Hi <br>Thanks for your suggestion. <br>I
forgot to mention the routing queue. I created the queue route which is
act as a routing queue, and my server default_queue was route. I
configured the queues through acl_hosts (not acl_users, because single
guy will submit all types of jobs).
<br>And finally onemore thing I was not clear about resources_max. In
the webpage it gave "set queue batch resources_max.nodect=16". Actually
what its saying, and what is mean by nodect. Can you please explain me
about resources_max. In webpage it given "specifies the maximum
resource limits for jobs submitted to the queue". Here resource limits
means?. Is it depends on memory or walltime like that? or anything
else?. <br><br>Thanks<br>Velan<br><br><div><span class="gmail_quote">On 1/30/07, <b class="gmail_sendername"><a href="mailto:garrick@clusterresources.com">garrick@clusterresources.com</a></b> <<a href="mailto:garrick@clusterresources.com">
garrick@clusterresources.com</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">On Mon, Jan 29, 2007 at 10:33:17PM +0530, Vadivelan Ranjith alleged:
<br>> Hi<br>> I'm sorry if it discussed earlier.<br>> I created new queues(long, short and interactive). Each queue having<br>> individual nodes. when I submitted using qsub -q <queue name> <job file>,
<br>> jobs are going to individual nodes correctly. But i am facing some<br>> difficulties.<br>> Assume following queues have the particular nodes<br>> queue long = node01, node02<br>> queue short = node03<br>
> queue interactive = node04, node05<br>><br>> 1. i submitted job interactively and job went to particular node04. i ran<br>> the job interactively using following command<br>> mpirun -machinefile hosts -np 2 ./jobfile.
<br>> hosts file having node04 & node05. Jobs are running fine. If i change the<br>> host file to node01 & node02, jobs are running on node01 & node02. But<br>> node01 & node02 are assigned for queue long.
<br>> I want to run particular queue to particular nodes. I dont want people to<br>> hack other nodes.<br><br>TORQUE can't prevent ssh/rsh logins, because those are out of its<br>control. If you are using an OS with PAM, you can do some fun things
<br>there. See src/pam/README.pam or contrib/README.pam_authuser for a first<br>direction.<br><br><br>> 2. I want to define the interactive queue only for interactive mode. I<br>> should not submit job, through interactive mode.
<br>> for example qsub -q interactive jobfile, now jobs started running<br>><br>> 3. if i submit job using, qsub -q long -I -l nodes=1:ppn=2, then jobs<br>> started in interactive mode using queue long. I want to use long and short
<br>> queue for submitting jobs(non-interactive mode) and interactive queue for<br>> only interactive jobs.<br><br>Job routing occurs based on resource requests and ACLs, not based on<br>other job attributes.<br><br>
Others have argued for routing based on interactive or batch, but I<br>still maintain that it is a bad idea.<br><br><br>> I dont know i did any mistake while configuring torque. I not yet configured<br>> maui except one command.
<br>> JOBNODEMATCHPOLICY EXACTNODE<br>> Can anybody help me how to configure queue as i mention above.<br><br>I don't see any misconfigs here. Maui doesn't know anything about job<br>routing in TORQUE.<br>
<br><br>_______________________________________________<br>torqueusers mailing list<br><a href="mailto:torqueusers@supercluster.org">torqueusers@supercluster.org</a><br><a href="http://www.supercluster.org/mailman/listinfo/torqueusers">
http://www.supercluster.org/mailman/listinfo/torqueusers</a><br></blockquote></div><br>