1235953783 Q * balbir_ Ping timeout: 480 seconds 1235953795 Q * balbirsingh Ping timeout: 480 seconds 1235954465 J * balbir_ ~balbir@122.172.58.101 1235954508 J * balbirsingh ~balbir@122.172.58.101 1235955219 N * sid4windr sid3windr 1235957861 Q * balbir_ Ping timeout: 480 seconds 1235957871 Q * balbirsingh Ping timeout: 480 seconds 1235958517 J * balbirsingh ~balbir@122.172.54.78 1235958539 J * balbir_ ~balbir@122.172.54.78 1235960332 Q * puck Ping timeout: 480 seconds 1235960737 Q * balbirsingh Ping timeout: 480 seconds 1235960742 Q * balbir_ Ping timeout: 480 seconds 1235961396 J * balbirsingh ~balbir@122.172.62.141 1235961419 J * balbir_ ~balbir@122.172.62.141 1235962851 J * aj__ ~aj@e180200066.adsl.alicedsl.de 1235963281 Q * derjohn_foo Ping timeout: 480 seconds 1235967892 Q * balbirsingh Read error: Connection reset by peer 1235967892 Q * balbir_ Read error: Connection reset by peer 1235968773 J * balbirsingh ~balbir@122.172.106.193 1235968787 J * balbir_ ~balbir@122.172.106.193 1235968910 Q * Piet Quit: Piet 1235970077 Q * balbir_ Ping timeout: 480 seconds 1235970082 Q * balbirsingh Ping timeout: 480 seconds 1235970700 J * balbir_ ~balbir@122.172.151.241 1235970750 J * balbirsingh ~balbir@122.172.151.241 1235971989 N * Bertl_zZ Bertl 1235971993 M * Bertl morning folks! 1235974788 Q * balbir_ Ping timeout: 480 seconds 1235974803 Q * balbirsingh Ping timeout: 480 seconds 1235975440 J * balbir_ ~balbir@122.172.51.179 1235975452 J * balbirsingh ~balbir@122.172.51.179 1235975727 Q * balbir_ Read error: Connection reset by peer 1235975751 Q * balbirsingh Read error: Connection reset by peer 1235976711 N * Bertl Bertl_oO 1235976716 M * Bertl_oO off for now .. bbl 1235979879 M * awk hi 1235980323 J * cluk ~cluk@p5B17F82D.dip.t-dialin.net 1235980844 J * davidkarban ~david@88.86.104.103 1235981859 J * esa ~esa@ip-87-238-2-45.static.adsl.cheapnet.it 1235982691 J * cga ~weechat@62.196.2.6 1235982921 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1235983227 J * doener ~doener@i577B95F1.versanet.de 1235983332 Q * doener_ Ping timeout: 480 seconds 1235983726 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1235985420 Q * aj__ Ping timeout: 480 seconds 1235985942 J * kir ~kir@swsoft-msk-nat.sw.ru 1235987644 J * BenG ~bengreen@92-238-216-179.cable.ubr20.aztw.blueyonder.co.uk 1235988914 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1235988971 Q * ensc Ping timeout: 480 seconds 1235989201 J * floyd_n_milan ~mrugesh@203.129.237.147 1235989226 M * floyd_n_milan hi 1235989413 M * floyd_n_milan wokay, solved the issue :) 1235989421 M * floyd_n_milan wanted to increase tmpfs size 1235989433 M * floyd_n_milan did it in /etc/vservers//fstab 1235989435 M * floyd_n_milan thanks 1235989437 Q * floyd_n_milan 1235989522 J * ensc ~irc-ensc@p57AA7D28.dip.t-dialin.net 1235989844 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net 1235992114 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it 1235994074 J * saulus ~saulus@d003237.adsl.hansenet.de 1235994099 N * ensc Guest203 1235994109 J * ensc ~irc-ensc@p57AA4DAE.dip.t-dialin.net 1235994183 Q * saulus_ Ping timeout: 480 seconds 1235994217 Q * Guest203 Ping timeout: 480 seconds 1235994364 A * h01ger waves good-bye for now. #517828 has been filed against the debian release notes to document the tagxid -> tag change. 1235994380 M * h01ger micah, ^^ - maybe you can come up with a good wording? i cant :-/ 1235994384 M * h01ger cu! 1235994395 P * h01ger & thanks for all the fish! 1235994900 J * balbirsingh ~balbir@122.172.59.249 1235994940 Q * mrfree Quit: Leaving 1235994950 J * balbir_ ~balbir@122.172.59.249 1235994983 M * Bertl_oO balbirsingh: please fix your client, two connects do not make much sense, and the permanent reconnect is annoying 1235995553 Q * BenG Quit: I Leave 1235995664 M * balbirsingh Bertl_oO: OK.. Thanks.. should get my isp to fix it 1235995698 Q * vamoquse Ping timeout: 480 seconds 1235995745 Q * balbirsingh Quit: Ex-Chat 1235995767 J * _Myself_ ~Myself@137.226.248.251 1235996515 M * _Shiva_ how dependent is util-vserver from Kernel? i.e. I have a running 32Bit vserver-host, that has some 32bit vserver guests. Now I want to upgrade the server board to a x86_64 capable one. To minimize the downtime of the services, I simply want to change the board/cpu and boot the old 32bit userland with a x86_64 Kernel - doing "ia32-emulation" all the time on the currently installed system 1235996592 M * _Shiva_ and then "migrate" the current system service by service into a x86_64 vserver .. 1235996597 M * Bertl_oO is supposed to work fine 1235996665 M * derjohn I use it that way (x86_64 host, x86 guest) for years now. 1235996666 M * _Shiva_ so util-vserver might be a 32Bit app that starts x86_64 guests on a x86_64 kernel? 1235996685 M * Bertl_oO to some extend 1235996686 J * vamoquse ~vamoquse@193.43.249.169 1235996694 M * derjohn _Shiva_, I would opt to use an "amd64" distribution 1235996695 M * Bertl_oO but it will definitely work for 32bit guests 1235996758 M * _Shiva_ derjohn: me too - but the current (old an running) setup is a 32bit one :-) 1235996809 M * _Shiva_ so it would be a 64Bit-Kernel starting a "pure32" userland that has "pure64" guests... 1235996831 M * Bertl_oO not the best idea, but as I said, will work to some degree 1235996846 M * _Shiva_ simply for migration purpose that is :-) 1235996862 M * Bertl_oO the better approach is to switch the mobo, use the 64bit kernel with 32bit userspace and guests (you don't have 64bit guests yet :) 1235996880 M * yang hi Bertl_oO, which IPv6 approach for guests would you recommend to me ? 1235996891 M * Bertl_oO _Shiva_: then, in the next step, switch host userspace to 64bit 1235996904 M * Bertl_oO _Shiva_: then create new 64bit guests 1235996921 M * Bertl_oO yang: hmm? 1235996969 M * yang Bertl_oO: i found this setup guide http://linux-vserver.org/IPv6 1235996985 M * yang Bertl_oO: do I need to restart guest to apply a new IPv6 address to it ? 1235997017 M * Bertl_oO nope 1235997061 M * yang so, how does it get loaded? 1235997101 M * Bertl_oO you configure it on the host, and add it to the guest with naddress 1235997174 M * _Shiva_ Bertl_oO: right - that would have been my roadmap - but I thought about mounting the new rootfs (which would be a new sw-raid1-volume) of the host as a vserver-chroot, then set up a gentoo stage3 to stage4 there, emerge'ing the services, stopping services in the 32bit host, starting them in the 64bit guest, and finally booting the server unsing the new rootfs which then is 64bit.. 1235997188 M * _Shiva_ minimizes downtime :-) 1235997207 M * yang Bertl_oO: so, I just write the IPv6 address under ipv4 address? 1235997244 M * Bertl_oO hmm? 1235997262 M * yang actually I have /interfaces/0/ directory 1235997272 M * yang do I create /interfaces/1 1235997278 M * Bertl_oO okay, that is only consulted on guest startup 1235997303 M * Bertl_oO if you want to configure it there (and restart the guest) you need to create a new dir in interfaces 1235997347 M * yang well according to that manual, its this how its explained...but I would not like to restart the guest 1235997354 M * _Shiva_ yang: just add another /interfaces/1/.. with "ip" holding your v6-address, "prefix" your v6-prefix and "dev" as your interface and restart the guest 1235997364 M * yang ok 1235997381 M * Bertl_oO yang: that's why I pointed you to the naddress 1235997403 M * Bertl_oO (which allows you to add the IP at runtime) 1235997460 M * yang ok i have done what shiva wrote 1235997500 M * yang bash: naddress: command not found 1235997501 M * _Shiva_ yang: Bertl_oO is right: http://linux-vserver.org/Frequently_Asked_Questions#How_do_I_assign_a_new_IP_address_to_a_running_guest.3F 1235997516 M * Bertl_oO how unexpected :) 1235997521 M * _Shiva_ Bertl_oO: :-P 1235997572 M * yang i guess its not in the debian version of LV 1235997593 M * yang then the only thing is to restart the guest, let me see if that works 1235997606 M * Bertl_oO depends on the debian version, you want 0.30.215 or later anyways 1235997646 M * yang i got .216 1235997661 M * yang or maybe i require a special packages called naddress 1235997675 M * Bertl_oO nope, is part of 0.30.216, definitely 1235997737 M * Bertl_oO rpm -qf `which naddress` 1235997737 M * Bertl_oO util-vserver-core-0.30.216-0.pre2824 1235997779 M * Bertl_oO yang: but if debian is missing that, and maybe other tools, please inform the maintainer(s) 1235997856 M * _Shiva_ Bertl_oO: do i *have to* set /etc/vservers//uts/machine and /etc/vservers//personality before switching from a 32bit kernel to a 64bit one? 1235997857 M * yang ok a restart works 1235997902 M * Bertl_oO _Shiva_: the personality stuff is unlikely to work, it is not intended to do 64bit switches on 32bit, it is doing the other way round 1235997926 M * Bertl_oO _Shiva_: but starting a 64bit binary should get you a 64bit personality anyways 1235997954 M * Bertl_oO for the 32bit guests, the personality should be set to 32bit 1235997955 M * _Shiva_ Bertl_oO: actualy I meant that for explicity setting 32_bit personality for the existing 32bit-guests.. ;-) 1235997989 M * _Shiva_ right now it's implicit - as the host is 32bit.. 1235997990 M * Bertl_oO yep, that is a good idea, although not strictly required (because of the beforementioned binary detection) 1235998051 M * _Shiva_ what's the difference between uts/machine an personality anyway? 1235998094 M * Bertl_oO the one is a value shown to userspace (can be arbitrary), the other is a set of flags/state, handled by the kernel 1235998121 M * Bertl_oO i.e. you can show sparc64 on a 32bit intel in uts, but you cannot assume such a personality 1235998147 M * _Shiva_ ok 1235998151 M * Bertl_oO so, I'm really off for now (that's what the _oO actually means) so bbl, have fun 1235998199 M * _Shiva_ . o 0 ( I thought "_oO" was "thinking about sth" and "_off" meant "away" ) ;-) 1235998786 Q * nou Ping timeout: 480 seconds 1235999283 A * blathijs thought that setting an away reason with /away meant away, without spamming any channels with nick changes :-) 1235999756 J * Piet ~piet@asteria.debian.or.at 1235999761 J * nou Chaton@causse.larzac.fr.eu.org 1236002335 J * huji ~huji@jem75-8-88-170-103-188.fbx.proxad.net 1236002342 M * huji hello ! 1236002479 M * huji I have a problem with vserver. 1236002486 J * jrdnyquist ~jrdnyquis@slayer.caro.net 1236002518 M * huji I have a custom kernel 2.6.28-7, patched for vserver. I do a vserver build with this command : 1236002527 M * huji vserver SRV_web build -m debootstrap --context 2 --hostname thana --interface eth0:192.168.20.234/24 -- -d lenny -m http://ftp.fr.debian.org/debian 1236002539 M * huji it works well, but when I do a vserver SRV_web start 1236002544 M * huji I have an error : 1236002554 M * huji # vserver SRV_web start 1236002554 M * huji /etc/init.d/rc: line 74: /etc/default/rcS: No such file or directory 1236002554 M * huji error: '/etc/init.d/rc' exited outside the expected code flow. 1236002566 M * huji what should I do ? 1236003210 M * Bertl_oO most likely complain to debian, looks like this is happening inside the guest 1236003257 M * Bertl_oO i.e. /etc/init.d/rc (inside the guest is looking for /etc/default/rcS, which isn't there after the debootstrap) 1236004263 Q * balbir_ Read error: Connection reset by peer 1236004869 M * huji I know. I've tried to copy the host rcS but it does nothing 1236006550 J * BenG ~bengreen@92-238-216-179.cable.ubr20.aztw.blueyonder.co.uk 1236006882 M * ard well, is there a valid debian filesystem installed in /var/lib/vservers/SRV_web ? 1236006931 A * ard build -m debootstrap last night 1236006942 M * ard and it is now running asterisk 1236007030 M * huji yes, (I've change the vdirbase link, but it must still work...) 1236007058 M * ard Hmmmm 1236007060 M * huji I've installed some server with vserver kernel with no problem... But it was debian kernel. 1236007078 M * huji But I don't think that the problem come from my custom kernel 1236007098 M * ard util-vserver 0.30.216~r2772-6 ? 1236007132 M * huji util-vserver 0.30.216~r2772-6 1236007133 M * huji yes 1236007145 M * ard but changing vdirbase links is practically undoable 1236007156 A * ard always mounts /var/lib/vservers 1236007174 M * ard either with -o bind, or as a seperate file system 1236007192 M * huji just have to change the target of the symbolic link in /etc/vserver/.default/vdirbase 1236007199 M * huji it works on my others servers 1236007231 M * ard ok 1236007244 A * ard remembers having more work than that on an older installation 1236007287 M * ard but it's weird. 1236007301 A * ard has 2.6.28.7-vs2.3.0.36.7 1236007345 M * ard and that actually is probably the only thing where your and my installation differ :-) 1236007382 M * huji But if I copy my host /etc/default/rcS in the guest, I don't have errors anymore 1236007389 M * huji but the guest doesn't start. 1236007399 M * ard it does start 1236007407 M * ard but there are no processes left 1236007428 M * ard it sounds like you have util-vserver 0.30.215 or another mixup 1236007452 M * ard look in the rc directories and see if any S script exists 1236007467 M * ard at least rsyslogd must be started to keep the vserver alive 1236007486 M * ard or the debootstrap source is broken 1236007497 M * ard I've used http://ftp.nl.debian.org/debian 1236007548 M * ard the missing rcS sounds like a failed debootstrap actually 1236007572 M * ard the not having any processes left can als be a debootstrap failure, or a util-vserver that's too old 1236007596 M * ard (the older versions wanted to symlink to syslogd, but that has been replaced by rsyslogd) 1236007869 M * huji so I must get the latest util-vserver ? 1236008104 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fe60:79d4 1236008152 Q * kwowt Ping timeout: 480 seconds 1236009413 M * huji so, no clues ? 1236009602 M * daniel_hozac did you verify ard's hints already? 1236009866 M * huji there is a rcS in /etc/default, /etc/rcX.d exist (X=0...6+S), the debootstrap is debian version, I use debian main repository (ftp.debian.org), util-vserver is the debian sid version. 1236009892 M * huji and it seems that /etc/init.d/rcX.d are empty 1236009905 M * huji (just a README) 1236010338 Q * harobed Ping timeout: 480 seconds 1236010351 M * daniel_hozac which means that either your debootstrap is incomplete, or your util-vserver version is too old. 1236011047 J * dowdle ~dowdle@scott.coe.montana.edu 1236011090 M * huji ok, I will compile the latest version of util-vserver. 1236011794 M * ard something is foul, but a fresh lenny util-vserver should work 1236011825 A * ard tries not to use sid for hosts 1236011938 M * ard although I've discovered the weird thing that if I mknod some devices, then chxid the devices, chxid will say the device does not exist, and lsxid will show Err! (which usually means a nontagging fs?) 1236012008 M * ard the tagging works though :-) 1236012033 M * daniel_hozac "works"? 1236012071 M * ard well, the device nodes can be opened by asterisk, and when I just bind mounted /dev/zap asterisk had no rights 1236012118 M * ard But maybe that's a point that I should examine the device stuff in vserver :-) 1236012154 M * daniel_hozac that just means you forgot to mount it with -o dev. 1236012169 M * ard asterisk only needs CAP_NICE to go realtime, so all in all it's a nice thing to put in a vserver 1236012173 A * ard sighs 1236012194 M * daniel_hozac all guests can access things belonging to tag 0. 1236012218 M * ard but is -o dev a vfs layer thing? 1236012232 M * daniel_hozac yes. 1236012243 A * ard sighs even louder :-( 1236012249 M * ard daniel_hozac : tanx :-) 1236012274 A * ard has gained another few points linux knowledge 1236012285 Q * cga Quit: WeeChat 0.2.6 1236012288 M * daniel_hozac well, this is Linux-VServer knowledge, technically. 1236012304 M * daniel_hozac -o dev doesn't exist outside of secure-mount. 1236012310 M * ard ah, ok 1236012319 M * daniel_hozac since we default to -o nodev. 1236012335 M * ard then I do not feel that ashamed :-) 1236012368 M * ard so I can fix another server too :-) 1236012381 M * ard although this is for my private pbx 1236012404 M * ard the old one is about 10 years old 1236012444 M * ard And since asterisk is a moving target, putting it in a vserver allows me to have some testing grounds ;-) 1236012874 Q * BenG Quit: I Leave 1236014272 Q * davidkarban Quit: Ex-Chat 1236014316 J * geb ~geb@157.4.82-79.rev.gaoland.net 1236014388 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1236014514 J * davidkarban ~david@88.86.104.103 1236014595 Q * davidkarban 1236014679 Q * huji Quit: Leaving. 1236014787 Q * Pazzo Quit: Ex-Chat 1236015977 J * balbir_ ~balbir@122.172.111.176 1236016068 Q * esa Quit: Coyote finally caught me 1236016239 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236016577 Q * Slydder1 Remote host closed the connection 1236016605 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236016874 N * pmenier pmenier_off 1236016965 Q * kir Quit: Leaving. 1236017622 Q * Slydder1 Remote host closed the connection 1236017648 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236017927 Q * Slydder1 Remote host closed the connection 1236017952 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236018159 Q * Slydder1 Remote host closed the connection 1236018209 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236018404 Q * Slydder1 Remote host closed the connection 1236018430 J * Slydder1 ~chuck@dslb-088-072-213-089.pools.arcor-ip.net 1236018807 J * aj__ ~aj@e180200066.adsl.alicedsl.de 1236019568 M * Bertl_oO nap attack ... off to bed .. maybe back later ... 1236019574 N * Bertl_oO Bertl_zZ 1236019982 J * huji ~huji@bgl93-2-82-226-41-76.fbx.proxad.net 1236019991 M * huji hello again ! 1236020145 M * huji I still have a problem with vserver : when I do a vserver build, it works but folder /etc/rcX.d (X=0..6,S) are empty. If I do a debootstrap, my deboostraped debian is ok, rcX.d are not empty. If I copy rcX.d from my deboostraped debian to my vserver guest (build with debootstrap method) it doesn't work : I can start my guest with no error, but if I do a vserver enter I have : 'vserver ... suexec' is supported for running vservers only; abor 1236020178 M * huji I have the latest util-vserver, I use a custom kernel 2.6.28-7 vserver patched. 1236020244 M * daniel_hozac so you have 0.30.216-pre2827? 1236020321 M * huji 0.30.216~r2772 1236020334 M * huji it's not the very last 1236020337 M * huji but it still must work, 1236020441 M * daniel_hozac it "must"? 1236020524 M * huji there are a lot of difference between 0.30.216~r2772 and 0.30.216-pre2827 1236020618 Q * _Myself_ Ping timeout: 480 seconds 1236020876 J * cga ~weechat@94.36.110.25 1236020890 Q * micah Ping timeout: 480 seconds 1236020941 M * huji and there is compilation's errors.. 1236021019 Q * nou Ping timeout: 480 seconds 1236021033 M * huji src/exec-remount.c: In function ‘main’: 1236021033 M * huji src/exec-remount.c:110: error: ‘MNT_DETACH’ undeclared (first use in this function) 1236021168 M * daniel_hozac which means your headers are broken. 1236021362 M * huji well... sorry I'm idiot. 1236021449 M * huji but, I still have this error... 1236021588 M * huji could it be because I have more than 1 kernel installed ? 1236021771 M * huji enven with 1 kernel it's not working... 1236021830 M * daniel_hozac it's not the kernel's fault. it's something on your system not including the definitions properly. 1236021855 M * huji ok, I will rebuild kernel headers and come back after 1236021873 M * daniel_hozac i think you need your system headers, not kernel headers. 1236021873 M * huji but I think headers are correct. 1236021883 M * huji system headers ? 1236021886 M * daniel_hozac the kernel headers have that definition protected by __KERNEL__, unfortunately. 1236021997 M * huji what could be their names ? 1236022066 M * daniel_hozac either look at the dietlibc or glibc headers. 1236022271 M * huji I have installed that : http://packages.debian.org/sid/linux-libc-dev 1236022277 M * huji but still don't work 1236022319 M * huji I try with libc6-dev 1236022417 M * huji same error... 1236022521 M * huji will see later. 1236022523 M * huji thanks. 1236022561 Q * huji Remote host closed the connection 1236022618 Q * balbir_ Ping timeout: 480 seconds 1236023439 J * balbir_ ~balbir@122.172.59.4 1236024221 Q * gnuk Quit: NoFeature 1236024278 J * micah__ ~micah@micah.riseup.net 1236024855 J * nou Chaton@causse.larzac.fr.eu.org 1236025664 J * steffan steffan@67.202.106.29 1236026430 Q * Piet Quit: Piet 1236026442 J * ktwilight_ ~ktwilight@102.89-66-87.adsl-dyn.isp.belgacom.be 1236026707 Q * ktwilight__ Ping timeout: 480 seconds 1236027623 Q * doener Read error: Connection reset by peer 1236027784 J * doener ~doener@i577B95F1.versanet.de 1236027792 P * steffan 1236028085 Q * cluk Quit: Ex-Chat 1236028704 Q * cga Quit: WeeChat 0.2.6 1236030097 J * kwowt ~quote@pomoc.ircnet.com 1236030098 M * kwowt hi 1236030123 M * kwowt how can i make a copy of a guest vserver? 1236030124 M * kwowt :) 1236030136 M * kwowt with rsync or is cp ok? 1236030158 M * daniel_hozac cp -a is okay too. 1236030178 M * derjohn i wopuld recomnend "rsync -axHDPS" if networking is involved. 1236030212 M * kwowt rsync -axHDPS /old /new ? 1236030228 M * derjohn on Debian you need /var/lib/vservers/foo and /etc/vservers/foo 1236030265 M * derjohn if you stay local you could "cp -pr" I think. 1236030279 M * derjohn daniel_hozac, what id about sparse files and devices with cp ? 1236030289 M * kwowt &then change context, vdir and ip? 1236030323 M * derjohn kwowt, yes, usually that is enough. (unless you have disk limits or other advanced stuff) 1236030340 M * kwowt thx 1236030381 M * daniel_hozac derjohn: -a takes care of devices, and --sparse=auto/always of sparse files. 1236032058 Q * awk Ping timeout: 480 seconds 1236032732 Q * hparker Quit: Quit 1236032873 N * Bertl_zZ Bertl 1236032878 M * Bertl back now ... 1236033023 M * kwowt hmm 1236033034 M * kwowt how would i copy everything BUT one directory inside guest? 1236033039 M * daniel_hozac exclude it. 1236033074 M * Bertl man rsync 1236033227 J * hparker ~hparker@2001:470:1f0f:32c:290:96ff:fe50:40fa 1236033434 Q * bonbons Quit: Leaving 1236035239 N * ensc Guest300 1236035249 J * ensc ~irc-ensc@p57AA4DAE.dip.t-dialin.net 1236035345 Q * Guest300 Ping timeout: 480 seconds 1236036330 J * imcsk8 ~ichavero@nat.ti.uach.mx 1236036341 M * imcsk8 hello, i'm trying to limit the cpu resource of my vserver gests but i'm not shure of the appropiate values. is there a guide for mangaging the vserver resources?? 1236036381 M * Bertl yep, but it really depends on your kernel version and what you want to do 1236036439 M * Bertl http://linux-vserver.org/Resource_Limits 1236036447 M * Bertl http://linux-vserver.org/CPU_Scheduler 1236036576 M * imcsk8 i'll check that 1236036590 M * Bertl what kernel do you use? 1236036845 M * imcsk8 2.6.22-vs2.2.0.7 1236036863 M * Bertl then the cpu scheduler page is of interest to you 1236036900 M * Bertl (if you want to limit cpu resources on a fine grained level) and the cpusets (for a coarse assignment) 1236036993 J * pusling ~pusling@li61-184.members.linode.com 1236037002 M * pusling hi peoples. 1236037009 M * Bertl hey 1236037024 M * pusling what does the following two things (on console and in dmesg) mean: 1236037025 M * pusling [32613.592858] Not cloning cgroup for unused subsystem ns 1236037026 M * pusling [48600.960741] vxW: [»ps«,7775:#4|4|4] did lookup hidden e1468638[#0,3] »/dev/pts«. 1236037058 M * Bertl the first one is a mainline message, and it can be considered harmless (should be removed in the future) 1236037060 M * imcsk8 i want set a token bucket hard limit but in the CPU_Scheduler page does not appear which file to edit 1236037078 Q * _nono_ Ping timeout: 480 seconds 1236037112 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1236037180 M * Bertl pusling: the second one means, that your ps (inside the guest) did look up a file which belongs to a different context, most likely due to the fact that you entered the guest half way, and brought your terminal with you :) 1236037279 Q * Slydder1 Quit: Leaving. 1236037304 M * pusling Bertl: the latter seems to be triggered by running vps on the host 1236037339 M * Bertl that's at least interesting, as that should not enter a guest, only the spectator context 1236037381 M * Bertl can you reproduce that? if so, what kernel and util-vserver? 1236037420 M * pusling oh yes. as many times as you want ;) 1236037465 M * pusling kernel is "debian lenny", util-vserver is "0.30.216~r2772-6" as in debian level 1236037490 M * Bertl okay, any guess on the version numbers for that kernel? 1236037500 M * pusling the kernel is 2.6.26 1236037519 M * Bertl ah, well, then I'd suggest to try with 2.6.27 or 2.6.28 1236037573 M * pusling is the message harmless ? 1236037595 M * Bertl yes, it is a warning, and just means that the access was blocked 1236037605 M * pusling okay. 1236037631 M * pusling I kind of like running distribution kernels, then "other peoples" take care of the updates 1236037662 M * Bertl no problem with that, just keep in mind that the 2.6.26 release debian picked was a little broken 1236037681 M * pusling security broken, crashing broken or something third ? 1236037693 M * Bertl something third :) 1236037708 M * pusling let us see if I can live with it then ;) 1236037735 M * Bertl just be careful, e.g. with unification, for example 1236037748 M * imcsk8 Bertl, i have set this files on the sched directory, does this settings make sense? http://pastebin.com/d78580dc7 1236037754 M * Bertl pusling: otherwise it should work 1236037763 M * pusling almost nothing to gain with unification in my setup. 1236037818 M * Bertl imcsk8: it will work, but not really good choices 1236037836 M * Bertl see the wiki page about choosing the values for more details 1236037841 M * pusling Bertl: but thanks for your time 1236037847 M * Bertl you're welcome! 1236037886 M * pusling (and the stock debian vserver kernel runs nicely as a xen guest ) 1236037931 M * imcsk8 Bertl i was setting them based on the wiki page, i have 4 processors 1236037963 M * imcsk8 and want to restrict the vserver for just one processor at a time 1236037996 M * Bertl you mean, you want to assign a cpu to each guest (or to a single guest?) 1236038013 M * Bertl or you want all guests to use a single cpu 1236038058 M * imcsk8 i want to restrict just this one guest to use one cpu 1236038088 M * Bertl then cpusets are probably the easiest way to go here 1236038125 M * Bertl /etc/vservers//cpuset 1236038202 M * imcsk8 i'll check that. thanks! 1236038208 M * Bertl np 1236038224 Q * imcsk8 Quit: This computer has gone to sleep