[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [HTCondor-users] Odd behaviour between pools?



Thanks.
I like the use of   -q   at the end of a condor_submit.
I'd also support the default to Queue 1 when no Queue line present in the file.
Regards
-Ian
-- 
Ian Cottam  | Head of Research Software Engineering | Research IT  | B43 Sackville Street Building  | The University of Manchester  |  M13 9PL  | +44(0)161 306 1851
 

ï-----Original Message-----
From: HTCondor-users <htcondor-users-bounces@xxxxxxxxxxx> on behalf of John M Knoeller <johnkn@xxxxxxxxxxx>
Reply to: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Date: Wednesday, 11 March 2020 at 14:57
To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: Re: [HTCondor-users] Odd behaviour between pools?

    IMO Dropping the -a is better.  This puts the qnum=20 statement at the *top* of your submit file
    so it can be referenced by *if* and *include* statements.
    While  -append puts qnum=20 it at the bottom, just before the queue statement.
    
    It's that 'just before' that's the issue here.  The submit parser doesn't *know* its gotten to the queue statement
    until it reads it, so there was a bug for a while in the 8.3 series where the $() expansion of the QUEUE statement
    happened before -append arguments were applied. 
    
    I would recommend that you get in the habit of using -append only if you are using such an old condor_submit that you don't have a choice,
    or when you need to override a statement in the submit file that has not been prepared to be overwritten.
    
    What I do with my submit files these days is to leave the queue statement off entirely, and then supply -queue as a command line argument
    
    condor_submit my.sub -q 1
    
    In support of this pattern, I have considered changing condor_submit so that it does not require a queue statement. 
    Instead it would treat any submit file without a queue statement as having an implicit queue 1 at the end.  
    
    I would be interested in feedback from the community on this idea.
    
    -tj
    
    
    HTCondor-users mailing list
    To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
    subject: Unsubscribe
    You can also unsubscribe by visiting
    https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users
    
    The archives can be found at:
    https://lists.cs.wisc.edu/archive/htcondor-users/
    
    _______________________________________________
    HTCondor-users mailing list
    To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
    subject: Unsubscribe
    You can also unsubscribe by visiting
    https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users
    
    The archives can be found at:
    https://lists.cs.wisc.edu/archive/htcondor-users/