Job stuck at 'submitted' and having problems at 'postproc': u-bh050

[cwc46]

Dear Helpdesk,

I am having trouble with one of my runs - u-bh050 is stuck currently at 'submitted', and I am sometimes getting error with 'postproc' when I kill the job and try to restart.

I am unable to solve the problem, and in particular I am questioning why it is not creating the required directory on my ARCHER /work, which is probably causing it to fail at 'postproc'..

Thank you for your help!

Glen

[ros]

Hi Glen,

I don't know what has happened but you have 7 atoms_main jobs for u-bh050 sitting in the queue on Archer (qstat -u glenchua).

Job ID          Username Queue    Jobname    SessID NDS TSK Memory Time  S Time
--------------- -------- -------- ---------- ------ --- --- ------ ----- - -----
6103300.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103903.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103912.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103919.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103923.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103950.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 
6103953.sdb     glenchua standard atmos_main    --   15 360    --  01:40 Q   -- 

Looks like you have been retriggering the atoms_main tasks and thus submitting multiple instances? So the reason postproc is failing is because the atoms_main for the cycle has not run yet hence being in status submitted.

I would suggest stopping your suite (choosing the option to kill all running tasks). Then on ARCHER use qdel (e.g. qdel 6103953.sdb) l to kill all your queuing jobs in the list above.

Then restart the suite, retrigger the atoms_main task in the 19910801 cycle and change the status of the associated postproc task to waiting. That should get it back on track.

Regards,
Ros.

[ros]

Hi Glen,

Overnight your suite has progressed fine the 6103953.sdb atmos_main finished fine and postproc has run successfully and the next 3 cycles have run.

Cheers,
Ros.

[cwc46]

Dear Ros,

Ah, thank you very much for your help!

Best wishes,
Glen