[Aces-support] Error with submitting job

aces-admin at techsquare.com aces-admin at techsquare.com
Mon Oct 1 10:26:08 EDT 2007


hello pop-

as you know, i really need a bit more data
here -- machine-name, job-id, something to 
go on.

thank you 

[greg]


> Date: Mon, 01 Oct 2007 10:22:07 -0400
> From: Chinnawat Surussavadee <surusc at MIT.EDU>
> Reply-to: pop at alum.mit.edu
> Cc: ACES-support at mitgcm.org
> MIME-Version: 1.0
> 
> Hello Greg,
> 
> Thank you for the reply.
> 
> 1. I just tried to qsub the pbs job again on geo and got the same 
> error.  By the
> way, the pbs job will read/write data from/to /net/ds-01/scratch-0/.  I also
> tried to submit similar pbs job on itrda as you suggested and got the same
> error attached below.
> 
> [surusc at itrda tbamsua_nadir]$ more February_12_9.e18899
> ModuleCmd_Load.c(199):ERROR:105: Unable to locate a modulefile for 'man'
> ??c(3):ERROR:105: Unable to locate a modulefile for 'modules'
> ??c(4):ERROR:105: Unable to locate a modulefile for 'rsh'
> ??c(5):ERROR:105: Unable to locate a modulefile for 'torque'
> $(7):ERROR:105: Unable to locate a modulefile for 'matlab'
> /var/torque/mom_priv/jobs/18899.itrda.SC: line 15: matlab: command not found
> 
> 2. I still could not get into /net/ds-0a/raid1/ as we can see error message
> below.
> 
> [surusc at geo ~]$ cd /net/ds-0a/raid1/
> -bash: cd: /net/ds-0a/raid1/: No such file or directory
> 
> Thanks,
> 
> Pop
> 
> 
> 
> Quoting aces-admin at techsquare.com:
> 
> > hello pop-
> >
> >> 1. I tried to submit a job using qsub, but I got some errors.  Could anyone
> >> please point out how I should fix this problem?  Both pbs job and 
> >> error message
> >> are shown below.  Please note that this same pbs job worked fine 
> >> before (some
> >> time last year)
> >
> > looks like transient network issue.
> >
> > for what it's worth, the itrda nodes are much
> > closer to ds-0a than the geo-nodes.
> >
> >> 2. I could not access /net/ds-0a/raid1/.  Could the admin please 
> >> check if the
> >> disk is mounted properly?
> >
> > yes, it is. if there was lots of network
> > activity, though, this may have timed-out.
> >
> > i have just verified that things at ds-0a
> > are well and will check out the several switches
> > throughout the day today.
> >
> > [greg]
> >
> >
> 
> 



More information about the Aces-support mailing list