[torqueusers] init_network failed dis and Unauthorized request

Jan jand at uvic.ca
Tue Feb 12 12:40:57 MST 2008


I forgot to mention: I am using version 2.2.1

Jan

Jan wrote:
> Thanks for the quick reply.
> 
> OK, so I uninstalled and reconfigured like this:
> ./configure --prefix=/usr/local --with-rcp=scp 
> --with-default-server=homathko1
> 
> Then I did
> pbs_server -t create
> 
> And finally:
> root at homathko:/usr/local/torque-2.2.1# qmgr homathko
> Max open servers: 4
> Qmgr: set server_name = homathko1
> qmgr: Illegal object type: server_name.
> 
> I am not sure what is going on. Both names (homathko and homathko1) are 
> resolved properly on the machine.
> 
> Thanks, Jan
> 
> 
> Garrick Staples wrote:
>> On Tue, Feb 12, 2008 at 09:40:08AM -0800, Jan alleged:
>>> Hello all,
>>>
>>> I ran into some more challenges. After I had Torque running on a 
>>> single machine for a while, I got a second machine and started 
>>> building a little cluster. I followed 
>>> http://debianclusters.cs.uni.edu/index.php/Main_Page which was great.
>>>
>>> My headnode has two ethernet interfaces:
>>> eth0 is the outside world and resolves to homathko
>>> eth1 resolves to homathko1 and is what the other node sees.
>>> I also created a tourque.cfg with the line
>>> SERVERHOST             homathko1
>>
>> Unless you are using an old version, pbs_server doesn't read 
>> torque.cfg.  Use
>> the 'server_name' server attribute in qmgr.
>>
>>  
>>> OK, so I compile torque with:
>>> ./configure --with-gnu-ld --prefix=/usr/local --with-rcp=scp 
>>> --with-default-server=homathko1
>>
>> --with-gnu-ld?  Which OS is this?
>>
>>  
>>> homathko1 resolves to 192.168.1.200 on eth1
>>> the other name is homathko which resolves to my outside address on eth0
>>>
>>> The I try to create the server database:
>>> root at homathko:/usr/local/torque-2.2.1#pbs_server -t create
>>> pbs_server: network: Address already in use
>>> PBS_Server: PBS_Server, init_network failed dis
>>>
>>> That sounds strange. But after killing the pbs_server and not 
>>> changing anything else, it starts:
>>
>> That's not strange.  Only one daemon can run at a time.
>>
>>  
>>> root at homathko:/usr/local/torque-2.2.1# ps -e |grep pbs
>>>  3248 ?        00:00:00 pbs_server
>>> root at homathko:/usr/local/torque-2.2.1# kill -9 3248
>>> root at homathko:/usr/local/torque-2.2.1# ps -e |grep pbs
>>> root at homathko:/usr/local/torque-2.2.1# pbs_server -t create
>>
>> Perfectly normal.  You have to kill the running daemon before you can 
>> run a new
>> daemon.
>>
>>  
>>> However, if I grep in $PBSHOME for my outside name, it seems as if 
>>> the server starts with my outside name:
>>>
>>> root at homathko:/var/spool/torque# grep -r seos * |less
>>> server_logs/20080212:02/12/2008 
>>> 09:29:30;0006;PBS_Server;Svr;PBS_Server;Server homathko.seos.uvic.ca 
>>> started, initialization type = 4
>>> server_logs/20080212:02/12/2008 
>>> 09:29:30;0004;PBS_Server;Svr;homathko.seos.uvic.ca;cannot open node 
>>> description file '/var/spool/torque/server_priv/nodes' in setup_nodes()
>>> server_logs/20080212:02/12/2008 
>>> 09:29:30;0004;PBS_Server;Svr;homathko.seos.uvic.ca;[continued]
>>>
>>> And I think that is why I get "Unauthorized Request":
>>>
>>> root at homathko:/usr/local/torque-2.2.1# qmgr
>>> Max open servers: 4
>>> Qmgr: set server scheduling=true
>>> qmgr obj= svr=default: Unauthorized Request
>>> Qmgr:
>>
>> 'qmgr homathko' and set server_name.
>>
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> torqueusers mailing list
>> torqueusers at supercluster.org
>> http://www.supercluster.org/mailman/listinfo/torqueusers
> 

-- 
Jan Dettmer, Postdoctoral Fellow
School of Earth and Ocean Sciences, University of Victoria	
Victoria, BC V8W 3P6
office: (250) 472-4342	email: jand at uvic.ca
http://web.uvic.ca/~jand/


More information about the torqueusers mailing list