[torquedev] [BUG] Torque 3.0.3 and Munge cannot submit jobs from other hosts

["Carlos M. Cámara"]

I reported a similar issue some weeks ago and it seems the problem persists:

My system configuration is:

- Torque 3.0.3
- Munge 0.5.10

Torque installed with this configuration line:

> ./configure --disable-privports --enable-munge-auth 
> --with-server-home=/var/spool/torque-3.0.3

Munge installed and properly configured and obviously testing according 
to the Munge installation doc. Server settings (acl and that stuff) 
properly configured in the pbs_server.

I have the pbs_server in one host and the pbs_mom in another host. When 
I submit a job from the pbs_server host everything works as expected and 
the job properly executes in the pbs_mom host.

Calling to qstat, momctl or any other commands from the pbs_mom works 
perfectly, momctl -d 3 -h pbs_host works too.

But, when I try to submit a job from the pbs_host I get this message:

> qsub: Invalid request MSG=no job owner specified

The same system configuration works as expected in torque 2.4.5.

I think this is a bug of the new versions but I have not found a bug 
tracker to report the issue. Sorry for the inconvenience.

Best Regards.

On 11/11/2011 11:13 PM, torquedev-request at supercluster.org wrote:
> Send torquedev mailing list submissions to
> 	torquedev at supercluster.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://www.supercluster.org/mailman/listinfo/torquedev
> or, via email, send a message with subject or body 'help' to
> 	torquedev-request at supercluster.org
>
> You can reach the person managing the list at
> 	torquedev-owner at supercluster.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of torquedev digest..."
>
>
> Today's Topics:
>
>     1. [PATCH][MINOR] enable-nvidia-gpus in torque.spec.in (Lloyd Brown)
>     2. [Bug 118] Dynamic Consumable Generic Resources dose not work
>        as documented (bugzilla-daemon at supercluster.org)
>     3. [Bug 118] Dynamic Consumable Generic Resources dose not work
>        as documented (bugzilla-daemon at supercluster.org)
>     4. [Bug 118] Dynamic Consumable Generic Resources dose not work
>        as documented (bugzilla-daemon at supercluster.org)
>     5. [Bug 118] Dynamic Consumable Generic Resources dose not work
>        as documented (bugzilla-daemon at supercluster.org)
>     6. [Bug 118] Dynamic Consumable Generic Resources dose not work
>        as documented (bugzilla-daemon at supercluster.org)
>     7. Re: TORQUE dinner at SuperComputing (Ken Nielson)
>     8. Re: TORQUE dinner at SuperComputing (Lloyd Brown)
>     9. Re: TORQUE dinner at SuperComputing (Glen Beane)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 10 Nov 2011 09:27:32 -0700
> From: Lloyd Brown<lloyd_brown at byu.edu>
> Subject: [torquedev] [PATCH][MINOR] enable-nvidia-gpus in
> 	torque.spec.in
> To: Torque Developers mailing list<torquedev at supercluster.org>
> Message-ID:<4EBBFB74.3070200 at byu.edu>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Attached should be a simple patch against Torque 2.5.9 release's
> buildutils/torque.spec.in, to enable the "--enable-nvidia-gpus" option
> during the configure stage when using "rpmbuild".  Basically, just do
> the normal "rpmbuild --with nvidia_gpus ...", which the updated spec
> file will translate to "./configure --enable-nvidia-gpus ..."
>
> This does not include anything for the NVML options as an alternative.
> I can work on that if it's needed, but I'll probably have to ask someone
> for RPM spec file advice.
>
> Any concerns with this patch?  As I said, it's pretty simple.  If not,
> what do we need to do to include it in future releases?
>
> Thanks,

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/torquedev/attachments/20111124/122d4cfe/attachment-0001.html