[torqueusers] unable to contact node, Connection refused

[Garrick Staples]

On Fri, Nov 04, 2005 at 08:50:08AM -0500, Alejandro Hurtado Turi?o alleged:
> Hi,
> I've installed a torque-1.1.0p6 on a cluster, but the jobs don't run 
> unless forced w/ qrun I'm not planning on installing Maui
> and just using the default fifo scheduler (pbs_sched)

1.1.0p6 is really old.  There have been countless improvements since
then.


> The pbs server log say at start up: 
> 10/31/2005 13:54:28;0006;PBS_Server;Svr;PBS_Server;Using ports Server:
> 15001  Scheduler:15004  MOM:15002
> 10/31/2005 13:54:28;0002;PBS_Server;Svr;PBS_Server;Server Ready, pid = 
> 2317
> 10/31/2005 13:54:28;0004;PBS_Server;Svr;WARNING;!!! unable to contact node 
> grid1 !!!
> 10/31/2005 13:54:28;0001;PBS_Server;Svr;PBS_Server;Connection refused 
> (111) in contact_sched, Could not contact Scheduler - port 15004

Is pbs_mom running on grid1?  Is pbs_sched running on the server?

> ---
> grid1 is the pbs server with pbsmon installed.
> no firewall
> my mom-priv/config
> $clienthost    grid1
> $logevent      255
> $restricted    grid1
> $usecp         *:/data /data

Is grid1 a node or server?  The information above is confusing.

The server logs indicate that it is a node.  The MOM config looks like
grid1 is the server.

And you don't need the $restricted line, that just weakens security.


> Looking for it in the web, i see the problem is common but notbody answer 
> it.
> could anybody helpme please!??
> thanks 
> ale

These kinds of things are just config errors that are hard to diagnose
over email.  Eventually the admin figures it out and doesn't tell anyone
:)

-- 
Garrick Staples, Linux/HPCC Administrator
University of Southern California
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20051104/bababcab/attachment.bin