> So it would appear that despite the registry setting being
> made I'm hitting a hard upper limit on the number of
> concurrent running vanilla jobs that can be supported by a
> single schedd instance in 6.7.3. Does that analysis seem apt?
Could all of this be network bandwidth issues? We run a 100 Mb
completely switched network. Our condor master machine is on GB fiber,
but our desktop machines are only 100 Mb copper. What kind of bandwidth
requirements could 80 condor_shadow processes for a vanilla job that
transfers little in the way of start up files and only transfers back a
33kb output file at the end of a job?
- Ian
|