[00:36] [S]ushi (Sushi@pD951276B.dip.t-dialin.net) joined #vserver. [00:36] <[S]ushi> hi [00:37] Tamama (~bgbgbg@a62-216-20-152.adsl.cistron.nl) left irc: Quit: one little two little three little piggies OINK! OINK! OINK! [00:44] kestrel (~athomas@dialup51.optus.net.au) joined #vserver. [00:52] [S]ushi (Sushi@pD951276B.dip.t-dialin.net) left irc: [01:12] JonB (~jon@129.142.112.33) left irc: Quit: zzzzz [01:45] Nick change: Bertl_oO -> Bertl [01:45] hi everyone ... [01:55] hi [01:55] hi dan! [01:55] how's it going? [01:56] fine, thanks ... [01:57] do you know why shaya is looking for me? [02:02] I do not [02:02] okay, thanks ... [02:37] hi [02:38] hi Alec! [02:51] whatcha doing herbert? [02:52] preparing the 1.3.0 release ... [02:54] shaya (~spotter@dyn-wireless-244-10.dyn.columbia.edu) joined #vserver. [02:54] Bertl: you here? [02:54] nope, just left ;) [02:54] ok, good finally [02:54] question for you [02:54] vserver supports INADDR_ANY binding to all ip's given to a vserver right? [02:54] hmm, basically it should somehow ... [02:55] do you know how? [02:55] I need to implement a similar thing [02:55] and couldnt figure it out [02:55] from the vserver code [02:55] have an idea of how it would work via the tcp hash buckets [02:55] well, don't tell, but it's a hack ... [02:55] yes, as is most of my code (we have something like vserver, but can migrate processes bw machines when run inside of it) [02:55] so can migrate X11/KDE for instance bw machines [02:56] also migrate network connections (so can be ssh'd to a different box) [02:56] hmm, sounds cool, what is this GPL stuff? [02:56] not released yet [02:56] probably will be GPLd [02:56] published once already in OSDI 2001 [02:56] 2002 [02:56] hmm, well probably isn't good enough if you want to use kernel code ;) [02:56] we do it all in kernel [02:57] so probably has to be gpld [02:57] I'm not against gpl, indian guy in our group sees fortune and not keen on gplk [02:57] anyways, eben moglen I think convinced our advisor that it has to be gpl'd [02:57] http://www.ncl.cs.columbia.edu/publications/osdi2002_zap.pdf <- first paper [02:57] okay, you had a look at the split-outs? [02:58] split outs? [02:58] hrm ... why am I doing this .... 8-) [02:58] ? [02:59] interesting thing about our work is that it works totally as a kernel module [02:59] okay, any vserver release, ... have you had a look at the release pages? [02:59] on solucorp [02:59] not on linux-vserver.org [02:59] okay, take a look at linux-vserver.org then ,... [03:00] what should I look at [03:00] the vs1.20 release for example ... and the patches of course ... [03:00] http://www.13thfloor.at/vserver/s_release/v1.20/split-2.4.21-vs1.20.tar.bz2 ? [03:00] yup, that is one of the split-outs ... [03:00] what does split-outs mean in vserver terminology? [03:01] that means that the large patch is splitted in several small patches, by logical categories ... [03:01] ahh [03:01] right [03:01] for linus in future [03:01] for example the 'network' is such a category ... [03:01] well basically it is for me, and other developers ... [03:02] so route_connect is the main func? [03:03] what do you call a main function in the kernel? [03:03] not main as in main() [03:03] but it seemed like an imp added function for this functinality [03:03] it just skimming through the patch [03:04] main as in impt addition [03:04] inet_bind is important ... [03:05] tcp_bind_hash, and the tcp_ipv4_addr_* [03:05] basically that was the idea i was having you bind the same bucket to more than one adapter [03:05] is that what you do? [03:05] instead of the catch all INADDR_ANY adapter [03:06] what happens if a program binds to INADDR_ANY on the host? [03:07] btw, once we release it might be worthwhile to try and feed you our checkpoint/restart infrastructure [03:07] that allows us to migrate between hosts [03:07] hmm, lets see ... a call to bind() with ip INADDR_ANY [03:07] netrose (~john877@CC3-24.171.21.47.charter-stl.com) left irc: Ping timeout: 480 seconds [03:07] can checkpoint VNCServer (X11) KDE, KWord, KSpread, Konq, kghostview and ssh connections to about 30-40MB of space uncompressed [03:08] gzips to about 8-10MB [03:08] currently the bind depends on the number of ips, and I'm not sure we do the right thing in any case ... [03:09] one ip, is easy to handle ... [03:09] glua (~glua@golderer.net) left irc: Ping timeout: 499 seconds [03:09] I mean bind(INADDR_ANY) outside of vserver [03:09] yes [03:09] i.e. does it affect vserver's? [03:09] hmm, outside the vserver, yes that would steal the ips ... [03:09] which is questionable if its the right semantic [03:10] that is the reason why we have the v_* wrappers ... [03:10] the semantic is clear in our case, as every vserver interface, _is_ a well defined interface on the host ... [03:10] except that the host can stomp on it if it wants [03:10] which I guess is fair [03:11] this will be changed, when we get a virtualized interface(s) per vserver ... [03:11] but to be clear, vserver's supports binding a single socket to an arbitrary # of real IPs [03:11] i.e. inaddr_any in a vserver [03:12] real kernel gets away with it, because of inaddr_any virtual adapter, which is a single catch all [03:12] yes, it supports it, but as a hack, as I said ... [03:12] ok, as I said most of our code is a hack too [03:12] benefits of research life :) [03:12] and being unreleased [03:12] but I see no problem in cleaning up those hacks ... [03:12] I'm so scared of releasing the code as i think I'm going to be very embarrased :) [03:13] you can handle this case via INADDR_ANY as long as required, and just look up the ips when necessary ... [03:14] hmm, I'll have to dig into the patch I guess and try to understand what you do [03:14] not I, actually jack ... [03:15] I'm just doing some cleanup, and trying to understand the details myself ... [03:15] :) [03:15] yea, I e-mailed him and didn't get a response [03:16] 6 years ago we e-mailed back and forth a lot over linuxconf [03:18] well the split patches should help you to concentrate on the changes ... [03:20] yes [03:20] I was trying to work off an incremental vserver patch and couldnt make heads or tails, this has all the network code in one place, much easier [03:20] thanks [03:20] anyways time to flee and go home [03:21] okay, good luck, and hope to hear soon of your project ... [03:26] btw, if anyone is interested in an insanely cheap Thinkpad X31 [03:26] https://www-132.ibm.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=-840&storeId=821258&productId=8682906&langId=-1&dualCurrId=201298 [03:35] shaya (~spotter@dyn-wireless-244-10.dyn.columbia.edu) left irc: Quit: Client exiting [03:49] Doener (~doener@pD9588AE1.dip.t-dialin.net) joined #vserver. [03:50] hi doener! [03:50] hi [03:51] checkpointing is a cool technology [03:52] object oriented design, too, what is your point? [03:53] it was merely a statement of my opinion, no specific point or implication [03:55] ahh okay, maybe I'm to tired ... I guess I go to bed now ... [03:56] have a nice whatever ... [03:56] Nick change: Bertl -> Bertl_zZ [05:00] netrose (~john877@CC3-24.171.21.47.charter-stl.com) joined #vserver. [05:08] netrose (~john877@CC3-24.171.21.47.charter-stl.com) left irc: Ping timeout: 480 seconds [05:17] serving (~serving@213.186.191.75) left irc: Ping timeout: 485 seconds [06:58] Nick change: riel -> surriel [07:00] netrose (~john877@CC3-24.171.21.47.charter-stl.com) joined #vserver. [07:10] serving (~serving@213.186.191.115) joined #vserver. [08:13] kestrelw (~athomas@o2rosock0a.optus.net.au) left irc: Remote host closed the connection [08:18] Doener (~doener@pD9588AE1.dip.t-dialin.net) left irc: Quit: Leaving [08:21] kestrel_ (~athomas@o2rosock0a.optus.net.au) joined #vserver. [08:25] herbert, just to let you know...netstat -ant on my sparc debian 2.4.22+vs1.1.3 locks up hard [08:25] i think this was discussed in the mailing list though, so i'll re-read that thread when i get home [08:27] Prakash (~Prakash@202.9.172.132) joined #vserver. [09:39] cpd (~fiona@h24-71-35-195.ok.shawcable.net) joined #vserver. [09:41] anyone online get vserver & gentoo to play nice? [10:06] history [10:07] cpd (~fiona@h24-71-35-195.ok.shawcable.net) left #vserver. [11:02] cpd: gentoo as host or vserver? [11:14] Ahh, he's already gone. I missed that. [11:17] Prakash (~Prakash@202.9.172.132) left #vserver (Client Exiting). [11:57] Doener (~doener@pD9588AE1.dip.t-dialin.net) joined #vserver. [13:50] Doener` (~doener@pD9E128B3.dip.t-dialin.net) joined #vserver. [13:57] Doener (~doener@pD9588AE1.dip.t-dialin.net) left irc: Ping timeout: 485 seconds [14:24] sylvio (sylvio@imk32.mb.uni-magdeburg.de) joined #vserver. [15:57] [S]ushi (Sushi@pD9E348AD.dip.t-dialin.net) joined #vserver. [16:16] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [16:16] MrBawb (abob@swordfish.drown.org) got netsplit. [16:17] MrBawb (abob@swordfish.drown.org) returned to #vserver. [16:17] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [16:36] ccooke (~ccooke@80.1.164.238) left irc: Quit: BRB [16:37] ccooke (~ccooke@80.1.164.238) joined #vserver. [16:42] serving (~serving@213.186.191.115) left irc: Ping timeout: 485 seconds [16:52] serving (~serving@213.186.191.58) joined #vserver. [17:02] JonB (~jon@129.142.112.33) joined #vserver. [17:15] [S]ushi (Sushi@pD9E348AD.dip.t-dialin.net) left irc: [17:16] riel (~riel@nat-pool-bos.redhat.com) joined #vserver. [17:24] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [17:24] MrBawb (abob@swordfish.drown.org) got netsplit. [17:25] MrBawb (abob@swordfish.drown.org) returned to #vserver. [17:25] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [17:57] say (~say@212.86.243.154) left irc: Ping timeout: 499 seconds [18:00] say (~say@212.86.243.154) joined #vserver. [18:44] Nick change: Bertl_zZ -> Bertl_oO [18:53] chrism (~chris@82-32-130-79.cable.ubr05.hawk.blueyonder.co.uk) joined #vserver. [18:53] hi all [18:54] does VServer 1.2 now natively include context quotas etc? [19:05] netrose (~john877@CC3-24.171.21.47.charter-stl.com) left irc: Ping timeout: 480 seconds [19:22] sylvio (sylvio@imk32.mb.uni-magdeburg.de) left irc: [19:29] i think so [19:30] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [19:30] MrBawb (abob@swordfish.drown.org) got netsplit. [19:31] MrBawb (abob@swordfish.drown.org) returned to #vserver. [19:31] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [19:32] maharaja (maharaja@ipax.tk) got netsplit. [19:32] Bertl_oO (~herbert@MAIL.13thfloor.at) got netsplit. [19:32] Bertl_oO (~herbert@MAIL.13thfloor.at) returned to #vserver. [19:32] maharaja (maharaja@ipax.tk) returned to #vserver. [19:53] chrism (~chris@82-32-130-79.cable.ubr05.hawk.blueyonder.co.uk) left irc: Quit: My damn controlling terminal disappeared! [20:15] Hmm. [20:16] Anyone had any trouble with 2.4.23? [20:16] (with or without vserver?) [20:18] i got one machine running 2.4.23 + vserver 1.2 no trouble so far, but it's just up for three days now [20:46] MrBawb (abob@swordfish.drown.org) left irc: Ping timeout: 480 seconds [20:46] MrBawb (abob@swordfish.drown.org) joined #vserver. [20:53] hmm [20:57] netrose (~john877@CC3-24.171.21.47.charter-stl.com) joined #vserver. [21:13] AGoe (~agoeres@Df8ae.d.pppool.de) joined #vserver. [21:14] AGoe (~agoeres@Df8ae.d.pppool.de) left irc: Client Quit [21:54] netrose (~john877@CC3-24.171.21.47.charter-stl.com) left irc: Ping timeout: 480 seconds [23:46] netrose (~john877@CC3-24.171.21.47.charter-stl.com) joined #vserver. [23:58] netrose (~john877@CC3-24.171.21.47.charter-stl.com) left irc: Ping timeout: 480 seconds [00:00] --- Thu Dec 11 2003