Login or Register to Ask a Question and Join Our Community


UNIX for Dummies Questions & Answers

g03 problems on rocks 5.3 with sge queue.

 
Thread Tools Search this Thread
Top Forums UNIX for Dummies Questions & Answers g03 problems on rocks 5.3 with sge queue.
# 1  
Old 02-02-2010
g03 problems on rocks 5.3 with sge queue.
Dear all,

I'm a new guy doing some quntum calculation with gaussian on my rocks clusters. Recently, my cluster upgraded to rocks 5.3, but something strange happened. Everything goes correct except a serious problem. when I summit my gaussian job with script to sge queue, the job does run, but with the following error message:
Code:
liuzhencc  /opt/gridengine/default/spool/compute-0-10/job_scripts/12: line 19: g03: command not found

However, the command "g03 gaussian.gjf" runs both on front node and all the computing nodes.

Actually, I reinstalled my cluster, but the problem still there.


Anyone could give some suggestions? Thanks!

PS: Before this upgration, everything goes correct.
Last edited by vbe; 02-03-2010 at 05:15 AM.. Reason: rm bad tags, added code tags
 
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. Shell Programming and Scripting

Submit a qsub script SGE

Hello, I'm working on a SGE cluster and trying to echo the variables passed to the qsub command. My script is as follows #!/bin/bash #$ -V #$ -cwd #$ -pe make 10 if ; then echo "Variable is not set" else echo "Variable says: $StatedUserName" fi and I run the... (1 Reply)
Discussion started by: nans
1 Replies

2. Shell Programming and Scripting

SGE submit script

Hi, I'm trying to edit my working bash script to an SGE script in order to submit it as a job to the cluster. Currently I have: #!/bin/bash # Perform fastqc on files in a specified directory. for ((j=1; j <=17; j++)) do directory=/data4/una/batch"$j"/ ... (0 Replies)
Discussion started by: una1992
0 Replies

3. UNIX and Linux Applications

Looking to reduce the number a cpus available for SGE

Hey all Im looking to reduce the number of cpus available on a certain node in our cluster available for jobs using SGE. i.e. we have one node that has 24 cpus available for jobs on SGE, i would like to reduce that to 16. Thanks (1 Reply)
Discussion started by: olifu02
1 Replies

4. AIX

Print queue problems

I am using AIX 5. Just swapped over an HP printer. The old printer was on a JetDirect box. The new printer connects directly to the network. Using the same IP address to print to, so the only alteration I made was to edit the etc/qconfig file to alter the port setting from 9103 to 9100. I can... (2 Replies)
Discussion started by: charding
2 Replies

5. Shell Programming and Scripting

Request for help with SGE submission script manipulation

Hi, I have the following SGE submission (HPC calculation) script, which is just a Bash script: #!/bin/bash -l #$ -S /bin/bash #$ -l h_rt=1:00:0 #$ -l mem=4G #$ -N XXX #$ -pe qlc 24 #$ -P XXX #$ -wd /home/uccaxxx/Scratch/222PdT/3vac/c0001/ mpirun -m $TMPDIR/machines -np $NSLOTS... (2 Replies)
Discussion started by: crunchgargoyle
2 Replies

6. High Performance Computing

HPL on Rocks 5.3

Hey I am new to cluster benchmarking wondering if there was any guide available that I missed in my google searchs for HPL. (0 Replies)
Discussion started by: futamill
0 Replies

7. AIX

Queue problems

Hi, We have a problem with remote printers installed on our server. Consistently, the printer status is ready but the impressions of the queue are waiting therefore blocked. To unblock the situation, we must launch stopsrc qdaemon-s-s and startsrc qdaemon to reactivate the queue. Thank you... (2 Replies)
Discussion started by: GBAGO
2 Replies

8. SCO

Sco openserver 5 print queue problems

Hwey I manage a Unix box where I ama having multiple printer problems. the only way to get printers to print again is by rebooting. I stop and start services with no luck. Any one have any ideas? It can go weeks with being OK but then any given printer can go and cause the entire print queue to... (6 Replies)
Discussion started by: uk2us88
6 Replies

9. Programming

IPC - queue problems

Hi, I´m having a lot of problems when working with message queues, both on HP-UX Systems and Sun Solaris. When we fill a queue with a messages, the system hangs and locks everything that relies on the use of IPC resources. Anyone knows how to eliminate this problem? Thanks, Haroldo Teixeira (2 Replies)
Discussion started by: haroldo
2 Replies

10. UNIX for Dummies Questions & Answers

Print Queue Problems

Hi All Does anyone know what is meant by PRINTER STATUS CONNECT concerning print queues? At work we run a Perle/Connectix JetStream 8500 (RJ45 Model) to connect our dot matrix printers and often we get the Connect Status. Nobody I speak to knows what the status means. When the printer... (0 Replies)
Discussion started by: nickbungus
0 Replies
Login or Register to Ask a Question

LEARN ABOUT HPUX

sgmgr

sgmgr(1M)																 sgmgr(1M)

NAME

       sgmgr - Serviceguard Manager

SYNOPSIS

       filename ] | COMserver username password cluster_name ...]]  COMserver2

   Remarks
       Serviceguard Manager is the graphical user interface for Serviceguard or Serviceguard Extension for RAC software, Version A.11.12 or later,
       Serviceguard products are not included in the standard HP-UX operating system.

DESCRIPTION

       The command starts Serviceguard Manager, the graphical user interface for Serviceguard clusters.  Serviceguard Manager can be installed	on
       HP-UX,  Linux,  or  Windows.  Serviceguard Manager can be used to view saved data files of a single cluster, or to see running clusters. To
       see the "live" cluster map, Serviceguard Manager connects to a Serviceguard node on the same subnet as those clusters,  specifically  to  a
       part of Serviceguard called the Cluster Object Manager (COM).

   Options
       supports  the  options  listed below. No options are required. If any option is not specified, the user will be prompted to supply it after
       the interface opens.

	      Open a previously saved or example object data file. The file will have the .sgm extension. It can display only  one  cluster.  This
	      option cannot be used with any other options.

	      Specify  the  Serviceguard node that will be the server. This node's COM will query cluster objects running on its subnets, and will
	      report their status and configuration.  Servers with Serviceguard Version A.11.12 or later can monitor clusters. Servers	with  Ser-
	      viceguard Version A.11.14 or later can also perform administrative actions. Servers with Version A.11.16 or later can also configure
	      clusters.  To specify multiple sessions, repeat the -s option.

	      The user login name for the COMserver node. Valid only if COMserver is specified.

	      The user password on the COMserver node. Valid only if username is specified.

	      In creating the map, the COMserver will include the cluster where it is a member.

	      In creating the map, the COMserver will report information
			     about the specified cluster_name(s). Specify clusters with the following cluster software installed:  MC/Serviceguard
			     Version  A.10.10  or  later,  MC/LockManager  Version A.11.02 or later, Serviceguard OPS or Extension for RAC Version
			     A.11.08 or later, and all versions of MetroClusters and ContinentalClusters.

	      To see several clusters, repeat the -c option.

	      If you specify this unused nodes option all COMservers will report information about nodes that have Serviceguard installed, but are
	      not currently configured in any cluster.

	      To connect to another COMserver for another session, repeat the -s options.

AUTHOR

       was developed by HP.

SEE ALSO

       See documents at http://docs.hp.com/hpux/ha including:

       Managing Serviceguard.
       Configuring OPS Clusters with Serviceguard OPS Edition.
       Using Serviceguard Extension for Real Application Cluster (RAC).

Series 700 or 800		    Works with Optional Serviceguard or Serviceguard Extension for RAC Software 			 sgmgr(1M)