IT came by my desk today (and you already know this is a story that doesn’t have a happy ending…), and rewired the network to a few of my machines.
One of them is our Condor central box, including the collector (and negotiator and credd).
In the course of doing that rewiring, my collector got a new IP assigned to it.
Now all my poor orphaned Condor machines have no idea where to send their classads; I’m reading their logs, and they’re pointed at the wrong (presumably old) IP address for the collector.
Here’s the surprising part: ipconfig /flushdns doesn’t seem to make a difference (it’s a Windows Condor pool).
Does Condor internally cache IP lookups beyond what the network stack does on its behalf?
Can I ask it not to do so? (Should I?)
To recover from this situation, I think I need to send a condor_reconfig to all my machines.
Unfortunately, since those machines are no longer announcing themselves to the collector, I can’t use Condor itself to do the reconfig on everything at once, and I think I’m stuck logging in everywhere and doing the reconfigs by hand.
Am I missing a better option?
Thanks!
------------------------------------------------------------------------------------------------------------------------------------------------
Jason Ross Intel Corporation
Graphics Architect FM5-79
VCG Architecture 1900 Prairie City Road
(916) 356-8964 Folsom, CA 95630