Re: [HTCondor-users] mpi job stuck as idle

[Jason Patton <jpatton@xxxxxxxxxxx>]

This output shows that your config changes haven't been reflected on
the execute nodes. Your modified
condor_config.local.dedicated.resource should go in
/etc/condor/config.d/ on the execute machine, which is where condor
looks for supplemental config files.

If you're on the execute machine that you're configuring, you can do
"condor_config_val -config" to see which config files are being picked
up. For example:

$ condor_config_val -config
Configuration source:
/etc/condor/condor_config
Local configuration sources:
/etc/condor/config.d/condor_config.local.dedicated.resource
/etc/condor/config.d/condor_config.local.dedicated.submit
/etc/condor/condor_config.local

If you see your condor_config.local.dedicated.resource listed, then
you should issue a condor_reconfig or restart condor.

Jason

On Mon, Jan 22, 2018 at 10:43 AM, Mahmood Naderan <nt_mahmood@xxxxxxxxx> wrote:
>
>>What if you run:
>>
>>condor_status -af:h Machine DedicatedScheduler
>>
>>This will show the value of DedicatedScheduler (and Machine) for each
>>slot on each execute machine.
>
>
> Please see the  output below.
>
>
>
>
> [root@rocks7 mahmood]# condor_status -af:h Machine DedicatedScheduler
> Machine           DedicatedScheduler
> compute-0-0.local undefined
> compute-0-0.local undefined
>
>
> Regards,
> Mahmood
>
>