1107993618 Q * prae Quit: Pwet 1107998737 Q * sod Ping timeout: 480 seconds 1107999850 Q * _are_ Ping timeout: 480 seconds 1108000512 J * _are_ ~are@dsl-084-056-132-190.arcor-ip.net 1108002800 Q * _are_ Ping timeout: 480 seconds 1108003584 M * Bertl okay folks, cya tomorrow ... 1108003601 M * Bertl night everyone! 1108003607 N * Bertl Bertl_zZ 1108010776 J * nox- ~nox@213.39.193.107 1108011123 Q * nox Ping timeout: 480 seconds 1108011136 N * nox- nox 1108012130 P * Vantage Kopete 0.9.1 : http://kopete.kde.org 1108016979 J * Vantage ~Vantage@www.toddcharron.com 1108017003 M * Vantage is there a document that describes how to use quotas and disk limits with vserver? 1108020573 J * witchdoc ~witchdoc@62.206.75.213 1108020579 M * witchdoc hi all 1108020816 M * witchdoc i use a 2.6.9-vs1.9.3n debian sarge hostsystem and the util-vserver-0.30.203. i can start a guest and enter into it, but all ssh connections to this guest goest to the host - why? 1108020994 Q * virtuoso Remote host closed the connection 1108021026 J * virtuoso ~s0t0na@tranq.dorms.spbu.ru 1108021733 M * hansi33 witchdoc, you have to bind the sshd from the host to hist ip-adress, for example with 1108021748 M * hansi33 ListenAddress 1.2.3.4 in /etc/ssh/sshd_config 1108021778 M * witchdoc hansi33: thx i try this 1108022007 M * witchdoc hansi33: doesnt work :( 1108022027 M * witchdoc maybe i should upgrade to http://vserver.13thfloor.at/Experimental/RC-1.9.4/ 1108022086 M * witchdoc i saw this in the a irc-log 1108022109 M * hansi33 witchdoc, i have the same version and no problems 1108022398 M * witchdoc hmmm 1108022421 M * witchdoc with debian as host ant 1.9.3 kernel? 1108022430 M * witchdoc -t+d 1108022740 Q * Snow-Man Read error: Operation timed out 1108022751 J * Snow-Man ~sfrost@snowman.net 1108022861 M * witchdoc hansi33: one question - edit the sshd_config of the host and/or the guest-system? 1108022895 M * hansi33 for the guest it is not necessary. 1108022954 M * hansi33 dig you restart you host, after edit sshd_config? 1108022968 M * witchdoc hansi33: jepp 1108023007 M * witchdoc okay - with the right settings in my hosts sshd_config it goes 1108023026 M * witchdoc thanks :-) 1108023555 J * Markus noid123@213-102-88-124.cust.tele2.de 1108023582 M * Markus good morning :-) 1108023611 M * Markus anyone time to help me with a vserver network problem on gentoo? 1108023699 M * hansi33 Markus, perhabs i can help? 1108023815 M * Markus i set up a debian woody vserver following a howto i found on the net, also set up ssh, but whenever i try to connect using ssh, i end up on the host machine instead of the vserver 1108023824 M * Markus maybe some missing kernel module or something 1108023834 M * Markus allready emerging iproute2 1108023856 M * hansi33 it was five minutes ago in this chat: 1108023858 M * hansi33 ListenAddress 1.2.3.4 in /etc/ssh/sshd_config 1108023862 M * witchdoc hehe 1108023867 M * hansi33 in your host! 1108023872 M * witchdoc shame on us :-) 1108023901 M * Markus could be it... i'll try it... should have results in a minute :-) 1108023907 M * witchdoc Markus: cause your host sshd listen to _all_ ips if you dont set this 1108023932 M * Markus i'll send a mail to the howto creator :-) 1108023941 M * Markus so i won't need NAT or something in the kernel? 1108023946 M * Markus it's very minimalistic 1108023968 M * witchdoc Markus: why nat? 1108024003 M * Markus thought it could have something to do with getting the packages where they should go... never mind :-) 1108024025 M * Markus *dau* it even says it in the manual ... sorry for wasting your time 1108024118 M * witchdoc Markus: what for a vserver-patch / kernel use you 1108024158 M * Markus the ones from: http://www.13thfloor.at/vserver/s_release/v1.2.10/ 1108024165 M * Markus didn't like the ebuilds 1108024195 M * hansi33 why not with the ebuilds, they worked for me. 1108024245 M * Markus Had the same network problem with them and somehow thought it could be due to the ebuilds 1108024268 M * Markus just in case i could get around forming english words... you speak german? :-) 1108024294 M * hansi33 yes 1108024303 M * witchdoc yes ist gut 1108024335 M * Markus besser :-D 1108024405 M * witchdoc allso muss man alle dienste des hostrechners an _eine_ ip binden, sonst bringt das nur ärger 1108024485 M * Markus gut zu wissen... bin gerade etwas verwirrt... unter gentoo gibt es keine /etc/xinetd.conf 1108024500 M * Markus nur nen ordner /etc/xinet.d/ 1108024510 M * hansi33 oder besser fast gar keine dienste mehr am hostrechner lassen 1108024521 M * Markus mit 2 files: cups-lpd swat 1108024532 M * Markus sowieso :-) 1108024612 M * witchdoc hansi33: ja das ist das ziel 1108025198 J * prae ~prae@212.11.15.34 1108025406 M * Markus dafür nen lösungsvorschlag: /usr/local/sbin/vserver: line 535: ulimit: max user processes: cannot modify limit: Invalid argument 1108025407 Q * flock Read error: Operation timed out 1108025415 M * Markus vproc habe ich eben noch dazu kompiliert 1108025769 Q * Markus Quit: 1108026400 J * flock ~restless@192.117.111.12 1108028186 Q * Pinnen Ping timeout: 480 seconds 1108028891 M * witchdoc hansi33: die option with-initrddir benutze ich bei /etc/init.d/ scripten wie? 1108028900 M * witchdoc wenn ich die tools erstelle? 1108028913 M * witchdoc habe kein /etc/rcd 1108028924 M * witchdoc habe kein /etc/rc.d 1108029003 M * hansi33 with-initrddir kenn ich gar nicht, habe mich ziemlich genau an http://dev.gentoo.org/~hollow/vserver/guide/ gehalten, da gibts eigenes ebuild für die guests 1108029114 M * witchdoc hmmm - ich hab nen debian host 1108029137 M * witchdoc wobei ich überlege nen gentoo host zu nehmen und debian als guests 1108029448 Q * monrad Quit: Leaving 1108029456 A * aba nutzt Debian als host und guest 1108029807 J * rs rs@ice.aspic.com 1108029809 M * rs hi 1108031123 M * witchdoc hi rs 1108032375 J * Pinnen ~pinnen@h194n2fls35o917.telia.com 1108033238 J * monrad ~monrad@213083190130.sonofon.dk 1108033269 Q * monrad Quit: 1108035296 M * Doener witchdoc / hansi33 : please speak english in this channel ;) 1108035550 M * witchdoc Doener: okay 1108036084 J * Vranarc ~vranarc@80.97.101.3 1108036134 M * Vranarc can i run a psyBNC on a linux dedicated server? 1108036388 M * witchdoc Vranarc: whats psyBNC? 1108036429 M * Vranarc ermm you know.. the irc thing thing that remains after you quit.. and log messages for you 1108036465 M * TheSeer what does a bouncer have to do with vserver? 1108036489 M * Vranarc does the vserver support a bouncer? 1108036499 M * witchdoc Vranarc: sure 1108036506 M * TheSeer it has nothing to do with vserver 1108036512 M * TheSeer vserver doesn't care 1108036524 M * TheSeer it's a service as any other 1108036532 M * Vranarc then how in blody hell do i run a psybnc... i don't have pico.. i use vi 1108036550 M * TheSeer this has NOTHING to do with vserver 1108036563 M * Vranarc ohh so u aren't going to help me.... 1108036565 M * TheSeer go to a psyBNC support channel for help on getting it configured 1108036569 M * Vranarc ok 1108036573 M * Vranarc thnks 1108036575 M * witchdoc Vranarc: thats your choice. install what you want in the guest-system 1108036576 M * TheSeer we can't help you here 1108036584 M * TheSeer even if we want to 1108036596 M * TheSeer since i just know WHAT psyBNC is but never used it 1108036597 M * Vranarc ok bbye 1108036599 P * Vranarc 1108036629 M * witchdoc maybe he should use irrsi :P 1108036637 M * TheSeer hehe 1108036709 M * witchdoc how can i separte pkgstate-Directory to /var/run/vserver ? (alpha-utils) 1108036742 M * witchdoc /configure --sysconfdir=/etc --with-vrootdir=/var/lib/vservers --with-initrddir=/etc/init.d --localstatedir=/var/run --prefix=/usr doesnt work right :( 1108036831 M * witchdoc this option result pkgstate-Directory: /var/run/run/vservers 1108036836 M * Doener witchdoc: i guess you got /var/run/run/vservers now? IIRC it should read --localstatedir=/var 1108038008 J * mutiman ~greenman@217.154.180.50 1108038357 M * mutiman Greetings. I am seeking a little advice - but it's about LVS. (I could find no dedicated LVS channel, so decided that you folks may know something anyway ) 1108038446 M * mutiman Are there any experienced LVS folks around, or should I seek elsewhere ? 1108038634 M * witchdoc Doener: if i run make with --localstatedir=/var configure, i got an error-msg 1108038658 M * witchdoc hi mutiman 1108038663 M * witchdoc LVS? 1108038676 M * mutiman Linux-Virtual-Server 1108038680 M * mutiman not vserevr 1108038682 M * Doener witchdoc: what error message? my crystalball seems to be broken... 1108038785 M * Doener mutiman: http://archive.linuxvirtualserver.org/html/lvs-users/2004-03/msg00137.html 1108038786 M * witchdoc Doener: mom 1108038789 M * Doener and http://archive.linuxvirtualserver.org/html/lvs-users/2002-11/msg00015.html 1108038791 M * witchdoc i run make ... 1108038851 M * witchdoc lib/syscall.c:28:26: linuxvirtual.h: No such file or directory 1108038862 M * witchdoc make[2]: *** [lib/lib_libvserver_la-syscall.lo] Fehler 1 1108038864 M * witchdoc make[2]: Leaving directory `/home/src/util-vserver-0.30.203' 1108038929 M * Doener hm, that shouldn't be related to --localstatedir settings... check if that file is in the base directory of the util-vserver source tree 1108038933 M * mutiman Thanks Doener - I'll peer in that direction 1108039170 M * witchdoc Doener: but if i use a longer --localstatdir setting it work 1108039173 M * witchdoc s 1108039568 M * witchdoc hmm linuxvirtual.h is a symlink 1108039577 P * mutiman 1108039590 M * Doener indeed 1108039697 M * Doener hm, works fine here, with --localstatedir=/var ... did you 'make clean' (or even distclean)? 1108039735 M * witchdoc hmmm sometimes 1108039829 A * witchdoc run make clean && make distclean now 1108039988 M * witchdoc same error 1108040000 M * witchdoc lib/syscall.c:28:26: linuxvirtual.h: No such file or directory 1108040015 M * witchdoc ls -la linuxvirtual.h 1108040015 M * witchdoc lrwxrwxrwx 1 root root 15 2005-02-10 13:48 linuxvirtual.h -> ./lib/virtual.h 1108040039 M * witchdoc ls -la ./lib/virtual.h 1108040040 M * witchdoc ls: ./lib/virtual.h: Datei oder Verzeichnis nicht gefunden 1108040546 M * witchdoc i try to untar the source fresh 1108041239 M * witchdoc Doener: hmmm what means src/vuname.c:203: warning: warning: your code still has assertions enabled! 1108041349 M * witchdoc Doener: the No such file or directory error gone with the wind, with fresh source 1108041537 M * bro hiya! 1108041635 M * bro ay1 got any idea if vserver is available with grsec? 1108041661 M * bro im interested in 2.4.29 1108042492 M * witchdoc hi bro 1108042514 M * witchdoc i saw some grsec patches for vserver, but i dont know the kernel-version 1108042556 M * DaCa 2.4.28, no .29 (yet) 1108042567 M * DaCa sandino.net/parches/vserver 1108042784 M * bro thanks ;> 1108043618 M * witchdoc vserver-copy in the alpha-utils doesn support the new configs ? 1108044164 N * Bertl_zZ Bertl 1108044172 M * Bertl morning folks! 1108044306 J * Pazzo ~thomas@213.21.136.175 1108044306 M * witchdoc morning Bertl 1108044331 M * Pazzo moinmoin 1108044335 M * Bertl hey witchdoc! morning Pazzo! 1108044400 M * Pazzo bertl: what about ngn? did you find a solution for your "communication between vservers"-problem? 1108044487 M * Bertl I have some ideas for that, but it will take some time, for now ng8.13 (which strictly isolates contexts) needs to be tested ... 1108044999 M * Pazzo is there a ngnet patch for 2.6.11-rc3-vs1.9.4-rc5? 1108045397 M * Bertl hmm, not yet I guess, but I can update ng8.13 ... 1108046230 M * Bertl okay, give a few minutes to test compile, and you have your patch ... ;) 1108046935 M * witchdoc hmmm, vserver beta build --force -m debootstrap --d sarge doesnt work. 1108046958 M * witchdoc /usr/lib/util-vserver/vserver-build: line 51: /dev/null: Datei oder Verzeichnis nicht gefunden 1108046959 M * Bertl --d looks weird ;) 1108046975 M * Bertl -- -d sarge would look better 1108047194 J * jfm ~jfm@57-71.244.81.adsl.skynet.be 1108047202 M * Bertl welcome jfm! 1108047214 M * jfm Hello, 1108047275 M * Bertl Pazzo: http://vserver.13thfloor.at/Experimental/NGNET/diff-2.6.11-rc3-vs1.9.4-rc5-ng9.0.diff 1108047276 M * jfm I forward a question : chcontext: vc_new_s_context(): Operation not permitted 1108047291 M * Bertl debian with old-style config? 1108047298 M * jfm This error is systematic trying to enter a vserver after a start 1108047306 M * jfm yes old style 1108047334 M * jfm vserver is working ok, all services (2.6.9-1.9.3) 1108047342 M * Bertl upgrade to the latest kernel (1.9.4-rc5) and it should go away 1108047356 M * jfm ok 1108047387 M * Bertl a, and btw, you should change a few things (if it is really debian) 1108047400 M * jfm No it is a mdk10.1 1108047426 M * Pazzo hehe... bertl, you're great! is http://vserver.13thfloor.at/Experimental/NGNET/diff-2.6.11-rc3-vs1.9.4-rc5-ng9.0.diff ready for being tested? 1108047431 Q * ndim Ping timeout: 480 seconds 1108047450 M * Bertl Pazzo: that's the idea, yes ;) 1108047473 M * Bertl jfm: okay, so I guess your .config is fine then ... 1108047520 M * jfm What do you mean? 1108047570 M * Bertl s/.config/*.conf/ 1108047609 M * jfm Do they have to be modified according to the kernel patch? 1108047609 M * Bertl I guess I add a wiki page for that ... 1108047618 J * ndim U2FsdGVkX1@helena.bawue.de 1108047735 M * jfm kernel to use is 2.6.11 ? 1108047792 M * Bertl it's not out yet, but 2.6.11-rc3, yes 1108047819 M * Bertl http://vserver.13thfloor.at/Experimental/RC-1.9.4/patch-2.6.11-rc3-vs1.9.4-rc5.diff 1108047842 M * witchdoc Bertl: hmm my /dev was /dev.X - so i got an error 1108047858 M * witchdoc ant the stuipd -d :P 1108047923 M * Bertl so I assume it's working now, right? 1108047932 M * witchdoc yes 1108047950 M * Bertl good ;) 1108048389 M * witchdoc the --interface options is confusing me 1108048426 M * witchdoc --interface eth0:192.168.0.45/24 maybe right? 1108048441 M * witchdoc if the guest becomes .45 1108048458 Q * jfm Remote host closed the connection 1108048707 M * Bertl witchdoc: what confuses you about that notation? 1108048802 A * Doener yawns 1108048805 M * Doener hey folks 1108048810 M * Bertl morning Doener! 1108048853 M * witchdoc Bertl: the ip/netmask part i understand 1108048869 M * Doener hm, and eth0 is the interface :) 1108048905 M * witchdoc but i am not sure if --interface eth0:192.168.0.45/24 is enough 1108048910 M * witchdoc for example 1108048985 M * witchdoc i use the alpha-utils and thet 1.9.3 patch 1108049027 M * witchdoc but what can i do with the [=] ? 1108049032 M * Doener it is enough to get a non-aliased ip address assigned to the vserver, being created on vserver startup and destroyed on vserver stop 1108049050 M * Bertl witchdoc: by default the ip will not be created as alias 1108049074 M * Bertl if you use hansi=eth0:.... the tools will make an alias eth0:hansi 1108050039 Q * Pazzo Quit: Download Gaim: http://gaim.sourceforge.net/ 1108050234 M * witchdoc aha. thx Bertl 1108050238 M * Vantage anyone know of a quick howto or guide to using quotas/disk limits once you've got all the patches applied? 1108050292 M * Bertl http://linux-vserver.org/index.php?page=Quota+and+Disk+Limits 1108050481 M * Vantage hrmm that explains how they work, but i'm still confused on how to use them. say i wanted to limit a vserver to 1Gig. how would I do that? 1108050546 M * witchdoc Vantage: lvm and a own partition ;-) 1108050618 M * Bertl Vantage: we are talking about 2.4 or 2.6 kernel? 1108050650 M * Vantage Bertl: 2.4.29 w/ 1.2.10 and patch-2.4.27-vs1.29-q0.14.diff applied 1108050778 M * Vantage witchdoc: the problem with that is then the total space of their quota is always allocated to them, even if it's not in use 1108050791 M * Bertl http://archives.linux-vserver.org/200402/0002.html 1108050822 M * Bertl (check the entire thread) 1108050869 M * Bertl Vantage: the basic priciple is: 1108050886 M * witchdoc Vantage: thats true 1108050893 M * Bertl - add a quota hash for the context/filesystem 1108050904 M * Bertl - set the appropriate limits 1108050933 M * Bertl - start the vserver 1108050967 M * Vantage i'll give it a read. thanks 1108051023 M * Bertl http://vserver.13thfloor.at/Linux2.6/index.php?page=Per+Context+Disk+Limits (good overview too) 1108051178 M * Vantage Bertl: do i need the experimental cq-tools? 1108051211 M * Vantage nm, seems to be the same version i have here already 1108051248 M * Vantage btw the link to cq-tools on that last url 404s 1108051254 M * Bertl cq-tools-0.06 is fine ... 1108051294 M * Bertl hmm, yes, thanks ... 1108051311 M * Bertl (the others probably do so too ;) 1108051325 M * Vantage yup ;) 1108051358 M * witchdoc time to go, thanks for help! 1108051361 M * witchdoc cya 1108051374 Q * witchdoc Quit: feierabend 1108051379 M * Vantage so all the cqtools are just symlinks to mquota? 1108051398 M * Bertl yes, it's a multi-purpose binary 1108052765 M * Vantage Bertl: what enables the mount option tagctx? that doesn't appear to work for me 1108053240 M * Bertl you specify it at mount time 1108053261 M * Bertl like mount -o tagxid /dev/whatever /vservers 1108053687 M * Vantage yeah, i had rebooted the machine and it booted the wrong kernel 1108053693 M * Bertl ;) 1108053747 M * Vantage ugh, and 2.4 doesn't like my lvm (no device-mapper module). how's building for 2.6? ;) 1108053828 M * Bertl well, if you just need disk limits, 2.6 will be fine 1108053842 M * Bertl per context quota support is not available for 2.6 ;) 1108053858 M * Bertl but you can also add the dm patch/module to 2.4 ;) 1108053939 M * Vantage ugh. decisions, decisions ;) 1108053997 M * Vantage per context quota support is just enabling user quotas from within the vserver, right? 1108054034 M * Bertl yep, precisely! 1108054051 M * Bertl will work on separate partitions (lvm) but not on shared ones 1108054133 M * Vantage so if I make say a 6gig lvm partition and dedicate that to 1 vserver i can get user quotas on that? but if I have another 8 gig partition that has 4 vservers on it, i can't? 1108054547 M * Bertl you can not have more than one working quota on a single partition ... 1108054590 M * Bertl so if you put 3 servers on a partition, you _can_ have 3 times disk limits, but just _one_ user/group quota 1108054629 M * Bertl (and this user/group quota will be across all vservers on that partition) 1108054823 M * Vantage hrmm. well i'm attempting to build another 2.4 w/ device-mapper support. we'll see how that goes 1108054824 J * JonB ~NoSuchUse@cpe.atm2-0-1031198.0x50a4ad0e.bynxx13.customer.tele.dk 1108054830 M * JonB hey Bertl 1108054847 M * JonB Bertl: your remote reset device... and the parport program 1108054849 M * Bertl hey JonB! long time no see! 1108054885 M * JonB Bertl: i've connected data 0 and data1 + data 2 and data 3 like your diagram 1108054897 M * JonB Bertl: one of them resets with this command 1108054913 M * JonB parport -p 0 2 1108054921 M * JonB and stops reseting with -p 0 0 1108054941 M * JonB and it doesnt matter which way i turn the cable on the 2 reset pins 1108054946 M * Bertl the other one will work with 1108054948 M * JonB but the other wont work 1108054961 M * Bertl parport -p 0 4 or parport -p 0 8 1108055019 M * JonB thanks it was 8 1108055024 M * JonB i only tried up to 7 1108055032 M * JonB how come it is 8 ? 1108055053 M * Bertl 2^0 = 1 (data 0) 1108055059 M * Bertl 2^1 = 2 (data 1) 1108055066 M * Bertl 2^2 = 4 (data 2) 1108055068 M * JonB 2^3 1108055070 M * JonB right 1108055074 M * Bertl ;) 1108055091 M * JonB and -p 0 0 always turns the reset signal off ? 1108055102 M * Bertl -p 0 0 sets all lines to off 1108055112 M * JonB okay 1108055116 M * Bertl which probably has the same result as -p 0 255 1108055123 M * Bertl (which sets all lines to on) 1108055144 M * JonB so i can reset multiple machines with that? 1108055191 M * JonB 255 wont work 1108055193 M * JonB 256 might 1108055254 M * Bertl no, 255 will have the same result as 0 and 256 is broken ;) 1108055277 M * JonB but 255 doesnt fit into 1 2 4 8 16 ... 1108055296 M * Bertl look, the basic principal here is trivial ... I'll explain: 1108055327 M * Bertl you have a opto-coupler attached with the LED side to two data lines (e.g. data 0 and data 1) 1108055390 M * Bertl now the LED will work if you attach power to the pins (but only if you do it in the 'right' way, i.e. +/- does only work in one direction) 1108055428 M * Bertl now you can set each data line (data 0 -8) with the parport command 1108055450 M * Bertl s/data 0 -8/data 0,1, ... 7/ 1108055489 M * Bertl if you write 17 to the parport, the data 0 and data 4 line will be '1' all others '0' 1108055520 M * JonB ahh 1108055520 M * Bertl if you attach a led to data0/1 and a different one to data4/5 for example 1108055546 M * Bertl you will be able to activate the first led with 2^0 but not with 2^1 1108055559 M * JonB right 1108055569 M * Bertl so 0,2 and 3 will turn it off, while 1 will turn it on 1108055602 M * Bertl similar for the other led, just shifted 0,32 and 48 will turn it off, 16 will turn it on 1108055617 M * Bertl 17 in this case will light both leds ... 1108055643 M * Bertl in your setup 10 will reset both machines ;) 1108055939 M * JonB mail is on the way 1108055946 M * JonB with a better explination 1108055953 M * JonB based on what you said 1108056040 M * Bertl s/explination/explanation/ s/added/ored/ 1108056062 M * JonB ored? 1108056073 M * JonB if you or 2 and 8 do you get 10 ? 1108056081 M * JonB you do if you add 1108056096 M * Bertl yes arithmetic or, if the values are powers of two, it's equal to the add 1108056107 M * JonB okay 1108056130 M * JonB maybe i can figure it out next time so i wont have to come running here ;-D 1108056146 M * Bertl hopefully ;) (at least via the irc logs then ;) 1108056171 M * JonB they are automatically logged? 1108056173 M * JonB how smart 1108056197 M * Bertl (and in realtime) 1108056200 M * JonB nice 1108056207 M * JonB hi mom 1108056212 M * JonB ;-D hahahaha 1108056226 M * Bertl lol 1108056568 M * JonB i dont think i will bring back the test server 1108056594 M * JonB i had a breakin where they stole the harddisks and ram in the machine that was connected to the testserver 1108056601 M * JonB annoying :/ 1108056613 M * Bertl yes, you mentioned that last time ... 1108056674 M * Bertl well, it was a good test place back then, but I guess I will not terribly miss it nowadays ...anyway, thanks again for putting up that machine! 1108056743 M * JonB i suppose you have more interresting machines 1108056752 M * JonB like quad cpu's 1108056756 M * JonB and such 1108056777 Q * rs Quit: leaving 1108056836 M * Bertl yes, atm, the quad PIII lycos provides is the testing platform .. I'm missing some non-x86 test machines but hey, they will come and test sooner or later ;) 1108056878 M * JonB single or dual cpu ? 1108056928 M * Bertl hmm? the non-x86 ones? well, whatever available (and somewhat maintained ;) 1108056951 M * JonB maintained? 1108057006 M * Bertl well, either somebody on site, who can isntall/reboot/etc the machine or via serial console (or similar) 1108057026 M * JonB ahh 1108057215 M * Vantage Bertl: ok, so i'm moving along (slowly). i've got 2.4.29 with device mapper going and created an lvm partition for vservers and mounted it -o tagctx. However, i'm not quite following how the next bit works. can you explain the activate quota hash for context N =100 for me? 1108057408 Q * JonB Quit: Leaving 1108057415 M * Vantage is 100 just an arbitrary unique number assigned to the vserver? 1108057449 M * Bertl you have to use static context ids for your vserver 1108057463 M * Vantage and then tied to the instance through S_CONTEXT in the config file? 1108057472 M * Bertl (i.e. S_CONTEXT=100 for example) 1108057496 M * Bertl changing the xid after tagxid is enabled will give strange results ... 1108057498 M * Vantage ok, so there's no significance to the 100, just that it's the same as assigned to S_CONTEXT in the config. gotcha 1108057513 M * Bertl (i.e. files belonging to different contexts and such) 1108057535 M * Bertl * 1108057536 M * Bertl o cqhadd -x 100 -v /dev/real/block/device 1108057563 M * Bertl this should add the quota hash for xid=100 and the device specified on the cmd line 1108057588 M * Vantage now you can have multiple xid's on the same device, right? 1108057590 M * Bertl once that is done, you can use the cqdlim to control the disk limits 1108057605 M * Bertl Vantage: yes, but you have to add a quota hash for each xid ... 1108057621 M * Bertl so if you ahve 3 vserver, with xid=100, 101, and 102 1108057624 M * Vantage so run cqhadd again with a different xid 1108057638 M * Vantage ok, i think i'm following now 1108057639 M * Bertl you have to add 3 quota hashed with -x 100, 101, and 102 respectively 1108057760 M * Vantage can s_context be set/modified after the vserver has been created? 1108057787 M * Bertl yes, but you have to 'move' the vserver's files into the new context (when tagxid is turned on) 1108057826 M * Bertl (otherwise they will be treated as belonging to a different context= 1108057834 M * Bertl s/=/)/ 1108057898 M * Vantage hopefully the debian newvserver tool increments s_context reasonably, since you can't specify it on the command line... 1108057952 M * Bertl no, idea, check the configs ;) 1108057982 M * Bertl (but it would surprise me if it would use the staic contexts at all ;) 1108058033 M * Vantage well at the start of creating the vserver it prints s_context=0 not sure what that would mean... 1108058070 M * Bertl check what your /etc/vservers/.conf file contains (or wherever that is on debian) 1108058131 M * Vantage yeah, it'll be in that dir. unfortunately i have to wait for it to recreate (removed my old vserver to put it on the lvm). Could you just add S_CONTEXT=0 to the config if it doesn't? 1108058161 M * Bertl well, S_CONTEXT=0 would be really bad ... you should use S_CONTEXT=100 or so ;) 1108058178 M * Vantage ok, so as you add would going 100, 101, 102, 103 be fine? 1108058186 N * BobR_oO BobR 1108058187 M * Bertl yep, right ... 1108058219 M * Vantage ok. thanks for all your help. i'll probably pop back once this vserver is finished installing (it'll take a bit, it's only on a PII 266..) 1108058240 M * Bertl you're welcome! have fun! 1108060605 M * Bertl Doener: you around? 1108060609 M * Doener yup 1108060628 M * Bertl I added two tiny fixes (agains -rc5) 1108060645 M * Doener usual location? 1108060650 M * Bertl yep 1108060657 M * Bertl delta-alpha-fix01.diff, delta-capcontext-fix01.diff 1108060745 M * Doener well, what should i say? *g* 1108060766 M * Bertl ;) 1108060835 Q * prae Quit: Client exiting 1108060836 M * Bertl guess the next thing we need is some kind of automated testing ... 1108060866 M * Bertl where you can add a test script and have them run on an arbitrary kernel/tool combination 1108060940 J * lars||| ~ledj@213-156-52-105.fastres.net 1108061002 M * lars||| I was in here some days ago asking about 64 bit Xeon and vserver... 1108061014 M * lars||| and what problems I might run into. 1108061016 M * Bertl welcome back lars|||! 1108061018 M * no_maam well 1108061023 M * no_maam this is an amd64 system, or? 1108061028 M * lars||| Bertl: thanks 1108061033 M * lars||| :) 1108061046 M * no_maam well, emt64 exactly, but this is the same kernel as on an amd64? 1108061083 M * lars||| yes, it is 1108061115 M * no_maam ok 1108061131 M * no_maam I got a real amd64 cpu here, running a 64 bit linux kernel and a 64 bit userland 1108061138 M * no_maam works fine 1108061140 M * Bertl you might encounter issues if you want to run mixed 32/64 bit user/kernel interface (regarding linux vserver) 1108061148 M * lars||| I've read that you could change a few lines, like syscall etc. 1108061171 M * no_maam at least with a 64 bit userland on the host it works fine 1108061177 M * Bertl linux-vserver has no 32bit emulation when compiled as 64bit kernel 1108061182 M * no_maam I got a 32 bit guest, which works fine most times 1108061196 M * no_maam some applications try to figure out on which system they are going to be compiled 1108061207 M * no_maam and then try to compile 64 bit binarys 1108061214 M * lars||| ok, so I won't be using Debian for this then... should I go out buy a Suse then? 1108061217 M * Bertl you can change all the utsname properties ... 1108061218 M * no_maam but I only got a 32 bit guest installed and only 32 bit compilers 1108061231 M * Loki|muh Bertl: you mean, a 64bit host wont accept a 32bit guest? for me, that works... 1108061252 M * Bertl no, I mean 64bit kernel won't work with 32bit userspace tools ... 1108061263 M * Loki|muh ah 1108061283 M * Bertl (you have to compile them 64bit for now) 1108061340 M * Bertl lars|||: what is the issue? 1108061348 M * Loki|muh Bertl: another ot question... how much costs a Liter 'Benzin' in Austria? 1108061396 M * lars||| Bertl: none yet, but I'm not sure if just installing Debian Sarge and patch the kernel will work, when using 64 bit Xeon 1108061453 M * lars||| BTW: it's Dual Xeon 64 bit, if that means anything... 1108061521 M * Bertl Loki|muh: too much! 'Diesel' currently at 0.86 and 'Super' around 1.10 1108061532 M * Bertl lars|||: sounds good ;) 1108061567 M * Loki|muh here it is 6 cents more... :/ 1108062538 M * lars||| Is there any free linux dist that support 64 bit properly+ 1108062640 M * Loki|muh ubuntu 1108062646 M * Loki|muh gentoo 1108062661 M * Loki|muh and another debian-port 1108062678 M * Loki|muh i think suse, too 1108062716 M * lars||| but suse costs money 1108062757 M * Doener hm? should be available at least as a netinstall thingy... 1108062873 M * lars||| ok, maybe I'm reading something wrong on their site :) 1108063029 J * mhepp ~mhepp@r72s22p13.home.nbox.cz 1108063035 M * Doener ftp://ftp.suse.com/pub/suse/i386/current/README.FTP 1108063049 M * Bertl welcome mhepp! 1108063329 J * hacklava ~marc@65.92.141.62 1108063361 M * hacklava Youppi! There's a lot of VServer friends here... ;-) 1108063376 M * hacklava I'd like to know if tcpd works on a vserver... 1108063591 M * Bertl welcome hacklava! 1108063615 M * Bertl hmm, what does tcpd do on a normal server? 1108063644 M * halorgium isn't tcpd just a wrapper for a server app? 1108063752 M * Bertl yes, looks like a tcp_wrappers thingy tho make tcp_wrapper unaware tools use them ... 1108063784 M * Bertl i.e. access control for non-tcp-wrapper based daemons 1108063797 M * halorgium yea 1108063801 M * halorgium like xinetd? 1108063913 M * Bertl hacklava: so I'd say, yes it should work fine ... 1108063967 M * lars||| how does suses UML (User-mode linux) compare to vserver? 1108063972 M * halorgium kernel-level access is restricted ... right? most apps/daemons will still run .... also correct? 1108064027 M * Bertl lars|||: UML has a separate kernel running for every 'guest' so it will add the overhead of that (resource wise) and the overhead added by double buffering stuff like files (or filesystems) 1108064051 M * hacklava I need to control access to iXvnc server instances. Without tcpd, So if a daemon doesn't start, it is because it is not compatible with tcpd? 1108064069 M * hacklava (Xvnc, not iXvnc) 1108064094 M * lars||| Bertl: ok, but it's still possible to use vserver on suse, right? 1108064115 M * Bertl lars|||: sure, if you do _not_ use the suse kernel, why not? 1108064131 M * Bertl hacklava: hum, please rephrase ... 1108064192 M * hacklava Oops, sorry... Xvnc works without tcpd. Does Xvnc need to be compiled with libwrap? 1108064247 M * Bertl no idea, depends on what you want it to do?! ... IMHO not vserver related ... 1108064281 M * Bertl you can use sshd with or without tcp-wrapper support inside a vserver ;) 1108064306 M * hacklava Berl: that's probably not a vserver issue then. I'll recompile Xvnc with libwrap support. Thanks! 1108064465 M * Bertl my pleasure! 1108064789 Q * hansi33 Quit: ChatZilla 0.9.61 [Mozilla rv:1.7.5/20041226] 1108064885 M * lars||| Ok later people 1108064892 Q * lars||| Remote host closed the connection 1108064909 M * Bertl Doener: what do you think, shall we release 1.9.4? 1108064967 M * halorgium Bertl: is that for 2.6.10? 1108064969 N * BobR BobR_oO 1108065000 M * Doener hm, i just remember tcpdiag_exact_match (or something like that...) did we care about that one? IIRC i didn't check that one... 1108065030 M * Bertl halorgium: 2.6.9/2.6.10 and 2.6.11-rc3 probably 1108065364 M * Doener hm, ss doesn't seem to allow such queries... or i'm too tired... anyway if that one is fixed, i'd go for a 1.9.4 release 1108065372 M * Doener off to bed now, cya 1108065390 M * Bertl night Doener! sweet dreams! 1108065396 M * Doener thanks 1108065404 N * Doener Doener|zZz 1108066277 J * prae ~prae@sherpadown.net 1108066333 M * Bertl evening prae! 1108066350 M * prae Mr Bertl :) 1108066445 M * Bertl prae: maybe you know, is there an amd64 64bit version of mandrake out there? 1108066545 J * Tbery ~tb@rt-pha-1.karneval.cz 1108066547 M * prae hmmm, available in ftp server on the net ? 1108066547 M * Tbery HI 1108066549 M * Tbery mans.. 1108066556 M * Bertl evening Tbery! 1108066564 M * Tbery to you.. 1108066581 M * Tbery tomorowe...I will install new serevr 1108066585 M * Tbery servers.. 1108066606 M * Tbery 2*3ghg 6g DDR 1.4T hdd 1108066617 M * Tbery on virtuals.. 1108066633 M * Tbery I hope that all will working well.. 1108066658 M * Bertl I hope that for you too ;) 1108066696 M * Tbery I can make there mirro.. 1108066699 M * Tbery mirror.. 1108066714 M * Tbery it is on 100M internet.. 1108066738 M * Bertl you mean for the wiki or so? 1108066744 M * Tbery do you have some examples with writing web linux programs.. 1108066756 M * Tbery no spcility for vservers,, 1108066777 M * Bertl ah, server (guest) images?! 1108066823 M * Tbery yes.. 1108066834 M * Bertl sounds good! 1108066839 M * Tbery it is on prague unoversit.. 1108066844 M * Tbery univerzit 1108066855 M * Bertl university ;) 1108066865 M * Tbery technical university 1108066869 M * Tbery yes 1108066878 M * Tbery did zou were in prague?? 1108066899 M * Bertl not yet ... 1108066925 M * Tbery you are from wien?/ 1108066957 M * Bertl yes Vienna (or to be precise in the vicinity now) 1108066980 M * Tbery I have frinds there.. 1108066985 M * Tbery friends.. 1108066999 J * hansi33 ~hansi33@193.80.53.173 1108067013 M * Bertl yes? at the university? or just in vienna? 1108067020 M * Bertl welcome hansi33! 1108067045 M * Tbery he is studing in berlin.. 1108067057 M * Tbery shotings films.. 1108067068 M * Tbery shoting films.. 1108067088 M * Tbery You have after school?? 1108067141 M * Bertl hmm, please rephrase ... 1108067183 M * Tbery Did you finished your studing? 1108067222 M * Bertl ah, well, I'm still working on my doctorial thesis ... 1108067235 M * Tbery woow.. 1108067243 M * Tbery me frinds too. 1108067248 M * Loki|muh Bertl: it is about vserver? ;) 1108067289 M * Tbery I think it is.. 1108067296 M * Bertl no, although it would be already done if it was ;) 1108067458 M * Tbery did you do some cgi with bash?? 1108067467 M * Tbery and forms?? 1108067503 M * Bertl better use php or perl for such things ... 1108067524 M * Tbery ok.. 1108067542 M * Tbery I borrow some book. 1108067558 M * Tbery do you know interrail?? 1108067588 M * Vantage Bertl: any idea why my vserver's when created now have 0 for owner and group and can't access anything? 1108067644 M * Bertl with tagxid? check with lsxid (on the files and dirs) 1108067693 M * Vantage from within the vserver or outside of it? 1108067703 M * Bertl on the host ... (i.e. outside) 1108067741 M * Vantage hrmm it occurs to me that I do not have this tool. what util package is it a part of? 1108067781 M * Bertl util-vserver (the alpha version) but there is a separate tool you can use instead ... 1108067791 M * Bertl (sec looking for the url) 1108067795 M * Tbery who make debian packages?? 1108067814 M * Bertl everybody makes debian packages ;) 1108067847 M * Tbery I had some trobles.. 1108067851 M * Tbery with them.. 1108067862 M * Bertl .. almost everybody has ... 1108067865 M * Tbery I use this.. 1108067952 M * Bertl Vantage: http://vserver.13thfloor.at/Experimental/TOOLS/vxid-0.02.tar.bz2 1108067971 M * Tbery http://www.zs-ln.cz/quido/kernel/util-vserver_0.30.196-1_i386.deb 1108068424 M * Bertl no ide who provides thos packages ... isn't there some readme/packager info? 1108068432 A * Bertl is no debian guy ... 1108068811 M * no_maam there is a info system 1108068832 M * no_maam http://packages.debian.org/unstable/net/util-vserver 1108069148 M * Vantage Bertl: how do you use vxid? i tried vxid -x 100 and got nothing back 1108069148 M * hacklava Anybody familiar with tcpd? Here's the syslog entry: libwrap refused connection to vnc-800x600x16 (libwrap=tcpd) from 1108069201 M * Bertl Vantage: first you use it like this: 1108069229 M * Bertl vxid /vservers/XXXX 1108069233 M * Bertl xid= 0, flags=0x00000000, mask=0x00000000, /vservers/XXXX 1108069245 M * Bertl to check (query) for xids 1108069268 M * Bertl might require -i (for legacy ioctls) 1108069354 M * Bertl if you do: 1108069363 M * Bertl vxid -x 100 /vservers/XXXX (or with -i) 1108069374 M * Bertl then the directory will be changed to context 100 1108069400 M * Vantage wow, i have no idea what i'm looking at here ;) 1108069417 M * Bertl if you want, I can explain ... ;) 1108069430 M * Vantage definitely :) 1108069433 M * Vantage here's what I entered 1108069441 M * Vantage ./vxid -i /var/lib/vservers/vserver1/ xid=0 flags=0x00000000, mask=0x00000000, /var/lib/vservers/vserver1/ 1108069458 M * Vantage where /var/lib/vservers/vserver1 is the path to the vserver i've created 1108069476 M * Vantage does that look right so far? 1108069489 M * Bertl okay, basically that means that the vserver dir is in xid=0 (host id) which is fine 1108069511 M * Vantage this is what i get back (hope it doesn't cause flood issues) 1108069545 M * Vantage getxflg:: Inappropriate ioctl for device 1108069545 M * Vantage xid= 0, flags=0x00000000, /var/lib/vservers/vserver1/ 1108069545 M * Vantage stat:: No such file or directory 1108069545 M * Vantage open:: No such file or directory 1108069545 M * Vantage getxid:: Bad file descriptor 1108069546 M * Vantage getxflg:: Bad file descriptor 1108069546 M * Vantage xid=-1073742296, flags=0xbffffe28, xid=0 1108069548 M * Vantage stat:: No such file or directory 1108069548 M * Vantage open:: No such file or directory 1108069550 M * Vantage getxid:: Bad file descriptor 1108069550 M * Vantage getxflg:: Bad file descriptor 1108069552 M * Vantage xid=-1073742290, flags=0xbffffe2e, flags=0x00000000, 1108069552 M * Vantage stat:: No such file or directory 1108069554 M * Vantage open:: No such file or directory 1108069554 M * Vantage getxid:: Bad file descriptor 1108069556 M * Vantage getxflg:: Bad file descriptor 1108069556 M * Vantage xid=-1073742272, flags=0xbffffe40, mask=0x00000000, 1108069558 M * Vantage getxflg:: Inappropriate ioctl for device 1108069558 M * Vantage xid= 0, flags=0x00000000, /var/lib/vservers/vserver1 1108069598 M * Bertl hmm, well, I thought the first line was your input _and_ output ;) 1108069615 M * Bertl just do: 1108069634 M * Bertl vxid -i /var/lib/vservers/vserver1 1108069638 M * Bertl for a start ;) 1108069665 M * Vantage heh, that might help, eh? ;) ok here it is 1108069668 M * Bertl (my example was input _and_ output) 1108069674 M * Vantage getxflg:: Inappropriate ioctl for device 1108069674 M * Vantage xid= 0, flags=0x00000000, /var/lib/vservers/vserver1 1108069694 M * Bertl okay, this basically means that xid tagging isn't enabled on that device ... 1108069716 M * Bertl did you mount it with tagxid? did you enable xid tagging at compile time? 1108069739 M * Vantage /dev/mapper/vg1-vservers on /var/lib/vservers type ext3 (rw,tagctx) 1108069743 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com 1108069744 M * Vantage is how it's mounted 1108069750 M * Bertl welcome ntrs! 1108069791 M * Bertl Vantage: sec, checking something ... 1108069857 Q * hansi33 uranium.oftc.net keid.oftc.net 1108069857 Q * Tbery uranium.oftc.net keid.oftc.net 1108069857 Q * ndim uranium.oftc.net keid.oftc.net 1108069857 Q * Pinnen uranium.oftc.net keid.oftc.net 1108069857 Q * BobR_oO uranium.oftc.net keid.oftc.net 1108069857 Q * SiD3WiNDR uranium.oftc.net keid.oftc.net 1108069857 Q * click uranium.oftc.net keid.oftc.net 1108069857 Q * matti uranium.oftc.net keid.oftc.net 1108069857 Q * ntrs_ uranium.oftc.net keid.oftc.net 1108069857 Q * eyck uranium.oftc.net keid.oftc.net 1108069857 Q * mcp uranium.oftc.net keid.oftc.net 1108069857 Q * bro uranium.oftc.net keid.oftc.net 1108069929 J * ndim U2FsdGVkX1@helena.bawue.de 1108069937 J * matti matti@linux.gentoo.pl 1108069943 J * SiD3WiNDR luser@bastard-operator.from-hell.be 1108069945 J * BobR_oO ~georg@MAIL.13thfloor.at 1108069947 J * Pinnen ~pinnen@h194n2fls35o917.telia.com 1108069980 J * click click@dsl-84-161.aal.tiscali.no 1108070046 Q * mhepp Quit: mhepp caught signal: Autobus error 1108070113 J * Tbery ~tb@rt-pha-1.karneval.cz 1108070174 J * tanjix tanjix@pD9FAC673.dip.t-dialin.net 1108070178 M * tanjix hi together 1108070243 M * Bertl welcome tanjix! 1108070264 J * eyck eyck@81.219.64.71 1108070272 M * Bertl evening eyck! 1108070318 J * bro ~vanity@lanparty.lv 1108070386 M * Bertl wb bro! 1108070645 J * mcp ~hightower@www.c-tera.de 1108070660 Q * Tbery Quit: Ukonèuji 1108070734 J * hansi33 ~hansi33@193.80.53.173 1108070800 J * shuri ~shuri@dsl.speedline209.226.electronicbox.net 1108070874 M * Vantage Bertl: thoughts? 1108070901 M * Bertl I'm working on it (building 2.4.29-vs1.2.10-q0.14 just finished ;) 1108070939 M * Vantage oh cool 1108070992 M * Bertl usually I have the kernels lying around, but I didn't build that combo yet ... 1108071229 M * Vantage would there be something to mark off in make menuconfig that I might have missed? 1108071280 M * Bertl we'll see .. not sure the vxid tool works with 2.4 (didn't test it yet) 1108071308 M * Vantage ah... 1108071374 M * Bertl ok, booting now ;) 1108071464 M * Bertl okay, I get this now (with a simple test) 1108071471 M * Bertl mkdir /vservers/X 1108071486 M * Bertl chcontext --ctx 100 touch /vservers/X/a 1108071496 M * Bertl $ /tmp/vxid -i /vservers/X/a 1108071496 M * Bertl getxflg:: Inappropriate ioctl for device 1108071496 M * Bertl xid= 100, flags=0x00000000, /vservers/X/a 1108071508 M * Bertl so you can safely ignore the getxflg message 1108071540 Q * click uranium.oftc.net arion.oftc.net 1108071540 Q * Pinnen uranium.oftc.net arion.oftc.net 1108071540 Q * matti uranium.oftc.net arion.oftc.net 1108071540 Q * ndim uranium.oftc.net arion.oftc.net 1108071540 Q * ntrs uranium.oftc.net arion.oftc.net 1108071540 Q * hacklava uranium.oftc.net arion.oftc.net 1108071540 Q * flock uranium.oftc.net arion.oftc.net 1108071540 Q * Loki|muh uranium.oftc.net arion.oftc.net 1108071540 Q * gaber uranium.oftc.net arion.oftc.net 1108071540 Q * logger uranium.oftc.net arion.oftc.net 1108071540 Q * Doener|zZz uranium.oftc.net arion.oftc.net 1108071540 Q * meebey uranium.oftc.net arion.oftc.net 1108071540 Q * ats uranium.oftc.net arion.oftc.net 1108071540 Q * no_maam uranium.oftc.net arion.oftc.net 1108071540 Q * stupidawy uranium.oftc.net arion.oftc.net 1108071547 J * gaber gaber@linux.gentoo.pl 1108071551 J * no_maam ~erik@datenzone.de 1108071552 J * hacklava ~marc@HSE-Montreal-ppp336055.sympatico.ca 1108071552 J * click click@dsl-84-161.aal.tiscali.no 1108071553 J * Pinnen ~pinnen@h194n2fls35o917.telia.com 1108071553 J * meebey meebey@meebey.net 1108071553 J * logger ~rs@vds.pas-mal.com 1108071557 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com 1108071557 J * stupidawy foo@you.wish.you.were.pimp.olicio.us 1108071564 Q * logger oxygen.oftc.net jupiter.oftc.net 1108071564 Q * gaber oxygen.oftc.net jupiter.oftc.net 1108071564 Q * SiD3WiNDR oxygen.oftc.net jupiter.oftc.net 1108071564 Q * Snow-Man oxygen.oftc.net jupiter.oftc.net 1108071564 Q * Vantage oxygen.oftc.net jupiter.oftc.net 1108071564 Q * DaCa oxygen.oftc.net jupiter.oftc.net 1108071564 Q * switch oxygen.oftc.net jupiter.oftc.net 1108071564 Q * micah oxygen.oftc.net jupiter.oftc.net 1108071564 Q * Bertl oxygen.oftc.net jupiter.oftc.net 1108071564 Q * boklm oxygen.oftc.net jupiter.oftc.net 1108071567 J * Doener|zZz doener@193.24.208.125 1108071570 J * matti matti@linux.gentoo.pl 1108071586 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il 1108071586 J * logger ~rs@vds.pas-mal.com 1108071586 J * gaber gaber@linux.gentoo.pl 1108071586 J * SiD3WiNDR luser@bastard-operator.from-hell.be 1108071586 J * Snow-Man ~sfrost@snowman.net 1108071586 J * Vantage ~Vantage@www.toddcharron.com 1108071586 J * DaCa ~danny@mail.limehouse.org 1108071586 J * switch ~switch@61.206.115.5.user.ad.il24.net 1108071586 J * micah micah@micha.hampshire.edu 1108071586 J * Bertl ~herbert@janus.mc.tuwien.ac.at 1108071586 J * boklm boklm@woper.mars-attacks.org 1108071612 J * ndim U2FsdGVkX1@helena.bawue.de 1108071630 M * Vantage Bertl: of course I can't cd into any of the other dirs (permission denied) 1108071645 M * Vantage would running chcontext on it help any? 1108071796 M * Bertl do not make haste ... first let's analyze what's the actual reason ... 1108071835 J * monrad ~monrad@213083190130.sonofon.dk 1108071844 J * Loki|muh loki@satanix.de 1108071852 J * ats ~as@lotes.vtu.lt 1108072000 M * Bertl wb folks! 1108072023 M * Bertl Vantage: okay, let's check the files inside your vserver in a more efficient way ... 1108072113 M * Bertl find /var/lib/vservers/vserver1 | xargs vxid -i 2>/dev/null 1108072137 M * Bertl that should give you a list of all files and the appropriate xid taggings ... 1108072326 M * Vantage nothin' 1108072423 M * Bertl ahem? 1108072453 M * Bertl try the find first ... it should list all your files 1108072460 M * Bertl (in this specific vserver) 1108072471 M * Vantage yup. plenty of files there 1108072527 M * Bertl okay, vxid -i works too? 1108072670 M * Vantage xid= 0, flags=0x00000000, /var/lib/vservers/vserver1/sbin/ipfwadm 1108072737 M * Bertl okay, and the xargs line doesn't work? 1108072775 M * Vantage ok, got it now (silly me didn't have vxid in my PATH) 1108072782 M * Bertl ;) 1108072801 M * Bertl now grep the output for xid values != 0 1108072808 M * Vantage i've also got some weird xid's ;) 1108072811 M * Vantage yup 1108072820 M * Bertl change those back to xid=0 1108072840 M * Bertl with the vxid -i -x 0 1108072840 M * Vantage xid=-1073744724, flags=0xbffff4ac, /var/lib/vservers/vserver1/proc/25063 1108072844 M * Vantage that can't be good ;) 1108072852 M * Bertl forget the procfs 1108072883 M * Bertl (thought you stopped the vserver first) 1108072891 M * Vantage i'll do that 1108072929 Q * shuri Quit: Leaving 1108072956 M * Vantage is there an easy way to change the xid all to 0? or one by one for that matter 1108072985 M * Bertl yes, with the chxid tool (which is also aprt of the alpha util-vserver) 1108073052 M * Vantage no other way, eh? ;) 1108073173 M * Bertl use the power of unix scripting ;) 1108073188 M * Bertl you can also make a complete copy of the vserver dir (on the host) 1108073202 M * Bertl and remove the original (all files will be moved to xid=0) 1108073226 M * Vantage ok, two questions then 1108073250 M * Vantage 1) the only way to change xid's is with chxid in the alpha utils, correct? 1108073264 M * Bertl nope, the vxid tool can do this too ... 1108073271 M * Vantage ah... 1108073277 M * Bertl 23:00 < Bertl> with the vxid -i -x 0 1108073307 M * Vantage ah... now that can work (with a bit of scripting) 1108073326 M * Vantage and you can change the xid to anything you want? (like say 100, 101, etc) 1108073417 M * Bertl yes, on the hsot, you can change it to any value between 0 and 65535 1108073549 M * Vantage hrmm should this work? 1108073562 M * Vantage find /var/lib/vservers/vserver1 | xargs vxid -i -x 0 2>/dev/null 1108074295 M * Bertl probably ... 1108074344 M * Vantage how important is the -i switch to vxid? 1108074382 M * Vantage this is weird 1108074391 M * Bertl the -i is the 'legacy ioctl' mode, without it, it works for 2.6 only 1108074402 M * Bertl 2.6/1.9.x that is 1108074414 M * Vantage if I use vxid to change xid=60412 to xid=0 on one file it changes 1108074415 M * Vantage however 1108074444 M * Vantage if I use 1108074444 M * Vantage find /var/lib/vservers/vserver1/ |xargs vxid -x 0 -i 2>/dev/null 1108074455 M * Vantage not only does it not change the others 1108074459 M * Vantage the one i fixed changes back! 1108074583 M * Vantage so i run 1108074583 M * Vantage vxid -i -x 0 /var/lib/vservers/vserver1/home/tcharron/.bashrc 1108074603 M * Vantage then 1108074603 M * Vantage find /var/lib/vservers/vserver1/ |xargs vxid -i 2>/dev/null 1108074610 M * Vantage and it's still xid 0 1108074615 M * Vantage followed by 1108074627 M * Vantage find /var/lib/vservers/vserver1/ |xargs vxid -x 0 -i 2>/dev/null 1108074632 M * Vantage and 1108074643 M * Vantage find /var/lib/vservers/vserver1/ |xargs vxid -i 2>/dev/null 1108074649 M * Vantage and it's no longer xid = 0 1108074672 M * Bertl what is it then? 1108074686 M * Vantage xid=60412 1108074704 M * Bertl do you have such an xid in your vserver-stat listing? 1108074764 M * Vantage vserver-stat has only a CTX of 0 1108074770 M * Vantage nothing matching that number 1108074783 M * Bertl okay, what xid tagging method did you select at kernel compile time? 1108074828 M * Vantage hrmm good question, what section would it be under? 1108074831 M * Bertl (you can check that with 'grep XID .config') 1108074866 M * Bertl you are using ext3, right? 1108074874 M * Vantage yeah, ext3 1108074906 M * Bertl okay, I switched my test setup to ext3 to reproduce ... 1108074909 M * Vantage funny, i've got only CONFIG_INOXID_GID16=y coming up with that grep 1108074938 M * Bertl hmm, okay, so this deviates from the default too ... 1108074955 M * Bertl anyway ... 1108075025 M * Bertl okay, let's do some tests in parallel (i.e. here and on your system) 1108075029 M * Vantage so you've got the same result on grepping your .config? 1108075030 M * Vantage k 1108075061 M * Bertl first let's make a /vservers dir and --bind mount you debian dir there, you know how to do that? 1108075114 M * Vantage just add the --bind switch to mount? 1108075129 M * Bertl mount --bind /var/lib/vservers /vservers 1108075197 M * Vantage done 1108075208 M * Bertl mkdir /vservers/X 1108075219 M * Vantage done 1108075219 M * Bertl (if it already exists, just remove it first) 1108075228 M * Bertl chcontext --ctx 100 touch /vservers/X/a 1108075240 J * duckx ~Duck@dyn-83-157-204-211.ppp.tiscali.fr 1108075243 M * Vantage done 1108075243 J * DuckMaster ~Duck@dyn-83-157-204-211.ppp.tiscali.fr 1108075257 M * Bertl chcontext --ctx 200 touch /vservers/X/b (this is supposed to fail with permission denied) 1108075325 M * Vantage yup 1108075328 M * Bertl vxid -i /vservers/X/a 1108075338 M * Vantage xid= 100, flags=0x00000064, /vservers/X/a 1108075353 M * Bertl okay, now 1108075356 M * Bertl vxid -i /vservers/X 1108075370 M * Vantage xid= 100, flags=0x00000064, /vservers/X 1108075391 M * Bertl vxid -i -x 200 /vservers/X 1108075402 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075412 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075469 M * Bertl vxid -i /vservers/X{,/a} 1108075487 M * Vantage xid= 100, flags=0x00000064, /vservers/X/a 1108075493 M * Vantage xid= 100, flags=0x00000064, /vservers/X 1108075499 M * Bertl hum? 1108075507 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075515 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075521 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075526 M * Vantage xid= 100, flags=0x00000064, /vservers/X/a 1108075550 M * Bertl okay, I guess the last two lines are the results there, right? 1108075583 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075587 M * Vantage xid= 100, flags=0x00000064, /vservers/X/a 1108075590 M * Bertl (at least I won't expect more than two lines ;) 1108075590 M * Vantage there 1108075592 M * Vantage those two 1108075602 M * Bertl okay, great! 1108075610 M * Bertl chcontext --ctx 200 touch /vservers/X/b 1108075616 M * Bertl (this is now supposed to work) 1108075622 M * Vantage yup 1108075657 M * Bertl vxid -i /vservers/X{,/a,/b} 1108075733 M * Vantage xid= 200, flags=0x000000c8, /vservers/X 1108075738 M * Vantage xid= 100, flags=0x00000064, /vservers/X/a 1108075743 M * Vantage xid= 200, flags=0x000000c8, /vservers/X/b 1108075775 M * Bertl okay, now let's unmount both mounts, first the --bind one then the real mount at /var/lib/vservers 1108075801 M * Bertl then remount both (the first with tagctx/tagxid as before ... 1108075812 M * Bertl s/remount/mount them again 1108075859 M * Vantage you mean mount -o tagctx -t ext3 /dev/device /var/lib/vservers ? 1108075900 M * Bertl yes 1108075903 M * Vantage ack. unfortunately I have to head out for a little bit. I'll be back by later though. thanks again for all your help 1108075914 M * Bertl hmm, okay ;) 1108075937 M * Vantage sorry, i'll be back later tonight though (and probably tommorrow). bye 1108075953 M * Bertl cya 1108077993 Q * DuckMaster Quit: Client exiting 1108078790 Q * tanjix Quit: 1108079228 Q * hacklava Remote host closed the connection 1108079595 Q * duckx Ping timeout: 480 seconds