Quoting Toad <[EMAIL PROTECTED]>: > On Wed, Jul 30, 2003 at 05:14:54PM -0500, J wrote: > > On Wed, Jul 30, 2003 at 01:31:15AM +0600, [EMAIL PROTECTED] wrote: > > >> Hi all, > > >>=20 > > >> I went from 5017 to 6117 and back to 5018 and have noticed that now > my=20 > > >> node is quite less specialized than when I was running 5017 according > to= > > >=20 > > >> the routing table graphic. I used to have decently thick bands and now > i= > > >t=20 > > >> looks like my node has forgotten almost all learned routing. I did > not= > > >=20 > > >> experience this upgrading 5016 to 5017, or any other upgrades. I have > > >> also noticed especially over the last two hours that my node has > started= > > >=20 > > >> having quite a harder time finding data. I have been running 5018 > for=20 > > >> almost 9 hours now... Any insight? > > > > >5018 ~=3D 6117. When you upgrade to 5018/6117, it starts handling load > > >substantially more efficiently, so it gets more traffic, with the result > > >that it despecializes temporarily, because there are few 5018/6117 > > >nodes... maybe. At least that's one theory. Another is that there is a > > >bug which is wiping out the routing table specializations on startup > > >sometimes (which I have tried and failed to reproduce btw)... > > > > That first thoery may be correct because today my RT looks significantly > better > > than it did last night. My thick dark bands are back and there seem to be > more > > of them than with 5017. > > Multiple specializations... about what we'd expect. > > > > > Not sure if this is related, but it appears that 90% of my outbound=20 > > > connections are to peers not in the RT. > > > > >That strongly suggests you are getting more traffic. What does your > > >Network Load page say? > > > > Network load page currently says: > > > > # Global mean traffic (queries per hour):10534.9 > > # Local mean traffic (queries per hour): 1540.2980476722246 > > # Current advertise probability: 0.18459220833731255 > > # Current proportion of requests being accepted: 0.69 > > That's not so good. 1540qph is pretty low, but 69% accepted is not good. > How many threads, connections?
As of right now, I'm looking like this: still build 5018, 35% load, up 1 day 8 hours 7 minutes Number of open connections: 410 Number of outbound connections: 113 Number of inbound connections: 297 Number of connections sending messages: 8 Number of connections receiving messages: 7 Bytes waiting to be sent: 2413090 Outbound connections that are to peers not in the routingtable: 91.150444% Global mean traffic (queries per hour):8096.988888888889 Local mean traffic (queries per hour): 905.8859942476239 Current advertise probability: 0.3936767243595679 Current proportion of requests being accepted: 0.87 Total pooled threads 55 Available pooled threads 10 Pooled threads in use 45 I've left every other option except outputBandwidthLimit as default. I've set that to 4096 to maintain decent performance on my crappy cable connection. Sorry to dump all this info, I'm not sure what all information would be useful. j > > > > Currently 94% of my open connections are to peers not in RT. Network load > is at > > 90%. > > Given the above load stats, that's somewhat troubling. > > > > >>=20 > > >> thx . j > > >>=20 > > >>=20 > > >> _______________________________________________ > > >> devl mailing list > > >> [EMAIL PROTECTED] > > >> http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl > > > > > > _______________________________________________ > > devl mailing list > > [EMAIL PROTECTED] > > http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl > > -- > Matthew J Toseland - [EMAIL PROTECTED] > Freenet Project Official Codemonkey - http://freenetproject.org/ > ICTHUS - Nothing is impossible. Our Boss says so. > _______________________________________________ devl mailing list [EMAIL PROTECTED] http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl
