>I'm going to make a very unpopular request. We really need more testing of the
>nio code on windows 2k/xp.


I've been running this week's unstable builds on Win2000 and they're looking good. Wednesday's build seemed to have a bad effect on the routing as Successes/Trials ratio dropped from its usual 60% to about 35%, but today's jar is fine (freenet-unstable-20030627.tgz - Build 6061 CVS Rev 1.254.)

The only thing I can't get working right with these builds is the outbound bandwidth limiting: No matter what I set for outputBandwidthLimit I get limiting of between 8500 and 9600 Bytes per second over 30 minutes(using outputBandwidthLimit settings between 2000 and 15000.) If I comment out the outputBandwidthLimit line, my line gets fully utilised (approx 14000 Bytes per sec), so bandwidth limiting is happening, just not at the rate configured.

Normal- and Minor-level logging doesn't reveal any problems - just plenty of "An existing connection was forcibly closed by the remote host" reports.

Processor utilisation seems stable at about 40% (900MHz AMD) running at "Below Normal" priority) and memory is stable around 83MB. None of the builds have crashed at all, but the longest I've let them run between restarts is 9 hours (test, test, sleep, test, test...!) Freenet.ini file is bog-standard except for rtMaxNodes=150

If there are any specific tests you want me to do, let me know, otherwise I'll just keep trying to break it.

-Kevin

_______________________________________________
devl mailing list
[EMAIL PROTECTED]
http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl

Reply via email to