>On Thu, Jan 16, 2003 at 10:16:49AM -0800, Chris Linstruth wrote:
>> Exactly what I'm seeing.  When I stop/restart it'll run for 
>> give-or-take a few hours, 48 hours then usage will go to 0, 0ms 
>> CurrentRoutingTime.
>What exact build number is this?
I have seen this phenomenon with builds 537,539 and 629. Running on
windows XP with Suns java-engine 1.4.1_01-b01

> And please send me a log file, with logLevel=debug, and a thread dump,
which you get by kill -QUIT'ing the node.
I've sent a thread-dump once before, see that one for an example. I
tried to generate debug-level logfiles but those became somewhere
between 5 and 10 gigabytes before the problem arised so I skipped that
idea.

I am currently running 544 and it just _might_ work better, I'm not
totally sure yet though. I have two 544 nodes running 300 resp 500
threads and they have been running for 49 hours and 43 hours resp. and
still seems to work as they should. I'll drop a mail if the problem
arises in either of them.

And, on a totally different subject, would it be possible to get
counters/graphs for bytes sent/bytes in sendqueue (thank you for that
one, whoever added it to the open connections page)/bandwidth consumed
in the diagnostics page? It would be a really nice feature to have I
think.

/N


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

Reply via email to