While updating my box, one of the hard drive on one of my older machines reached capacity. This machine was running distccd, and was working on compile jobs for other machines. When this occoured, the emerges on the other machines b0rked because the distcc job failed. I reproduces this once, cleaned HD on old machine, and the problem went away. Reproducible: Always Steps to Reproduce: 1. Have box running Distccd with full hd 2. Distribute compile job to it Actual Results: distcc builed failed @ [ip of full machine] Expected Results: I believe the system should have tried to compile locally instaed of just failing.
would also be nice if when running distcc and it fails to find the executable on the remote host it falls back to local ...
what version is the daemon?
to follow up to my previous comment ... crosscompiling distcc hosts is something ive been messing around with of late ... building a crosscompiling gcc is easy ... usually the hard part is building g++ it'd be nice if i could run a 'half' distcc host in that it supports C compiles (gcc) and when given a command involving g++ it'll currently fail with 'g++: command not found' because ... well because it doesnt exist :) so it'd be nice if the local host doing the actual build falls back to local compiling when the remote host goes 'hey now, i dont got that binary over here'
Lisa, I kinda like seeing what's going on with distcc, I don't know about the rest of us on cluster@. Maybe you could leave us assigned or at least CC'd in the future.
The reason I assign it to myself is so that it shows up in my bug list. CCing cluster@ sends me double mails. :( I guess assigning bugs to cluster@ and letting me reassign to myself. The rest of people on cluster alias should get an email anyways, so they can opt to CC themselves.
distccd 2.11.1 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) built Jan 20 2004 23:04:57
this is an upstream problem. send a mail to the distcc mail list. distcc@lists.samba.org and free up disk space
