Timestamps are in GMT/BST.
[0:47] * goatee_ is now known as goatee
[0:55] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 60 (Operation timed out))
[1:33] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[1:44] * addos (~addos@cpe-065-184-151-014.ec.rr.com) Quit (Read error: 110 (Connection timed out))
[2:41] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) has joined #freenet
[2:43] * nextgens (~nextgens@d213-103-64-133.cust.tele2.fr) has joined #freenet
[2:56] * mazzanet (~mazzanet@mazzanet.user) Quit (Read error: 110 (Connection timed out))
[2:56] * m[a]zzanet (mazzanet@mazzanet.user) has joined #freenet
[2:57] * m[a]zzanet is now known as mazzanet
[2:59] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[3:20] * TLF (~francisco@82.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[3:44] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[4:03] * m[a]zzanet (~mazzanet@mazzanet.user) has joined #freenet
[4:11] * mazzanet (mazzanet@mazzanet.user) Quit (Read error: 113 (No route to host))
[4:11] * m[a]zzanet is now known as mazzanet
[4:21] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[5:12] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) has joined #freenet
[5:14] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[5:19] * hirvox (~hirvox@cs181027153.pp.htv.fi) has joined #freenet
[5:19] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[5:35] * Fisu (~fisu@hoas-50dd08c8.hoasnet.inet.fi) has joined #freenet
[5:39] * Toast (~Toast@82-34-170-133.cable.ubr03.maid.blueyonder.co.uk) has joined #freenet
[5:44] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[6:10] * TLF (~francisco@82.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[6:14] * TLF (~francisco@85.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[6:14] * nextgens (~nextgens@d213-103-64-133.cust.tele2.fr) Quit (Read error: 54 (Connection reset by peer))
[6:21] * fimp (~fimp@0x50a436f4.aaanxx1.adsl-dhcp.tele.dk) has joined #freenet
[6:24] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) Quit ("leaving")
[6:24] <fimp> hi
[6:26] <fimp> what political ideology do most freenet-people support?
[6:28] <TLF> fimp: as I can see there are people of many ideologies
[6:36] <fimp> TLF, sure, but I guess that, for instance, there are more anarchists than fascists?
[6:36] <TLF> fimp: don't know
[6:37] <fimp> TLF, okay. it could be interesting to know, though
[6:37] <TLF> fimp: ;)
[6:38] * TLF (~francisco@85.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[6:39] <- *fimp* hi
[6:46] * TLF (~francisco@208.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[6:57] * nextgens (~nextgens@d80-170-169-58.cust.tele2.fr) has joined #freenet
[7:27] * kers (~kers@183.ppp132.rsd.worldonline.se) has joined #freenet
[7:41] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has joined #freenet
[8:50] * TLF (~francisco@208.Red-81-40-116.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[8:58] * Fisu (~fisu@hoas-50dd08c8.hoasnet.inet.fi) Quit ("leaving")
[9:01] * TLF (~francisco@35.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[9:06] * nextgens (~nextgens@d80-170-169-58.cust.tele2.fr) Quit (Read error: 104 (Connection reset by peer))
[9:08] * Toast (~Toast@82-34-170-133.cable.ubr03.maid.blueyonder.co.uk) has left #freenet
[9:30] * moskau23 (~Miranda@dsl-082-082-233-230.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[9:34] * hirvox (~hirvox@cs181027153.pp.htv.fi) Quit (Read error: 54 (Connection reset by peer))
[9:45] * nextgens (~nextgens@d83-177-154-236.cust.tele2.fr) has joined #freenet
[9:51] * TLF (~francisco@35.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[9:51] * kers (~kers@183.ppp132.rsd.worldonline.se) Quit (Read error: 104 (Connection reset by peer))
[10:46] * TLF (~francisco@202.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[10:46] * nextgens (~nextgens@d83-177-154-236.cust.tele2.fr) Quit (Read error: 54 (Connection reset by peer))
[10:53] * kers (~kers@116.ppp133.rsd.worldonline.se) has joined #freenet
[11:04] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) has joined #freenet
[11:25] * jabawok_w (~jabawok_w@203-59-90-210.dyn.iinet.net.au) has joined #freenet
[11:32] * moskau23 (~Miranda@p50850055.dip0.t-ipconnect.de) has joined #freenet
[11:33] <sanity> join #debian
[11:33] <sanity> crap
[11:41] <hobx_> hey
[11:42] <hobx_> (if that was recent. I have no tiemstamps in this client)
[11:43] <cbreak> 8 min delta t
[11:44] <hobx_> as irc clients go, chatzill pretty much stinks
[11:44] <hobx_> Maybe they'll make a "Dodobird" irc client some time...
[11:45] <cbreak> they have firefox and thunderbird. hmm... waterfish maybe...
[11:45] <cbreak> or ice lizzard
[11:45] <hobx_> but the calendar is "sunbird"
[11:46] <cbreak> oh, yes. I remember that one now.
[11:46] * hobx_ hasn't tried sunbird
[11:46] <hobx_> perhaps it is time I started organizing my life.
[11:47] <cbreak> I already have iCal. Enough for me :)
[11:48] * TLF (~francisco@202.Red-81-40-116.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[11:48] <hobx_> I'll be dead before I use macos
[11:48] <cbreak> :)
[11:48] * kers (~kers@116.ppp133.rsd.worldonline.se) Quit ("Leaving")
[11:49] <hobx_> I have a calendar on printed out A4s. But it is very sparse, often I sit and fill in things that I have already done so it looks like I have a bit of a life.
[11:50] <cbreak> All I need at the moment is a plan of my next exam...
[11:50] <hobx_> giving or taking
[11:50] * Hory (~Miranda@81.196.25.110) Quit (Read error: 104 (Connection reset by peer))
[11:51] <cbreak> Unfortunately I am one of those who has to find the solution to unfair problems :(
[11:51] <hobx_> take it from me. Making up unfair problems is harder.
[11:52] <cbreak> maybe it is harder. but no one complains if the problems are not hard enough :)
[11:52] <hobx_> Grading is fun though. Such laughter!
[11:52] <cbreak> hmm...
[11:53] <hobx_> So what are you taking?
[11:54] <cbreak> my next test is tomorrow, electronics and digital technology.
[11:54] <hobx_> hmm
[11:55] <hobx_> I know nothing about electronics.
[11:55] <cbreak> then I have left system programming, Scientific Calculus, Network systems, Communication Theory, and something I don't remember yet.
[11:55] <hobx_> Except that it is shiny and I want more of it!
[11:55] <cbreak> Those are all lectures in the base class, all mandatory for computer science here.
[11:55] <hobx_> Scientific Calculus? Instead of what, artistic calculus?
[11:56] <hobx_> "A surrealist curve does not have a derivative, it is at once without being."
[11:56] <hobx_> Sounds mostly like flummery.
[11:57] <cbreak> its about how to use the limited precision of computers FPU to find solution numericaly.
[11:57] <hobx_> Computer science should be 90% math, but they water it down for all the crybabies.
[11:57] <cbreak> I had 90% Math in my first two semesters.
[11:58] <cbreak> the third and fourth only had one math lecture each.
[11:59] <cbreak> Not that I don't like math, but some things are just more important.
[11:59] <hobx_> ha!
[11:59] <hobx_> he speaks in ignorance!
[12:00] <cbreak> Theoretical Computer Science (that stuff with P=NP) or System Programming (Sparc ASM) for example.
[12:00] <cbreak> Not every computer scientist wants to develop some kind of maple replacement :)
[12:00] <hobx_> Well, try proving anything about P and NP without math...
[12:01] <cbreak> math is the base for everything. but you have to go over it, stop thinking about math and just use it as a tool :)
[12:02] <hobx_> what you have said means nothing
[12:02] <hobx_> you have to know it to use it as a tool.
[12:02] <cbreak> yes, but endless proofs of nice theories are not helpfull.
[12:02] * Ash-Fox (Hal-9000@aag97.neoplus.adsl.tpnet.pl) Quit (Client Quit)
[12:02] * cparker (cparker@jcvmfl-as-2-209-208-82-253.dyn.atlantic.net) has joined #freenet
[12:03] <hobx_> Yeah, I know.
[12:03] <hobx_> Aweful stuff like Huffman's optimality proof.
[12:04] <hobx_> A real computer scientist hand waves!
[12:04] <cparker> Is there any delphi programmers in here? (please don't kick me :)
[12:05] <hobx_> You will have the green tree, but not the brown programing language.
[12:06] <hobx_> Many a things you will see, but lost is the significance of pottery.
[12:06] <hobx_> Programmer, I say onto thee. The gods have goto!
[12:06] <cbreak> Oracle of Delphi imitation...
[12:07] <hobx_> Oracle of Delphi would be SQL in Pascal I guess.
[12:07] * hobx_ shudders
[12:07] <cbreak> pascal... I had to learn oberon. Much worse...
[12:08] <cparker> The gods of delphi are uncertain about com and collections
[12:09] <cbreak> All I know about delphi is, that it more or less... well... that it's not a language I want to learn :)
[12:10] <cparker> darn...I know Java is probably a favorite here. I was hoping mxbee was here. Oh well, my needs are way off topic here :)
[12:11] <hobx_> we only need to be on topic when toad is around
[12:11] <cparker> That might not be for much longer if people don't open their wallets
[12:12] <hobx_> Java is like a drug. Everytime I'm gonna code something I tell myself to write it in something else, but java is so easy and inticing.
[12:12] <hobx_> I have basically forgotten how to code in any other language.
[12:12] <cbreak> try objective-c, the power of C with the OO Concepts of Smalltalk :)
[12:12] <cparker> I know, thats the way I feel about csharp.
[12:13] * Ash-Fox (Hal-9000@abe180.neoplus.adsl.tpnet.pl) has joined #FreeNET
[12:13] <hobx_> yes, objective-c is nice.
[12:14] <hobx_> nicer than c++ anyways.
[12:14] <cbreak> not hard :)
[12:14] <cparker> Well, my job HAS to be delphi :( I hate it because I love the "for each object in collection" loops
[12:15] <hobx_> C-cash is just java with some sugary additions and unnecessary changes to make it seem different.
[12:15] <cbreak> Obj-C is the best language I know, but on the other hand, I barely know any other.
[12:15] <hobx_> I wrote something in ruby not so long ago. Ruby is nice.
[12:16] <cparker> is it flexible for MS programming?
[12:16] <hobx_> I have said to myself: Ruby shall hitherfrom be my perl!
[12:16] <cbreak> :)
[12:16] * TLF (~francisco@149.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[12:17] * cbreak still scripts on sh
[12:18] <hobx_> me names my ruby scripts .sh and then blames it on the users when they don't work
[12:19] <cparker> later! (still searching for the answers :)
[12:19] * cparker (cparker@jcvmfl-as-2-209-208-82-253.dyn.atlantic.net) has left #freenet
[12:19] <hobx_> gl
[12:19] <cbreak> what has the name to do with that? If the first #!/usr/bin/ruby or whatever is set right, it should work.
[12:20] <hobx_> I'm going to stay at work until seven.
[12:20] <hobx_> I have a lab to prepare, but I can do it tomorrow.
[12:21] * cbreak goes back to learn about cables
[12:35] * jabawok_w (~jabawok_w@203-59-90-210.dyn.iinet.net.au) Quit (Read error: 110 (Connection timed out))
[12:49] * TLF (~francisco@149.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[12:51] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) has joined #freenet
[12:52] * nextgens (~nextgens@d80-170-97-139.cust.tele2.fr) has joined #freenet
[13:00] <fimp> anyone who knows of libertarian or anarchistic material on freenet?
[13:18] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) Quit ("leaving")
[13:30] * jabawok_w (~jabawok_w@203-59-100-254.dyn.iinet.net.au) has joined #freenet
[13:37] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[13:37] * pupok (~r00t@81-178-112-130.dsl.pipex.com) Quit (Read error: 54 (Connection reset by peer))
[13:37] * pupok (~r00t@81-178-112-130.dsl.pipex.com) has joined #freenet
[13:40] * TLF (~francisco@246.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[13:51] * jabawok_wrk (~jabawok_w@203-59-216-170.dyn.iinet.net.au) has joined #freenet
[13:55] * jabawok_w (~jabawok_w@203-59-100-254.dyn.iinet.net.au) Quit (Read error: 113 (No route to host))
[14:02] * TLF (~francisco@246.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[14:05] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) Quit (Remote closed the connection)
[14:06] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) has joined #freenet
[14:06] * unknown_ (~unknown@dsl-213-023-233-040.arcor-ip.net) has joined #freenet
[14:06] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) Quit (Client Quit)
[14:06] * unknown_ (~unknown@dsl-213-023-233-040.arcor-ip.net) Quit (Client Quit)
[14:06] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) has joined #freenet
[14:45] * nextgens (~nextgens@d80-170-97-139.cust.tele2.fr) Quit ("Download Gaim: http://gaim.sourceforge.net/")
[15:01] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[15:16] * greycat (rfc1413@wooledge.org) has joined #freenet
[15:23] * moskau23 (~Miranda@p50850055.dip0.t-ipconnect.de) Quit (Read error: 54 (Connection reset by peer))
[15:30] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) Quit (Read error: 54 (Connection reset by peer))
[15:30] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) has joined #freenet
[15:40] * jabawok_w (~jabawok_w@203-59-216-170.dyn.iinet.net.au) has joined #freenet
[15:40] * jabawok_wrk (~jabawok_w@203-59-216-170.dyn.iinet.net.au) Quit (Read error: 104 (Connection reset by peer))
[15:41] * greycat fires up a node for the first time in a while.
[15:41] <greycat> unfortunately my data store is empty (the 28gigger is on the Maxtor disk which is quite deceased)
[15:45] <greycat> "Fred version 0.5, protocol version STABLE-1.50 (build 5095, last good build 5089)" ... hmm. Shouldn't that be "oldest good build" instead of "last good build"?
[15:50] <hobx> depends on your perspective
[15:55] <greycat> <newbie brand="hypothetical">But if 5089 is the last good build, why do I want 5095? I want 5089!</newbie>
[16:02] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) Quit (Connection reset by peer)
[16:02] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) has joined #freenet
[16:12] <d-ArkAngel> bah who needs a good node, join the dark side!
[16:13] <greycat> good point. I don't mind having a naughty node.
[16:15] <d-ArkAngel> makes it all so much more fun :-)
[16:23] <hobx> your Jessica Alba, aren't you?
[16:23] <hobx> I always knew you dug me!
[16:24] <greycat> women hitting you in the head with shovels don't count as "chicks who dig me"
[16:24] <hobx> it's all relative
[16:24] <hobx> I try to look at it from the positive side.
[16:24] <hobx> I mean, hate is not the opposite of live, indifference is. So if they hate me, I'm half way there!
[16:26] <greycat> and in the "not bad news" deparment, freenet hasn't managed to turn my internet connection into cottage cheese yet.
[16:26] <greycat> then again, it's still early
[16:28] * fimp (~fimp@0x50a436f4.aaanxx1.adsl-dhcp.tele.dk) Quit ("Leaving")
[16:28] <hobx> gonna watch the final table of wsop now!
[16:33] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) Quit (Remote closed the connection)
[16:41] * pupok (~r00t@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[16:41] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[16:41] * pupok (~r00t@81-178-112-130.dsl.pipex.com) has joined #freenet
[16:42] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[16:46] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) Quit ()
[17:02] * KenMan (~chaziller@pcp403292pcs.mntcrm01.md.comcast.net) Quit ("ChatZilla 0.9.61 [Mozilla rv:1.7b/20040324]")
[17:06] * moskau23 (~Miranda@dsl-082-082-233-230.arcor-ip.net) has joined #freenet
[17:25] <greycat> starting to see IRC lag now.
[17:40] * Zv2S (~rfc1413@squirrel.owl.de) has left #freenet
[17:40] * Zv2S (~rfc1413@squirrel.owl.de) has joined #freenet
[17:54] * moskau23 (~Miranda@dsl-082-082-233-230.arcor-ip.net) Quit (Read error: 54 (Connection reset by peer))
[17:54] * moskau23 (~Miranda@dsl-082-082-233-230.arcor-ip.net) has joined #freenet
[18:14] * moskau23 (~Miranda@dsl-082-082-233-230.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[18:29] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[18:34] <greycat> Transfers active (Transmitting/Receiving)99 (23/76)
[18:34] <greycat> Grrrrr.
[18:35] <greycat> Transfers active (Transmitting/Receiving)110 (29/81)
[18:36] <greycat> looks like there's been no improvement in this respect since the last time I jumped ship..
[18:36] <d-ArkAngel> what's the problem?
[18:36] <d-ArkAngel> it's eating all your bandwidth?
[18:36] <greycat> the actual problem is that I get severe lag on my IRC connection
[18:37] <greycat> my hypothesis is that this is symptom of having too many simultaneous transfers going on, but I can't prove it
[18:38] <d-ArkAngel> have you tried adding input and output BandwidthLimits to the conf file.
[18:38] <greycat> I do have outbound bandwidth limits set, but not inbound
[18:39] <d-ArkAngel> I've set both mine to be about 75% of my full limit.
[18:39] <d-ArkAngel> before I did it was crippling my total internet connection too.
[18:39] <d-ArkAngel> what have you got your incomming limit?
[18:40] * greycat kills the node. let's see what we need to tweak in freenet.conf..
[18:40] <d-ArkAngel> mine are set at 5000 out and 30000 in
[18:41] <greycat> oh, 5000 is really low...
[18:41] * greycat restarts with 24000 in and 12288 (default) out.
[18:41] <d-ArkAngel> well that's in bytes per second isn't it?
[18:42] <greycat> wow.... uptime 1 minute, and Connections open (Inbound/Outbound/Limit)58 (55/3/200)
[18:42] <d-ArkAngel> 5000 bytes, but I've only got 128kbits/second as my actual limit (factor of 8 difference)
[18:42] <greycat> is that sane?
[18:44] <d-ArkAngel> yeah I think that's the right way
[18:45] <greycat> I'm stunned at how many inbound connections I got in such a short period of time.
[18:45] <greycat> Lag: 1.11
[18:46] <greycat> wow... http://209.142.155.49/greycat.yi.org/graphs.cgi
[18:46] <greycat> the amount of inbound bandwidth that freenet was sucking up is very interesting
[18:47] <greycat> Lag: 2.17
[18:48] <d-ArkAngel> how's the CPU usage on your machine?
[18:48] <greycat> I run dnetc also, so a load average of 1.00 would be "idle/normal"
[18:48] <greycat> 19:07:45 up 7 days, 9:20, 12 users, load average: 1.05, 1.30, 1.30
[18:48] <greycat> it's an Athlon XP 2000+ (1.666 GHz), so freenet isn't CPU-bound
[18:50] <d-ArkAngel> what IRC client are you using?
[18:50] <greycat> irssi
[18:53] <greycat> the next symptom is that my clock starts drifting, probably because the ntpd packets don't come through in a timely manner
[18:53] <greycat> 2004-09-19 19:12:49.620156500 reply from 192.168.2.1: offset -2.465910 delay 0.000221, next query 30s
[18:53] <greycat> and on 192.168.2.1: 2004-09-19 19:12:59.302421500 reply from 128.2.136.71: offset 1.018845 delay 0.486561, next query 30s
[18:54] <greycat> the final symptom (which I haven't seen yet today) is that DNS lookups fail completely.
[18:57] <d-ArkAngel> mine did that, but I was running it on my DNS server on my local network.
[18:57] <d-ArkAngel> rather than on my client.
[19:07] <greycat> 2004-09-19 19:26:16.447007500 reply from 128.2.191.71: offset 1.935449 delay 4.467766, next query 30s
[19:07] <greycat> Lag 6.97
[19:09] <greycat> Lag 13.57
[19:09] <greycat> 2004-09-19 19:28:55.168831500 reply from 128.2.129.21: offset 3.310871 delay 7.159694, next query 30s
[19:14] * greycat gives up (again).
[19:19] <mazzanet> you need to be a blue cat to make it owrk
[19:19] <mazzanet> work
[19:33] * soros (~soros@d57-37-118.home.cgocable.net) has left #freenet
[19:38] <ejhuff> greycat: try output limit 8000, and add -x to ntpd options (always slew clock, never step)
[19:39] <greycat> I'm not using that ntpd. I'm using the one from www.openntpd.org.
[19:40] <greycat> and this has nothing to do with the bandwidth limits. My upstream is 256kbps.
[19:40] <greycat> it suddenly started happening around july when toad did something to the builds that caused them to do more transfers
[19:41] <ejhuff> Windows or Linux?
[19:41] <greycat> before that it was fine with my rather aggressive PF setup on the firewall
[19:41] <greycat> openbsd firewall, linux desktop box running freenet
[19:42] <ejhuff> I had lots of clock step messages in log, slewing let the clock stay fairly stable.
[19:42] <ejhuff> Don't know when it started, might have been in July.
[19:42] <greycat> the problem was the "delay 4.467766" part. If it takes 4.4 seconds to get the replies from the upstream NTP source, it's not going to matter how you slew the clock.
[19:42] <ejhuff> Do you run ntpd on openbsd?
[19:43] <greycat> yes
[19:43] <greycat> that's the box that the "delay 4.4..." was issued by
[19:43] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) Quit (Connection reset by peer)
[19:43] <ejhuff> This ntpd gives values in seconds?
[19:44] <greycat> whatever freenet is doing to my internet connection just *kills* *everything*. Lag spirals out of control. I'd love to know what's going on, but I've never been able to figure it out.
[19:45] <greycat> I've got giftd spewing ogg/mp3/mpg/avi/ogm/etc. files all day long without any problems.
[19:45] <greycat> PF can handle that; I just prioritize ACK and DNS and NTP and ssh and IRC packets higher than the giftd packets.
[19:46] * salahx (salahx@sc1-24.217.174.147.charter-stl.com) has joined #freenet
[19:46] <ejhuff> Well, how about trying output limit at 8000? It might well help.
[19:47] <ejhuff> Whatever freenet is doing, it can't do it as much with lower quota.
[19:47] <greycat> can't toad just give us a "maximum simultaneous data transfers" setting instead? Then I could set that to 16 or something.
[19:47] <greycat> instead of 110.
[19:47] <salahx> where doy ou see 110 ?
[19:48] <ejhuff> Oh. Maybe something to do with packet reassembly on openbsd?
[19:48] <ejhuff> Is openbsd thrashing? How much RAM?
[19:49] <greycat> 18:54 greycat> Transfers active (Transmitting/Receiving)110 (29/81)
[19:49] <greycat> that was over an hour ago
[19:49] <greycat> no, the openbsd box is fine. 512 MB RAM and it never even uses that much.
[19:49] <salahx> THe number of tranfers can't be limited; otherwise an attack coudl sue them all up by "truickling" data through
[19:50] <greycat> then maybe the tricklers should just be dropped
[19:50] <salahx> you'd need some kind of "minumum reanfer rate" thing for that
[19:50] <salahx> which was a suggested idea
[19:50] <greycat> well, just to make you happy, I'll set outputBandwidthLimit to 8000 and restart it
[19:50] <ejhuff> That will be interesting...
[19:50] <salahx> OF course; we all know how good Freenet is at estimating bamndwisdth usage..
[19:52] <ejhuff> On general info page is actual usage. I haven't compared to real actual usage, but changing it from 12000 to 8000 definitely improves my browsing performance.
[19:52] <ejhuff> I think Fred is better at bandwidth limiting than it used to be.
[19:53] <d-ArkAngel> yeah it seems fairly acurate in my bandwidth limit tests.
[19:54] <greycat> I never had a problem with bandwidth utilization per se
[19:54] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[19:55] <greycat> ok, I've got fred running on the linux box, and I've killed dnetc on both linux and openbsd boxes.
[19:55] <greycat> we'll see what happens...
[19:56] <ejhuff> BTW transfers are multiplexed. You won't have more connections to have more transfers, just more packets per connection.
[19:57] <d-ArkAngel> well I hope you get sorted out, I'm off to get some sleep. night all.
[19:58] <ejhuff> what is dnetc?
[19:59] <salahx> TThink its part of Distributed.net
[19:59] <greycat> distributed.net client for using spare CPU cycles to do semi-useless shit :)
[19:59] <d-ArkAngel> what nice level was that running at?
[20:00] <ejhuff> Not much network traffic?
[20:00] <greycat> dnetc's network traffic is meaningless. a trickle.
[20:00] <greycat> it fetches some blocks over http about once a day and then crunches them
[20:01] <d-ArkAngel> what nice level? lower than freenet?
[20:01] <greycat> correct
[20:01] <ejhuff> freenet at nice 10, dnetc at nice 19 shouldn't hurt...
[20:01] <ejhuff> How many open connections?
[20:02] <d-ArkAngel> well catch you all later.
[20:02] <d-ArkAngel> night
[20:02] <greycat> the lag symptoms haven't started for me yet
[20:02] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) Quit (Remote closed the connection)
[20:02] <greycat> Connections open (Inbound/Outbound/Limit)27 (11/16/200)
[20:03] <ejhuff> I think it must be something to do with freebsd limits...
[20:03] <ejhuff> What were connections before?
[20:03] <greycat> I'm more inclined to suspect a limitation of my ADSL modem or my ISP at this point.
[20:04] <ejhuff> Hmm, that's possible. Do they do bandwidth shaping?
[20:05] * proteusguy (proteusguy@68.191.242.147) has joined #freenet
[20:05] <greycat> it's a 512/256 connection (512 kbps incoming, 256 kbps outgoing)
[20:07] <ejhuff> Do you really get 256 up? Have you tried one of the bandwidth test sites?
[20:08] <greycat> I run giftd and serve up mp3/ogg/etc. files 24x7.
[20:09] <greycat> I've got PF set up to permit 190kbps out of the tun0 interface (a fair bit less than the 256 kbps figure they quote, but pppoe overhead eats up a lot of bits)
[20:09] * jabawok_w (~jabawok_w@203-59-216-170.dyn.iinet.net.au) Quit (niven.freenode.net irc.freenode.net)
[20:09] * mazzanet (~mazzanet@mazzanet.user) Quit (niven.freenode.net irc.freenode.net)
[20:09] <ejhuff> Well, if giftd is using up the bandwidth, then freenet's additional traffic would go over the limit.
[20:09] <salahx> PPPoE overhead is tiny
[20:09] * mazzanet (~mazzanet@mazzanet.user) has joined #freenet
[20:09] * jabawok_w (~jabawok_w@203-59-216-170.dyn.iinet.net.au) has joined #freenet
[20:10] <salahx> compaortf to the ATM ovehead anyway at least 10% of the bandwidth - ouch
[20:10] <ejhuff> Do you have statistics on PF to see how many packets get dropped?
[20:11] <greycat> ejhuff: no, freenet packets are higher priority than giftd's. I rely on fred's bandwidth limiting to control freenet's usage. giftd gets whatever's left over, and even when fred's turning my connection to mush, I still get about 4-5 kB outgoing reported by gift (until the final stage of the disorder in which *NOTHING* works at all any more)
[20:12] <ejhuff> How is it doing now?
[20:12] <greycat> Connections open (Inbound/Outbound/Limit) 40 (19/21/200)
[20:12] <greycat> Transfers active (Transmitting/Receiving) 28 (14/14)
[20:12] <greycat> no lag symptoms yet
[20:13] <ejhuff> With less bandwidth permitted, your node will be less popular...
[20:13] <greycat> queue fn priority 2 priq( red )
[20:13] <greycat> [ pkts: 343412 bytes: 59697768 dropped pkts: 3555 bytes: 717039 ]
[20:13] <greycat> [ qlength: 7/ 50 ]
[20:13] <greycat> [ measured: 51.0 packets/s, 48.23Kb/s ]
[20:13] <greycat> queue gift priq( red )
[20:13] <greycat> [ pkts: 16514444 bytes: 17064860881 dropped pkts: 842045 bytes: 826632165 ]
[20:13] <greycat> [ qlength: 12/ 50 ]
[20:13] <greycat> [ measured: 14.1 packets/s, 86.49Kb/s ]
[20:14] <greycat> but I believe some of freenet's packets end up in the default queue because I can't match them (variable ports)
[20:15] <greycat> oh, here we go. lag symptoms are starting now.
[20:15] <greycat> Lag: 1.72 on the irssi display
[20:15] <greycat> 2004-09-19 20:34:22.942751500 reply from 128.2.191.71: offset 1.030656 delay 1.905699, next query 30s
[20:15] <greycat> Connections open (Inbound/Outbound/Limit) 48 (25/23/200)
[20:15] <greycat> Transfers active (Transmitting/Receiving) 31 (16/15)
[20:15] <ejhuff> Then use netfilter on linux to filter by username, and change the QOS.
[20:19] <greycat> Connections open (Inbound/Outbound/Limit) 59 (35/24/200)
[20:19] <greycat> Transfers active (Transmitting/Receiving) 58 (37/21)
[20:19] <greycat> Lag: 2.12
[20:20] <greycat> load average is ~1.0 on the openbsd box, but CPU is > 90% idle
[20:20] <greycat> biggest CPU consumers on the openbsd box are giftd, ppp, pppoe
[20:20] <ejhuff> My ntpq -p offets are mostly under 50ms.
[20:20] <greycat> under normal conditions mine are under 100 ms
[20:21] <greycat> at the moment I'm getting 1500 ms
[20:21] <ejhuff> s/offset/delay
[20:22] <ejhuff> no, wait. delay is around 200ms
[20:24] <greycat> Connections open (Inbound/Outbound/Limit) 61 (37/24/200)
[20:24] <greycat> Transfers active (Transmitting/Receiving) 82 (59/23)
[20:24] <greycat> Lag: 1.06
[20:24] <ejhuff> Ok, so even with limit 8000 you are getting bad performance?
[20:24] * proteusguy (proteusguy@68.191.242.147) Quit (Read error: 60 (Operation timed out))
[20:24] <greycat> it's not *as* bad... yet...
[20:24] <ejhuff> Maybe you should reduce max connections to 40.
[20:25] <ejhuff> Since it was ok with 40 open...
[20:26] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) has joined #freenet
[20:28] <ejhuff> I tried using the linux bandwidth shaping about a year ago, decided it just didn't work. If you drop packets, they get retransmitted, causing more load.
[20:29] * greycat kills node again
[20:30] <greycat> Lag: 2.78 (still not as ruinous as it had been before, but I didn't want to let it get too bad before trying the lower max connectiopns idea)
[20:30] <ejhuff> Good.
[20:30] <ejhuff> You might also reset bandwidth limit to 12000
[20:30] <ejhuff> (Only change one thing at a time...)
[20:31] <greycat> outputBandwidthLimit=12000
[20:31] <greycat> maxNodeConnections=40
[20:32] <greycat> what the hell? 20:51:27 up 7 days, 11:04, 12 users, load average: 7.49, 3.37, 1.54
[20:32] <greycat> why's my load average going up through the roof with freenet down?
[20:32] <greycat> Oh. Duh. The insertion client.
[20:32] <ejhuff> Right. Let's see what happens. Wonder if there is any connection between queue length 50 and performance crapping out at 48 connections...
[20:34] <ejhuff> Yes. That happens... BTW load average is just the number of processes ready to run. High load average does not necessarily mean overload if the processes are niced.
[20:34] * greycat restarts
[20:36] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) has joined #freenet
[20:36] <greycat> Connections open (Inbound/Outbound/Limit) 16 (1/15/40)
[20:36] <greycat> Transfers active (Transmitting/Receiving) 2 (1/1)
[20:37] <KenMan> that is 16 more connections than me :)
[20:37] <ejhuff> Hi KenMan. greycat is trying to find out why network performance goes to crap when Fred is running, even if Fred uses only 50% of available upstream.
[20:38] <ejhuff> We suspect a bandwidth shaper either PF on his openbsd or his ISP
[20:38] <ejhuff> s filter
[20:39] <ejhuff> Reduced maxNodeConnections to 40 b/c performance was ok when connections were 40 but went bad after 48...
[20:42] <greycat> just FYI I used to run giftd as an openft search node with 300 children (300 open connections), but I haven't done that in several months. It's possible my ISP has changed something since then, or that a new bug in openbsd has popped up....
[20:42] <KenMan> perf was okay when MAXconns>=48 or ACTIVEconns>=48 ??
[20:43] <ejhuff> Active
[20:43] <ejhuff> that is ok with active < 48, started to go bad at 48.
[20:43] <ejhuff> roughly...
[20:44] <greycat> Connections open (Inbound/Outbound/Limit) 27 (1/26/40)
[20:44] <greycat> Transfers active (Transmitting/Receiving) 14 (10/4)
[20:48] <ejhuff> The problem might be with failure to id freenet packets on outgoing connections. Tagging packets from userid freenet in netfilter (iptables) and changing QOS bits, then filter on QOS in freebsd NF.
[20:48] <greycat> openbsd, but anyway...
[20:49] <ejhuff> right...
[20:50] <ejhuff> I suppose giftd uses a fixed port?
[20:52] <greycat> yes and no... but more importantly, it runs on the same box as PF, so I can tag by user id.
[20:54] <greycat> RNF with Attempts were made to contact 0 nodes.
[20:54] <greycat> Connections open (Inbound/Outbound/Limit) 26 (1/25/40)
[20:54] <greycat> Transfers active (Transmitting/Receiving) 27 (19/8)
[20:55] <ejhuff> Problem with fewer open connections is they get backed off...
[20:55] <ejhuff> What you need is a limit on number of connections which are actively transferring.
[20:59] <ejhuff> Maybe there is a limit on the number of simultaneous packet reassembly buffers for incomming packets, or something.
[21:08] <ejhuff> Hopefully this will work... bbl.
[21:11] <greycat> I seem to be holding steady at 27 open connections for some reason
[22:11] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[22:11] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) Quit (Read error: 104 (Connection reset by peer))
[22:18] * greycat (rfc1413@wooledge.org) Quit ("Lick Bush in '04")
[23:36] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.