On Monday 30 April 2001 21:50, you wrote:
> >>>>> "JT" == Jay Tamboli <[EMAIL PROTECTED]> writes:
>
> JT> Same deal here. A quick test shows that about thirty percent
> JT> of the nodes referenced in my datastore aren't reachable. Of
> JT> course, I've never tested it before, so I don't know what's
> JT> normal. I get 27 out of 39 responding to pings.
>
> Is it at all possible that a) that many nodes have already installed
> 0.3.9 or 0.3.9.1, even without an announce on the Freenet project
> page? and b) that there's such grave bugs in 0.3.9.x that they make
> these nodes unusable?
>
> Neither seems probable to me, but I thought I'd throw that out as a
> possibility.
>
> ~Mr. Bad
I don't think it's the the new release.
My bet is on the known inbound thread limit /routing issue in 0.3x.
SUMMARY:
When the network gets stressed by lots of inserting or requesting, nodes
start to hit their inbound thread limits so they stop answering requests, so
the nodes making the requests drop the saturated nodes
from their routing.
It's probably more of a problem now because people actually are inserting
lots of data for web pages and other people are fetching that data a) to read
the pages and b) to keep them alive. "Sponsorship" in the vernacular of
the anime corner.
This is just speculation. I don't have data to back it up. Maybe one of the
Python/Perl jockeys on this list could work over their logs and see if it can
be confirmed or refuted.
-- gj
--
Web page inside Freenet:
freenet:MSK@SSK@enI8YFo3gj8UVh-Au0HpKMftf6QQAgE/homepage//
_______________________________________________
Devl mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/devl
