Author |
Message |
01/06/2010 12:06:59
|
nkparimi
neo
Joined: 01/06/2010 12:05:12
Messages: 5
Offline
|
Hi,
I’m using Quartz 1.6 with CronTriggers. I’d like to know the following -
Say a trigger t1 is scheduled for 10pm, and trigger t2 is scheduled for 10:05pm. If there are no available workers at 10pm, then t1 is misfired. If workers next get freed up only at 10:10pm, is there a way to ensure t1 is fired before t2?
In short, is there a way to retain the ordering of misfired triggers? Per http://www.quartz-scheduler.org/docs/tutorial/TutorialLesson06.html, it looks like the only two options are either:
1. MISFIRE_INSTRUCTION_FIRE_ONCE_NOW, which appears to be the default. It doesn’t seem to guarantee retaining the ordering
or
2. MISFIRE_INSTRUCTION_DO_NOTHING, I haven’t tried this yet
Is there a way to guarantee ordering of misfired triggers?
thanks,
Nagender
ps: I posted this question to users@quartz.dev.java.net yesterday but don't see my message on the archive page yet (https://quartz.dev.java.net/servlets/BrowseList?listName=users&by=date&from=2010-01-01&to=2010-01-31&first=1&count=3), is that the correct users email list?
|
|
|
01/07/2010 19:49:01
|
jhouse
seraphim
Joined: 11/06/2009 15:29:56
Messages: 1703
Offline
|
Are you using JDBC JobStore ?
That (ordering) seems to have been broken in recent releases within that JobStore.
Please confirm, and I'll file an issue to fix it ASAP.
james
|
|
|
01/07/2010 19:49:41
|
jhouse
seraphim
Joined: 11/06/2009 15:29:56
Messages: 1703
Offline
|
Also, I didn't see your message come through on the mail list, so I'm guessing it was eaten by the mailing list server's spam filters or something.
|
|
|
01/08/2010 00:33:26
|
nkparimi
neo
Joined: 01/06/2010 12:05:12
Messages: 5
Offline
|
Yes, this was using JDBC JobStore on MySQL.
Was the ordering of misfired triggers ever guaranteed? Based on what I found in the tutorial and Javadoc, I didn't see a misfire instruction for CronTrigger that would provide this feature.
thanks,
Nagender
|
|
|
01/08/2010 08:26:50
|
jhouse
seraphim
Joined: 11/06/2009 15:29:56
Messages: 1703
Offline
|
Yes, it was, in the sense that the most outstanding misfired trigger would be updated first, and would hence get the earliest new fire time (depending upon its misfire instruction of course).
|
|
|
01/08/2010 08:50:25
|
nkparimi
neo
Joined: 01/06/2010 12:05:12
Messages: 5
Offline
|
I would love to see this feature restored in Quartz. Is there a bug report I should file? Any ETA for when this feature may be restored?
thanks
|
|
|
01/12/2010 11:36:35
|
nkparimi
neo
Joined: 01/06/2010 12:05:12
Messages: 5
Offline
|
Hi James,
Should I file a bug report for this?
thanks,
Nagender
|
|
|
01/16/2010 09:02:42
|
jhouse
seraphim
Joined: 11/06/2009 15:29:56
Messages: 1703
Offline
|
Sorry, for the late response. I filed a bug a few days ago QUARTZ-827.
Will try to have a 1.7.1 release with the fix soon.
|
|
|
01/16/2010 16:34:04
|
nkparimi
neo
Joined: 01/06/2010 12:05:12
Messages: 5
Offline
|
Fantastic, thanks! I look forward to the 1.7.1 release.
|
|
|
|