[torqueusers] s s submit_hosts

Albino A. Aveleda bino at coc.ufrj.br
Fri Nov 6 08:31:10 MST 2009


Hi Michel,

I have a queue in cluster-1 called batch and I have created a routing queue 
batch2 to batch at cluster-1 in cluster-2.

[bibo at adm test]$ qstat -Q
Queue              Max   Tot   Ena   Str   Que   Run   Hld   Wat   Trn   Ext T         
----------------   ---   ---   ---   ---   ---   ---   ---   ---   ---   --- -         
batch                0     0   yes   yes     0     0     0     0     0     0 E         
batch2               0     0   yes   yes     0     0     0     0     0     0 R   

I put in /etc/hosts.equiv both clusters and I tested with command "rsh cluster-1 ls"
and it works.
When I try this command in cluster-2 "qsub -q batch2 job" I received this message in cluster-1.

11/06/2009 12:58:04;0080;PBS_Server;Req;req_reject;Reject reply code=15023(Bad UID for job execution MSG=ruserok failed validating bibo/bibo from cluster-2.mydomain), aux=0, type=QueueJob, from PBS_Server at labcin-cluster-1.mydomain

Why the ruserok failed if the command "rsh cluster-1 ls" in cluster-2 works.

Regards,
Bibo


----- Mensagem original -----
De: "Michel Béland" <michel.beland at rqchp.qc.ca>
Para: "Albino A. Aveleda" <bino at coc.ufrj.br>
Enviadas: Quinta-feira, 5 de Novembro de 2009 14:21:00 (GMT-0300) Auto-Detected
Assunto: Re: [torqueusers] s s submit_hosts

Hello,

> I have two clusters (cluster-1 and cluster-2) and I would like to submit
> jobs to run in cluster-1 from cluster-2.
> Then, I put this configuration in cluster-1 by qmgr:
> 
> set server submit_hosts = cluster-1.mydomain
> set server submit_hosts += cluster-2.mydomain
> 
> But, when I try to create a queue in cluster-2 I don´t have authorization.
> 
> [root at adm pbs]# qmgr
> Max open servers: 4
> Qmgr: c q batch at cluster-1 <mailto:batch at cluster-1>
> qmgr obj=batch svr=cluster-1: Unauthorized Request
> Qmgr:

You are not trying to create a queue on cluster-2, you are trying to 
create a queue on cluster-1 from cluster-2. What you should do instead 
is create a routing queue on cluster-2 and have as sole destination a 
queue on cluster-1.

-- 
Michel Béland, analyste en calcul scientifique
michel.beland at rqchp.qc.ca
bureau S-250, pavillon Roger-Gaudry (principal), Université de Montréal
téléphone : 514 343-6111 poste 3892     télécopieur : 514 343-2155
RQCHP (Réseau québécois de calcul de haute performance)  www.rqchp.qc.ca



More information about the torqueusers mailing list