I have had to grapple with this and ended up adding a parallel class that destroys the JobEntry after it has run the once. It does this when the SchedulerJob subclass associated with the JobEntry runs and tells it to do it. The user still has a record of the CRON details so can run again by putting in a new time (in the future) to run and the parallel class re-creates the Job Entry object.
But this is pretty ugly as keeping a track of whether the related JobEntry exists or not is a pain. I think I noticed that you can set some illegal values for some of the CRON attributes, like day of month is set to 32 and it remains in the database but does not run, some other time values however cannot be used for this because the SchedulerService complains. So then you just need the ScheduledJob subclass to change the oneshot JobEntry CRON attribute value after running the job to makes sure it does not run again. The only problem is cleaning up these non running jobs at some point. Hope this helps David -----Original Message----- From: Pugh, Eric [mailto:[EMAIL PROTECTED]] Sent: 27 December 2001 22:54 To: 'Turbine Users List' Subject: One Time run of Scheduler Job? Hi, I want to be able to queue up a job to run exactly once, sometime in the future... I would like to use the Scheduler framework that comes in TDK 2.1... I can see how I would create a jobentry that maybe was set to sometime in the next month based on putting in a day_mo value, and populating the min and hour values. However, that would cause it to run every month... I am using the TurbineNonPersitentSchedulerService, so as long as the server gets bounced in the next 31 days, I would be in the clear, but is there any other more elegent way? Eric -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
