Sorry I forgot to post my queue script
# cat qsub
#---------------rareCluster bdimer03job sample script --------
#
#PBS -S /bin/csh
#PBS -q workq
#PBS -N bdimer
#PBS -l nodes=1:fast:nm
#PBS -e bdimer.err
#PBS -o bdimer.out
#
# Insert proper node and walltime specifications above making sure ppn is
# set to either 1 or 2 so that you get either all single processor nodes
# or all dual processors nodes instead of a mixture
# Other node specifications such as memory can also be added
#
# Put in proper WORKDIR, FILENAME in the first two lines below, respectively
# Set SCRFLAG to either GLOBAL or LOCAL in third
# Set NODES, PPN (either 1 or 2) in a manner consistent to the PBS -l
line above
# Put in USER ID for use in scratch directory path
#
# Also make sure %nprocs and/or %nprocl and %mem are set properly in
the com file
# Allow at least 64MB of the memory available per node for the OS
# On multiple node runs use %nprocl; max %mem is the lower value of 512MB or the
# actual amount of memory minus 64MB
# On single node runs use %nproc when using ppn=2; max %mem is based on memory
# of node, again allowing 64MB; the 512MB restriction does not apply
#
setenv WORKDIR /home/vuser/pdg/2007fall/1124
setenv SCRFLAG GLOBAL
setenv NODES 1
setenv USER vuser
#
#
# nothing should need changed below here to run unless you do not
# want to use the default version of Gaussian; in this case the
# g03root path will need to be changed
#
cd $WORKDIR
setenv g03root /home/chemtools
set path = ($path $g03root/g03/linda7.1/intel-linux2.4/bin)
setenv MP_NEWJOB yes
setenv LINDA_CLC network
setenv LINDA_FLC network
cat $PBS_NODEFILE | uniq > $WORKDIR/Gau.hosts
set NODELIST = `xargs echo < Gau.hosts`
cat Gau.hosts | xargs >tsnet.nodes
echo $NODELIST
if ("$SCRFLAG" == "GLOBAL") then
setenv GAUSS_SCRDIR /scratch/$USER/$PBS_JOBID
source $g03root/g03/bsd/g03.login
mkdir -p /scratch/$USER/$PBS_JOBID
endif
if ("$SCRFLAG" == "LOCAL") then
setenv GAUSS_SCRDIR /scratch/$USER/$PBS_JOBID
source $g03root/g03/bsd/g03.login
foreach HOST (`cat $PBS_NODEFILE`)
ssh $HOST mkdir -p /scratch/$USER/$PBS_JOBID
end
endif
echo $GAUSS_SCRDIR
setenv GAUSS_LFLAGS "-nodelist '$NODELIST'"
echo $GAUSS_LFLAGS
# Run the program
foreach FILENAME (bdimer)
if ("$NODES" == "1") then
g03 <$WORKDIR/$FILENAME.com >$WORKDIR/$FILENAME.log
else
g03l <$WORKDIR/$FILENAME.com >$WORKDIR/$FILENAME.log
endif
end
# Clean up scratch space
if ("$SCRFLAG" == "GLOBAL") then
rm -r /scratch/$USER/$PBS_JOBID
endif
if ("$SCRFLAG" == "LOCAL") then
foreach HOST (`cat $PBS_NODEFILE`)
ssh $HOST rm -rf /scratch/$USER/$PBS_JOBID
end
endif
On Wed, Mar 19, 2008 at 10:49 AM, Chunwei Han <[EMAIL PROTECTED]> wrote:
> Thank you for your kindly reply
>
> Right now no user can run any job on the server. And I can ssh any
> nodes without password. I tried to reboot all the cluster when the
> server is running, but it doesn't seem to work.
>
> Here is some debug info as you suggested, I hope this helps :)
>
> # checkjob 21208
>
> checking job 21208
>
> State: Idle (User: vuser Group: relab)
> WallTime: 0:00:00 of INFINITY
> SubmitTime: Wed Mar 19 09:47:04
> (Time Queued Total: 0:02:48 Eligible: 0:00:00)
>
> Total Tasks: 1
>
> Req[0] TaskCount: 1 Partition: ALL
> Network: [NONE] Memory >= 0 Disk >= 0 Swap NC 0
> Opsys: [NONE] Arch: [NONE] Class: [workq 1] Features: [nm][fast]
>
>
> IWD: [NONE] Executable: [NONE]
> QOS: DEFAULT Bypass: 0 StartCount: 0
> PartitionMask: [ALL]
> Flags: RESTARTABLE
>
> job is deferred. Reason: NoResources (exceeds available partition procs)
> Holds: Defer
> PE: 1.00 StartPriority: 2
> cannot select job 21208 for partition DEFAULT (job hold active)
>
>
> # qstat -an
>
> parellel:
> Req'd Req'd
> Elap
> Job ID Username Queue Jobname SessID NDS TSK Memory Time S
> Time
> --------------- -------- -------- ---------- ------ --- --- ------ ----- -
> -----
> 21208.parellel vuser workq bdimer -- 1 1 -- 10000 Q
> --
> --
>
> # qstat -f
> Job Id: 21208.parellel
> Job_Name = bdimer
> Job_Owner = [EMAIL PROTECTED]
> job_state = Q
> queue = workq
> server = parellel
> Checkpoint = u
> ctime = Wed Mar 19 09:47:04 2008
> Error_Path = parellel:/home/vuser/pdg/2007fall/1124/bdimer.err
> Hold_Types = n
> Join_Path = n
> Keep_Files = n
> Mail_Points = a
> mtime = Wed Mar 19 09:47:04 2008
> Output_Path = parellel:/home/vuser/pdg/2007fall/1124/bdimer.out
> Priority = 0
> qtime = Wed Mar 19 09:47:04 2008
> Rerunable = True
> Resource_List.cput = 10000:00:00
> Resource_List.ncpus = 1
> Resource_List.nodect = 1
> Resource_List.nodes = 1:fast:nm
> Resource_List.walltime = 10000:00:00
> Shell_Path_List = /bin/csh
> Variable_List = PBS_O_HOME=/home/vuser,PBS_O_LANG=en_US.UTF-8,
> PBS_O_LOGNAME=vuser,
> PBS_O_PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bi
>
> n:/usr/X11R6/bin:/opt/env-switcher/bin:/opt/mpich-1.2.5.10-ch_p4-gcc/bi
>
> n:/opt/hdf5-oscar-1.6.0/bin/:/opt/kernel_picker/bin:/opt/pvm3/lib:/opt/
>
> pvm3/lib/LINUX:/opt/pvm3/bin/LINUX:/opt/c3-4/:/opt/pbs/bin:/opt/pbs/lib
>
> /xpbs/bin:/home/chemtools/g03.c02/g03/bsd:/home/chemtools/g03.c02/g03/p
>
> rivate:/home/chemtools/g03.c02/g03:/usr/pgi/linux86/bin:/usr/pgi/linux8
> 6/lib:/usr/pgi/linux86/include:/opt/maui/bin:/root/bin,
> PBS_O_MAIL=/var/spool/mail/root,PBS_O_SHELL=/bin/bash,
> PBS_O_HOST=parellel,PBS_O_WORKDIR=/home/vuser/pdg/2007fall/1124,
> PBS_O_QUEUE=workq
> etime = Wed Mar 19 09:47:04 2008
>
>
>
> On Wed, Mar 19, 2008 at 3:09 AM,
> <[EMAIL PROTECTED]> wrote:
> > Send Oscar-users mailing list submissions to
> > [email protected]
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> > https://lists.sourceforge.net/lists/listinfo/oscar-users
> > or, via email, send a message with subject or body 'help' to
> > [EMAIL PROTECTED]
> >
> > You can reach the person managing the list at
> > [EMAIL PROTECTED]
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of Oscar-users digest..."
> >
> >
> > Today's Topics:
> >
> > 1. How to make job run? (Chunwei Han)
> > 2. Re: How to make job run? (Michael Edwards)
> > 3. Re: How to make job run? (Greenseid, Joseph M.)
> > 4. Re: How to make job run? (arun shankar)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Tue, 18 Mar 2008 11:10:37 +0800
> > From: "Chunwei Han" <[EMAIL PROTECTED]>
> > Subject: [Oscar-users] How to make job run?
> > To: [email protected]
> > Message-ID:
> > <[EMAIL PROTECTED]>
> > Content-Type: text/plain; charset=ISO-8859-1
> >
> > Hey guys
> >
> > Recently I take over the job as the cluster administrator. Unluckily
> > the former administrator did not leave any documents and I have no
> > experience at all. Right now the job can not be submitted. Here is
> > some debugger info:
> >
> > # qstat
> > Job id Name User Time Use S Queue
> > ---------------- ---------------- ------------------ -------- - -----
> > 21205.parellel bdimer vuser 0 Q workq
> > 21207.parellel gram0 vuser 0 Q workq
> >
> > # tracejob -n 10 21205
> >
> > Job: 21205.parellel
> >
> > 03/17/2008 22:23:13 S Job Queued at request of [EMAIL PROTECTED],
> > owner = [EMAIL PROTECTED], job name = bdimer, queue = workq
> > 03/17/2008 22:23:13 A queue=workq
> >
> >
> > So, how to make it run?
> > The version of OSCAR is 3.0
> >
> >
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Tue, 18 Mar 2008 07:25:30 -0400
> > From: "Michael Edwards" <[EMAIL PROTECTED]>
> > Subject: Re: [Oscar-users] How to make job run?
> > To: [email protected]
> > Message-ID:
> > <[EMAIL PROTECTED]>
> > Content-Type: text/plain; charset=ISO-8859-1
> >
> > It could be any number of things. Are other users able to run jobs on
> > the cluster?
> > When you started the cluster, did you boot the head node completely
> > before booting the cluster nodes?
> >
> > Other than that, it would probably be an issue with either your user
> > permissions (can you ssh to a node without entering a password?) or
> > your queue script. If you wanted to send your queue script it might
> > help.
> >
> > On Mon, Mar 17, 2008 at 11:10 PM, Chunwei Han <[EMAIL PROTECTED]> wrote:
> > > Hey guys
> > >
> > > Recently I take over the job as the cluster administrator. Unluckily
> > > the former administrator did not leave any documents and I have no
> > > experience at all. Right now the job can not be submitted. Here is
> > > some debugger info:
> > >
> > > # qstat
> > > Job id Name User Time Use S Queue
> > > ---------------- ---------------- ------------------ -------- - -----
> > > 21205.parellel bdimer vuser 0 Q workq
> > > 21207.parellel gram0 vuser 0 Q workq
> > >
> > > # tracejob -n 10 21205
> > >
> > > Job: 21205.parellel
> > >
> > > 03/17/2008 22:23:13 S Job Queued at request of [EMAIL PROTECTED],
> > > owner = [EMAIL PROTECTED], job name = bdimer, queue = workq
> > > 03/17/2008 22:23:13 A queue=workq
> > >
> > >
> > > So, how to make it run?
> > > The version of OSCAR is 3.0
> > >
> > >
> -------------------------------------------------------------------------
> > > This SF.net email is sponsored by: Microsoft
> > > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > > _______________________________________________
> > > Oscar-users mailing list
> > > [email protected]
> > > https://lists.sourceforge.net/lists/listinfo/oscar-users
> > >
> >
> >
> >
> > ------------------------------
> >
> > Message: 3
> > Date: Tue, 18 Mar 2008 08:10:30 -0500
> > From: "Greenseid, Joseph M." <[EMAIL PROTECTED]>
> > Subject: Re: [Oscar-users] How to make job run?
> > To: <[email protected]>,
> > <[email protected]>
> > Message-ID:
> > <[EMAIL PROTECTED]>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > If you're using Maui for scheduling, try running the command `sudo
> checkjob [job-id]` (or run as root if you don't have sudo set up). This
> gives more info than tracejob, and may tell you something a little more
> helpful, like "not enough available resources," etc.
> >
> > --Joe
> >
> > ________________________________
> >
> > From: [EMAIL PROTECTED] on behalf of Chunwei Han
> > Sent: Mon 3/17/2008 11:10 PM
> > To: [email protected]
> > Subject: [Oscar-users] How to make job run?
> >
> >
> >
> > Hey guys
> >
> > Recently I take over the job as the cluster administrator. Unluckily
> > the former administrator did not leave any documents and I have no
> > experience at all. Right now the job can not be submitted. Here is
> > some debugger info:
> >
> > # qstat
> > Job id Name User Time Use S Queue
> > ---------------- ---------------- ------------------ -------- - -----
> > 21205.parellel bdimer vuser 0 Q workq
> > 21207.parellel gram0 vuser 0 Q workq
> >
> > # tracejob -n 10 21205
> >
> > Job: 21205.parellel
> >
> > 03/17/2008 22:23:13 S Job Queued at request of [EMAIL PROTECTED],
> > owner = [EMAIL PROTECTED], job name = bdimer, queue = workq
> > 03/17/2008 22:23:13 A queue=workq
> >
> >
> > So, how to make it run?
> > The version of OSCAR is 3.0
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > Oscar-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/oscar-users
> >
> >
> >
> >
> >
> > ------------------------------
> >
> > Message: 4
> > Date: Tue, 18 Mar 2008 07:48:13 -0700 (PDT)
> > From: arun shankar <[EMAIL PROTECTED]>
> > Subject: Re: [Oscar-users] How to make job run?
> > To: [email protected]
> > Message-ID: <[EMAIL PROTECTED]>
> > Content-Type: text/plain; charset="us-ascii"
> >
> > Try using the command ''qstat -atn'', which will give you more debugging
> information ( extra comment line ), like ''Not Enough Resources Available''
> or ''License not vaild''. Also try using the command ''qstat -f <jobid>'' to
> get more info.
> >
> > If the comment says ''license not valid'', check for the license file ( i
> guess it should be under <HOME directory>/server_priv/license_file if i am
> not wrong )
> >
> > Check for the Queue Settings, # qmgr ( Enter )...Check for ''s q workq
> started=true and s q workq enabled=true'', if neither of this is false, make
> it true.
> >
> > Above are some steps you can perform if jobs go into queue state. Hope
> this will be useful.
> >
> > Regards
> > Arun
> >
> >
> >
> > ----- Original Message ----
> > From: "Greenseid, Joseph M." <[EMAIL PROTECTED]>
> > To: [email protected]; [email protected]
> > Sent: Tuesday, March 18, 2008 9:10:30 PM
> > Subject: Re: [Oscar-users] How to make job run?
> >
> > If you're using Maui for scheduling, try running the command `sudo
> checkjob [job-id]` (or run as root if you don't have sudo set up). This
> gives more info than tracejob, and may tell you something a little more
> helpful, like "not enough available resources," etc.
> >
> > --Joe
> >
> > ________________________________
> >
> > From: [EMAIL PROTECTED] on behalf of Chunwei Han
> > Sent: Mon 3/17/2008 11:10 PM
> > To: [email protected]
> > Subject: [Oscar-users] How to make job run?
> >
> >
> >
> > Hey guys
> >
> > Recently I take over the job as the cluster administrator. Unluckily
> > the former administrator did not leave any documents and I have no
> > experience at all. Right now the job can not be submitted. Here is
> > some debugger info:
> >
> > # qstat
> > Job id Name User Time Use S Queue
> > ---------------- ---------------- ------------------ -------- - -----
> > 21205.parellel bdimer vuser 0 Q workq
> > 21207.parellel gram0 vuser 0 Q workq
> >
> > # tracejob -n 10 21205
> >
> > Job: 21205.parellel
> >
> > 03/17/2008 22:23:13 S Job Queued at request of [EMAIL PROTECTED],
> > owner = [EMAIL PROTECTED], job name = bdimer, queue = workq
> > 03/17/2008 22:23:13 A queue=workq
> >
> >
> > So, how to make it run?
> > The version of OSCAR is 3.0
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > Oscar-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/oscar-users
> >
> >
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > Oscar-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/oscar-users
> >
> >
> >
> ____________________________________________________________________________________
> > Be a better friend, newshound, and
> > know-it-all with Yahoo! Mobile. Try it now.
> http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> >
> > ------------------------------
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> >
> > ------------------------------
> >
> > _______________________________________________
> > Oscar-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/oscar-users
> >
> >
> > End of Oscar-users Digest, Vol 22, Issue 7
> > ******************************************
> >
>
--
+---------------------------------------------------------------------
| Chun-Wei Han
|
| College Of Chemistry and Molecular Engineering, Peking University
| Address : Room 809, New South Chemistry Building
| College of Chemistry and Molecular Engineering
| Peking University, Beijing 100871,
| P.R.China
| Office : +86-10-62751723
| E-Mail : [EMAIL PROTECTED]
+---------------------------------------------------------------------
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Oscar-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/oscar-users
