Re: [HTCondor-users] Condor Watch Queue Announcement

[Stefano Dal Pra <stefano.dalpra@xxxxxxxxxxxx>]

    Hello David, 
    
    maybe you can adapt to your preferences or needs the example awk
    script in attachment, producing something similar: 
    
    Â~$ condor_q -glob -all -cons 'Member(Jobstatus,{1,2})' -af Owner
    '{"0","P","R","X","C","H",">","S"}[JobStatus]' AcctGroup
    RequestCpus RequestMemory | ./aggr.awk 
    ÂÂÂÂÂÂ QUEUEÂÂ SC_RUNÂÂ MC_RUNÂÂÂÂ PENDÂ MEM(GB)
    ÂÂÂÂ tristanÂÂÂÂÂÂ 45ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 668ÂÂÂÂÂÂ 43
    ÂÂÂÂÂÂ aliceÂÂÂÂ 5976ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 840ÂÂÂ 11952
    ÂÂÂ borexinoÂÂÂÂÂ 148ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂÂ 5ÂÂÂÂÂÂÂ 2
    ÂÂÂÂÂÂ virgoÂÂÂÂ 5148ÂÂÂÂ 5148ÂÂÂÂÂ 970ÂÂÂ 17905
    ÂÂÂ clas12voÂÂÂÂÂ 400ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 819
    ÂÂÂÂÂÂÂÂ cmsÂÂÂÂ 6177ÂÂÂÂ 6176ÂÂÂ 43200ÂÂÂ 24706
    ÂÂÂÂ limadouÂÂÂÂÂÂ 42ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 392ÂÂÂÂÂ 328
    ÂÂÂÂÂÂ augerÂÂÂÂÂÂ 11ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂÂ 6ÂÂÂÂÂÂ 10
    ÂÂÂÂÂÂ cuoreÂÂÂÂÂÂ 79ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 186ÂÂÂÂÂ 195
    ÂÂÂÂÂÂÂ na62ÂÂÂÂÂÂ 12ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂ 24
    ÂÂÂ darksideÂÂÂÂÂ 261ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 534
    ÂÂÂÂÂÂÂÂ ctaÂÂÂÂÂ 439ÂÂÂÂÂÂÂ 0ÂÂÂÂ 1767ÂÂÂÂ 1868
    ÂÂÂÂÂÂ xenonÂÂÂÂÂ 128ÂÂÂÂÂ 128ÂÂÂÂÂ 240ÂÂÂÂÂ 512
    ÂÂÂÂÂÂ vligoÂÂÂÂÂ 112ÂÂÂÂÂ 112ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 448
    ÂÂÂÂÂÂÂ lhcbÂÂÂÂ 7487ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 905ÂÂÂ 14974
    ÂÂÂÂÂÂ dampeÂÂÂÂ 1012ÂÂÂÂÂÂÂ 0ÂÂÂÂÂÂ 31ÂÂÂÂ 2024
    ÂÂÂÂÂÂ belleÂÂÂÂ 1403ÂÂÂÂÂÂÂ 0ÂÂÂÂÂ 574ÂÂÂÂ 2806
    ÂÂÂÂÂÂ atlasÂÂÂÂ 6938ÂÂÂÂ 5592ÂÂÂÂ 2496ÂÂÂ 14215
    Â----------
    ÂÂÂÂÂÂ TOTALÂÂÂ 35818ÂÂÂ 17156ÂÂÂ 52280Â 93369.8
    
    The condor_q query only select running or pending jobs, and the awk
    script does not consider username (Owner) only the associated
    AcctGroup, printed as QUEUE. 
    
    Stefano

Il 08/12/20 13:02, David Cohen ha scritto:

Hi,

condor_status -submitter

Gives very partial understanding on how much the cluster is utilized, as there is no way to know from it how many of the jobs are multicore, if any.

There is an unmaintained tool "condor_wn" that gives summary tables, I was hoping to find a way to accomplish the same with an integrated tool.

https://indico.cern.ch/event/686369/contributions/2815630/attachments/1583413/2502661/Htcondor_-_condor_wn.pdf

-David

On Mon, Dec 7, 2020 at 7:07 PM Greg Thain <gthain@xxxxxxxxxxx> wrote:

Hi,

Is there a chance that condor_watch_q will be backported to the 8.8 branch?

I'm in the process of migrating from torque to HTCondor and I'm missing a tool like that to get overall status of system load.

Today I'm using an homemade script that displays other the number of jobs, also how many cores and memory are requested by the jobs:

It doesn't show all the information your current script shows, but you may also want to consider

condor_status -submitter

which is a quick way to get running/idle/held job counts per submitter.

-greg

_______________________________________________
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/

Attachment: aggr.awk
Description: application/awk