Re: [help-gnubatch] ( [PATCH] Bump version in debian/changelog to 1.4 also ) + newbie question

[Reuti]

Am 06.06.2012 um 02:10 schrieb Peter Valdemar Mørch:

> Simple enough patch...
> 
> Is there a searchable archive anywhere?
> 
> Newbie questions about whether GNUBatch is for us:
> 
> We're looking for a scheduler similar to the one built-in to nagios
> that allows us to run many small jobs where the scheduler takes care
> of determining when to actually run them, we just want to be
> guaranteed that they all get run at least every 15 minutes (as an
> example).  And that as jobs get added/removed, the scheduler
> automatically re-arranges the scheduling to smoothe out the load. I'd
> rather we didn't get into the specifics of "job 458 needs to run at
> 15:23:33", but more "these 1200 jobs all need to run every 15
> minutes".
> 
> Other jobs are more "normal" and need to run exactly at specific
> times. These other "normal" jobs should not interfere with the "small
> job pool" above.
> 
> We'd like later to be able to view the output ($? + stdout + stderr)
> of specific job runs, see when they started and examine how long they
> took. Because of the number of jobs, this info needs to expire /
> rotate / be deleted after some time.
> 
> Is GNUBatch the tool for us? If so, how do we achieve the "small job
> pool"? I'm guessing with variables, but how?
> 
> Also, I read in the GNUbatch reference manual, that "The time at which
> a job is scheduled to start can be specified by date and time to the
> nearest minute" :-( The lack of second-specific scheduling leads me
> towards "not for us" - is that a hard limit, can it be overcome

You could put a sleep with the necessary amount of seconds in the jobscript and 
start it at the full minute before, so that you can start it at any second.

Is it to run in a cluster of machines or only on a local one? To me it sounds 
like you need some features from GridEngine (like the accounting which you can 
access later), and on the other hand some kind of "real-time-feature" from 
GNUbatch.

Do you have more machines than jobs, i.e. all 1200 jobs should run at the same 
time on a bunch of machines?

-- Reuti


> somehow, or is it solvable by some "small job pool" pattern above?
> 
> If GNUBatch isn't really for us, can anybody recommend something else
> given my outline above?
> 
> Peter
> 
> -- 
> Peter Valdemar Mørch
> http://www.morch.com
> <0001-Bump-version-in-debian-changelog-to-1.4-also.patch>