On November 18, 2003 10:06 pm, Toad wrote:
> On Tue, Nov 18, 2003 at 09:55:54PM -0500, Ed Tomlinson wrote:
> > On November 18, 2003 09:48 pm, Toad wrote:
> > > On Tue, Nov 18, 2003 at 09:47:03PM -0500, Ed Tomlinson wrote:
> > > > On November 18, 2003 05:29 pm, Toad wrote:
> > > > > On Tue, Nov 18, 2003 at 11:00:27PM +0100, Thomas Themel wrote:
> > > > > > Hi,
> > > > > >
> > > > > > my node is currently running off an ISDN link, and it seems to
> > > > > > build up pretty high values in the connection manager's 'Data
> > > > > > waiting to be transfered' counter. It's constantly around 30MiB,
> > > > > > and that's more than the link can do in an hour under optimal
> > > > > > conditions - I don't believe this makes sense. Or does it?
> > > > > >
> > > > > > Is there any way to get more sensible numbers?
> > > > > >
> > > > > > I've tried opening a lot of freesites, and they always seem to
> > > > > > restart. However, some give me the "The request couldn't even
> > > > > > make it off of your node." message, on a node that wasn't
> > > > > > reseeded for months.
> > > > > >
> > > > > > Is this related?
> > > > > >
> > > > > > I get the feeling that the behaviour has worsened in recent
> > > > > > versions, it used to work nicely around 632x.
> > > > >
> > > > > I suspect this is because we reinstated the QR based on bandwidth
> > > > > code.
> > > >
> > > > Without the QR based on bandwidth code I had about 200MB queued in
> > > > the OCM on 100-200 connections - all going thru a 10K pipe...  It
> > > > will take it a long long long time to send all that data.
> > >
> > > It doesn't matter. Slow data is better than no data. And NGRouting is
> > > quite capable of routing around slow nodes. What is HARD is dealing
> > > with QueryRejected's.
> >
> > I will agree with this to some extent.  However it _does_ matter when we
> > need 5 mins to send a 16K file and this is happening on lots of nodes. 
> > Is not this what happened when the switch was off, so its not the
> > solution either.
>
> No. As I said, NGRouting will be able to fix it. Look at your
> successTransferRate for the days when it was disabled.

4K-6K per second...  With a sendData value of bwtween .13-.29

So its still not doing that well.  Mind you, you did fix a bug that
might just help this...

Maybe (just maybe) the best way to handle the massive number 
of sends it to time them out if they do not achieve a high enough 
rate.  ie. accept a low sendData number when the sends do not 
progress fast enough and we end up with a giant queue.  Given
that there are lots of DSL type connections out there, IMO we
have to limit sends somewhere.

Ed
_______________________________________________
Devl mailing list
[EMAIL PROTECTED]
http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to