Hello,
I am not new to HTCondor as a user, but this is my first time setting up HTCondor. I have several Mac Minis (Apple silicon, so having to run HTCondor 25.11.0 x86_64 using Rosetta) that I want to build a multi-machine pool with. I set up a head node (CM + submit) and one worker on a private LAN and attempted to run a test job.
Collector, negotiator, and schedd communication work normally. The shadow starts and communicates with schedd fine, but then it gets the following error:
attempt to connect to <ip:port> failed: No route to host (connect errno = 65).
nc and bash /dev/tcp both connect successfully from head node to
worker as root and as condor user. The routing table shows
correct route and there are no firewall rules. Is this a known issue with Rosetta? Is there a workaround or a native
ARM build planned?
Any help would be greatly appreciated.
Thanks,
Katie