[00:04] a lot of output :) [00:04] really want to see it ? [00:04] just the line about proc [00:05] proc on /proc type proc (rw) [00:05] seems fine... [00:16] cool, looks like my comment (with link here) is the 3rd down from the /. article ;) [00:17] JonB (~NoSuchUse@129.142.112.33.ip.tele2adsl.dk) joined #vserver. [00:46] dilox (~dilox@host43-6.pool8249.interbusiness.it) joined #vserver. [00:46] ciao tutti [00:47] dilox (~dilox@host43-6.pool8249.interbusiness.it) left irc: Client Quit [00:47] oO( www.gelid.de ) [01:01] loger joined #vserver. [01:12] Termin4t0r (Termin4t0r@pD904A83C.dip.t-dialin.net) left irc: Quit: [01:37] Nick change: Bertl_oO -> Bertl [01:37] evening everyone ... [01:38] hey Bertl [01:38] wb Bertl [01:39] hi Jon! [01:39] hi Doener_! [01:39] Bertl: ive been playing around with union filesystems under FreeBSD. it seems to act like it should. the inodes do change for the copy on write but if a process already has the dile descripter open for reading it sees the changes due to how the stackable filesystem support works. [01:39] hi talon! [01:40] you can for example mount union over a read only filesystem. [01:40] and it pushes itself onto the vfs stack. [01:40] and stores all changes ina backing directory you specify at mount time. [01:40] yeah I know those stackable filesystems ... [01:41] only thing is there is no easy way to get files back form the original filesystem after youve deleted them. it apears to do this with special directory entrys in the backing store shadow. [01:42] Bertl: odes linux have any mechinisim similar ot stackable filesystems? [01:42] nope, but actually this kind of layering isn't that complicated [01:42] but it gives some troubles ... [01:46] im probably going ot stick with the bme patch. but its somthing id love to look into further if i ever get the time. [01:47] sure, remind me to search for the vfs filesystem sources I did some years? ago ... [01:49] ok, i will probably send an email about it. [01:51] i know nothign liek that will be perfect. but i would like to find/possibly create somthing that is at least as good as union, or good enough for most purposes. like everyone having theri own personal copys of a shared source tree, or sharing system librarys and binarys that dont change often. [01:51] well you saw the approaches on the TBVFS page ... [01:53] Bertl: still sifting through it. while i do other things. i just took a look at union when i didnt see it listed. thats why i provided a link to the stackable filesystem stuff since you seemd to be wanting to collect links to past attempts and concepts. [01:57] ahh its ff the main page on the past implementations. i thgouth you might want to put a link to teh paper onthe knoweldge page. [01:58] hmm, i forgot its all wiki. but i wasnt sure if that was the propper place for such things anyway. [01:59] did you want a list of published papers and books on the knoweldge and findings page? [02:04] sure [02:05] the UCLA paper and teh design and implementation of the 4.4BSD operating system are the only two i know about. [02:06] good, add them ... [02:06] ok. is this linked off the main site anywhere? [02:11] hmm, sorry lost you there, what do you want to know? [02:17] JonB (~NoSuchUse@129.142.112.33.ip.tele2adsl.dk) left irc: Quit: Leaving [02:17] paul (~irssi@muedsli-wan118.citykom.de) left irc: Quit: leaving [02:17] Nick change: riel -> unriel [02:18] riel: thanks forthe /. posting ;) [02:26] Bertl: ok how does that look? [02:27] ah okay, found it (Publications) [02:28] talon: hmm, maybe you should add yourself to the participants ;) [02:30] if you do so, please add a \\ after the first line which ends in 'and want' [02:30] (removing the paragraph space between first and second line [02:34] sigh ofcourse i read this after i did the change. bu ti will add that too i guess. [02:35] ok done. [02:37] i might see if there are other papers written on the other approaches listed later and add them. [02:47] jes_ (~jes@cpc1-leed5-3-0-cust196.ldst.cable.ntl.com) joined #vserver. [02:47] evening all [02:48] i probably wont be of much use in writing vfs code for a good long while though. [02:48] hi jes! [02:49] talon: well as I said, the 'example' code is somewhere, so that shouldn't be too hard ... [02:49] hiya Bertl [02:50] Bertl: oh im talking about my doing kernel code in general regardless of the existance of example code to work from. [02:50] you just need a good testing setup (like with qemu for example) [02:51] _shur1 (~shushushu@vserver.electronicbox.net) left irc: Ping timeout: 480 seconds [02:51] talon: so you are 'bgm-66-24-16-23.stny.rr.com' [02:52] oh, heh forgot about the ip feature. [02:52] not used ot uising wikis at all. [02:53] was http://vserver.13thfloor.at/Experimental/delta-2.6.3-vs0.09-vs-0.09.1b.diff generated with the help of the Ultra 10? if so im glad its been helpful. (im reading through the ml backlog) [02:55] ExpiryJames (~james@h24-71-63-164.ok.shawcable.net) left irc: Quit: Leaving [02:59] talon: no IIRC this was the compile test report from Gerald Leier (Subject: [Vserver] doesnt compile :( sparc64 - linux-2.6.3 + [03:00] yeah that was it. [03:01] but your sparc can account for the last few (not mentioned) rlimit format/size corrections ... [03:11] okay, I'm off to bed now ... cu tomorrow ... [03:12] Nick change: Bertl -> Bertl_zZ [03:12] Doener_ (~doener@pD9E120C3.dip.t-dialin.net) left irc: Ping timeout: 499 seconds [04:09] jes_ (~jes@cpc1-leed5-3-0-cust196.ldst.cable.ntl.com) left #vserver (Leaving). [04:59] Nick change: cdub -> cgone [05:15] soor_ (as@217.81.171.67) joined #vserver. [05:16] Doener (~doener@p5082D5BD.dip.t-dialin.net) joined #vserver. [05:17] soor (as@pD951A5FE.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [05:58] Mcleod (~altec@202.9.60.199) left irc: Read error: Connection reset by peer [06:40] Doener_zZz (~doener@p5082DA0A.dip.t-dialin.net) joined #vserver. [06:45] Doener (~doener@p5082D5BD.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [08:05] Doener_zZz (~doener@p5082DA0A.dip.t-dialin.net) left irc: Quit: Leaving [08:09] Doener (~doener@p5082DA0A.dip.t-dialin.net) joined #vserver. [09:23] talon (talon@host-63-149-223-100.irwinresearch.com) got netsplit. [09:23] unriel (~riel@riel.netop.oftc.net) got netsplit. [09:23] talon (talon@host-63-149-223-100.irwinresearch.com) returned to #vserver. [09:23] unriel (~riel@nat-pool-bos.redhat.com) joined #vserver. [09:26] hmm... there are about a dozen of unionfs in linux, [09:27] problem is - most of them are no longer maintained, [09:27] and none is going to get to mainline' [09:27] :( [11:22] Nick change: Bertl_zZ -> Bertl [11:22] morning everyone! [11:30] okay, cu all later ... [11:30] Nick change: Bertl -> Bertl_oO [11:57] Doener_zZz (~doener@p5082DA83.dip.t-dialin.net) joined #vserver. [12:01] Doener (~doener@p5082DA0A.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [12:03] stubbsd (~stubbsd@217.206.216.194) joined #vserver. [12:03] moning all, [12:25] rs (rs@ice.aspic.com) joined #vserver. [12:25] hi guys [12:45] _Axu_ (~axu@mail.mcmarketing.at) joined #vserver. [12:45] <_Axu_> hi folks :) [12:48] DazeDark (~chris@82-32-130-79.cable.ubr05.hawk.blueyonder.co.uk) joined #vserver. [12:48] hmm [12:48] I fear I have broken something ;) [12:56] deadguy (deadguy@bananajoe.big.du.se) left irc: Read error: Connection reset by peer [13:20] mcp (~hightower@wolk-project.de) joined #vserver. [13:29] pollar (bc3@dievai.net) left irc: Ping timeout: 499 seconds [13:44] Immunixy (~linux@dsl81-215-12613.adsl.ttnet.net.tr) joined #vserver. [13:45] DazeDark (~chris@82-32-130-79.cable.ubr05.hawk.blueyonder.co.uk) left irc: Quit: ..(cyp): [BX] Tickle-Me Elmo uses BitchX. *giggle* *giggle* *giggle* [13:47] Immunixy (~linux@dsl81-215-12613.adsl.ttnet.net.tr) left irc: Quit: Leaving [14:42] is there any info on how the setup memory limits ect in the 1.3.8 release? [14:46] chaosle (~yvan@bragi.fh-brandenburg.de) joined #vserver. [14:46] hi all [14:52] deadguy (deadguy@bananajoe.big.du.se) joined #vserver. [15:14] monrad (~monrad@213083190243.sonofon.dk) joined #vserver. [15:16] Nick change: Bertl_oO -> Bertl [15:16] hi everyone! [15:20] pollar (nobody@dievai.net) joined #vserver. [15:21] hi pollar! [15:23] _Axu_ (~axu@mail.mcmarketing.at) left irc: Quit: Client exiting [15:33] * Solaris has long had the "translucent" file system. [15:33] thats not actually true. [15:33] Solaris 2.x doesnt have that. [15:33] hehe [15:33] in 2.x it was so very translucent, you didn't even see it ;) [15:34] heh :) [15:34] i find the wrapfs stuff interesting. [15:37] _Axu_ (~axu@mail.mcmarketing.at) joined #vserver. [15:37] hi _Axu_! [15:38] <_Axu_> hi Bertl :) [15:38] there is quite a list of customers on the mcmarketing page ;) [15:39] <_Axu_> Bertl: jepp [15:39] looks good for an austrian company ... [15:39] <_Axu_> Bertl: its a marketingcompany...so dont believe everything you see there ;) [15:39] ah, good point 8-) [15:42] awesome. [15:42] check this page out. [15:42] http://www.filesystems.org/ [15:42] written by the wrapfs guy. [15:43] if what he says about ti cratign solaris modules is true... i will be a very happy man. also quite happy if it creates valid linux modules too. [15:44] talon: be careful ... [15:45] Bertl: heh, i dont plan to use any of this in production. but it does look like a rather cool concept. [15:45] no, what I mean is, Erez (Zadok) is a special kind of guy ... [15:47] Bertl: special in what way? [15:48] in the way, that Fist (which is partially crap, sorry) is the ultimate solution for any problem ... [15:50] heh, ive ran into those sorts before :) [15:50] one thing he for sure does, is publishing ... [15:50] so what is particularly crap about fist ? [15:51] so i know what im getting into. [15:55] well it didn#t even compile, and wasn't updated for quite a while, just the moment it was discussed (as being crap) a new version appeared ... [15:56] the idea is not so bad, and you can use it for testing ... [15:56] but that's it ... [15:57] Action: talon nods [15:58] if anything it will generate a template i can refine into real code. [15:58] assuming i have time to play with it. [15:58] the main thing that caught my attention is that it supported making filesystem modules for solaris. which ive been wanting to see a good non encumbered example for. [16:05] that and the mention of patches no longer being required ofr 2.4.x kernels. [16:05] yeah, it can't be wrong to check it out ... so let me know what you find ... [16:06] Bertl: will do for sure. at least once things calm down a bit around here. definatly somthing im interested in. [16:07] Bertl: im still interested inthe vfs code you were working on for the template fs as well ofcourse. [16:08] well, im off to bed. ttyl [16:13] cu [16:34] frealek (~nobo@guigeek.org) joined #vserver. [16:34] hi [16:34] hi frealek! [16:34] I havent dig up much of the docs, but I didn't see that mentionned... [16:35] what is the diff between vs / uml ? [16:35] you know uml? [16:38] I runned 1 uml w/ skas, 2.4.20~~ [16:38] zev_ (~zev@masya.aviaserv.com.ua) joined #vserver. [16:38] still some glitches w/ 2.6 umls [16:38] hey [16:38] Nick change: zev_ -> zev [16:38] okay, if you open a file, what happens behind the scenes ... [16:38] hi zev! [16:39] simple: the open is a syscall into the uml ekrnel, which activates a special block device driver, which is emulated by the uml program, which in turn calls a syscall into the host kernel, which reads data from a file, and passes this back all the way [16:40] on vserver, there is no in between kernel ... [16:40] Berti is rebootmgr proveded to reboot|halt commands working inside vserver? [16:40] provided [16:40] hmm so could it be painted as "less secure" ...? [16:40] frealek: the syscall is handled exactly like it would happen on a normal systema [16:41] frealek: hmm, well it is secure, and much faster ... [16:41] I bet then that there are performance gain [16:41] s [16:42] zev: should be in the tools ...somewhere ... but there is also the vshelper interface ... [16:42] frealek: fact is, you can put about 30-60 vservers on a dual 1GHz pentium with 1.5GB ram ... [16:42] oh my :) [16:43] each complete with postfix/apahce/etc [16:43] Berti i installed util-vserver-0.29, than vserver build vs01, than vserver vs01 start, than sshed to vs01, than run reboot, but got: [16:43] "init: timeout opening/writing control channel /dev/initctl" [16:43] also, service rebootmgr start... :) [16:44] zev, for the rebootmanager you need a modified vreboot ... with the vshelper you need reboot -f [16:44] ok, now what's the diff between vs / virtual server (in 2.6.* kernels) ? I don't know this at all [16:44] Berti what is vshelper, and where i can find info about? which vshelper: /usr/bin/which: no vshelper i [16:46] Berti sorry, i'm turbing you, when you discuss something important [16:48] zev: np ... [16:51] Berti so, to reboot/halt inside VS, i need to run vreboot, not reboot ? [16:51] depends ... [16:51] Berti thats all? [16:51] Berti from what? [16:51] if you use the rebootmanager [16:51] then you need a modified reboot [16:51] Berti just rename vreboot, to reboot inside vs? [16:52] yes, that should work ... [16:52] hey Bertl, unificatoin works for me with 1.3.8, thanks for the help [16:52] or unification even [16:52] ben: great, np [16:52] Berti btw, why one should use vshelper? and what is that ? [16:52] vshelper is called from the kernel in behalf of the reboot [16:52] so if a process in the vserver executes a sys_reboot [16:53] then vshelper is started on the host, and can take down the vserver [16:53] Berti where can i read about? [16:53] basically on the mailing list (or the archives) [16:54] give me a second, I might have a link# [16:54] http://archives.linux-vserver.org/200311/0115.html [16:55] 10x [16:59] Action: frealek is intensively waiting for 2.6 vservers :)) [17:00] Berti moved v[halt|reboot] on not 'v' done reboot/halt, works great, thanx for the tip :) [17:00] vservers seem more flexible than UMLs towards networking [17:01] zev: ok .. np [17:01] frealek: yes, but UML has also some advantages ... [17:01] for example you can use a different kernel ... [17:01] and of course you can combine them ... [17:01] vserver on uml, uml on vserver ... [17:02] Bertl: oh you mean, vservers are the same version of current kernel ? [17:02] actually they are not even that ... [17:03] vserver is a kind of partitioning/process separation ... [17:03] ok [17:04] Berti but there is one more trouble, i assigned for my VS 2 ip's one is dummy0:XX IP and second from eth0:XX real ip. they brings when i boot up my host(main) system, i don't want vserver script to do it, i commented all "ifconfig" lines in /usr/sbin/vserver and it works for me, except for that when i run ifconfig inside VS i see ALL dummyX, eth0:X, ppp0:X ifaces that is on my MAIN system, is it right? [17:08] well, partially, you should see that there is something ... but you should not see the ip for example ... [17:14] Berti hmmm... i currently see all of them. [17:15] then something is wrong ... [17:15] Berti and i think so, but don't understand why [17:15] should look like this: [17:15] eth0 Link encap:Ethernet HWaddr 00:E0:81:21:22:B1 [17:15] UP BROADCAST MULTICAST MTU:1500 Metric:1 [17:15] RX packets:0 errors:0 dropped:0 overruns:0 frame:0 [17:15] TX packets:10186 errors:0 dropped:0 overruns:0 carrier:10186 [17:15] collisions:0 txqueuelen:1000 [17:15] RX bytes:0 (0.0 b) TX bytes:611160 (596.8 Kb) [17:15] Interrupt:19 Base address:0x3400 [17:16] Berti ahhh. sorry. yes i see all my "root" ifaces but ip's [17:17] Berti is there any way to hide them fully [17:17] ? [17:17] frealek (~nobo@guigeek.org) left irc: Quit: No need to understand now. Soon you will understand it. Others will do. [17:18] Berti i have 6 pppX vpn ifaces on my root host, and don't want inside VS them totally. not only them without ip's :) [17:18] paul (~irssi@195.202.59.212) joined #vserver. [17:19] hi [17:20] hi paul! [17:20] Nick change: unriel -> riel [17:28] serving (~serving@213.186.190.121) left irc: Read error: Connection reset by peer [17:35] \quit [17:35] zev (~zev@masya.aviaserv.com.ua) left irc: Quit: Trillian (http://www.ceruleanstudios.com) [17:46] <_Axu_> i am a little confused with the new vserver utils.... how iss the official configfile directorystructure of vserver /etc/vservers or /etc/util-vserver ... or is it both and you put the vserverconfigs into /etc/vservers and some stuff i dont know into /etc/util-vservers ? [17:51] please define 'new vserver utils' ... [17:51] <_Axu_> util-vserver-0.29 [17:52] stable util vserver uses the same config files as the older tools do ... [17:54] <_Axu_> Bertl: hmmm, ok... so its still /etc/vservers/serverone.config and /etc/vservers.config + the vserver and rebootmgr startupscript ? [17:57] yep, there is an additional config file, which knows the tool location, but that is 'configured' automatically on build/install [17:57] but enrico (ensc) knows more about it (he wrote it) [17:58] <_Axu_> Bertl: i c thanx [17:59] maharaja (maja@ipax.tk) left irc: Remote host closed the connection [18:05] maharaja (maja@ipax.tk) joined #vserver. [18:06] <_Axu_> how can i check if ctx iss working without a vserver ? [18:07] <_Axu_> vserver-stat still gives me : unable to switch in context security #1 [18:07] chaosle (~yvan@bragi.fh-brandenburg.de) left irc: Ping timeout: 480 seconds [19:04] chaosle (~yvan@bragi.fh-brandenburg.de) joined #vserver. [19:15] _shur1 (~shushushu@vserver.electronicbox.net) joined #vserver. [19:17] Nick change: Bertl -> Bertl_oO [19:22] <_Axu_> ok, i had mix of different vserver tools on this machine but cleaned it up. but still, the new tools seem not to recognize ctx. maybe tomorrow :) by folks [19:26] grisha (~grisha@ip68-100-225-22.dc.dc.cox.net) joined #vserver. [19:26] serving (~serving@213.186.190.121) joined #vserver. [19:28] So what is the best way to measure network io per vserver? [19:34] nm - i think i found an answer in a faq - using iptables - going to investigate [19:34] grisha (~grisha@ip68-100-225-22.dc.dc.cox.net) left irc: Quit: Leaving [19:36] _shur1 (~shushushu@vserver.electronicbox.net) left irc: Ping timeout: 480 seconds [20:04] Nick change: cgone -> cdub [20:36] stubbsd (~stubbsd@217.206.216.194) left irc: Remote host closed the connection [21:31] rs (rs@ice.aspic.com) left irc: Quit: pula [21:41] serving (~serving@213.186.190.121) left irc: Read error: Connection reset by peer [21:49] chaosle (~yvan@bragi.fh-brandenburg.de) got netsplit. [21:49] maharaja (maja@ipax.tk) got netsplit. [21:49] paul (~irssi@195.202.59.212) got netsplit. [21:49] pollar (nobody@dievai.net) got netsplit. [21:49] monrad (~monrad@213083190243.sonofon.dk) got netsplit. [21:49] deadguy (deadguy@bananajoe.big.du.se) got netsplit. [21:49] pollar (bc3@dievai.net) joined #vserver. [21:49] deadguy (deadguy@bananajoe.big.du.se) returned to #vserver. [21:49] chaosle (~yvan@bragi.fh-brandenburg.de) returned to #vserver. [21:49] maharaja (maja@ipax.tk) returned to #vserver. [21:49] monrad (~monrad@213083190243.sonofon.dk) returned to #vserver. [21:54] paul (~irssi@195.202.59.212) returned to #vserver. [22:08] ben (~ben@bengrimm-host225.dsl.visi.com) left #vserver. [22:20] monrad (~monrad@213083190243.sonofon.dk) left irc: Remote host closed the connection [22:38] chaosle (~yvan@bragi.fh-brandenburg.de) left irc: Quit: Leaving [00:00] --- Thu Mar 4 2004