-
1. Re: TimerService weird behavior
wdfink Jun 6, 2013 9:35 AM (in response to pawtur)From the minute expression it looks wrong
The tutorial show:
The following expression represents every 2 hours starting at noon: hour="12/2"
But this is a weird explanation for the result, as the 0/5 are calculated as from 'current time' every 5 minutes
Also from the expanation in the tutorial it might be wrong as '0'=='*' (from my understanding)
-
2. Re: TimerService weird behavior
pawtur Jun 6, 2013 1:22 PM (in response to wdfink)I think the minute expression is also correct, as minute="0/5" should mean run every 5 minutes starting from minute 0.
Anyway I'm not sure this is related in anyway to the issue.
-
3. Re: TimerService weird behavior
wdfink Jun 6, 2013 2:45 PM (in response to pawtur)What I meant is that this might be a bug.
What if you use */5 in your test, is it than working as expected?
-
4. Re: TimerService weird behavior
pawtur Jun 7, 2013 1:58 AM (in response to wdfink)Unfortunately the result is the same: the first execution time is scheduled in relation to the current time.
-
5. Re: TimerService weird behavior
wdfink Jun 7, 2013 2:41 AM (in response to pawtur)I've re-test that with EAP6.1 and the problem is solved there
-
6. Re: TimerService weird behavior
pawtur Jun 7, 2013 2:45 AM (in response to pawtur)Issue reported https://issues.jboss.org/browse/WFLY-1468
-
7. Re: TimerService weird behavior
wdfink Jun 7, 2013 3:24 AM (in response to pawtur)Forget the last reply
I've overlooked that the start is in the past.
If you use a start date in the future it will work as expected. If the start date is in the past the calculation of the first invocation is wrong.
If you set the start date to the future it will work
I'vechecked that with WildFly upstream and the problem is still present.
Could you file a JIRA?
-
8. Re: TimerService weird behavior
wdfink Jun 7, 2013 3:31 AM (in response to pawtur)Thanks for create the JIRA, I've added some relevant informations.