You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Apologies if I'm not following "protocol", I'm trying. Clue me in to the group norms and I'll gladly follow them.
I've filed issue 1453, asking how do I submit a fix for the TimeoutOverflowWarning issue. Haven't heard anything since then. Was I supposed to post here instead? In addition? Something else?
TL;dr on issue: agenda can get into a state where a job outside the current period is locked. All locked job's are processed every time period. Even if that time period is more than 2^31 millis in the future. Agenda schedules jobs at the start of each time period by use setTimeout. setTimeout does not like a duration greater than 2^31 millis. This is the proximate cause. Fix for this is to not process locked jobs that are not in the current time period.
It looks like the root cause is agenda not cleaning up properly when nodejs exits in some atypical scenarios. I haven't characterized those exactly, but perhaps things like unhandled exceptions or promise rejections. Possibly more likely: when AWS decides to kill a container. For sure if nodejs is kill 9ed. If possible, agenda should clean up as often as it can. When it can't (kill 9) it needs to recover. Either at startup, or with a fix like mine above.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Apologies if I'm not following "protocol", I'm trying. Clue me in to the group norms and I'll gladly follow them.
I've filed issue 1453, asking how do I submit a fix for the TimeoutOverflowWarning issue. Haven't heard anything since then. Was I supposed to post here instead? In addition? Something else?
TL;dr on issue: agenda can get into a state where a job outside the current period is locked. All locked job's are processed every time period. Even if that time period is more than 2^31 millis in the future. Agenda schedules jobs at the start of each time period by use setTimeout. setTimeout does not like a duration greater than 2^31 millis. This is the proximate cause. Fix for this is to not process locked jobs that are not in the current time period.
It looks like the root cause is agenda not cleaning up properly when nodejs exits in some atypical scenarios. I haven't characterized those exactly, but perhaps things like unhandled exceptions or promise rejections. Possibly more likely: when AWS decides to kill a container. For sure if nodejs is
kill 9
ed. If possible, agenda should clean up as often as it can. When it can't (kill 9) it needs to recover. Either at startup, or with a fix like mine above.Thanks
Beta Was this translation helpful? Give feedback.
All reactions