Say I have a job that runs once a week at a configurable time, which
is stored in a DB. During application startup, I use the scheduler to
schedule the job for next Monday at noon.
Then, someone uses a web interface to indicate the job should run
tomorrow at noon instead.
I'd like to be able to ask the scheduler "give me the job that's
responsible for X", then unschedule that job and reschedule it at the
new time.
So, my feature request is to be able to tag jobs with an identifier so
that I can request them by that identifier instead of by the job ID I
get when I schedule it, since I'm not stashing that away anywhere.
Yes, I can stash the ID away, but this seems like something that would
be useful. Does anyone agree? :-)
I'm undecided whether you can have multiple jobs with the same tag or
not, and if so, does the "give me the job with this tag" returns an
array, etc.
Thanks,
Lee
Hey, it's an interesting idea.
But one day or the other, someone will end up asking for this "tag
database" to be persisted and then the scheduler will slowly become
something very fat... I exaggerate...
In OpenWFEru (the workflow and BPM engine), the object that scheduled
a task keeps the job id reference (to unschedule it at will). You
might think of your job as resources linked to a scheduler trigger
(via the job id). Right, this task resource is out of the scheduler,
but well, the scheduler only cares about scheduling stuff...
Somehow, it's an interesting idea... I would implement it "memory
only", no persistence (like the whole scheduler), String tags.
job_ids = scheduler.find_job_ids 'elephant'
What do you think ?
Best regards,
--
John Mettraux -///- http://jmettraux.openwfe.org
Thanks,
Lee
On Oct 3, 7:17 pm, "John Mettraux" <jmettr...@openwfe.org> wrote:
http://rubyforge.org/tracker/index.php?func=detail&aid=14439&group_id=2609&atid=10026
OK, best regards,
Hi Lee,
it's implemented and available via gem.
http://jmettraux.wordpress.com/2007/10/07/tags-for-scheduled-jobs/
Best regards and thanks for the suggestion,