Re: [Condor-users] Condor pool - switched some laptops to 10.99

[Alan De Smet <adesmet@xxxxxxxxxxx>]

So in summary: You have machines with IPs in a RFC1918 range
(10.x.x.x).  You've got DNS setup such that the names resolve to
these addresses as expected.  The machines show up in
condor_status (and thus, are in your collector), but you don't
end up with jobs running on the machines.

In short, this should work fine.  Check your logs.  Is the
NegotiatorLog even matching to these machines?  If so, is the
ScheddLog or ShadowLog associated with the match reporting any
complaints?  How about the StartLog/StarterLog?

The thought that immediately leaps to mind is routing.  Can your
negotiator and schedds successfully initiate connections to your
10.x.x.x laptops?  Also, Condor generally binds to one network
interface.  If your schedd (for example) has an interface for
10.x.x.x, and a different interface for the public internet, and
it's using the public interface, it may never try to use the
10.x.x.x interface and thus be unable to route there.

-- 
Alan De Smet                              Condor Project Research
adesmet@xxxxxxxxxxx                 http://www.condorproject.org/