1158365610 J * s0undt3ch_ ~s0undt3ch@bl7-242-14.dsl.telepac.pt 1158366027 Q * s0undt3ch Ping timeout: 480 seconds 1158366027 N * s0undt3ch_ s0undt3ch 1158366372 M * morfoh gn8 * 1158367399 Q * Blissex Remote host closed the connection 1158368435 Q * yarihm Quit: Leaving 1158368704 M * Bertl okay, I'm off to bed now .. have a good one everyone and cya tomorrow! 1158368713 N * Bertl Bertl_zZ 1158374562 Q * fluor- Ping timeout: 480 seconds 1158374657 J * ensc_ ~irc-ensc@p54B4EF3C.dip.t-dialin.net 1158374716 N * ensc_ Guest1043 1158374772 Q * ensc Ping timeout: 480 seconds 1158380517 J * spq___ ~spq@dslb-084-063-013-059.pools.arcor-ip.net 1158380522 M * spq___ gn8 1158380539 M * spq___ 6:22am :/ 1158380556 Q * spq___ Quit: spq___ 1158381535 N * nokoya _nokoya 1158383297 Q * matti Ping timeout: 480 seconds 1158387024 J * meandtheshell ~markus@85-125-231-157.dynamic.xdsl-line.inode.at 1158390936 Q * mire_ Quit: Leaving 1158390960 J * litage ~nick@203.220.55.70 1158391064 M * litage hi guys. an app (imapproxy) is being told to listen on 127.0.0.1, however it listens on the public ip address. could the fact that the app is running in a vserver be causing this? 1158391455 Q * Loki|muh Ping timeout: 480 seconds 1158391539 J * Loki|muh loki@satanix.de 1158392579 M * Hollow litage: there is no 127.0.0.1 in vservers 1158392613 M * litage Hollow: ah i see. how come? 1158392652 M * Hollow 99% of the time you don't need it ;) 1158392694 M * Hollow it has something to do with network isolation instead of virtualization, although i don't know the details... there will be a lo interface in vservers in the future though 1158392859 M * litage thanks Hollow! 1158392880 M * Hollow yw! 1158393660 J * bonbons ~bonbons@83.222.36.111 1158394263 J * dna_ ~naucki@p54BCF0EB.dip.t-dialin.net 1158394471 J * Monte PJIRCUser@j110074.upc-j.chello.nl 1158394476 M * Monte 'hallo 1158394495 Q * Monte 1158394801 J * dna___ ~naucki@p54BCF0EB.dip.t-dialin.net 1158395145 Q * dna_ Read error: Operation timed out 1158396130 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl 1158398403 M * phreak`` Hollow: wow, the patchlist is getting even smaller :-) 1158398787 M * Hollow phreak``: yup, regarding the syscall patch, did you read my comit message? ;) 1158398789 J * fluor- ~fluor@tanneries.squat.net 1158398856 M * phreak`` Hollow: now, yeah :) 1158398925 M * Hollow phreak``: i.e. i'll wait for the bug reports :P 1158398929 M * phreak`` bah, darn autotools DEPNEDs :) 1158398937 M * phreak`` *DEPENDs 1158398988 Q * coocoon Quit: KVIrc 3.2.0 'Realia' 1158399104 M * phreak`` Hollow: current util-vserver/trunk is broken ;) 1158399137 J * _mcp ~hightower@wolk-project.de 1158399137 Q * mcp Read error: Connection reset by peer 1158399141 N * _mcp mcp 1158399162 M * phreak`` or not .. 1158399798 M * Hollow phreak``: did you see the svn ebuild in our overlay? 1158399823 M * phreak`` Hollow: haven't updated that in a while :) 1158403467 Q * ||Cobra|| Remote host closed the connection 1158403470 J * coocoon ~coocoon@p54A0581A.dip.t-dialin.net 1158403588 J * dna_ ~naucki@p54BCE1A7.dip.t-dialin.net 1158403979 Q * dna___ Ping timeout: 480 seconds 1158405292 M * morfoh moin moin 1158407791 M * daniel_hozac anyone familiar with VLANs here? 1158407847 M * daniel_hozac has /proc/net/vlan changed from a file to a directory? 1158407937 Q * fluor- Ping timeout: 480 seconds 1158408372 Q * derjohn2 Ping timeout: 480 seconds 1158408416 J * derjohn2 ~aj@dslb-084-058-222-088.pools.arcor-ip.net 1158409818 M * daniel_hozac Bertl_zZ: http://people.linux-vserver.org/~dhozac/p/uv/experimental/vlan.patch does it look ok to you? 1158412469 J * shedii ~siggi@inferno.lhi.is 1158412707 Q * shedi Read error: Operation timed out 1158413704 M * daniel_hozac Hollow, phreak``: why are there two files in the bash completion script? what do they do? 1158413725 M * Hollow iirc one is for bash and one for zsh? 1158413737 M * daniel_hozac ah, ok. 1158413773 M * Hollow we just copied it from the wiki, i'm just a bashcomp user, it's syntax is aweful 1158413882 M * phreak`` daniel_hozac: what hollow said :) 1158413896 J * dna___ ~naucki@p54BCFF27.dip.t-dialin.net 1158414018 M * Wonka naucki with a long tail... 1158414091 M * harry daniel_hozac: you mean: spare:~# cat /proc/net/vlan/config 1158414117 M * harry i never knew it in a different way 1158414119 M * daniel_hozac harry: nevermind, i decided the test was dubious. 1158414130 M * daniel_hozac it has to be a directory. 1158414132 M * harry but i only just started using vlan trunking :) 1158414136 Q * dna_ Read error: Operation timed out 1158414151 A * harry gotta run now 1158414154 M * harry cya'll later 1158415486 J * mire ~mire@102-167-222-85.COOL.ADSL.VLine.verat.net 1158415852 N * Bertl_zZ Bertl 1158415857 M * Bertl morning folks! 1158415867 M * daniel_hozac good morning! 1158415901 M * Wonka run, harry, un 1158415903 M * Wonka +r 1158415906 M * Wonka wb Bertl daniel_hozac 1158415981 M * Bertl daniel_hozac: regarding the vlandev, so it actually checked for the existance of vlan (before) 1158415990 M * daniel_hozac it still does. 1158416012 M * daniel_hozac a few lines above that, it has a test -d /proc/net/vlan || { ; return 1 } 1158416037 M * Bertl ah, okay, so it was a duplicate, right` 1158416040 M * daniel_hozac thus the test -f /proc/net/vlan is doomed to fail, anyhow. 1158416070 M * Bertl okay, then it's fine for me :) 1158416558 M * daniel_hozac my list is now down to bash/zsh completion, fstab.xid, scheduler config, raising the host's ulimits, and dynamic contexts in userspace. 1158416574 M * Hollow morning Bertl 1158416629 P * marcfiu 1158416735 M * daniel_hozac Bertl: btw, what were your thoughts on including testme/testfs? 1158416741 M * morfoh moin Bertl 1158416746 M * morfoh Hi Hollow 1158416779 M * Hollow hey morfoh 1158416826 M * morfoh Hollow: still in "Beckstein County" ? :) 1158416839 M * Hollow yep, nearly finished with packaging stuff ;) 1158416846 M * morfoh good ;) 1158416889 J * marcfiu ~mef@targe.CS.Princeton.EDU 1158416917 M * Bertl wb marcfiu! 1158416974 M * Bertl daniel_hozac: not sure .. they are part in debian and gentoo? (at least the testme) and this results in old scripts to be used and reported back ... I'd prefer a central location and an url, with a comment to get the newest one 1158417091 M * phreak`` Bertl: at least not in Gentoo (last time I checked) 1158417097 M * daniel_hozac Bertl: how about a wrapper that does the downloading? 1158417120 M * daniel_hozac Bertl: could we then get an update so testme doesn't require legacy networking anymore? :) 1158417174 M * Hollow what about letting the testme script sync itself via getopt switch? 1158417236 M * phreak`` Hollow: that is bad :) it would break the checksum'ing (at least for us) and would probably create a blocker with FEATURES=collision-protect 1158417262 A * Hollow yawns 1158417265 M * Hollow ;) 1158417300 M * Bertl daniel_hozac: hmm, auto download would be an interesting idea ... 1158417326 M * Hollow Bertl: did you notice my query? 1158417329 M * Bertl daniel_hozac: and yes (regarding update), where is the tested patch? 1158417345 M * daniel_hozac http://daniel.hozac.com/vserver/testme-legacynet.diff 1158417366 M * Bertl Hollow: which one? 1158417395 M * Hollow -10 minutes 1158417416 M * Bertl 16:26 < Hollow> hey morfoh 1158417424 M * Hollow /query 1158417429 M * Bertl ah, sec 1158417462 M * morfoh HEY! 1158417466 M * morfoh :) 1158417501 M * Bertl Hollow: nope, no indication here ... 1158417508 M * Hollow uhm.. 1158417527 M * Hollow i'll send you an email :) 1158417534 M * Bertl okay :) 1158417569 M * Hollow ok, sent.. probably even better instead of volatile irc messages ;) 1158417606 M * Bertl ah, okay, thanks for the info! 1158417627 M * Hollow :) 1158417655 M * Hollow there are some open/free WLANs, so i can come online if it is needed (server outage or sth like that) 1158417698 M * Bertl okidoki, the email is the best solution for that anyway (i.e. non-volatile :) 1158417705 M * Hollow yep :) 1158417779 M * morfoh ack ... if the MUA is not volatile at all 1158417792 M * Hollow imap doesn't care about MUA :) 1158417797 A * morfoh claws crashed already 1158417804 M * morfoh using IMAP 1158417830 M * morfoh yet another issue to file in their bugzilla 1158417834 M * morfoh :) 1158418018 P * marcfiu 1158418681 M * daniel_hozac Hollow, phreak``, micah: is there a reason the zsh completion script isn't in zsh itself? 1158418696 M * Hollow no idea, sorry 1158418703 J * Adrinael adrinael@hoasb-ff0edd00-43.dhcp.inet.fi 1158418706 M * Hollow as i said, just copied it from the wiki 1158418717 M * daniel_hozac ok. 1158418914 M * daniel_hozac ah, found it. the Debian maintainer of zsh has forwarded it. 1158418963 M * Bertl do we have bash completion too? 1158418985 M * daniel_hozac yes. 1158419056 M * doener daniel_hozac: any changes in your fixed testme over mine? 1158419063 M * daniel_hozac doener: that is yours ;) 1158419069 M * doener ok :) 1158419074 M * daniel_hozac i just diffed it. 1158419088 M * doener ah, didn't notice the diff suffix 1158419714 M * michal` you've got zsh completion for vserver tools? 1158419926 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh-0.16 1158420157 M * michal` hm, found it...good to know 1158421125 J * pisco hiddenserv@tor.noreply.org 1158421139 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1158421281 M * Bertl welcome pisco! Blissex! Adrinael! 1158421316 M * Blissex hi! 1158421400 M * micah daniel_hozac: I submitted it to the debian package maintainer and its included in there, it was forwarded to the upstream zsh developers 1158421431 M * pisco Bertl: what do you think about http://www.vmware.com/vmtn/blog/console/2006/09/#benchmarking 1158421570 M * Blissex pisco: but VMware is an alternative to things like UML and Xen, not to VServer... 1158421578 M * michal` exactly 1158421615 M * michal` vserver is for running multiple virtual servers on a single machine under a single kernel 1158421624 M * michal` (what, of course, outpefroms them all :) 1158421640 M * pisco yes, but many users following the vmware hype, in many cases vserver can also be an alternative 1158421666 M * michal` and vserver is more comfortable too - it has a lot of features that other virtualisation techniques won't have 1158421691 M * michal` like - vserver is very resource friendly - think about shared memory,buffers,unification...to name a few 1158421849 M * pisco thats the thing, if you can compare to other technologies and can say 'hey if you need *virtualisation* for linux systems with very smal overhead ...' you can make a point by people wich loves benchmarks 1158422103 M * pisco sure, it's a maketing foobla, but IMHO dont hide skills if you have some or many ;) 1158422253 M * doener daniel_hozac: hm, how does the symlink for env look like in the case that makes it fail? 1158422266 M * daniel_hozac ../../bin/env from /usr/bin/env. 1158422317 M * doener just tried that here, doesn't fail for me 1158422324 M * doener (namespace with bindmount) 1158422331 M * daniel_hozac does your util-vserver-vars contain /usr/bin/env? 1158422333 M * Blissex the big question as to VMware vs. VServer is: do you want one single 'root' and config, or multiple independent ones? 1158422365 M * doener _ENV="/usr/bin/env" 1158422384 M * doener oh, the symlink is inside the vserver only, right? 1158422402 M * daniel_hozac no, the symlink is on the host 1158422406 M * Blissex in many many cases one wants a single 'root' user and config, and isolation between the _users_, and then VServer is particularly good. 1158422454 M * doener daniel_hozac: ah, so we have, host: /usr/bin/env -> ../../bin/env, and vserver /usr/bin/env only, right? 1158422499 M * daniel_hozac i'm not sure about the guest's setup... i believe it was an Ubuntu guest i was testing with (i don't have that system available right now, i'll start duplicating it locally). 1158422518 M * doener ubuntu should be like debian 1158422524 M * doener and yes, it fails with that setup 1158422539 M * daniel_hozac so that's why :) 1158422557 M * doener I'll dig around a bit 1158422673 M * daniel_hozac seems very odd that it's accessing the host for /usr/bin/env, and then the guest to resolve the symlink... 1158422698 M * daniel_hozac in general i find it very strange that the vserver binaries are even accessible during start, since the rbind happens so early. 1158422812 M * doener it doesn't use the guest to resolve it 1158422821 M * doener it hits the bind mount 1158422826 M * daniel_hozac well, right. 1158422841 M * daniel_hozac but why doesn't it hit it when accessing /usr/bin/env? 1158422852 M * daniel_hozac or /usr/sbin/vcontext, etc. 1158422888 M * doener probably the same reason why cd / doesn't hit it 1158422898 M * doener I still don't understand how you can hit it at all 1158422918 M * daniel_hozac forgive my ignorance, but why doesn't cd / hit it? :) 1158422932 M * daniel_hozac is / specially resolved using current->root or whatever? 1158422939 M * doener yep 1158422949 M * daniel_hozac ok. 1158422958 M * doener and AFAICT .. uses the parent mount, but WTF has the bindmount as parent?! 1158422986 M * daniel_hozac well, / would be the bind mount, right? 1158423055 M * doener the hierarchy of the mount structs themselves is not changed, the old / is still the parent of /proc 1158423068 M * doener the bindmount is a stacked mount 1158423082 M * doener you have to go there from its parent to ever reach it 1158423088 M * daniel_hozac doesn't the rbind remount /proc too? 1158423121 M * doener that's another thing we wonder about... cat /proc/mounts shows only the / mount 1158423164 M * doener and if you ever enter the bindmount area, there's nothing mounted below it 1158423173 M * doener /proc etc. are empty 1158423186 M * daniel_hozac so rbind doesn't work? 1158423221 M * doener I'll see what happens if I do the stuff manually 1158423318 M * doener daniel_hozac: ok, cd ../../../.. is useless for testing, bash does its own simplification 1158423340 M * daniel_hozac there needs to be a C-shell. 1158423347 M * doener tcsh? 1158423364 M * daniel_hozac not quite what i meant, more along the lines of a shell that executes actual C :) 1158423369 M * doener heh :) 1158423376 M * daniel_hozac without messing with it. 1158423395 M * daniel_hozac i suppose it would suck to actually use though :) 1158423405 M * doener ok, using C I can hit the bindmount, and I see /proc mounted there 1158423415 M * doener and /proc/mounts agrees with me 1158423457 M * doener (done manually using vnamespace -n bash ... mount -t proc none /vservers/foo/proc; mount --rbind /vservers/foo/ / ) 1158423565 M * doener secure-mount is broken it seems 1158423604 M * doener yup, passes MS_MGC_VAL|MS_NODEV|MS_BIND 1158423656 M * daniel_hozac hmm, really? no MS_REC even for rbind? 1158423667 M * doener I did test with --rbind only 1158423947 M * doener daniel_hozac: hm, it called /bin/mount, and has no support for passing rbind to it, it just passes --bind 1158423958 M * doener it should never actually do that though 1158423982 M * daniel_hozac hmm, src/secure-mount.c:canHandleInternal is broken. 1158423996 M * doener yeah, if mnt!=0 1158424006 M * doener should rather be ==, right? 1158424015 M * daniel_hozac yeah, i guess. 1158424021 M * daniel_hozac i don't see where it would ever be NULL though. 1158424044 M * doener it would crash anyway 1158424050 M * doener (in mountSingle) 1158424056 M * daniel_hozac yeah. 1158424064 M * daniel_hozac hmm. 1158424065 M * daniel_hozac if ((flag & MS_NODEV)!=0) flag |= MS_NODEV; 1158424072 M * daniel_hozac isn't that rather redundant as well? 1158424089 M * daniel_hozac (mountSingle) 1158424114 M * doener yeah, it says "if it's there, add it" :) 1158424139 M * doener should be == as well I guess 1158424174 M * doener though != and == 0 is quite ugly... a plain !(flag & MS_NODEV) looks a lot better IMHO 1158424222 M * doener it actually took me some seconds to grasp the meaning of !=0, the other syntax seems more natural 1158424339 M * daniel_hozac hehe, yeah. 1158424435 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/secure-mount.diff 1158424446 M * daniel_hozac umm, ! in the NODEV, of course. 1158424471 M * doener hm, !mnt? 1158424496 M * daniel_hozac ok. 1158424502 M * doener mnt->type == 0 is probably fine, but for mnt I'd do it with ! 1158424531 M * daniel_hozac yeah, i agree. 1158424684 M * doener hm, ls does actually open("../../../..") if I ask it to, but does not hit the bindmount *starts to go crazy* 1158424714 M * daniel_hozac lol 1158424769 M * doener ARGH! 1158424799 M * doener try this: vnamespace -n bash 1158424805 M * doener then: mount --bind /vserver/foo / 1158424808 M * doener then: ls .. 1158424813 M * doener and then: ls ../.. 1158424821 M * doener and then throw away your brain 1158424880 M * daniel_hozac what the... 1158424901 M * daniel_hozac ls .. shows the /vservers/..., ls ../.. shows the host's /. 1158424909 M * doener indeed... 1158424916 M * daniel_hozac that's just nuts. 1158424920 M * doener this isn't even funny anymore 1158424924 M * daniel_hozac agreed. 1158425458 J * Snow-Man_ ~sfrost@kenobi.snowman.net 1158425461 Q * Snow-Man_ 1158427352 M * ebiederm Someone forgot the chroot? 1158427532 M * Bertl hey ebiederm! how is it going! 1158427628 M * daniel_hozac doener: i committed the secure-mount fixes. 1158427677 M * doener ebiederm: hm? the symlink thing is pre-chroot 1158427682 M * ebiederm Bertl: Ok. 1158427736 M * doener there's an additional security hack that bind mounts /vservers/foo to /, which helps in someway to keep the user in his filesystem even if he escaped the chroot 1158427759 M * doener but we found some problems with that. 1158427763 M * ebiederm doener: Sorry I saw a part of the context and must have misuderstood. I was remember that when you set up a mount namespace yout have to chroot into it or else your current working directory and the root directory wind up outside of the new mount namespace. 1158427841 M * ebiederm doener: I would have thought just properly setting up a mount namespace would have been enough to guarantee the user couldn't escape. 1158427876 M * doener no, the namespace is just a copy of the host's namespace which gets the additional mounts 1158427905 M * doener so he's captured inside the namespace, but all the good stuff is available in that namespace 1158427960 M * doener ebiederm: but maybe you can help us to understand the whole thing... 1158427974 A * ebiederm Uh-oh! 1158427999 M * daniel_hozac doener: hmm, i cd'd to /vservers/.../root, .. showed the guest, ../.. a list of guests, ../../.. showed the guest again, and ../../../.. was the host. 1158428005 M * Bertl okay, I'm off for now .. will be back later! 1158428007 M * Hollow doesn't the rbind only prevent breakage after you have chrooted? 1158428012 N * Bertl Bertl_oO 1158428014 M * doener we see rather funny things with the bind mount (see the instructions above, about an hour ago) 1158428048 M * doener the vnamespace -n bash thing just gets you a new namespace so you don't mess with your box, the rest is plain vanilla stuff 1158428074 M * doener and the results don't match with how I read the kernel source 1158428178 M * ebiederm doener: I think I roughly read the situation right. 1158428205 M * ebiederm After you do the bind mount you haven't updated the current working directory to be a child of the new / 1158428222 M * ebiederm So you have your old current working directory. 1158428237 M * ebiederm At least that is what is happening if I have read the logs correctly. 1158428354 M * ebiederm So if you do cd / before you tring the ../.. thing it should fail. 1158428360 M * daniel_hozac it doesn't. 1158428371 M * daniel_hozac .. still shows the host. 1158428387 M * ebiederm Hm... 1158428434 M * ebiederm And you need to change the root directory, in a similar manner. 1158428445 M * ebiederm I believe. 1158428446 M * Hollow well, IMO you still need chroot fot it to work correctly.. 1158428456 M * daniel_hozac yeah, you do. 1158428522 M * ebiederm The difference is that once you have done the chroot, it should become impossible to escape. 1158428570 M * doener the bind mount and the chroot are not really connected, except for the fact that you end up in the bind mount after you escape from the chroot (assuming that there's no barrier) 1158428605 M * ebiederm Let me put this a silight different way. 1158428626 M * ebiederm In struct fs_struct you have: struct vfsmount * rootmnt, * pwdmnt 1158428686 M * ebiederm Since those hold onto their location in the namespace tree, the must be manually changed once you have updated the namespace if they are below the mount that was performed. 1158428692 M * ebiederm Since everything is below / 1158428696 M * ebiederm You need it. 1158428702 M * ebiederm Does that make sense? 1158428752 M * ebiederm Basically after you mount a new / you need to chroot / and chdir /. 1158428855 M * ebiederm doener: ? 1158428898 M * doener I'm setting up a test case 1158428964 M * ebiederm Ok. 1158428978 M * doener ebiederm: using chroot / (which does chroot("/") and chdir("/")) does not put me into the bind mount 1158428997 M * doener / is resolved to the current rootmnt 1158429033 M * daniel_hozac if you do chroot /vservers/..., ls ../* doesn't work though 1158429062 M * doener yeah, because it stops at rootmnt, there's a check in follow_dotdot 1158429088 M * doener that's why I don't get how we into the bindmount area at all 1158429108 M * ebiederm Ok. I see the problem. 1158429135 M * ebiederm When I wrote code to do this I did it two different ways. 1158429143 M * doener I actually _can_ get there, but I don't see _how_ that works actually 1158429152 M * ebiederm 1) I changed to the directory I wanted to become root and then did: 1158429182 M * ebiederm mount(".", "/", NULL, MS_MOVE, NULL); 1158429203 M * ebiederm Then: chroot("."); 1158429213 M * ebiederm 2) I used pivot_root. 1158429268 M * ebiederm After converting the directory I wanted to be my new root to a mount point with: 1158429269 M * doener pivot_root is a no-go for util-vserver (atm) as it relies on still being able to access the host's stuff 1158429282 M * ebiederm mount(root, root, NULL, MS_BIND | MS_REC, NULL); 1158429353 M * ebiederm doener: I I guess I can see that. Although just keeping a filedescriptor to the hosts filesystem around should achieve that I believe. 1158429385 M * ebiederm But I don't remember exactly what pivot_root does to update everything. 1158429386 M * doener ebiederm: not if you still want to be able to mount sth. inside the vservers 1158429423 M * doener pivot_root actually _exchanges_ mount points, the plain rbind (and probably the move in some way) only creates a stacked mount 1158429457 M * ebiederm doener: I see what you mean. Yes you don't want to remove that earlier mounts. 1158429485 M * ebiederm What I did was to set everything up in a subdirectory that I wanted and then exchanged root mount points and then unmounted the old root. 1158429503 M * ebiederm But goals differ, which is why flexible kernel tools are nice :) 1158429505 M * doener yeah, that's clean and easy, but doesn't work for us :( 1158429998 M * ebiederm Later. 1158430503 M * Hollow doener: basically the rbind is useless, right? 1158434097 Q * Blissex Remote host closed the connection 1158434619 Q * somegeek Ping timeout: 480 seconds 1158435146 J * somegeek hiddenserv@tor.noreply.org 1158437447 J * fluor- ~fluor@tanneries.squat.net 1158440770 Q * lilalinux_ Remote host closed the connection 1158443981 Q * dna___ Quit: Verlassend 1158444093 J * marl ~matt@84.92.193.226 1158444405 M * marl hi, sounds like a daft question :) but if im compiling kernel 2.6.17.11 with vserver patches (ubuntu dapper) do i just need the vmlinuz file in my grub or do i need to create an extra file with mkinitrd? most docs ive found so far dont mention mkinitrd can anyone give me a definate answer? 1158444443 M * Wonka marl: it works just the same as without vserver patches 1158444475 M * daniel_hozac and you really should use 2.6.17.13 with the latest patch. 1158444519 M * Wonka ack bzgl .13 1158444529 M * Wonka corrects some uuugly errors 1158444530 M * marl i was goin gby the website 2.6.17.11 1158444600 M * marl where do i get the .13 patch files? 1158444604 M * daniel_hozac the website is missing a few releases, as you can tell by the topic :) 1158444624 M * daniel_hozac i'm adding them to the FTP now, but http://vserver.13thfloor.at/Experimental/ has them at the bottom. 1158444636 M * marl thanks daniel_hozac 1158444648 M * marl typical just started another make of .11 LOL 1158444755 M * marl anyone know if theres any pre-compiled kernels based on ubuntu dapper with vs already patched in? 1158444804 M * marl been googleing, but havnt found any yet :( and would prefer one that is know to people here anyway 1158444804 M * daniel_hozac http://ubuntu.uni-klu.ac.at/ubuntu.uniklu/ has some, i think. 1158444837 M * daniel_hozac there's a wiki page on Ubuntu's wiki too, IIRC. 1158444856 M * daniel_hozac https://help.ubuntu.com/community/VServer 1158445018 M * marl lol, thanks, i could have saved myself an hours worth of googling if i asked here first :( 1158445115 M * daniel_hozac that was like hit number 4 if you googled on ubuntu vserver... 1158445138 M * marl no idea what i typed then, but i couldnt find it :( 1158445166 M * marl thanks again btw :) 1158445181 M * daniel_hozac np. 1158446050 M * morfoh can I bur a bit your 1158446051 M * morfoh can I bur a bit your 1158446227 M * morfoh ok ... burnin' bro's bandwidth :) 1158446575 M * morfoh 1/4 of De_La_Soul-The_Grind_Date done 1158446750 J * Laibsch ~leggewie@V33e6.v.pppool.de 1158447120 P * Laibsch 1158447182 A * mnemoc wonders if morfoh is drunk 1158447395 J * vrwttnmtu ~eryktyktu@82-69-161-137.dsl.in-addr.zen.co.uk 1158447465 M * vrwttnmtu I'll ask just in case anyone is here - I'm getting errors in various apps related to Disk quota exceeded. 1158447492 M * vrwttnmtu I don't use quotas, although it is on an XFS filesystem mounted usrquota,grpquota 1158447543 M * daniel_hozac did you try remounting it without those options? 1158447548 M * vrwttnmtu This is what I see in Postfix: Sep 16 23:55:02 asdfg postfix/local[10646]: 01624E499E16: to=, orig_to=, relay=local, delay=0.13, delays=0.09/0.01/0/0.04, dsn=5.2.2, status=bounced (maildir delivery failed: create maildir file /var/www/xxxx/.maildir/tmp/1158447302.P10646.x.y.z: Disk quota exceeded) 1158447558 M * vrwttnmtu No - it would need a reboot for that 1158447579 M * daniel_hozac hmm, mount -o remount? 1158447587 M * vrwttnmtu Didn't seem to work 1158447625 M * vrwttnmtu There's no real reason to have it mounted with those options on the host admittedly 1158447738 M * vrwttnmtu Just tried remounting it again, and it still didn't do it. 1158447895 M * vrwttnmtu daniel_hozac, On another note, have you heard of Ghost Rider? 1158447925 Q * ag- Remote host closed the connection 1158447937 J * ag- ~ag@caladan.roxor.cx 1158447945 M * daniel_hozac nope. 1158448008 M * vrwttnmtu He's this Swedish (?) guy (?) that rides motorbikes on the public roads over there. Absolutely insane. Stockholm to Upsalla in 15 minutes, or something crazy. 1158448066 M * vrwttnmtu http://video.google.co.uk/videosearch?q=Ghost+Rider But if you're not into bikes, it's not going to be interesting. :) 1158448179 M * doener *lol* according to wikipedia, videos are worth nothing in court, so the police would have to capture the guy while he's doing his stuff... 1158448191 M * doener (in Sweden that is) 1158448212 M * vrwttnmtu Really? 1158448214 M * daniel_hozac ah, well, that's sort of common here. 1158448286 M * vrwttnmtu I thought Scandinavians were supposed to be not aventurous, not risk takers..? 1158448326 M * vrwttnmtu This is the one. http://video.google.co.uk/videoplay?docid=5984187290687814499&q=ghostrider 1158448354 M * vrwttnmtu I want a Hayabusa. 1158448396 M * daniel_hozac there's probably at least one street race per month around here. 1158448459 M * vrwttnmtu I like the sound of Sweden. Decent, progressive social policies (although that could be changing), women (of course), and now this? 1158448470 M * vrwttnmtu You are in the EU, aren't you? 1158448481 M * vrwttnmtu I can just move over when I like? 1158448489 M * vrwttnmtu Got tag, or something 1158448519 M * vrwttnmtu What's not so great about Sweden, Daniel? 1158448708 M * daniel_hozac yep. 1158448730 M * daniel_hozac the high taxes are probably peoples biggest gripe. 1158448759 M * vrwttnmtu But if high taxes = decent public transport, roads, health care, social security, that's good, no? 1158448793 M * daniel_hozac yeah, but at least health care is in dispute... 1158448811 Q * Johnnie Remote host closed the connection 1158448825 M * vrwttnmtu Bet it beats the NHS though... :) 1158448866 M * daniel_hozac it's too bad that movie is so low res, i don't recognize anything. 1158448866 J * Johnnie ~jdlewis@jdlewis.org 1158448921 M * vrwttnmtu There is a high res version of it - which is just mesmerising. Sure, a race at night with no-one else around is one thing, but in the daytime, with loads of cars around...! 1158449015 P * Johnnie 1158449022 J * Johnnie ~jdlewis@jdlewis.org 1158449168 J * renihs ~renihs___@193.170.52.125 1158449176 M * renihs narf :) 1158449921 Q * meandtheshell Quit: bye bye ... 1158450139 Q * bonbons Quit: Leaving 1158450619 Q * vrwttnmtu Quit: Leaving