[torqueusers] Scheduler problem in Torque or maui?

Anna Jonna Armannsdottir annaj at hi.is
Thu Dec 20 06:00:50 MST 2007


Hi!
I am trying out the maui scheduler version maui-3.2.6p16 . 
The problem is that somehow it can not schedule jobs 
to run. 

checkjob 41349 gives: 
job is deferred.  Reason:  RMFailure  (job cannot be started - cannot
set hostlist)

Checking the Torque logs, reveals that the maui scheduler 
requests are coming from localhost.localdomain, which is not
authorized in my configuration. 
I would rather not change Torque configuration to allow this, 
so how can I get maui to bind to the hostname jotunn.rhi.hi.is, 
which is the real hostname?

In maui.cfg is:

SERVERHOST		jotunn.rhi.hi.is
ADMIN1			root 
#ADMIN1			adm
ADMIN3			ALL

ADMINHOST		jotunn.rhi.hi.is

Root is not registered in torque as administrator or operator, but
adm is. There are, however technical problems in getting maui to run 
under the adm username. 

Here is an excerpt from Torque Server logs: 

12/20/2007 12:34:05;0100;PBS_Server;Job;41349.jotunn.rhi.hi.is;enqueuing
into short, state 1 hop 1
12/20/2007 12:34:05;0008;PBS_Server;Job;41349.jotunn.rhi.hi.is;Job
Queued at request of user at jotunn.rhi.hi.is, owner =
user at jotunn.rhi.hi.is, job name = ORTE_nwchem, queue = short
12/20/2007 12:34:05;0040;PBS_Server;Svr;jotunn.rhi.hi.is;Scheduler sent
command new
12/20/2007 12:34:06;0100;PBS_Server;Req;;Type StatusNode request
received from root at localhost.localdomain, sock=10
12/20/2007 12:34:06;0100;PBS_Server;Req;;Type StatusQueue request
received from root at localhost.localdomain, sock=10
12/20/2007 12:34:06;0100;PBS_Server;Req;;Type StatusJob request received
from root at localhost.localdomain, sock=10
12/20/2007 12:34:06;0100;PBS_Server;Req;;Type ModifyJob request received
from root at localhost.localdomain, sock=10
12/20/2007
12:34:06;0020;PBS_Server;Job;41349.jotunn.rhi.hi.is;Unauthorized
Request, request type: 11, Object: Job, Name: 41349.jotunn.rhi.hi.is,
request from: root at localhost.localdomain
12/20/2007 12:34:06;0080;PBS_Server;Req;req_reject;Reject reply
code=15007(Unauthorized Request ), aux=0, type=ModifyJob, from
root at localhost.localdomain
12/20/2007 12:34:06;0100;PBS_Server;Req;;Type DeleteJob request received
from root at localhost.localdomain, sock=10
12/20/2007
12:34:06;0020;PBS_Server;Job;41290.jotunn.rhi.hi.is;Unauthorized
Request, request type: 6, Object: Job, Name: 41290.jotunn.rhi.hi.is,
request from: root at localhost.localdomain
12/20/2007 12:34:06;0080;PBS_Server;Req;req_reject;Reject reply
code=15007(Unauthorized Request ), aux=0, type=DeleteJob, from
root at localhost.localdomain

-- 
Kindest Regards, Anna Jonna Ármannsdóttir,       %&   A: Because people read from top to bottom.
Unix System Aministration, Computing Services,   %&   Q: Why is top posting bad?
University of Iceland.



More information about the torqueusers mailing list