fanchun at pku.edu.cn
Thu Apr 13 20:35:46 MDT 2006
Would like to test if you can rsh each nodes correctly? I got the same =
when I could just ssh each nodes but not rsh. After I changed come =
to be able to rsh, it run well.
Peking University, P.R.China
=B7=A2=BC=FE=C8=CB: torqueusers-bounces at supercluster.org
[mailto:torqueusers-bounces at supercluster.org] =B4=FA=B1=ED Neelesh Arora
=B7=A2=CB=CD=CA=B1=BC=E4: 2006=C4=EA4=D4=C213=C8=D5 7:59
=CA=D5=BC=FE=C8=CB: torqueusers at supercluster.org
=D6=F7=CC=E2: [torqueusers] job submission from remote hosts
I have a question regarding the default behavior of acl_host_enable and=20
acl_hosts parameters in torque server and queue config. According to the =
manuals, default value of acl_host_enable is false, which allows all=20
remote hosts to submit jobs to the server. But this isn't working.
We have torque-2.0.0p2 and without changing the default values of these=20
parameters, a remote host (with pbs client commands only) fails to=20
submit jobs to our main pbs_server. The error returned is "qsub: Bad UID =
for job execution". The passwd file entries are the same on these two=20
machines, so the UID does match.
Am I missing something here?
There are references to the ALLOWCOMPUTEHOSTSUBMIT and SUBMITHOSTS=20
parameters in the torque.cfg file, at other places on this forum. Are=20
these torque.cfg entries required, even with the above mentioned pbs=20
configs? What's the relation between these two sets of config =
torqueusers mailing list
torqueusers at supercluster.org
More information about the torqueusers