We have a central manager, submit and execute machine. The submit
and execute machine can not see each other because there is a
private network. Each machine can see the central manager and doing
a status command on any machine shows the execute machine fine.
I can submit jobs from the submit machine to run on the central
manager, and can run jobs on the execute machine if they are
submitted from the central manager. Jobs that are submitted from the
submit machine to run on the execute machine fail and the submit
machine gets an error in it's log file that it's schedd daemon can't
connect to the startd on the execute machine.
We use a vanilla universe, and in reality the jobs are simple
enough that we do not need any info sent back to the submit machine,
but from my research it appears condor is made to work this way
always and is not configurable to not need a shadow daemon and so on.
It appears that GCB would not be a good option as it looks like it
doesn't run on MAC OSX which some of our machines are MACs and
condor-C is for a grid universe and we just wanted to use vanilla.
Based on that and what I've been able to find with google, it
appears I need to look at either Dynamic Port Forwarding or IP
Tunneling ? If there is a better option or I've gotten some of this
wrong please let me know or if someone can confirm that what I've
said is correct that would be helpfull and appreciated.
Thanks
_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx
with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users
The archives can be found at:
https://lists.cs.wisc.edu/archive/condor-users/
![[Computer Systems Lab]](http://www.cs.wisc.edu/pics/csl_logo.gif){kind=logo}