamic behavior why don't you use a persistent store? In-memory store is best for applications that know beforehand their tasks and can safely restart on server reboots etc. |
Beta Was this translation helpful? Give feedback.
-
I do understand there is not much for you guys to go on. However, we do have logging enabled and there is nothing special going on. It just happens sporadically from time to time. We have added additional logging just to be safe, I will report them here once it happens again. |
Beta Was this translation helpful? Give feedback.
-
Another note. If there's not triggers left the job will be removed from scheduler if it doesn't have |
Beta Was this translation helpful? Give feedback.
-
Interesting, what could cause cronjobtriggers to disappear ? |
Beta Was this translation helpful? Give feedback.
-
Well if they don't have any firings left 🙂 That would be a cron trigger with fixed year or an end time. |
Beta Was this translation helpful? Give feedback.
-
I don't believe we do have an end time specified. But I will look into that path sometime next week. |
Beta Was this translation helpful? Give feedback.
-
Converting this to a discussion, we can open an issue if we find something actionable. |
Beta Was this translation helpful? Give feedback.
-
Describe the bug
Once in a while (give or take once a month) all of our jobs disappear from the scheduler. We have not found a root cause for it yet. We create / update jobs 'dynamically' at runtime. When looking for the jobkeys, they are suddenly lost. Without rescheduling these jobs 'manually' they are no longer triggered.
Version used
3.0.5
To Reproduce
At a regular interval we do something like:
At some point, this returns false, even if we know that the job was scheduled (it was running for a month before suddenly being gone). Even more, all the jobs are gone and/or no longer triggering.
We do not see any exceptions / issue in our logs. The application does not suffer from memory / cpu pressure at the point of failure. Only that the scheduled jobs no longer fire
Expected behavior
Scheduled jobs not to disappear
Additional context
We are using the RAM store
Beta Was this translation helpful? Give feedback.
All reactions