[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] job goes held

On Oct 12, 2005, at 4:30 PM, Ariff wrote:

1) yes, it's a personal gatekeeper, however i don't
find this error while submitting job with

[arif@Cern1-8 scratch]$ globus-job-run
Cern1-8:40611:/O=Grid/OU=GlobusTest/OU=simpleCA- localhost.localdomain/OU=localdomain/CN=arifsattar
Wed Oct 12 14:28:30 PKST 2005
[arif@Cern1-8 scratch]$

and this is the entry of globus scheduler from submit
file ; it's the same

globusscheduler=Cern1-8:40611:/O=Grid/OU=GlobusTest/OU=simpleCA- localhost.localdomain/OU=localdomain/CN=arifsattar:/jobmanager-fork

No it's not. :-)

The jobmanager name goes before the credential DN.

2) ofcourse, i amn't interested in matchmaking or
glide-in at this point ; i think that i checked (or
perhaps might added) that entry in config file after i
gone through a similar type of poblem description on
condor-users mailing archives.

One thing to watch out for if you don't start a collector: If CONDOR_HOST is set to a machine not running a collector, then the schedd will spend some time in a blocking connect that will timeout. During that time, it won't respond to commands like condor_submit and condor_q. You can set CONDOR_HOST to nothing to stop these connects, but then condor_on/off/restart won't work unless you change HOSTALLOW_ADMINISTRATOR (which is $(CONDOR_HOST) by default).

|            Jaime Frey            |  Public Split on Whether        |
|        jfrey@xxxxxxxxxxx         |  Bush Is a Divider              |
|  http://www.cs.wisc.edu/~jfrey/  |         -- CNN Scrolling Banner |