Hello,
I solved the path issue. It had to do with the fact that I had different
compiler paths on my local machine and the cluster, and that our build
system used an absolute path for the compiler. I had forgotten about
this snag.
Once I fixed that, the trick of using a wrapper script to cause
"distccd" to be migrated to a worker node seems to work very well. From
the external network, it appears as if we only have the head node as a
worker, but the wrapper makes sure that each daemon runs on a separate
worker node.
The "beo_distccd" script I ended up using is this:
#!/bin/sh
DISTCCD=/opt/scratch/distcc/bin/distccd
# Currently use only nodes 3 and 4 for testing
HOST=`beomap --no-local --exclude
'0:1:2:5:6:7:8:9:10:11:12:13:14:15:16:17:18'`
bpsh -L $HOST $DISTCCD $*
On my desktop:
export
[EMAIL PROTECTED]:/opt/scratch/distcc/bin/beo_distccd
-- Marcio
__
distcc mailing list http://distcc.samba.org/
To unsubscribe or change options:
https://lists.samba.org/mailman/listinfo/distcc
