1243641894 Q * dowdle Remote host closed the connection 1243647133 J * imcsk8 ~ichavero@evdomip-32-194.iusacell.net 1243649550 J * scientes ~scientes@174-21-88-43.tukw.qwest.net 1243649656 Q * imcsk8 Ping timeout: 480 seconds 1243650234 M * Bertl off to bed now ... have a good one eveyone! 1243650238 N * Bertl Bertl_zZ 1243650322 Q * scientes Remote host closed the connection 1243650352 J * scientes ~scientes@174-21-88-43.tukw.qwest.net 1243653550 Q * nou Ping timeout: 480 seconds 1243653781 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1243655028 Q * nenolod Ping timeout: 480 seconds 1243655062 J * nenolod nenolod@petrie.dereferenced.org 1243656518 Q * urbi Ping timeout: 480 seconds 1243657602 J * Net147 Net147@c211-30-18-226.rivrw2.nsw.optusnet.com.au 1243657787 M * Net147 harrydg: i've tested 2.3.0.36.14 and it seems to fix the issues I had with 2.3.0.36.12/13. feel free to create a grsec patch for it 1243665224 J * derjohn_mob ~aj@e180193173.adsl.alicedsl.de 1243666492 J * doener ~doener@i59F5782F.versanet.de 1243666594 Q * doener_ Ping timeout: 480 seconds 1243667686 M * Net147 hmm... when I run vserver-info I get "VS-API: ???" 1243668680 M * Net147 nevermind, ran command using wrong user 1243669733 J * imcsk8 ~ichavero@8.14.226.138 1243670286 Q * imcsk8 Quit: This computer has gone to sleep 1243670551 J * dna ~dna@p54BCC682.dip.t-dialin.net 1243671844 J * geb ~geb@AOrleans-253-1-52-135.w92-140.abo.wanadoo.fr 1243671875 J * nou Chaton@causse.larzac.fr.eu.org 1243672182 M * geb hi 1243672885 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1243674727 P * phedny 1243675602 J * derjohn_foo ~aj@e180195167.adsl.alicedsl.de 1243676033 Q * derjohn_mob Ping timeout: 480 seconds 1243676483 Q * derjohn_foo Ping timeout: 480 seconds 1243677060 N * Bertl_zZ Bertl 1243677065 M * Bertl morning folks! 1243677175 M * Net147 morning Bertl 1243677287 M * Bertl thanks for the testing and feedback on vs2.3.0.36.14! 1243677425 M * Net147 no problem. i'm using vserver now to build some 32-bit Arch packages on a 64-bit vserver host. 1243677515 M * Net147 strange thing is, when I start up vserver /dev/null has u=rw,g=r,o=r permissions 1243677536 M * Net147 have to chmod 666 /dev/null in rc startup script 1243677641 M * Bertl maybe some 'security' script which runs inside the guest? 1243677677 M * Bertl is /dev/null a device? maybe it was missing and ended up a file? 1243677679 M * Net147 it doesn't do that on a normal system 1243677721 M * Net147 ahh you're right it's a file for some reason... 1243677746 M * Bertl because it was missing, and the first 'bla >/dev/null' created it 1243677828 M * Net147 yea the bootstrap must have overwrote it and it became a file 1243677855 M * Bertl it might make sense to create a 'skeleton' guest, and copy the /dev over 1243678050 M * Net147 it's strange, when I start the guest it becomes a file again 1243678066 M * Bertl then something removes it on startup 1243678082 M * Bertl i.e. some scripts you want to clean up (inside the guest) 1243678343 J * pmenier ~pmenier@ACaen-152-1-68-98.w83-115.abo.wanadoo.fr 1243678428 M * Net147 it does this: /bin/cp -a /lib/udev/devices/null /dev/ 1243678454 M * Bertl so better keep it from doing that :) 1243678464 M * Net147 well it says /bin/cp: cannot create special file `/dev/null': Operation not permitted 1243678487 M * Bertl expected, unless you use the device mapper, no devices may be created inside a guest 1243678502 M * Net147 so why is it becoming a file... 1243678540 M * Bertl 12:02 < Bertl> because it was missing, and the first 'bla >/dev/null' created it 1243678563 M * Net147 yes but the reason it is missing 1243678580 M * Net147 is because after it errors out trying to create it by copying, the file doesn't exist anymore 1243678585 M * Bertl because the cp first removes the original, and then tries to create it 1243678602 M * Net147 hmm... 1243678727 M * Net147 does udev add any extra functionality to a vserver guest? 1243678751 M * Bertl as there are no changes in devices (for a guest) and udev is all about that, no 1243678767 M * Bertl on the contrary, udev will not run inside a guest without 'special' permissions 1243678819 M * Net147 maybe I should create a udev-vserver package that can be used as a substitute for the normal udev then 1243678839 M * Bertl why not just remove it or disable it? 1243678852 M * Net147 it's required by initscripts package 1243678869 M * Bertl dummy entry in package database? 1243678898 M * Net147 if I can figure out how to add a dummy entry 1243679065 J * simNIX ~simNIX@156-60.bbned.dsl.internl.net 1243679079 M * Net147 I think a dummy package is probably easier 1243679111 M * Net147 it's not that hard to make one 1243679131 M * Bertl yeah, but you end up with a bunch of them, for no reason 1243679171 M * Net147 if I have a udev dummy, it will get upgraded on next release of udev 1243679181 M * Net147 if I provide a udev replacement though, it won't get upgraded 1243679199 M * Bertl IMHO distros should slowly start to realize that they can be installed in a chroot/guest too and do not need the dependancies to hardware related stuff 1243679296 M * Net147 a package can replace multiple packages. so only need one package to handle the vserver-related stuff. 1243679313 M * Bertl okay 1243679346 M * Net147 maybe it can be called vserver-additions or vserver-stubs or something 1243679366 M * Bertl stubs sounds good 1243679374 M * Net147 yea 1243679379 M * Net147 would be nice if this was documented on the wiki 1243679406 M * Bertl you could even name it 'guest-stubs' or chroot-stubs (doesn't need to be Linux-VServer specific) 1243679415 M * Bertl go ahead, add/change it 1243679454 M * Net147 yea, after I create the package 1243679479 M * Net147 are there any other people using arch with vserver? 1243679524 M * Bertl I'm pretty sure, but probably they use guest distros which can be created via util-vserver 1243679538 M * Net147 gentoo guests are pretty nice 1243679543 M * Bertl that's why I suggested to integrate arch build stuff into util-vserver 1243679565 M * Net147 well the boot fixes are first steps 1243679993 M * Net147 i'm just concerned whether udev programs are used anywhere else in the system 1243680088 J * thierryp ~thierry@home.parmentelat.net 1243680191 M * Net147 if devices aren't allowed to be created in guest, should they not be allowed to delete devices too? 1243680233 J * cga ~weechat@82.84.189.184 1243680381 M * Bertl that would be an option actually, and easy to implement as flag too 1243680410 M * Net147 it seems strange that guest can't create devices but can delete them 1243680440 M * Bertl Linux-VServer gives you enough rope to shoot yourself in the foot :) 1243680481 M * Net147 is there a wishlist I can add it to? 1243680524 M * Bertl you can make me an argument why that would be a GoodThing(tm), and volunteer to test it (if we agree that it makes snese :) 1243680546 M * Net147 well yea I will test it 1243680550 M * Net147 if it is implemented 1243680585 M * Bertl now, why would you want that :) 1243680633 J * docelic ~docelic@78.134.198.23 1243680649 M * Net147 because it causes headache for vserver admin when a guest deletes their devices by accident but can't recreate them. 1243680672 M * Bertl same goes for files inside the guest, no? 1243680684 M * Net147 probably 1243680685 M * Bertl i.e. what if a guest deletes /usr 1243680724 M * Net147 then it's probably enough of their fault 1243680735 M * Net147 I don't know anyone that has deleted their /usr 1243680811 M * Bertl and I don't know anyone who deleted their devices :) 1243680822 M * Net147 sure but in this case it isn't the user at all deleting their devices 1243680846 M * Bertl so, the purpose is to work around a distro deficiency? 1243680852 M * Net147 it's the system trying to ensure the devices properly exist on boot 1243681174 M * Net147 at least with user deleting normal files they can create them again 1243681175 M * Bertl there is no 'boot' (hardware related) inside a guest, and if the distro wants to 'ensure' that the device properly exists, it should first check 'if the device already exists' or at least if it is possible to create that device 1243681227 M * Bertl it's not the best approach to remove the device first, just to try to recreate and fail an instant later :) 1243681242 M * Net147 well it's cp that does it 1243681266 M * Net147 could it be cp's fault then? 1243681281 M * Bertl sure, but in this case, it is the wrong tool anyway 1243681324 M * Net147 and the right tool to check if a device properly exists and has the correct type, major, minor? 1243681342 M * Net147 and permissions 1243681368 M * Bertl stat? 1243681451 M * Bertl stat --printf "%a-%t:%T" /dev/null 1243681664 M * Net147 does vserver restrict any device from being created? 1243681943 M * Bertl by default, all of them for security reasons 1243681959 M * Net147 is there a configuration option to change this? 1243682013 M * Bertl you can give the mknod capability to a guest, but that will harm security, or you can setup devmapper to allow specific 'secure' devices 1243682107 M * Net147 where can I find devmapper documentation? 1243682178 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html (look for vdevmap) 1243682313 M * Net147 well if I can get vdevmap to make it work, then we don't need such an option then eh? 1243682582 M * Bertl I guess we do not need that anyway, after all, you could make the directory immutable (for example) 1243682593 M * Bertl or do a read-only --bind mount 1243682791 M * Net147 immutable sounds like a good option 1243682840 J * geos_one ~chatzilla@chello084115149052.4.graz.surfer.at 1243682932 M * Net147 well immutable works. thanks Bertl. 1243682945 M * Bertl you're welcome! 1243682969 M * Bertl so we can scratch the 'disallow device removal' request? 1243682981 M * Net147 well immutable does same thing 1243682983 M * Net147 for devices 1243682995 M * Net147 so yea 1243683073 M * Net147 oh doesn't work cuz can't write to null now 1243683108 M * Bertl you did put it on the directory, yes? 1243683144 M * Net147 yes I did 1243683158 M * Net147 crw-rw-rw- 1 root root 1, 3 2009-05-30 21:24 null 1243683181 M * Net147 permission denied when trying to write to null 1243683225 M * Bertl that sounds like a mainline bug :) 1243683246 M * Net147 depends how you define immutable 1243683262 M * Bertl immutable on a directory means, you cannot change any entries 1243683290 M * Bertl accessing a device, which solely happens through the device major/minor is not part of that 1243683307 M * Bertl i.e. it will never change anything in the dir entries 1243683324 M * Bertl btw, just tested that here, works fine on 2.6.27.x 1243683329 M * Net147 but for files, you can't append? 1243683339 M * Net147 or overwrite? 1243683339 M * Bertl correct, but device nodes are no files 1243683350 M * Net147 hmm k 1243683352 M * Net147 so new bug 1243683371 M * Bertl what kernel version is that? 1243683384 M * Net147 2.6.29.4 1243683400 M * Net147 with vs2.3.0.36.14 1243683507 M * Bertl hmm, testing with 2.6.29.3 here this sequence works flawlessly: http://paste.linux-vserver.org/12957 1243683531 M * Net147 okay i'll try that sequence 1243683621 M * Bertl haven't tested inside a guest though 1243683629 M * Net147 I was using setattr 1243683631 M * Net147 is that different? 1243683648 M * Bertl what options? 1243683653 M * Net147 --immutable 1243683669 M * Bertl should be fine 1243683687 M * Net147 chattr +i null gives chattr: Operation not supported while reading flags on null 1243683703 M * Bertl as I said, on the dir, not on the device :) 1243683708 M * Net147 oh ok 1243683736 M * Net147 well that works on host 1243683764 M * Net147 --immutable on device is supposed to prevent writing though? 1243683804 Q * SauLus Ping timeout: 480 seconds 1243683882 M * Bertl strictly speaking, immutable on devices is not supported by mainline at all 1243683890 M * Net147 hmm k 1243683893 M * Net147 well setting immutable on dev directory seems to work 1243683896 M * Bertl it is possible to set it through Linux-VServer tools 1243683989 M * Net147 if it's not supported, shouldn't it be disallowed? 1243684003 J * saulus ~saulus@d047007.adsl.hansenet.de 1243684003 M * Bertl as you can see, it works :) 1243684018 M * Net147 well... immutable on devices works but is not supported? 1243684035 M * Bertl by mainline (i.e. kernel.org kernels) 1243684046 M * Net147 ohhh I get it 1243684055 M * Net147 I didn't know what mainline means 1243684075 M * Net147 well, thanks 1243684079 M * Bertl np 1243684236 J * chun ~sonicvari@158.37.56.5 1243684260 M * chun hi all 1243684265 M * Bertl hey 1243684284 M * chun i am trying to rename a vserver that i have, and is following what is on the FAQ, but ran into a problem 1243684289 M * chun it says: update link: /etc/vservers//run → /var/run/vservers/ 1243684312 M * chun but how could i have /var/run/vservers/ before i start the vserver? 1243684320 M * chun if that make sense 1243684345 M * Bertl if it isn't there, no need to update it 1243684363 M * chun and the same quesion goes to updating /etc/vservers//cache too 1243684374 M * chun Bertl: thx 1243684388 M * chun Bertl: then where do i update the run link to? 1243684402 M * Bertl those are links to the respective dirs, if they are missing, and everything works fine, no need to add or change them 1243684449 M * chun atm, its still pointing to /var/run/vservers/, that's ok? 1243684466 M * Bertl nope, you want to adjust that 1243684486 M * chun yes, then that's my quesiton 1243684526 M * chun where do i adjust that link to? i mean, i don't have a file that is called /var/run/vservers/ 1243684539 M * chun you see what i mean? 1243684549 M * Bertl isn't required for a symlink 1243684572 M * chun what isn't? 1243684582 M * chun oh i see 1243684588 M * chun ok, i try 1243684710 M * chun all looks good 1243684713 M * chun Bertl: thx again 1243684719 M * Bertl you're welcome! 1243684725 Q * balbir Read error: Connection reset by peer 1243684760 M * blathijs Is there any separate place to report util-vserver bugs, or is the mailing list fine for that? 1243684780 M * Bertl we have a bug tracker for util-vserver, but the ML should be fine too 1243685138 J * simontwo ~simon@78.129.201.122 1243685142 M * simontwo hello! 1243685245 M * simontwo I just upgraded to Debian Lenny and am now curious about what the difference is between the packages linux-image-2.6.26-[12]-vserver-686 (1 and 2) -- the package information does not disclose that. 1243685269 M * Bertl well, I tell you, I do not have a clue :) 1243685274 M * simontwo haha, ok 1243685285 M * Bertl you might ask the maintainer, who should know that 1243685318 M * Bertl the problem is, as the debian kernels do not even include the Linux-VServer patch name/number anywhere, it is hard to tell what was used (not to speak of changes) 1243685350 M * Bertl what I know is, that you should avoid the 2.6.26 kernels from debian, as they will break quite a number of things 1243685365 M * simontwo aw 1243685383 M * Bertl best go for 2.6.27 or for a recent 2.6.29.x 1243685485 M * simontwo thanks for that bit of advice. 1243685547 M * simontwo I think the linux-image-2.6.26-1-vserver-686 image uses a vserver version compatible with the util-vserver package (which is 0.30.216~r2772-6) 1243685584 M * simontwo wait, actually my current linux-image-2.6.18-6-vserver-686 kernel uses that version of the util-vserver package now, too. 1243685622 M * Bertl just be careful with 2.6.26, you have been warned :) 1243685946 M * Net147 is vlogin enabled by default for vserver guests? 1243686250 J * JohnnyPoo 182eb911@webchat.mibbit.com 1243686271 M * JohnnyPoo error: '/etc/init.d/rc' exited outside the expected code flow. 1243686291 M * JohnnyPoo ^^ Error I get when trying to use a Ubuntu Jaunty Guest on Debian Lenny 1243686601 M * Bertl yes, known issue, you are missing a script (empty one) in your guest 1243686617 M * Bertl IIRC, it specifies what script is missing at some point 1243686683 M * JohnnyPoo :/ 1243686690 M * JohnnyPoo /etc/default/rcS 1243686776 M * Bertl yeah, just touch it inside the guest 1243686838 M * JohnnyPoo I then start it, nothing happens. 1243686853 M * JohnnyPoo 8.10 hangs at the kernel logging daemon, too :( 1243686913 M * Bertl that's expected too ... 1243686930 M * Bertl theree is no point in running a kernel logger inside a guest, just disable it 1243686953 M * JohnnyPoo Ok 1243687037 M * Bertl if that's not possible, you can work around the problem by enabling a dummy kernel syslog 1243687055 Q * hijacker Remote host closed the connection 1243687159 M * JohnnyPoo When I try to start the vservers, no output is displayed and it doesn't start 1243687201 M * Bertl it probably starts, but no service is selected, so the last process exits, and the context is disposed 1243687215 M * Bertl make sure something keeps running inside the guest, e.g. syslog or cron 1243687242 M * Bertl (or switch to plain init style, and have 'init' running inside) 1243687371 M * JohnnyPoo I'll start an SSHd 1243687393 M * Bertl for example 1243688006 M * JohnnyPoo Er. 1243688015 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net 1243688051 M * JohnnyPoo Odd 1243688082 M * Bertl hmm? 1243688287 M * JohnnyPoo My client was being funky. 1243688561 J * docelic_ ~docelic@78.134.192.103 1243688719 M * JohnnyPoo The Jaunty debootstrap is way too basic. 1243688946 M * geb that may be normal if you are on lenny 1243688948 M * geb mercure:/srv/home/geb# LANG=; ls /usr/lib/util-vserver/jaunty 1243688948 M * geb ls: cannot access /usr/lib/util-vserver/jaunty: No such file or directory 1243688953 Q * docelic Ping timeout: 480 seconds 1243688987 M * geb oups it is /usr/lib/util-vserver/distributions/ 1243689032 M * geb there is now jaunty dir , so postinst scripts are only doing minimal work 1243689066 M * geb i guess than if you copy one from another recent ubuntu distribution it will work better ... but not perfectly 1243689070 M * JohnnyPoo When I debootstarp Intrepid, I have all the apps, but wit jaunty, only basics (no apt) 1243689140 M * geb hum so, it is a debootstrap problem 1243689144 M * geb see http://packages.debian.org/squeeze/all/debootstrap/filelist 1243689144 M * Net147 bootstrapping jaunty works fine for me. it installs apt-get. 1243689163 M * Bertl JohnnyPoo: what util-vserver version? 1243689194 M * simontwo Bertl, I've read up on some IRC logs from this channel. it seems like a pretty serious bug, the one with 2.6.26 and xattr flags. I wonder why that's not updated in debian. 1243689227 M * Bertl simontwo: yeah, no idea, let me know when/if you find out 1243689228 M * geb you may want to install a more recent deboostrap package or try to copy a ubuntu script to /usr/share/debootstrap/scripts/jaunty 1243689235 M * JohnnyPoo I did 1243689241 M * JohnnyPoo It doesn't install apt 1243689241 M * Net147 i've used debootstrap 1.0.13 and it includes jaunty support 1243689279 M * Net147 JohnnyPoo: what command you use to boot strap? 1243689295 M * geb yeah sid version seems to include jaunty support 1243689311 M * JohnnyPoo Well, I am using it in conjuction with the program to make the vserver 1243689327 M * Net147 this worked fine for me last I tried: vserver ubuntu build -m debootstrap --hostname ubuntu -- -d jaunty -m http://au.archive.ubuntu.com/ubuntu -- --arch=amd64 1243689345 M * Bertl JohnnyPoo: the program is 'vserver - build ..'? 1243689455 M * JohnnyPoo vserver vserver1 build -m debootstrap --context 42 --hostname Ubuntu --interface eth1:192.168.2.1/24 -- -d jaunty -m http://us.archive.ubuntu.com/ubuntu/ 1243689472 M * geb JohnnyPoo, please try with a more recent debootstrap version 1243689495 M * Net147 this one works - http://ftp.debian.org/debian/pool/main/d/debootstrap/debootstrap_1.0.13_all.deb 1243689664 M * JohnnyPoo Since Intrepid works, I use it and uprade to Jaunty 1243690490 Q * nenolod resistance.oftc.net osmotic.oftc.net 1243690791 J * nenolod nenolod@petrie.dereferenced.org 1243691291 M * Bertl okay, off for now .. bbl 1243691295 N * Bertl Bertl_oO 1243691899 Q * Net147 1243692541 Q * scientes Ping timeout: 480 seconds 1243693378 M * JohnnyPoo Why won't nmap work with vserver 1243693391 M * simontwo in what context? 1243693474 M * JohnnyPoo socket troubles in Init: Operation not permitted (1) 1243693568 Q * Mr_Smoke Server closed connection 1243693591 J * Mr_Smoke smokey@layla.lecoyote.org 1243693684 Q * geb Quit: Quitte 1243693922 J * ghislainocfs21 ~Ghislain@adsl2.aqueos.com 1243693953 M * mnemoc JohnnyPoo: give the context enough capabilities, and it will work 1243694232 Q * ghislainocfs2 Ping timeout: 480 seconds 1243695007 M * JohnnyPoo How do you mean 1243695524 M * mnemoc JohnnyPoo: http://linux-vserver.org/Capabilities_and_Flags 1243695661 Q * itess Quit: KVIrc 3.4.0 Virgo http://www.kvirc.net/ 1243697544 Q * dna Quit: Verlassend 1243698089 J * hparker ~hparker@linux.homershut.net 1243700152 P * dutchy_ 1243701094 J * cga_ ~weechat@82.84.189.184 1243701247 Q * cga Ping timeout: 480 seconds 1243701549 N * cga_ cga 1243701973 Q * Pazzo Quit: Ex-Chat 1243704411 Q * pmenier Quit: Konversation terminated! 1243705550 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1243705852 Q * thierryp Remote host closed the connection 1243706101 Q * JohnnyPoo Quit: http://www.mibbit.com ajax IRC Client 1243708279 J * derjohn_foo ~aj@e180195167.adsl.alicedsl.de 1243711460 J * cga ~weechat@82.84.189.184 1243713784 Q * sid3windr Ping timeout: 480 seconds 1243714264 J * sid3windr luser@bastard-operator.from-hell.be 1243714537 J * ktwilight_ ~keliew@136.88-240-81.adsl-dyn.isp.belgacom.be 1243714837 Q * ktwilight__ Ping timeout: 480 seconds 1243714988 J * scientes ~scientes@174-21-88-43.tukw.qwest.net 1243715533 Q * geos_one Quit: ChatZilla 0.9.84 [Firefox 3.0.10/2009050120] 1243715533 N * Bertl_oO Bertl 1243715541 M * Bertl back now ... 1243720802 Q * derjohn_foo Ping timeout: 480 seconds 1243725272 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1243725285 J * hparker ~hparker@2001:470:1f0f:32c:290:96ff:fe50:40fa 1243726056 Q * bonbons Quit: Leaving 1243726063 Q * FireEgl Quit: Leaving... 1243726314 Q * scientes Ping timeout: 480 seconds 1243726807 J * scientes ~scientes@174-21-88-43.tukw.qwest.net