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

[Condor-users] Job remaining idle when adding SEC_DEFAULT_AUTHENTICATION

I have a submit host that is submitting to a local pool and flocking to others. The local pool is composed by a host with the schedd; a host running collector, negotiator and some startds (central host); Other hosts running startds.

Jobs run fine on the clusters where the schedd is flocking.

If I enable security on the central host:
then jobs don't run on the pool.

From the schedd the jobs report "Request has not yet been considered by
the matchmaker" (condor_q -analyze)

The schedd and the daemons on the central host (collector, negotiator, startd) report that the job was matched (is in the MatchLog) but then there is no track of it in any StarterLog* file.
I don't see error messages.

If I comment the two SEC_... lines above then jobs run OK.

The nodes are in the same networks (both have 2 NICs), share the home directories, have same UID_DOMAIN (the submit host FQDN) and FILESYSTEM_DOMAIN

An inconsistency I noted is that both these hosts are dual homed and even if I set NETWORK_INTERFACE to the public IP, condor_status is printing the name associated with the private IP.

Any suggestion on what is going on or how to find out more?
Thank you,