1116374422 M * daniel_hozac ciphernaut: and it is still running?
1116374453 M * Martin ciphernaut: http://linux-vserver/sparc64 - description, scroll to the bottom, under "SUCCESS"
1116374688 M * Doener_ /usr/sbin/vserver: line 79:  4069 Killed -- this looks strange...
1116374741 M * ciphernaut daniel_hozac: yes, I can still enter the vserver
1116374745 Q * rs Quit: rs
1116374764 M * daniel_hozac ciphernaut: so what's still running? and why isn't the vserver cleaned up?
1116374787 M * SEAwolfx hmm .. how would u build a new vserver instance which is running on unionfs?
1116374819 M * ciphernaut there are a few 'reboot -d -f -i' processes hanging around, in 'D' (diskio? daemon?) mode
1116374963 M * Martin Doener_: This is now fixed, the file's a bit of a mess as I've been using it as a notepad
1116374980 M * ciphernaut http://deadbeefbabe.org/paste/636
1116375065 M * Doener_ Martin: hm? i've been talking to ciphernaut, if you were commenting on my last sentence... ;)
1116375214 M * Martin Yes I was and sorry for intruding
1116375234 M * Doener_ no problem, i've just been wondering what you meant ;)
1116375244 M * Martin Wiki now actually contains the information
1116375655 M * ciphernaut I have been using the dietlibc from Debian package version 0.28-3
1116375676 M * Doener_ Martin: thanks :)
1116376085 M * ciphernaut is the svn repo for util-vserver being updated?
1116376127 M * Doener_ regarding the stuff enrico does, only when a new release is ready
1116376136 M * daniel_hozac really?
1116376145 M * daniel_hozac is that due to the internet connection thing?
1116376150 M * daniel_hozac he used to update more frequently.
1116376183 M * Doener_ svn <-> cvs ;) the svn repository is the one at OF, while enrico uses savannah
1116376206 M * daniel_hozac well, i thought the "stuff enrico does" implied savannah's CVS ;)
1116376270 M * Doener_ uhm, what i was trying to say was: the svn repo only gets updated with what enrico does, when a new version of util-vserver is released
1116376298 M * Doener_ i.e. enrico may have commited to his cvs repo, but that's not in the svn repo yet (as no new release was done yet)
1116376300 M * daniel_hozac yeah, i got that now.
1116376303 M * Doener_ ok :)
1116376614 M * ciphernaut sweet
1116376942 M * Martin Please ignore previous doubts about networking, it seems to work - it's just the test servers I was using seem to have very heavily loaded connections.
1116376962 M * Martin What should ifconfig (from one of the vservers and from host) output
1116376964 M * Martin ?
1116376997 M * Doener_ depends on whether you use aliases or not
1116377012 M * Doener_ if your build command didn't change, you don't use any
1116377021 M * Martin OK
1116377024 M * ciphernaut this problem I am having, affects only the 2.0rc1 release of vserver, 1.9.5 patched against the same debian kernel source 2.6.11 (.7?) works fine
1116377034 M * Martin I didn't /explicitly/ request any
1116377054 M * Doener_ then ifconfig on the host shouldn't change, and in the vserver, it should show an empty (i.e. no ip address) entry for eth0
1116377070 M * Martin then it does work -cool :-D
1116377086 M * Doener_ to see the non-aliased ip addresses, install iproute, and then use "ip addr show" or just "ip a"
1116377120 M * Martin indeed.
1116377126 M * Doener_ iproute uses netlink, which can handle non-aliased ip addresses, while ifconfig uses ioctls IIRC and can't do that
1116377172 M * Martin Cool
1116377179 M * Martin And this is a build option?
1116377216 M * Martin I was having some issues with two hosts getting the same alias device.
1116377243 M * Doener_ you can do: --interface alias=device:ip/mask
1116377255 M * Doener_ but you can also change that later on in the configuration for that vserver
1116377283 M * Doener_ the flower page should have the details, i don't use aliases, so i don't remember which file you need to create/change
1116377333 M * Martin OK, I'll probably avoid them but with this info I can ammend the bug report I sent.
1116377371 M * Doener_ usually you either specified the same alias twice, or the alias was simply too long
1116377413 M * Doener_ the interface address name (=the alias) can take up to 15 characters, including the ethX:, i.e. you have 10 characters for your alias
1116377426 M * Doener_ everything else is simply truncated
1116377520 M * Martin Yes, this seemed to be the problem I was running into however it seemed to truncate to four characters from the vserver name.
1116377540 M * Martin eth0:test  for test-vserver-1 and test-vserver-2
1116377555 M * Martin May be that it was a 2.4 series kernel
1116377637 M * Doener_ hm, that's a ifconfig limitation, it only shows up to 9 characters
1116377659 M * Doener_ doener ~ # ifconfig | grep test
1116377660 M * Doener_ eth0:test Link encap:Ethernet  HWaddr 00:26:54:08:28:D8  
1116377660 M * Doener_ doener ~ # ip a | grep test
1116377660 M * Doener_     inet 1.2.3.4/32 scope global eth0:test1-test
1116377714 M * Doener_ but if you have been using stable tools back then, they have automatically created aliases (as they use ifconfig)
1116377740 M * Doener_ and as the first 10 characters of both vservers match, you get a conflict there
1116377764 M * Doener_ s/vservers/vserver names/
1116377876 M * Martin OK, will note this on update to the bug report
1116377935 M * Doener_ well, there's nothing that can be done to this with the stable tools, and basically a user error with the alpha tools (well, the alias limitation isn't well known ;)
1116377977 M * Doener_ s/ to / about /
1116378036 J * eXplasm2 explasm@p549FF30A.dip.t-dialin.net
1116378115 M * Martin I take it the long term solution is to move to using the non-alias / netlink solution
1116378156 M * Doener_ yep, as done by the alpha tools. you can still use alias, but now you have to take care of that yourself (and thus can avoid conflicts)
1116378167 M * Martin Definately cancel any adverse comments about networking, I just got a sustained connection of over 1 MByte/s international so I'm happy :-D
1116378324 Q * explasm__ Read error: Operation timed out
1116379378 T * services.oftc.net http://linux-vserver.org/ | latest stable 1.2.10, devel 1.9.5, 2.0-rc1, ng9.4  -- He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 
1116380106 M * mep_ how can i specify a special mtab for a vserver?
1116380729 M * Martin OK, vservers now works (with some prodding) on sparc64, I'll try to get back later this week (Thursday probably) to work on the last few bugs (debootstrap not working properly and requiring ----no-namespace to start the server)
1116380736 M * Martin Thanks all for all the help.
1116380970 M * Martin Night all
1116380974 Q * Martin Quit: using sirc version 2.211+KSIRC/1.1
1116381185 Q * nox Remote host closed the connection
1116382657 Q * flock Read error: Operation timed out
1116383195 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il
1116383634 N * Bertl_zZ Bertl
1116383642 M * Doener_ morning Bertl!
1116383647 M * Bertl morning Doener_!
1116383649 M * Doener_ unable to sleep?
1116383670 M * Bertl well, I did sleep, and not too bed, but the rain* woke me up ...
1116383695 M * Bertl *) rain, something evil, happening outside ;)
1116383706 M * Doener_ :)
1116383733 M * Bertl say, what did Martin mean with the 'requiring ----no-namespace'?
1116383804 M * Doener_ hm, the 'normal' call failed with "wait: no child" or sth. like that, he then did just the 'internal' call and that seems to have worked for him
1116383841 M * Doener_ but now that you ask, he had to setup the vshelper first (using the 'internal' call), so maybe the 'normal' call would now also work for him
1116383887 M * Doener_ btw, you pointed me to the lkml thread mentioning chroot'ing to /proc/NNNN/root some time ago, right?
1116384080 M * Doener_ i had a discussion in #kernelnewbies today and got back to that thread... looking at what some mails say is supposed to work, i'm living in a parallel kernel universe... ;)
1116384125 M * Doener_ ('normal' call above mean "vserver xxx start", 'internal' call means "vserver ----nonamespace xxx start")
1116384400 M * Bertl yeah, was a funny thread ...
1116384421 M * Bertl oh my god, I just survived ;)
1116384454 M * Bertl 04:42 -!- speciesx [speciesx@144.139.79.194] has joined 
1116384454 M * Bertl 04:43 -!- speciesx [speciesx@144.139.79.194] has left []
1116384629 M * Doener_ ?
1116384670 M * Bertl (insider joke, not even a good one, nevermind)
1116384686 M * Doener_ Bertl: anyway, apart from feeling like living in a parallel kernel universe, that thread gave me some ideas...
1116384718 M * Bertl let's hear then!
1116384762 M * Doener_ it seems like the cross namespace mount has a bug, because mnt->mnt_namespace is not adjusted in this case (it doesn't have to in all other cases ;)
1116384793 M * Doener_ but that's just a thing we'd need to fix if we want to implement a vmount
1116384844 M * Doener_ vmount would be a tool that let's you cross mount directly into a vserver's namespace, without having the mount source available in the vserver's namespace
1116384876 M * Doener_ it just has to be available on the host (as the current kernel-side stuff only works for simple bind mounts [locking reasons i guess])
1116384898 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com
1116384937 M * Doener_ we could possibly do that as we have namespace as 1.5th class citizens (1st class for contexts, 2nd class for processes)
1116385015 M * Doener_ hm, wait... maybe i'm thinking the wrong way... sec..
1116385045 M * Doener_ no, should work...
1116385093 M * Bertl IMHO is would be a 'dangerous' thing ...
1116385159 M * Doener_ hm... it's already possible, you just have to write a tool... but it would create unremovable mounts atm...
1116385178 M * Doener_ what exactly would you consider dangerous?
1116385191 M * Bertl well, every 'namespace' has his 'own' vfsmnt entries
1116385203 M * Bertl (currently)
1116385210 M * Doener_ yep
1116385220 M * Bertl and they separate the 'view' into the filesystem quite nicely
1116385260 M * Bertl if you would really do what you meant (i.e. make 'mount' from an existing xid=0 vfsmnt into xid!=0)
1116385268 M * Bertl s/meant/said/
1116385289 M * Bertl then you would start mixing up namespaces pretty soon ...
1116385298 M * Doener_ http://www.inf.bme.hu/~mszeredi/nstest.c -- mounts from child namespace to parent namespace
1116385352 M * Doener_ (as namespaces are 2nd class citizens in vanilla kernels, you can only do it this way or using a process that exposes it's file descriptors via a unix socket with SCMRIGHT
1116385365 M * Bertl yes
1116385396 M * Bertl but now the host and the 'guest' would shar this path, no?
1116385402 M * Bertl *share
1116385407 M * Doener_ with current vanilla kernels, this actually mixes up namespaces, because mnt->mnt_namespace is 'wrong' afterwards (the mnt in namespace A still has mnt_namespace==B)
1116385437 M * Doener_ they would both have a vfsmount for this path...
1116385446 M * Bertl well, if you are looking for a clean vmount implementation, that should be easy
1116385456 M * Bertl (but without sharing ;)
1116385521 M * Doener_ what do you mean by "sharing" exactly?
1116385578 M * Bertl having mixed up namespace entries and/or common vfsmounts 
1116385639 M * Doener_ hm? do_loopback() calls clone_mnt()... still a common vfsmount?
1116385654 A * Doener_ is confused...
1116385750 M * Bertl you would make a clone_mnt_namespace() which supplies the namespace
1116385772 M * Bertl then 'copy' the vfsmounts as usual, and stick them into the 'new' namespace
1116385792 M * Bertl would take the old->mnt_namespace as default argument
1116385877 M * Bertl no?
1116385923 M * Doener_ that would need other enhancements, where would you get the new namespace from?
1116385951 M * Doener_ doing "mnt->mnt_namespace = current->namespace;" seems to make more sense to me
1116385955 M * Bertl passed as argument ...
1116385978 M * Bertl you would then use it in non_default way in a vmount syscall command, no?
1116386005 M * Doener_ uhm, i don't want a vmount syscall... just a userspace tool ;)
1116386007 M * Bertl i.e. start with the fs 'normal' then do the lookup inside the context's namespace, and pass that to the copy
1116386039 M * Bertl yes, but that's exactly what you 'try' to accomplish by doing that dir passing, no?
1116386068 M * Doener_ yes, but i don't see the difference...
1116386134 M * Bertl the difference is, that in one case you know _exactly_ what you do and when you allow it (and what namespace to use), in the other, you build bridges between the namespaces
1116386151 M * Bertl (which is not what we really want in linux-vserver, or am I wrong?)
1116386420 M * Bertl the namespaces are not handled in vanilla properly ... right?
1116386426 M * Doener_ right
1116386449 M * Bertl and folks like viro haven't figured a 'smart' way to 'specify' a namespace yet, right?
1116386468 M * Bertl especially because namespaces are invisible right now
1116386486 M * Bertl now we add an unique number to 'our' namespaces ...
1116386487 M * Doener_ yep (not sure if they [viro] want that at all)
1116386520 M * Bertl 'our' being the 'main' namespaces for vserver guests ;)
1116386560 M * Bertl using the 'current' namespace for mounting into is what is done right now, and works reasonably well ...
1116386590 M * Bertl mounting across namespaces requires to 'specify' the target namespace somehow, no?
1116386596 M * Doener_ yep
1116386611 M * Bertl so how would you like to specify that in userspace right now?
1116386739 M * Doener_ for vservers? by xid...
1116386854 M * Bertl so how does that differe from what I suggested?
1116386887 M * Doener_ sec, preparing some small c program
1116387816 M * Doener_ http://www.13thfloor.at/~doener/vserver/tools/vmount.tar.bz2
1116387841 M * Doener_ except for the kernel not correctly setting mnt_namespace and thus creating an unremovable mount, this works
1116387948 M * Doener_ hm, last upload still contained some useless debug output, did a new one...
1116387968 M * Bertl too late, already worked through the annoying process of downloading and unpacking ;)
1116387975 M * Doener_ heh ;)
1116387982 M * Bertl (would have been easier to provide 'just' the vmount.c)
1116388003 M * Bertl so and how is that 'better' than doing:
1116388046 M * Bertl vserver(VCMD_bind_mount, xid, from, to)
1116388155 M * Bertl (except that you do not have fds which cross the namespace boundary?)
1116388296 M * Bertl Doener_: do you know how to get a 'normal' util-vserver package from the cvs repository at savannah?
1116388310 M * Doener_ no need to reimplement do_loopback
1116388314 M * Doener_ no, i don't know
1116388426 Q * ntrs Quit: Leaving
1116388795 M * Bertl yum and apt somehow hate me ...
1116388889 M * Doener_ hm, we could switch the current->namespace between 'current' and 'context' namespace, then we'd only have to reimplement some parts of do_mount...
1116388911 M * Bertl imho we do not have to reimplement anything ...
1116388925 M * Bertl we just extend it by passing the destination namespace
1116388950 M * Bertl for all do_*_mount()
1116388970 M * Bertl the default calls 'just' pass the current->namespace
1116389194 M * Doener_ hm, guess i'll have to try to implement this to know if it works ;)
1116389240 M * Bertl go ahead ... idea is to propagate the namespace up into the do_*mount() functions first
1116389272 M * Bertl then pass current->namespace from all current callers, then make a new syscall command
1116389441 M * Bertl Doener_: any ideas what to do in this case?
1116389451 M * Bertl yum install autoconf automake
1116389466 M * Bertl .. bla bla bla .. wait .. bla bla ...
1116389474 M * Bertl Getting m4-1.4.1-15.i386.rpm
1116389480 M * Bertl retrygrab() failed for:
1116389481 M * Bertl   http://ayo.freshrpms.net/fedora/linux/2/i386/core/RPMS/m4-1.4.1-15.i386.rpm
1116389489 M * Bertl Error getting file http://ayo.freshrpms.net/fedora/linux/2/i386/core/RPMS/m4-1.4.1-15.i386.rpm
1116389492 M * Bertl [Errno 4] IOError: HTTP Error 403: Forbidden
1116389509 M * Bertl (and I did yum update just a minute before that)
1116389596 M * Bertl just switch to fedora.us?
1116389633 M * Doener_ probably... gave up on yum and switched to apt-rpm for my fc tests
1116389718 M * Bertl I really don't get it, why are all yum/apt repositories broken somehow?
1116389836 M * Doener_ hm, i'm stuck... will try again once i got some sleep, i.e. tomorrow...
1116389845 M * Bertl k, good night then!
1116389872 M * Doener_ no, the 'real' tomorrow ;) got to go to university in two hours
1116389912 M * Bertl ah, okay, then 'good university' ;)
1116389920 M * Doener_ *g*
1116389935 M * Doener_ will get some breakfast now... have a good day!
1116389939 N * Doener_ Doener|gone
1116396532 Q * yarihm Quit: Leaving
1116399967 M * SEAwolfx good morning @all
1116400051 M * Bertl morning SEAwolfx!
1116400223 M * SEAwolfx Bertl: could u imagine to use a LVM managed partition in combination with unionfs to get vserver-context quota to work?
1116400641 M * Bertl define context quota ;)
1116400994 M * SEAwolfx Bertl: the root of the vserver should be possible to give quotas to his users. ;)
1116401008 M * Bertl works fine with LVM volumes
1116401043 M * SEAwolfx Bertl: ok, next thing to work out for me. ;)
1116401101 M * Bertl you use the vroot device to proxy it  securely
1116401116 M * Bertl (it being the quota ioctls)
1116401127 M * matti Bertl: :]
1116401129 M * matti romke: :)
1116401132 M * matti albeiro: :)
1116401136 M * matti * :)
1116401167 M * SEAwolfx Bertl: is there any further documentation about LVM with vserver and quota?
1116401193 M * Bertl yes, there are one or two howtos (older ones)
1116401236 M * SEAwolfx Bertl: could you please point me to them?
1116401316 M * Bertl http://linux-vserver.org/Quota+and+Disk+Limits
1116401360 M * SEAwolfx thank you Bertl. ;)
1116401366 M * Bertl you're welcome!
1116401370 M * matti ;]
1116401526 J * sukria ~sukria@213.223.184.193
1116401567 M * Bertl morning sukria!
1116401576 M * sukria hey Bertl :)
1116401630 M * Bertl ad xfree: why do you want to start X inside a vserver (with custom modules?)
1116401689 M * SEAwolfx you could enable a terminal server this way?
1116401717 M * Bertl no, not really, you would use xvnc or so for that ...
1116401729 M * Bertl (i.e. a headless server)
1116401737 M * sukria Bertl: in fact I like the idea to use vserver for having several desktop at the same time (ubuntu, debian sarge, debian sid...)
1116401758 M * sukria my laptop has a ATI Radeon Mobility card, and I compiled the ATI module 
1116401759 M * Bertl on different vcs?
1116401764 M * sukria yes
1116401788 M * sukria the idea is to have the main host on tty7, the ubuntu vserver on tty8....
1116401788 M * Bertl k, well, IIRC X requires quite some permissions ...
1116401807 M * Bertl you should check out the problematic programs section
1116401826 M * sukria yes, at least : CAP_SYS_RAWIO
1116401836 M * sukria but my problem is solved
1116401852 M * Bertl ah, okay, fine ... nevermind then ;)
1116401855 M * sukria I just don't use the ATI module in my vserver, I use the internal vesa driver
1116402811 Q * SiD3WiNDR Ping timeout: 480 seconds
1116403016 J * SiD3WiNDR luser@bastard-operator.from-hell.be
1116403442 M * Hollow morning
1116403736 M * Bertl morning Hollow!
1116403742 M * Hollow hey Bertl 
1116405018 Q * lilo Remote host closed the connection
1116405159 J * lilo ~lilo@lilo.usercloak.oftc.net
1116408359 J * rs ~rs@staff.lycos.fr
1116408965 M * Bertl morning rs!
1116408978 M * rs morning bertl, how are you ?
1116408985 M * Bertl tired ;)
1116408998 M * rs is it evening for you ? :)
1116409006 M * Bertl very late evening, yep ;)
1116409012 M * rs hehe
1116409035 M * rs how many time did you see the light of the sun those last years ? :)
1116409071 M * Bertl every second day, sunrise, just before I go to bed ;)
1116409124 M * rs like I thought :)
1116410639 M * hellekin motivés motivés
1116410724 M * Hollow haha
1116411096 M * Bertl okay, folks ... I'm off to bed now ... cya later
1116411112 N * Bertl Bertl_zZ
1116411839 J * TeraX ~terax@195.37.209.41
1116411849 M * TeraX Good Morning
1116412059 M * TeraX i have a problem with my kernel compilation, after i'm running the testme.sh the following errors are printed: http://nopaste.php-q.net/134637
1116412067 M * TeraX can anybody help me?
1116412768 M * SEAwolfx hmm, a missing lib?
1116412811 M * TeraX i don't know which lib
1116412814 M * SEAwolfx hm, it seems to be a vserver-specific function that is missed.
1116412903 M * SEAwolfx this problem occurs during kernel compilation?
1116412978 M * TeraX no there are no problems at the kernel compilation (sorry for my bad english)
1116413001 M * SEAwolfx I also suffer under a bad english ;) dont mind.
1116413038 M * SEAwolfx well .. dont really know how to debug this.
1116413121 M * TeraX mh :/
1116413906 Q * rs Quit: rs
1116415316 J * knoppix_ ~knoppix@dsl-213-023-128-204.arcor-ip.net
1116415519 J * nox ~nox@noxlux.de
1116416883 J * rs ~rs@staff.lycos.fr
1116418105 J * JohnL ~JohnL@nsnoc.plus.com
1116418789 J * Tera-X ~terax@195.37.209.41
1116418789 Q * TeraX Read error: Connection reset by peer
1116418809 N * Tera-X TeraX
1116419111 M * sukria is it possible to specify a gateway in a vserver conffile?
1116420643 M * SEAwolfx imho this is not needed.
1116420655 M * SEAwolfx u only have to set a snat for the vserver.
1116420739 M * daniel_hozac the vserver will use your host's gateway.
1116420829 M * SEAwolfx but why its sufficient to only set a snat-rule. isnt a additional dnat rule to substitute masquerading? 
1116420848 M * SEAwolfx +necessary
1116420983 M * SEAwolfx on snat i set the source-ip to the ip of the host using ppp0 for internet connection, so what is when the response comes back?
1116421974 Q * DuckMaster Quit: Leaving
1116422307 M * daniel_hozac i guess SNAT'ing takes care of that for you.
1116422777 M * SEAwolfx hmm .. i cant imagine that. this is are normal snat/dnat iptables rules:
1116422784 M * SEAwolfx iptables -t nat -I POSTROUTING 1 -o eth0 -s 10.0.1.2  -j SNAT --to x.y.z.k
1116422784 M * SEAwolfx iptables -t nat -A PREROUTING -i eth0 -d x.y.z.k -j DNAT --to 10.0.1.2
1116422803 M * romke dsoul: bêdê mia³ dla ciebie screeny jak nazwa.pl ma zrobione wszelkie kwestie z ns'ami + zmian± danych do whois
1116422813 M * SEAwolfx what u said?
1116423866 Q * rs Quit: rs
1116423990 J * rs ~rs@staff.lycos.fr
1116424728 Q * cereal Remote host closed the connection
1116425758 J * DuckMaster ~duckx@195.75.27.158
1116426184 Q * TeraX Read error: Connection reset by peer
1116426613 M * romke SEAwolfx: oops, sorry, wrong window :P
1116426742 Q * DuckMaster Quit: Leaving
1116426766 J * DuckMaster ~duckx@195.75.27.158
1116426910 J * mep__ mep@p5091FFF1.dip.t-dialin.net
1116427096 Q * mep_ Read error: Operation timed out
1116428767 J * TeraX ~terax@195.37.209.41
1116429171 N * Doener|gone Doener
1116429180 M * Doener evening folks!
1116429192 M * Doener TeraX: hm, you didn't boot a vserver-enabled kernel i'd say...
1116429410 M * TeraX hi Doener
1116429428 M * TeraX yes i see, now a vserver kernel is booted and i'm testing install a vserver :)
1116429448 M * Doener ok :)
1116429474 M * TeraX mhh
1116429529 M * TeraX the script say following errors: [000]# failed.; [031]# failed. and [202]# failed.  the other's are succeeded 
1116429626 M * Doener hm, for #031 and #202 i suspect the dynamic context bug in util-vserver, don't remember which version fixed that... i'd suggest upgrading to the latest util-vserver (0.30.207)
1116429660 M * Doener anyways, i'm gone again, too tired... good luck TeraX!
1116429664 N * Doener Doener|gone
1116429686 M * TeraX okay :)
1116429693 M * TeraX good night
1116430134 J * floyy ~dream@195.222.40.202
1116430139 P * floyy 
1116431040 Q * infowolfe Quit: Leaving
1116432107 J * mep_ mep@p5091F10D.dip.t-dialin.net
1116432544 Q * mep__ Ping timeout: 480 seconds
1116434107 J * infowolfe infowolfe@66-230-103-130-cdsl-rb1.nwc.acsalaska.net
1116434561 Q * tchan Quit: leaving
1116434596 Q * infowolfe Ping timeout: 480 seconds
1116435604 J * infowolfe infowolfe@209-112-216-98-cdsl-rb1.nwc.acsalaska.net
1116435889 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net
1116436094 Q * infowolfe Ping timeout: 480 seconds
1116437682 Q * rs Quit: rs
1116438009 Q * TeraX Quit: 
1116438834 Q * atsab Ping timeout: 480 seconds
1116441825 P * JohnL 
1116442394 Q * SEAwolfx Remote host closed the connection
1116442416 J * SEAwolfx debian-tor@kisangani.ruwenzori.net
1116442785 N * Bertl_zZ Bertl
1116442803 M * Bertl morning folks!
1116443176 M * Johnsie Hi, Bertl.
1116443230 M * Bertl hey Johnsie!?
1116443236 M * Johnsie I'm new.
1116443237 M * Johnsie haha
1116443307 M * Bertl yeah, it's interesting .. there are a lot of new folks around recently .. any specific reason for that? (not that I would not appreciate it ;)
1116443375 M * Johnsie Well...
1116443383 M * Johnsie I found out from Beave.
1116443386 M * Johnsie I know him...
1116443400 M * Johnsie He setup Gentoo as a vserver for me on a server hosted by ThePlanet in Dallas, TX.
1116443406 M * Johnsie So, I'm just here to sponge and learn.
1116443416 M * Johnsie And, hopefully, give something back. ;)
1116443484 M * Bertl sounds good to me .. so basically hearsay ...
1116443490 M * Johnsie Yes sir.
1116443500 M * SiD3WiNDR Setting up util-vserver (0.30.204-5) ...
1116443500 M * SiD3WiNDR chmod: changing permissions of `/var/lib/vservers/': Operation not permitted
1116443500 M * SiD3WiNDR dpkg: error processing util-vserver (--configure):
1116443506 M * SiD3WiNDR ola: please fix this..
1116443522 M * Bertl and please update to 0.30.207 + syscalls ;)
1116443533 M * SiD3WiNDR please update debian sarge to 0.30.207 :p
1116443752 M * Bertl Johnsie: so how do you like linux-vserver in the 'age' of Xen ;)
1116443772 M * Johnsie I love it, actually.
1116443787 M * Johnsie I'm sort of an intermediate Linux user in general...
1116443797 M * Johnsie And I've found it very easy to use.
1116443824 M * Johnsie We have a system with about 25 virtual private servers in use.
1116443838 M * Bertl you haven't used UML before, right?
1116443842 M * Johnsie No.
1116443848 M * Johnsie I understand UML ...
1116443867 M * Johnsie There are some things about UML that make me laugh, though.
1116443881 M * Johnsie Building an empty file of a particular size for a file system to reside, etc.
1116443917 M * Bertl well, you can do all that with linux-vserver too, it's just so that folks which used to use UML, are somewhat confused with the networking ...
1116443938 M * Johnsie Well, I started playing and reading and what not and figured stuff out.
1116443946 M * Bertl they try hard to use tun/tap devices or so ...
1116443980 M * Johnsie I wasn't sure how to add IPs and figured I'd try copying the interface directory in /etc/vservers ...
1116444004 M * Johnsie Then changed the ip file.
1116444009 M * Johnsie Restarted and it worked.
1116444017 M * Johnsie I mean, it's pretty straight forward.
1116444021 M * Johnsie And I'm a new guy on the block.
1116444057 M * Bertl good to hear that you like it ... some folks are used to the older config (one file config) and do not like the tree based config very much (yet)
1116444058 M * SiD3WiNDR running a uml from a loopback is not that bad of a plan Johnsie :p
1116444074 M * SiD3WiNDR bleh
1116444077 M * Johnsie heh
1116444080 M * SiD3WiNDR stupid util-vserver debian package
1116444084 A * SiD3WiNDR hits ola 
1116444087 M * SiD3WiNDR 184938 times
1116444110 M * Bertl not on the head! not on the head! hmm ...
1116444157 M * Bertl wasn't there an up-to-date package floating around?
1116444181 M * SiD3WiNDR I don't know
1116444188 M * SiD3WiNDR I just upgraded my sarge to the latest
1116444194 M * SiD3WiNDR I guess the uptodate one is in unstable
1116444202 M * SiD3WiNDR but I bet it still contains chmod 000 /var/lib/vservers
1116444243 M * Bertl heh, lol
1116444280 M * Bertl the funny part is, this wasn't even right with old 2.4/1.2 ;)
1116444285 M * SiD3WiNDR could be
1116444291 M * SiD3WiNDR this is breaking the postinstall though
1116444298 M * SiD3WiNDR luckily you can just edit that in /var/lib/dpkg
1116444300 M * SiD3WiNDR so I commented it out
1116444319 M * SiD3WiNDR I'm not sure why it gets operation not permitted
1116444326 M * SiD3WiNDR is that because of the +t attribute?
1116444355 M * SiD3WiNDR yea, seems it is
1116444367 M * SiD3WiNDR if ola really WANTS that 000 he should chmod -t first
1116444371 M * SiD3WiNDR but I dislike the 000 :p
1116444396 M * SiD3WiNDR it also breaks "df" output as regular user
1116444462 M * Bertl doesn't make any sense either ...
1116444481 M * Bertl 000+t was used in 2.4 not used/recognized in 2.6 anymore
1116444498 M * Bertl and the tools get it right on 2.4 _and_ 2.6
1116444527 M * Bertl so the _only_ sane thing to do is setattr --barrier
1116444552 M * SiD3WiNDR ah
1116444556 M * SiD3WiNDR so the +t isnt needed either?
1116444567 M * Bertl no, on 2.6 it's just bogus ...
1116444594 M * SiD3WiNDR okay. :)
1116444619 M * Bertl well, it means tail merging for directories ;)
1116444647 M * Bertl (so it's at least not vserver related)
1116444695 M * SiD3WiNDR it's a bug
1116444699 M * SiD3WiNDR which ola closed as unreproducible
1116444700 M * SiD3WiNDR *sigh*
1116444715 M * Bertl yeah, it's hard to test without installing ;)
1116445116 M * DaPhreak hmmms :)
1116445130 M * DaPhreak is noexec inside a vserver even supported / doable ?
1116445158 M * SiD3WiNDR even better
1116445165 M * SiD3WiNDR util-vserver is not even installable on unstable atm
1116445182 M * DaPhreak heh ...
1116445210 M * SiD3WiNDR   util-vserver: Depends: beecrypt2 but it is not installable
1116445214 M * SiD3WiNDR beecrypt2 doesn't exist :p
1116445219 M * DaPhreak heh :D
1116445232 M * SiD3WiNDR well
1116445235 M * SiD3WiNDR in stable and testing it does
1116445238 M * SiD3WiNDR but not in unstable
1116445240 M * SiD3WiNDR for whatever reason
1116445250 M * Bertl you do not 'require' the beecrypt
1116445261 M * Bertl it is only required for vhashify
1116445278 M * DaPhreak Bertl: heh, but the depend chain requires him to need it ;)
1116445313 M * DaPhreak SiD3WiNDR: wasn't it possible also on debian to "fake" the package manager that a package is installed ?
1116445349 M * Bertl good choice of wording ... ;)
1116445369 M * DaCa DaPhreak: why would you care? try running /lib/ld-linux.so 
1116445394 M * DaPhreak DaCa: you mean the noexec ?
1116445400 M * SiD3WiNDR DaPhreak: everything is possible
1116445409 M * SiD3WiNDR Bertl: well, it's required on debian :)
1116445409 M * DaCa DaPhreak: yes, mounting noexec is worthless
1116445416 M * SiD3WiNDR Bertl: so, 2 more bugs in the debian bts
1116445424 M * SiD3WiNDR mounting noexec isn't worthless
1116445425 M * DaPhreak heh poor ola ;)
1116445428 M * SiD3WiNDR I mount /tmp noexec
1116445440 M * SiD3WiNDR a) it breaks lots of preinstall/postinstall debian scripts
1116445449 M * SiD3WiNDR b) it breaks regular apache/phpbb/... worms
1116445455 M * SiD3WiNDR ofcourse it doesn't break determined users ;)
1116445464 M * DaCa :p
1116445466 M * SiD3WiNDR but a makes me do bugreports!
1116445473 M * SiD3WiNDR I never did it, but I always think I should ;)
1116445477 J * Doener` ~doener@p54873E6D.dip.t-dialin.net
1116445480 M * SiD3WiNDR "do not assume you can put shit in /tmp and make it run"
1116445496 M * DaPhreak hmmm man 8 mount
1116445534 M * Bertl welcome Doener` ?!
1116445554 M * DaPhreak DaCa: noexec Do not allow direct execution  of  any  binaries  on  the mounted  file system.  (Until recently it was possible to run binaries anyway  using  a  command  like  /lib/ld*.so /mnt/binary.  This  trick  fails  since  Linux  2.4.25  / 2.6.0
1116445576 M * SiD3WiNDR 2 doeners
1116445579 M * SiD3WiNDR overdose!
1116445581 M * DaPhreak well as I said .. I doesn't know ..
1116445587 M * DaPhreak heh overkill!
1116445630 M * Bertl DaPhreak: so much about 'recently' ;)
1116445651 M * DaPhreak heh ;) yeah
1116445697 M * DaPhreak well the man-page is quite _a_ bit old :D (2004-12-16)
1116445723 M * DaCa oh well, its good to know that they finally solved that 
1116445904 Q * Doener|gone Ping timeout: 480 seconds
1116446996 Q * duckx Remote host closed the connection
1116448184 Q * SEAwolfx Ping timeout: 480 seconds
1116450449 Q * SiD3WiNDR Read error: Operation timed out
1116450495 J * SiD3WiNDR luser@bastard-operator.from-hell.be
1116451738 J * ntrs ~ntrs@Dardeene-68.188.50.87.charter-stl.com
1116452039 N * BobR_oO BobR
1116452293 N * BobR BobR_zZ
1116453081 Q * ntrs Quit: Leaving
1116454377 J * duckx ~Duck@mna75-1-81-57-39-234.fbx.proxad.net
1116454499 Q * knoppix_ Quit: Verlassend
1116458862 M * Bertl okay, night folks!
1116458901 N * Bertl Bertl_zZ
1116459529 Q * Hollow Remote host closed the connection