[FIX] Drop Blackout Window after First Try #162
Conversation
There was a problem hiding this comment.
@AlexanderKanakis I'm quite not understanding how cancelling event scheduler (timer) using event name cancelAllEventSchedulerJobForSessionAndName fixes the issue.
Can you please help me understand? Did you find any root cause? Thanks...
I have not been able to find a root cause for the issue. What I did notice during the meeting on Thursday is that the blackout window timeout scheduler works perfectly multiple times if the timer is longer than the event scheduler timer. Which means that something in the event scheduler interrupts it. So I tried killing the event scheduler and it worked (at least locally). |
There was a problem hiding this comment.
@AlexanderKanakis I think the issue is still there. In the mongo DB I'm not seeing a continue blackout timeout timer scheduled in the second test.
Steps:
- DF sends timer event and App schedules both jobs( timeout:120 s, timer event: 180 s)
- after 120 s, App cancels the event-scheduler job.
- When DF again sends a timer event with 180 seconds, it clears past completed jobs and schedules only the event scheduler job. So in the second time testing it there is no timeout job scheduled.
@AlexanderKanakis Can you please check at your end?
|
NOTE: Maybe this design needs to be reconsidered...... i.e. if we are trying to protect from a scheduler issue, it does not make sense to use the scheduler? chicken/egg? Need to think about this |
For issue: WideChat/Rocket.Chat#1225