1100568437 J * rs rs@imhotep.rhapsodyk.net
1100568971 Q * tchan Quit: leaving
1100569031 J * tchan tchan@c-24-13-81-164.client.comcast.net
1100571400 Q * mhepp Ping timeout: 480 seconds
1100573521 M * Doener 2.6.10-rc2 is out
1100575088 Q * rs Ping timeout: 480 seconds
1100575891 J * Shuri Dew@dsl.speedline209.226.electronicbox.net
1100578540 J * anonymous-coward nwalsh@shaggy.internode.com.au
1100578541 M * anonymous-coward n
1100578698 J * monrad monrad@213083190130.sonofon.dk
1100578854 M * mugwump p
1100578869 M * Doener r
1100578885 M * monrad morning guys
1100578924 M * Doener morning monrad 
1100579393 M * mugwump pinky, I forgot to ask - are you an infobot or a dumb logger?
1100579417 A * mugwump slaps pinky on the side of its cabinet
1100580599 J * _no_x vps@c135165.adsl.hansenet.de
1100580691 Q * no_x Ping timeout: 480 seconds
1100580699 N * _no_x no_x
1100581450 M * Shuri gnite all
1100581457 M * Shuri 24:04
1100581459 M * Shuri :P
1100581464 M * Shuri cya tomorow
1100581493 M * Shuri this is  crazy Doener juste wake up!
1100581514 M * Shuri im going to bed
1100581519 M * Shuri have fun
1100581527 Q * Shuri Read error: Connection reset by peer
1100585731 N * Bertl_zZ Bertl
1100585743 M * Bertl morning folks!
1100585835 M * monrad morning
1100585861 M * Bertl mugwump: so was pinky a nice boy^Ht?
1100585931 M * Bertl hey monrad, how are you?
1100585937 M * infowolfe Bertl: why is there a 16 ip limit per vserver?
1100585956 M * Bertl because nobody ever need more than 16 ips ;)
1100585966 M * infowolfe well, what if i do?
1100585978 M * Bertl .oO( or was that 640k of memory?)
1100585988 M * infowolfe lol
1100585997 M * infowolfe so Bertl what do i do if i need more than 16 ips?
1100586006 M * Bertl you can raise the limit, but be careful, it has a cost
1100586016 M * infowolfe what kind of cost?
1100586044 M * monrad i am fine, i am about to learn how to design my own cpu today (and the next 8-9 weeks)
1100586047 M * Bertl the list of ips assigned to a vserver is checked with O(n) at every connection/action/etc
1100586093 M * infowolfe Bertl: what kind of performance hit would i be looking at?
1100586137 M * Bertl monrad: hey great! ASIC design? or hardware architecture in general?
1100586149 M * infowolfe Bertl: i'm needing either 23 or 24
1100586164 M * Bertl probably not that problematic ... compared to 16
1100586190 M * infowolfe Bertl: where do i change that limit?
1100586198 M * Bertl and it solely depends on the number of _active_ ips ...
1100586201 M * monrad Bertl, i can tell you that when i get home :)
1100586215 M * infowolfe well, i've been given a /27 (32 ips)...
1100586232 M * infowolfe and the biggest traffic hog is on their own vps... so i don't think it's that big of a problem
1100586262 M * infowolfe Bertl: so i'd like to raise that 16 ip max limit to 24, can you point me in the right direction to change it?
1100586277 M * monrad infowolfe, have you seen hollows gentoo guide?
1100586316 M * infowolfe monrad: haven't had a chance, i haven't been home lately
1100586317 M * Bertl infowolfe: include/linux/vserver/network.h
1100586322 M * infowolfe as soon as i get home
1100586327 M * infowolfe Bertl: so i need to recompile?
1100586333 M * monrad ok its quite nice
1100586337 M * Bertl #define NB_IPV4ROOT     16
1100586351 M * Bertl infowolfe: yes, but you can raise it to 32
1100586382 M * infowolfe Bertl: is that enrico's tools or kernel-space?
1100586408 M * Bertl you have to change the kernel, and modify the tools IIRC
1100586415 M * infowolfe ok
1100586682 M * infowolfe thanks for the help Bertl 
1100586688 M * Bertl you're welcome!
1100587328 Q * monrad Quit: Leaving
1100590427 J * zzyber andreb@h111n3c1o1022.bredband.skanova.com
1100590437 M * Bertl welcome zzyber!
1100590451 M * zzyber Bertl, Hi
1100590673 M * zzyber I have an idea to offer virtual server automaticly thrugh a website for a couple of $:s. Is vserver the best for this if i want to make linux servers virual? Does anyone here know of any success stories with online offers like this?
1100591275 M * Bertl hmm ... good question ...
1100591290 M * Bertl what are the alternatives for you?
1100592940 J * monrad monrad@shop3.diku.dk
1100593071 M * zzyber Bertl, i really dont know if there is any alternatives other than vmware or vm for s390
1100593121 M * Bertl hmm, okay, well, vmware adds a large overhead (because it emulates the entire system)
1100593139 M * Bertl this is not the case with linux-vserver, which is basically about sharing resources
1100593174 M * zzyber hmm, is this good or bad from the customers side?
1100593209 M * Bertl really depends ... how many vmwares do you plan to put on one server?
1100593401 M * zzyber as many as i can, i was thinking to have 3 levels of service. Nr1 with good bandwith and maybe a cluster, Nr2 with less banwith and no cluster and Nr3 with less evrything, How many vm server do you think its possible to run on a Intel P4x2 2Ghz 2Gb memory and a lot of disk?
1100593445 M * Bertl don't know about vmware there, but I guess the limit will be somewhere around 10 or maybe 15?
1100593458 M * zzyber i must set up a test enviroment soon
1100593480 M * zzyber 10 or 15 with vserver?
1100593484 M * Bertl lycos has tested linux-vserver with up to 100 vservers
1100593494 M * zzyber nice!
1100593497 M * Bertl (although I consider that a little extreme)
1100593505 M * zzyber are you running vserver?
1100593510 M * Bertl yep
1100593521 M * zzyber private or in work?
1100593529 M * Bertl both
1100593597 M * zzyber i will read up a bit on vserver and alternatives but im sure i will be back. Thanks for the chat
1100593618 M * Bertl you're welcome! btw, there is a nice paper about linux-vserver
1100593629 M * Bertl http://linux-vserver.org/
1100593635 M * zzyber thanks
1100593638 M * Bertl http://linux-vserver.org/Linux-VServer-Paper
1100593651 P * zzyber Lämnar
1100594625 J * Val val@tatooine.loria.fr
1100594660 M * Bertl welcome Val!
1100594666 M * Val Hi Bertl :)
1100594862 M * Val 'm trying to package last alpha vserver utils for debian for use with 2.6.X dev. kernel patches, of course another unofficial package
1100594870 M * Val ...
1100594912 Q * monrad Ping timeout: 480 seconds
1100594933 J * monrad monrad@shop3.diku.dk
1100594949 M * Val btw it works for me but many friends needs packages that works out of the box
1100595043 M * Bertl well, _another_ unofficial debian package is welcome, I guess
1100595059 M * Val yes
1100595088 M * Val shity stuff
1100595095 M * Val shity staff :)
1100595123 M * Loki|muh is it normal, that chkrootkit gave the following output?
1100595124 M * Loki|muh |>Checking `lkm'... You have    14 process hidden for readdir command
1100595124 M * Loki|muh |>You have    14 process hidden for ps command
1100595124 M * Loki|muh |>Warning: Possible LKM Trojan installed
1100595196 M * Loki|muh because I have a load at 1.0 and vtop and vps shows _no_ Process does cause this
1100595224 M * Loki|muh :(
1100595252 M * Bertl hmm, well, no, not exactly what you want to get from chrootkit ...
1100595391 M * Loki|muh shit.
1100595441 M * Loki|muh any recommendations for proceeding?
1100595456 M * Bertl system shutdown, then analysis ...
1100595482 M * Bertl (i.e. putting the disk into a clean system and searching for evil stuff)
1100595628 M * Loki|muh how can I let me show this 14processes? do you have an idea?
1100595654 M * Bertl well, it's probably a kernel module doing that ... or a ps wrapper
1100595667 M * Bertl first you could check with /proc
1100595725 M * chand morning all
1100595736 M * Bertl morning chand!
1100595904 M * chand how ya doin ?
1100595917 M * Bertl fine thanks! and you?
1100595946 M * Bertl Loki|muh: another option would be to use magic-sysreq if enabled to dump the task list
1100595961 M * chand pretty good now that the vds seems to behave correctly at last :)
1100595968 M * chand and also from other stuff :)
1100595975 M * Bertl good to hear!
1100595983 Q * monrad Ping timeout: 480 seconds
1100596002 M * Val hi chand :)
1100596055 M * Loki|muh Bertl: ah! du you now more about this? it is enabled in the kernel
1100596086 M * Bertl okay, you can activate it by echoing commands to the proc entry
1100596095 M * Loki|muh i do not find any process number in /proc that is not in ps :(
1100596119 M * Bertl for example 'echo h >/proc/sysrq-trigger'
1100596136 M * Bertl that should write the help message to the dmesg/klog
1100596143 M * Loki|muh thx
1100596628 J * monrad monrad@shop3.diku.dk
1100596633 M * Loki|muh maybe the vserver kernel has a security problem? i'm on it
1100596657 M * Bertl probably the same security problem the vanilla kernel has ...
1100596708 M * Bertl or do you imply, that the host was compromised from within a well configured vserver?
1100596886 M * Loki|muh i do not know, i know anything :(
1100596913 M * Loki|muh the 2.6.9 has a security problem?
1100596940 M * Bertl well, not that I know of ... but maybe?
1100597185 M * chand hey val
1100597215 M * Loki|muh seems not so
1100597226 M * Loki|muh seems to be a buggy backup client
1100597229 M * Loki|muh :/
1100597232 M * Loki|muh sucks.
1100597328 J * rs rs@ice.aspic.com
1100597334 M * rs hi folks
1100597428 M * Bertl welcome rs!
1100597510 M * rs hey bertl
1100597513 M * rs how are you?
1100597533 M * rs I saw you managed to fix the uptime bug ?
1100597534 M * Bertl fine, thanks, and you?
1100597544 M * Bertl no, that was Doener!
1100597549 M * rs not enought sleep, but fine :)
1100597570 M * rs you was for the team :)
1100597587 M * Bertl ah, okay, of course 'we' fixed it ;)
1100597654 M * rs would be possible that a vserver with bogus uptime oddly behave ?
1100597677 M * Bertl yes, it might cause some 'unexpected' side effects ...
1100597678 M * rs like being always out of tokens
1100597692 M * Bertl well, no actually the tokens are not related
1100597724 M * rs anyway, those vserver are really slow doing nothing
1100597973 M * Val Bertl : got 3 rejects when applying 2.4.27-vs1.29-q0.14 on debian patched ctx kernel (stable one : debianized vs1.29 patch) : in include/linux/ext2_fs.h line 294 #define EXT2_MOUNT_XATTR_USER 0x4000 instead of #define EXT2_MOUNT_TAG_XID 0x4000, same thing in include/linux/ext3_fs.h line 320 and in include/linux/fs.h line 115 #define MS_POSIXACL 65536 instead of #define MS_TAGXID (1<<24) ...  What should i choose between EXT2_MOUNT_TAG_XID and EXT3_MOUNT_TAG_XI
1100597990 M * Val And ... should i preserve MS_POSIXACL with MS_TAGXID define ?
1100598011 M * Bertl hmm, don't we have a debian version?
1100598020 M * Val yes we do
1100598027 M * Val without q0.14 :)
1100598033 M * Bertl ah, okay, missed that one
1100598038 M * Val arf :)
1100598051 J * id_ id@pD9E61B9D.dip.t-dialin.net
1100598057 Q * monrad Quit: Leaving
1100598096 M * Bertl Val: any new debian kernel sources/patches?
1100598112 M * Val none
1100598140 M * Bertl linux-2.4.27-3/ <-- current?
1100598148 M * Val i used patch you made for me for 2.4.27 that were packaged by Ola
1100598156 M * Val yes i use current one
1100598172 M * Bertl what I meant was, this is the current debian 2.4.27 version, right?
1100598178 M * Val am just trying to apply 2.4.27-q0.14 on it
1100598182 M * Val right
1100598186 M * Bertl okay, sec
1100598189 M * Val thx :)
1100598251 J * jsambrook jsambrook@aelfric.plus.com
1100598259 M * Bertl welcome jsambrook!
1100598470 Q * id__ Ping timeout: 480 seconds
1100598542 M * Val gona take a coffee... back in a few
1100598584 M * rs Bertl: in the meantime (waiting for uptime fix apply) is there a way to live fix uptime of buggy vservers ?
1100598687 M * Bertl not really, it is code which was left over in the soruce
1100599408 M * Bertl Val: looks like debian is adding code just for fun, or do you have any other explanation for this:
1100599411 M * Bertl grep -r EXT2_MOUNT_XATTR_USER *
1100599413 M * Bertl include/linux/ext2_fs.h:#define EXT2_MOUNT_XATTR_USER0x4000/* Extended user attributes */
1100599417 M * Bertl - nada
1100599420 M * Val :)
1100599444 M * Bertl so I'd say we remove that flag completely, but you have to test
1100599450 M * Val ok so let's trash EXT2_MOUNT_XATTR_USER
1100599457 M * Val i'll test it
1100599469 M * Bertl okay, patch is available in a minute
1100599487 M * Val may be it's part of POSIX extended attributes patch
1100599738 Q * cetex neutron.oftc.net jupiter.oftc.net
1100599738 Q * brc neutron.oftc.net jupiter.oftc.net
1100599738 Q * Doener neutron.oftc.net jupiter.oftc.net
1100599738 Q * Bertl neutron.oftc.net jupiter.oftc.net
1100599738 Q * Val neutron.oftc.net jupiter.oftc.net
1100599738 Q * Loki|muh neutron.oftc.net jupiter.oftc.net
1100599738 Q * mcp neutron.oftc.net jupiter.oftc.net
1100599738 Q * matti neutron.oftc.net jupiter.oftc.net
1100599738 Q * Plug neutron.oftc.net jupiter.oftc.net
1100599738 Q * BWare neutron.oftc.net jupiter.oftc.net
1100599738 Q * Zoiah neutron.oftc.net jupiter.oftc.net
1100599738 Q * Hollow neutron.oftc.net jupiter.oftc.net
1100599738 Q * click neutron.oftc.net jupiter.oftc.net
1100599738 Q * no_maam neutron.oftc.net jupiter.oftc.net
1100599738 Q * dsanta neutron.oftc.net jupiter.oftc.net
1100599738 Q * eyck neutron.oftc.net jupiter.oftc.net
1100599738 Q * virtuoso neutron.oftc.net jupiter.oftc.net
1100599738 Q * sebd neutron.oftc.net jupiter.oftc.net
1100599738 Q * Shotygun neutron.oftc.net jupiter.oftc.net
1100599738 Q * TheSeer neutron.oftc.net jupiter.oftc.net
1100599738 Q * ntrs neutron.oftc.net jupiter.oftc.net
1100599738 Q * cereal neutron.oftc.net jupiter.oftc.net
1100599738 Q * sannes neutron.oftc.net jupiter.oftc.net
1100599738 Q * lilo neutron.oftc.net jupiter.oftc.net
1100599738 Q * albeiro neutron.oftc.net jupiter.oftc.net
1100599738 Q * jsambrook neutron.oftc.net jupiter.oftc.net
1100599738 Q * ensc neutron.oftc.net jupiter.oftc.net
1100599738 Q * flock neutron.oftc.net jupiter.oftc.net
1100599738 Q * ndim neutron.oftc.net jupiter.oftc.net
1100599738 Q * pusling neutron.oftc.net jupiter.oftc.net
1100599738 Q * chand neutron.oftc.net jupiter.oftc.net
1100599738 Q * stupidawy neutron.oftc.net jupiter.oftc.net
1100599738 Q * Medivh neutron.oftc.net jupiter.oftc.net
1100599738 Q * mugwump neutron.oftc.net jupiter.oftc.net
1100599738 Q * v00dY neutron.oftc.net jupiter.oftc.net
1100599738 Q * daniel_hozac neutron.oftc.net jupiter.oftc.net
1100599738 Q * maharaja neutron.oftc.net jupiter.oftc.net
1100599738 Q * infowolfe neutron.oftc.net jupiter.oftc.net
1100599738 Q * sladen neutron.oftc.net jupiter.oftc.net
1100599738 Q * logger neutron.oftc.net jupiter.oftc.net
1100599738 Q * Snow-Man neutron.oftc.net jupiter.oftc.net
1100599738 Q * UFOczek neutron.oftc.net jupiter.oftc.net
1100599738 Q * berni neutron.oftc.net jupiter.oftc.net
1100599808 J * Zoiah Zoiah@matryoshka.zoiah.net
1100599808 J * BWare bware@212.26.196.41
1100599808 J * Plug plug@datadot.net
1100599808 J * matti matti@linux.gentoo.pl
1100599808 J * mcp hightower@81.17.110.148
1100599808 J * Loki|muh loki@satanix.de
1100599808 J * Val val@tatooine.loria.fr
1100599808 J * berni berni@obelix.ipv6.birkenwald.de
1100599808 J * jsambrook jsambrook@aelfric.plus.com
1100599808 J * ensc ircensc@ultra.csn.tu-chemnitz.de
1100599808 J * flock restless@l192-117-111-12.broadband.actcom.net.il
1100599808 J * ndim hun@helena.bawue.de
1100599808 J * virtuoso s0t0na@spb.sot.com
1100599808 J * sannes ace@home.skarby.no
1100599808 J * cereal cereal@ns1.starhosting.de
1100599808 J * logger rs@vds.pas-mal.com
1100599808 J * UFOczek ufoczek@hood.openbug.net
1100599808 J * albeiro albeiro@lexx.eu.org
1100599808 J * sladen paul@starsky.19inch.net
1100599808 J * ntrs ntrs@SP2-24.207.228.55.charter-stl.com
1100599808 J * eyck eyck@81.219.64.71
1100599808 J * dsanta santa@c68.190.156.105.roc.mn.charter.com
1100599808 J * no_maam erik@datenzone.de
1100599808 J * click click@dsl-84-161.aal.tiscali.no
1100599808 J * Hollow bene@home.xnull.de
1100599808 J * TheSeer theseer@border.office.salesemotion.net
1100599808 J * lilo lilo@elided.info
1100599808 J * Shotygun shotgun@shotygun.com
1100599808 J * sebd konversat@lns-th2-4f-81-56-247-131.adsl.proxad.net
1100599808 J * infowolfe infowolfe@ns30831.ovh.net
1100599808 J * maharaja maharaja@ipax.at
1100599808 J * daniel_hozac daniel@h212n1fls33o829.telia.com
1100599808 J * v00dY v00dy@62.241.52.143
1100599808 J * mugwump sv@210-54-92-188.ipnets.xtra.co.nz
1100599808 J * Medivh ck@paradise.by.the.dashboardlight.de
1100599808 J * stupidawy foo@you.wish.you.were.pimp.olicio.us
1100599808 J * chand chand@cancun.aspic.com
1100599808 J * Snow-Man sfrost@snowman.net
1100599808 J * pusling pusling@cpe.atm4-0-7285.0x50c44806.boanxx19.customer.tele.dk
1100599808 F * jupiter.oftc.net +o mugwump
1100599864 J * cetex cetex@h128n2fls31o259.telia.com
1100599864 J * Bertl herbert@janus.mc.tuwien.ac.at
1100599864 J * brc bruce@200165178107.user.veloxzone.com.br
1100599864 J * Doener doener@193.24.208.131
1100600737 Q * sebd uranium.oftc.net iridium.oftc.net
1100600737 Q * Shotygun uranium.oftc.net iridium.oftc.net
1100600737 Q * TheSeer uranium.oftc.net iridium.oftc.net
1100600737 Q * ntrs uranium.oftc.net iridium.oftc.net
1100600737 Q * cereal uranium.oftc.net iridium.oftc.net
1100600737 Q * sannes uranium.oftc.net iridium.oftc.net
1100600737 Q * virtuoso uranium.oftc.net iridium.oftc.net
1100600737 Q * lilo uranium.oftc.net iridium.oftc.net
1100600737 Q * albeiro uranium.oftc.net iridium.oftc.net
1100600805 J * virtuoso s0t0na@spb.sot.com
1100600805 J * sannes ace@home.skarby.no
1100600805 J * cereal cereal@ns1.starhosting.de
1100600805 J * albeiro albeiro@lexx.eu.org
1100600805 J * ntrs ntrs@SP2-24.207.228.55.charter-stl.com
1100600805 J * TheSeer theseer@border.office.salesemotion.net
1100600805 J * lilo lilo@elided.info
1100600805 J * Shotygun shotgun@shotygun.com
1100600805 J * sebd konversat@lns-th2-4f-81-56-247-131.adsl.proxad.net
1100600814 F * ChanServ -o mugwump
1100601380 Q * Val Quit: back in a few
1100601816 M * Bertl okay, off for dinner now ...
1100601828 N * Bertl Bertl_oO
1100602123 J * Val val@tatooine.loria.fr
1100603635 Q * jsambrook Ping timeout: 480 seconds
1100603935 J * jsambrook jsambrook@aelfric.plus.com
1100604274 P * jsambrook 
1100604371 N * Bertl_oO Bertl
1100604380 M * Bertl back now ...
1100604395 M * albeiro hey Bertl :)
1100604406 M * albeiro matti: are you with us ?
1100604444 M * eyck or are you against us?
1100604454 M * albeiro Bertl: matti came up with nice idea of localy 'shadowed' files
1100604471 M * Bertl hmm, please elaborate!
1100604474 M * albeiro what could be nice in vserver virtualization
1100604478 M * albeiro sure sir !
1100604511 M * albeiro basicaly what can you see in directory depends on context you are in
1100604532 M * albeiro so, user A can make file file1 in some_dir
1100604556 M * albeiro and other user B (in different context) can also make file1 in this dir
1100604573 M * albeiro but any of them would only see their own version of tyhis file
1100604600 M * Bertl hmm, well, that's neither easy nor really useful IMHO
1100604603 M * albeiro possible or just nightmare ?
1100604610 M * albeiro why it is not ?
1100604629 M * Bertl the unix filesystems are based on inodes
1100604649 M * Bertl and special inodes, called directories, contain a list of 'named' inodes
1100604659 M * Bertl that is what we know as files ...
1100604694 M * Bertl so having a different 'view' for each context, would actually mean storing one and the same file entry more than once in a dir inode
1100604719 M * Bertl which of course could be done by adding a prefix for example
1100604725 M * Bertl <xid>_<filename>
1100604752 M * Bertl but that in turn would mean that if somebody removes that dir, it would be gone for all contexts
1100604764 M * Bertl so you would actually need a dir per context
1100604784 M * Bertl and of course as dirs are hierarchical, you end up with a separate mount for each context ;)
1100604786 M * albeiro yeah, looks so
1100604799 M * Bertl which is something you can already do ;)
1100604809 M * albeiro it is not as simple as it seemed to be ;]
1100604826 M * Bertl now for the second part: the usefullness
1100604854 M * Bertl what would it buy you to have a common dir, when you want each context to have a differing view?
1100604890 M * albeiro well, i could have / dir common for all my vservers
1100604899 M * albeiro but with a different view
1100604939 M * Bertl which would help you in what way?
1100604956 M * albeiro i would not have to replicate "/" for all contexts
1100604976 M * Bertl what does replicate mean, and when do you do that?
1100605048 M * albeiro well (is has been a long time since using vserver) i mean that i would need to create separete dir structure for diferent vservers, much like normal chroot
1100605072 Q * berni uranium.oftc.net plasma.oftc.net
1100605085 M * Bertl and if somebody in context A removes an important file from your / view?
1100605150 M * albeiro he would remove its own version
1100605151 M * Bertl I guess what you are looking for is a lightweight 'copy' or 'copy on write' solution
1100605172 M * Bertl and the linux-vserver unification just does that
1100605179 M * albeiro duno exactly know what copy on write means :(
1100605187 J * berni berni@obelix.ipv6.birkenwald.de
1100605222 M * Bertl in the future a real copy on write (i.e. when you write to a 'shared' file it is automatically copied) solution will be available
1100605223 Q * Plug Quit: changing servers
1100605243 J * Plug plug@datadot.net
1100605264 M * albeiro and now it is ... ?
1100605282 M * Bertl unification i.e. hardlinks with a special protection
1100605311 M * Bertl do you know how hardlinks/unification works?
1100605349 M * Loki|muh is it possible, that the user in the server un-unifies itself?
1100605349 M * albeiro sorry, but not
1100605354 A * Bertl is just asking before you have to listen to my elaborations ...
1100605382 M * Bertl okay, everybody interested in unification, here is a simple example:
1100605382 M * Loki|muh 12:34 < Bertl> but that in turn would mean that if somebody removes that dir, it would be gone for all contexts
1100605385 M * Loki|muh oh
1100605387 M * Loki|muh ic. ;)
1100605408 M * Bertl get you favorite tar archive of whatever is big enough to be noticeable
1100605422 M * Bertl for this example, I take the kernel source ;)
1100605461 M * Bertl # df .
1100605468 M * Bertl                        31G   17G   14G  55% /opt
1100605498 M * Bertl                       31926160  17331076  14595084  55% /opt
1100605528 M * Bertl # tar xjf linux-2.6.9.tar.bz2
1100605597 M * Bertl                       31926160  17598088  14328072  56% /opt
1100605632 M * eyck ok
1100605669 M * albeiro so far so good ;)
1100605706 M * matti Hmm...
1100605732 M * Bertl okay .. now we do
1100605771 M * Bertl # cp -la linux-2.6.9 linux-2.6.9-copy
1100605780 M * Bertl                       31926160  17602220  14323940  56% /opt
1100605805 M * Bertl 4132k difference?
1100605827 M * albeiro small, because it was making hard links :)
1100605842 M * Bertl exactly ... that's the reason
1100605852 M * eyck so far so good...
1100605864 M * Bertl the directories are normal directories and the file entries are referring to the same inode
1100605885 M * Bertl # ls -la linux-2.6.9/MAINTAINERS 
1100605885 M * Bertl -rw-r--r--    2 500      500         54549 Oct 18 23:54 linux-2.6.9/MAINTAINERS
1100605905 M * Bertl the '2' shows that there are 2 references to the file
1100605951 M * Bertl writing to one reference (one 'view') of this inode will overwrite the data itself
1100605971 M * Bertl to protect against this, the immutable flag can be used
1100605990 M * Bertl # chattr +i linux-2.6.9/MAINTAINERS
1100605995 M * Bertl # lsattr linux-2.6.9/MAINTAINERS
1100605995 M * Bertl ----i-------- linux-2.6.9/MAINTAINERS
1100606010 M * Bertl # rm -f linux-2.6.9/MAINTAINERS
1100606010 M * Bertl rm: cannot remove `linux-2.6.9/MAINTAINERS': Operation not permitted
1100606047 M * Bertl now that will protect the contents of that file, but will not allow to remove and/or replace it ...
1100606077 M * Bertl that is the reason for the IUNLINK flag, which is part of linux-vserver kernel patch
1100606116 M * Bertl it basically inverts the effect of the IMMUTABLE flag for removal (unlink)
1100606147 M * Bertl so files with IMMUTABLE and IUNLINK set, can not be modified, but removed, and therefore replaced ...
1100606167 M * Bertl (removing one hardlink doesn't affect the other)
1100606216 M * albeiro (yeah, file is not "deleted" till you remove all of its hardlinks...)
1100606222 M * Bertl that is the method which is used to unify linux-vservers across different (but originally identical) vservers
1100606251 M * Bertl of course you can 'unify' arbitrary servers too, as long as there are 'common' files at all ...
1100606311 M * Bertl in the future, the limtation that you still can not write to such a file, will be removed by adding the 'copy on write', which will copy such a link when a change is likely to happen
1100606423 M * albeiro nice...
1100606737 M * albeiro and separate mount for each context ? how does it work ?
1100606781 M * Bertl well, you can mount a separate filesystem for each vserver easily ...
1100606796 M * Bertl (for example using lvm volumes or similar)
1100606894 M * albeiro ok, thank you for elaborate :)
1100606905 M * Bertl you're welcome!
1100606941 M * Bertl sidenote: tools like vunify and various scripts support unification of vservers and/or arbitrary dirs
1100606996 M * albeiro one more question: files tagging if only to protect them from beeing modyfied in another context ? but they are visible ?
1100607040 M * matti Heh.
1100607042 M * Bertl is already there, it's called tagxid and works for ext2/3 and reiser
1100607065 M * Bertl it's also used to keep track of the used diskspace per vserver
1100607447 M * matti Bertl: So, how about to use VFS for make such local namespace for files/directories? It can be "fs" independent?
1100607521 M * Bertl yes, that is correct, but the problem is that VFS uses dentries which, when created, instanciate the inode from the udnerlying filesystem
1100607556 M * Bertl so while it's perfectly fine to have different 'views' of the same filesystem
1100607573 M * Bertl it would not help with making those views persistent
1100607597 M * eyck hmm, 
1100607612 M * eyck doesen't LVM support copy-on-write volumes?
1100607618 M * Bertl examples of such uses are files which where removed, while still being used ...
1100607630 M * matti Yes.
1100607634 M * matti I know.
1100607634 M * Bertl eyck: yes, they do, but on the block level
1100607657 M * Bertl and this pretty soon ends up with a separate copy of everything ...
1100607674 M * eyck huh, that was my argument against using vunify ;) 
1100607697 M * Bertl but you can re-unify, how to do that with lvm?
1100607697 M * eyck hmm, so this lvm thing doesen't work that well in practice?
1100607925 M * albeiro lvm's copy on write is for snapshothing rathe thatn something else
1100608002 M * matti Bertl: But, if we do some scary magic with VFS, we can tell "fs" what we only want to tell :-)
1100608032 M * albeiro ;p
1100608038 M * Bertl sure, you could do arbitrary remapping, but it would be a hell of work, with not much gain ...
1100608051 M * matti Bertl: But it is possible, right?
1100608057 M * albeiro with some black magic and deep hackery ;]
1100608066 M * Bertl (for example the userfs does this, and erez zadoks stuff too)
1100608081 M * matti :-)
1100608100 M * matti So, my almost stupid idea is not so really stupid?
1100608121 M * eyck what is erez zadoks?
1100608192 M * Bertl erez is the guy who 'developed' FiST
1100608205 M * Bertl (a stackable filesystem template generator)
1100608269 M * Bertl http://www.filesystems.org/docs/fist-lang/fist.html
1100608279 M * eyck that's the one that creates modules on the fly?
1100608291 M * albeiro Bertl: have you heard (some time ago) that Hans Reiser is willing to implement "views" in his fs ?
1100608294 M * Bertl btw: http://www.filesystems.org/docs/ is very interesting ...
1100608298 M * eyck we're getting very close to self-compiling kernels...
1100608303 M * albeiro for amazing amount of money of course
1100608342 M * Bertl albeiro: yeah, for _amazing_ amounts of money hans is willing to implement your right shoe into reiser X
1100608351 M * matti Gezz... Hans Reiser is little bit crazy :-)
1100608361 M * eyck hmm, I think for amazing amount of money I might try running his 'fs' again
1100608365 M * matti Hehehehehehhehe.
1100608374 M * albeiro he is smoking some bad crack, no doubt
1100608387 M * albeiro reiserfs is very good on /tmp ;p
1100608398 M * albeiro it is like tmpfs ;p
1100608402 M * matti albeiro: Yep... And spender tooo -)
1100608425 M * albeiro matti: yeah, but you see - sspender's grsecurity mostly works :)
1100608429 M * albeiro oposite to reiserfs
1100608443 M * eyck here here,
1100608446 M * matti :]
1100608450 M * eyck or maybe hear hear?
1100608456 M * matti albeiro: It a really small difference :P
1100608457 M * Bertl the basic idea of reiserfs is not that bad .. but what it needs most is stabilization ...
1100608470 M * albeiro and cleaning of code, and rewriting...
1100608483 M * Bertl and it seems that the hunt for benchmarks doesn't help there ...
1100608487 M * eyck and getting rid of the 'leader'
1100608488 M * albeiro talking about reiserfs4
1100608496 M * matti eyck: Hehehehhe.
1100608517 A * albeiro is using xfs
1100608524 M * eyck who isn't ;) 
1100608524 M * Bertl eyck: have I introduced you to Pinky?
1100608540 M * eyck Bertl: Pinky is your new bot without brain?
1100608544 M * matti Pinky?
1100608553 M * Bertl eyck: yep, no Brain yet ;)
1100608568 M * eyck what does it do?
1100608568 M * albeiro heh
1100608582 M * matti Heheheh.
1100608599 M * Bertl eyck: http://irc.13thfloor.at/LOG/
1100608642 M * matti albeiro: Eh, I need a job...
1100608703 M * albeiro matti: sure you do. lets write hans "views" a bit cheaper than he
1100608727 M * matti LOL
1100608884 M * eyck Bertl: does this mean I will be on google?
1100608894 M * eyck saying bad things about reiserfs?
1100608898 M * eyck COOL.
1100608906 A * albeiro too !
1100608909 M * matti Hehehehe.
1100608970 M * albeiro some offtopic question: if my biosb shows properly amount of ram (i have 256 MB) and linux reports 254180k what can be wrong ?
1100608994 M * Bertl sounds fine ... did you consider the kernel memory?
1100609011 M * albeiro nein ;p
1100609029 M * albeiro why so much ?
1100609040 M * matti :-)
1100609067 M * Bertl check with 'dmesg | grep kernel'
1100609075 M * albeiro 256 - 248 = 8 MB for kernel
1100609109 M * albeiro (standard input):Memory: 251980k/262080k available (2054k kernel code, 7672k reserved, 303k data, 156k init, 0k highmem)
1100609112 M * eyck 8M ain't that much..
1100609130 M * eyck unless you're trying to boot on 8M machine
1100609146 M * albeiro er...ok
1100609166 M * eyck in that case you should consider 2.0. kernel,
1100609182 M * eyck for which new version was released recently
1100609183 M * albeiro i always wondered what this where can i find detailed infromation about kernekl memory managment
1100609194 M * albeiro wrr, lags
1100609197 M * Bertl well, it's just a matter of configuration ...
1100609209 M * eyck not really, 
1100609209 M * Bertl my iPAQ boots fine with 2.6.9
1100609220 M * albeiro what have you done to him ? ;p
1100609223 M * eyck you can tear some stuff off 2.6.x , 
1100609234 M * eyck but it'll still be bulkier then 2.0
1100609287 M * albeiro ok, second try, without lags: do you mayby know where can i find detailed infromation about kernekl memory managment ? (except source of course ;)
1100609323 M * eyck there are books...
1100609337 M * Bertl are you interested in vm memory management of the linux-kernel, or what direction are you heading?
1100609382 M * albeiro rather generic knowledge of how it all works
1100609400 M * albeiro and after generic, some more specyfic too...
1100609731 M * Bertl http://kerneltrap.org/node/view/3160
1100609731 Q * cetex Remote host closed the connection
1100609753 M * Bertl http://home.earthlink.net/~jknapka/linux-mm/vmoutline.html
1100609841 M * Bertl http://www-106.ibm.com/developerworks/linux/library/l-inside.html
1100609886 M * Bertl but the 2.6 VM is still changing ...
1100609969 M * maharaja hi folks
1100609979 M * Bertl welcome maja!
1100609991 M * maharaja noticed your kernel talk ... did any1 of you notice a performance decrease with 2.6.8 and higher?
1100610018 M * maharaja somehow, i get the feeling that the other kernels used to give a little more performance
1100610322 M * Bertl hmm, not really ... 
1100610557 M * maharaja ok
1100610568 M * maharaja (btw. im not talking vserver specific, but in general)
1100610587 M * albeiro thx for links, have to go out nw
1100610668 M * Bertl cya
1100610694 M * Bertl maharaja: yeah, but I didn't notice any regression in performance so far ...
1100611483 M * Bertl Val: any results so far?
1100612458 M * matti Eh.
1100612471 A * matti need some motherboard with SLOT1.
1100612472 M * matti :]
1100612712 M * Bertl http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&category=1244&item=6721094675&rd=1
1100612788 J * mhepp mhepp@r72s22p13.home.nbox.cz
1100612808 M * Bertl welcome mhepp!
1100612826 Q * Val Quit: My damn controlling terminal disappeared!
1100612892 M * mhepp Bertl: hello
1100613567 P * ni 
1100613808 M * maharaja bertl: thnx
1100613839 M * Bertl np
1100615175 Q * mhepp Remote host closed the connection
1100616154 M * matti :]
1100617891 J * Val val@gj403.loria.fr
1100617901 M * Bertl wb Val!
1100617924 M * Val sorry lots of meetings today :(
1100617933 M * Val another one in a ew minutes...
1100617936 M * Val +f
1100617973 M * Val kernel is done but i have no time now to test it
1100617988 M * Val i'll continue this night :)
1100618060 M * Bertl okay, I have a patch for q0.14 too, so you might compare
1100618108 M * Val ok
1100618138 M * Val (meeting time...)
1100619225 J * meebey meebey@meebey.net
1100619235 M * Bertl welcome meebey!
1100619246 M * meebey hiya
1100619252 Q * Val Quit: My damn controlling terminal disappeared!
1100619334 M * Doener morning folks!
1100619346 M * Bertl morning Doener!
1100619452 J * Shuri Dew@dsl.speedline209.226.electronicbox.net
1100619827 M * Shuri hi
1100619923 M * Bertl welcome Shuri!
1100620185 M * Doener hm, porting to rc2 may take some time...
1100620196 M * Bertl ah?
1100620202 M * Bertl haven't had a look yet ...
1100620409 M * Doener doener@doener ~/kernel/linux-2.6.10-rc2-vs1.9.3 $ patch -p1 --dry-run -F0 -s < ../patch-2.6.10-rc2-vs1.9.3.diff | wc -l
1100620409 M * Doener 26
1100620440 M * Bertl and without -F0 ?
1100620458 M * Bertl and -l instead?
1100620508 M * Doener 16
1100620525 M * Doener -l makes no difference
1100620557 M * Bertl okay ...
1100620800 J * Val val@gj403.loria.fr
1100620802 Q * Val Quit: 
1100620834 J * Val val@gj403.loria.fr
1100622943 J * monrad monrad@213083190130.sonofon.dk
1100622953 M * Bertl welcome monrad!
1100622961 M * monrad hi
1100623005 M * monrad i should be ready to design a RISC (MIPS) cpu in 14 days
1100623043 M * Bertl a great, when will you be able to _build_ one?
1100623083 M * monrad well we are going to implement it some software
1100623112 M * Bertl ah, mips style cpu in software emulation?
1100623175 M * monrad yes in kreds some software made at our university
1100623183 M * monrad based on mosml
1100623197 M * Bertl hmm, maybe you start working on qemu later ;)
1100623224 M * rs re
1100623229 M * Bertl wb rs!
1100623266 M * monrad well we are going to make our own OS later so i might be a kernel dev some day too
1100623280 M * Bertl excellent ...
1100623523 M * monrad i think its quite hard when one of the first slides just says: "1000110010100000" and so on
1100623935 M * Bertl well computer sciences is not just kernel hacking ...
1100624019 M * monrad and it is adding two registres together
1100624263 M * Doener Bertl: hmm... vanilla asm-m68k/unistd.h defines __NR_vserver as 278, while we were using 273
1100624280 M * Bertl yep, that is korrect
1100624283 M * Bertl correct even
1100624301 M * Doener ok, so that hunk can be removed, right?
1100624305 M * Bertl vserver syscall was recently added for m68k
1100624332 M * Bertl yes, that and probably the syscalls.* stuff has to be adjusted
1100624657 M * Doener what number to use for m68knommu? 273 is now taken by something else, NR_syscalls is 282
1100624691 M * Doener i.e. do you have any preferences there, or just use 282?
1100624880 M * Bertl 278 should be assigned for all m68k*
1100624893 M * Bertl but let me check that once again ...
1100624930 M * Doener is setaltroot is vanilla for m68knommu
1100625129 M * matti :]
1100625170 M * Bertl Doener: I only have a definite assignment for m68k in my mailbox ... so maybe we should remove it for m68knommu for now?
1100625235 M * Doener ok. i guess 278 will be free again in 2.6.10, they started removing setaltroot (which was added in 2.6.10-rc1)
1100625434 M * Bertl from Geert:
1100625443 M * Bertl It's 278, cfr. the patch below.
1100625443 M * Bertl Greg: I decided to put if after waitid, since that's the newest syscall
1100625443 M * Bertl in 2.6.9.
1100625481 M * Doener i meant m68knommu
1100625492 M * Doener hm, we never added sys_vserver to entry.S for m68k
1100625499 M * Doener at least not in 1.9.3
1100625511 M * Bertl probably because nobody tested on amiga yet ;)
1100625547 M * Doener shall i add it?
1100625560 M * Bertl yeah, why not ...
1100625647 M * Bertl cross compiling for m68k is working here, so we can check (if we find a suitable .config)
1100626755 J * DuckMaster Duck@dyn-83-155-18-154.ppp.tiscali.fr
1100626790 M * Bertl welcome DuckMaster!
1100627182 Q * DuckKing Ping timeout: 480 seconds
1100628282 M * Hollow lo all!
1100628305 M * Bertl hey Hollow!
1100628424 M * Hollow fuckin' dentist.... my wisdom teeth must be operated :(
1100628845 M * Bertl don't talk about dentists to me, I'm overdue ...
1100628856 M * Hollow heh
1100630472 M * Bertl okay, off now, back later ...
1100630483 N * Bertl Bertl_oO
1100631404 Q * rs Quit: leaving
1100631599 M * Loki|muh anyone used a gentoo inside a vserver here?
1100631653 M * Doener i bet Hollow has ;)
1100631747 M * monrad i got one working from following Hollow's guide
1100631761 M * Loki|muh monrad: do you have the url?
1100631777 M * Loki|muh or: is there an url for the guide? ;)
1100631800 M * monrad http://oss.croup.de/vserver/guide/
1100631811 M * Loki|muh thx
1100631820 M * monrad np
1100631830 M * monrad time for some sleep night guys
1100631891 Q * monrad Quit: Leaving
1100632122 M * Hollow Loki|muh: i'm just working on some nice init script for gentoo :)
1100632220 M * Loki|muh Hollow: fine :) im trying just now :)
1100632322 M * Loki|muh they are nice ;)
1100632402 M * Loki|muh Failed to start vserver 'vserver44'
1100632404 M * Loki|muh :(
1100632420 M * Hollow you did my howto?
1100632454 M * Loki|muh yeah
1100632461 M * Hollow any error message?
1100632497 M * Hollow but wait... how could you do my howto in 10 minutes? :)
1100632528 M * Loki|muh i have a ready ubuntu host here
1100632536 M * Loki|muh there are other vservers running
1100632538 M * Hollow ah...
1100632549 M * Loki|muh and i have a prepared gentoo-chroot ;)
1100632618 M * Loki|muh so I did only re-emerge of the baselayout and etc-update
1100632620 M * Hollow so again: any error message?
1100632623 M * Hollow mhm
1100632699 M * Hollow what does ls /vservers/vserver44/var/lib/init.d/|wc -l say?
1100632764 M * Loki|muh 6
1100632785 M * Hollow then first do rm -rf /vservers/vserver44/var/lib/init.d/*
1100632908 M * Loki|muh and then another re-emerge?
1100632921 M * Hollow no
1100632924 M * Hollow try again
1100632953 M * Loki|muh same error
1100632982 M * Hollow *shrug* if theres no error message from vserver script i dunno
1100633007 M * Hollow i'm relativly new to vserver, but this sounds not like a gentoo-guest configuration issue...
1100633014 M * Hollow perhaps some host configuration issue
1100633029 M * Loki|muh |>root@pc13:/vservers # vserver vserver44 start
1100633029 M * Loki|muh |>No command given; use '--help' for more information.
1100633033 M * Loki|muh hmmm
1100633050 M * Hollow hu? which version of util-vserver?
1100633147 M * Loki|muh util-vserver-0.30.195
1100633186 M * Loki|muh hmmm
1100633193 M * Doener looks like start command is not given...
1100633196 M * Loki|muh where does is_server_guest come from?
1100633208 M * Doener vserver --debug vserver44 start
1100633208 M * Hollow /sbin/functions.sh
1100633215 M * Doener and make the output available
1100633215 M * Hollow erm
1100633231 M * Hollow you mean is_vserver_guest?
1100633340 M * Loki|muh yeah, found it, thanks
1100634203 J * tribble foobar@u6-59.dsl.vianetworks.de
1100634206 M * tribble hi
1100634260 M * tribble i've a problem, my load is around 15, klogd uses much of the cpu power. i'm running 2.6.9-vs1.9.3 with util-vserver-0.30.196
1100634700 M * Loki|muh Hollow: in debian /etc/init.d/rc is used for starting a vserver. which would be the appropriate command to start a gentoo vserver?
1100635245 M * Loki|muh /sbin/rc default i guess
1100635387 M * Doener tribble: did you unhide proc entries? if not, debian klogd goes crazy...
1100635401 M * Doener it's not necessary/useful to run klogd in a vserver anyways
1100635494 M * Doener Loki|muh: if you set initstyle to gentoo, the right one should be used
1100635697 M * Loki|muh Doener: where can I set it to gentoo? I searched in /usr/sbin/vserver but there is no STARTCMD
1100635733 M * Doener vserver calls/uses other scripts. 
1100635734 M * Doener http://www-user.tu-chemnitz.de/~ensc/util-vserver/doc/conf/configuration.html
1100635747 M * Hollow Loki|muh: echo "gentoo" > /etc/vservers/vserver44/apps/init/style
1100635838 M * Loki|muh |>Give root password for maintenance
1100635838 M * Loki|muh |>(or type Control-D for normal startup): 
1100635841 M * Loki|muh yeah :)
1100635854 M * Loki|muh I guess now I can proceed with the howto
1100635902 M * Hollow hehe
1100635965 M * Loki|muh I didnt find this in http://oss.croup.de/vserver/guide/
1100636073 M * Doener that's part of the skeleton creation
1100636095 M * Doener Code listing 4.1 line 4
1100636234 M * Loki|muh ah! thx :)
1100636423 M * Loki|muh it works :)
1100636432 M * Loki|muh Hollow: you're great
1100636437 M * Loki|muh Doener: thx for help :)
1100636459 M * tribble Doener: hm, yep. but after i started the vservers ;)
1100636472 M * tribble so restarting klogd should help?
1100636482 M * Doener stopping it as well ;)
1100636503 M * Doener it does pretty much nothing in a vserver AFAIK
1100636916 Q * infowolfe Ping timeout: 480 seconds
1100637470 J * infowolfe infowolfe@ns30831.ovh.net
1100639931 M * tribble hm
1100639961 M * tribble i saw, that on my server some "servers" or something like that are running
1100639972 M * tribble but they've no name but a klogd inside :)
1100639980 M * tribble how can i kill them?
1100640067 M * tribble contextnumbers are between 45001 and 45006
1100640098 M * Doener "vps -A" to get the pids, then just kill from context 0
1100640355 M * tribble thx
1100641041 J * kalou kalou@AToulon-201-1-24-32.w80-13.abo.wanadoo.fr
1100641044 M * kalou Hi
1100641084 M * kalou snif .. Chief Project Manager Bertl sleeping :)
1100641141 M * tribble Doener: 49xxx contexts belongs to the rootsystem, nor?
1100641202 M * tribble or is the rootsystem only one context?
1100641214 M * Doener all above 49xxx (don't remember the exact number) are dynamic context, i.e. you get one of those if you don't specify a context number.
1100641243 M * Doener they are no less secure than static contexts, but IMHO a little harder to administer
1100641264 M * tribble hm
1100641369 A * tribble doesn't if he should kill the processes in it to minimize the cpu load and after that start all the vservers again without rebooting the rootsystem
1100641400 M * Doener please rephrase
1100641426 M * Doener kalou: do you need anything?
1100641460 M * tribble i'll kill all processes that belongs to vservers, my vservers all have static ids. but i don't know to handle 49152-49156
1100641476 M * tribble i'll minize the cpu load without rebooting
1100641769 Q * ntrs Read error: Connection reset by peer
1100641810 M * kalou Hi Doener.. In fact I'm struggling with "Operation not permitted" (chattr +iu in the host vserver, 2.4.27 vs1.28, util-vserver 0.30)
1100641825 M * tribble got it :)
1100641831 M * tribble thx4all Doener, good night :-)
1100641833 Q * tribble Quit: gn8
1100641897 A * Doener wonders if +iu is the right thing to do...
1100641909 M * Doener anyways, check current settings with lsattr
1100641940 M * Doener i guess you're trying to unify, right?
1100641959 M * kalou yep, by hand at the first time
1100641999 M * kalou someone told me that chattr +iu was not supported, and that maybe I could only use vfiles for this ..
1100642004 M * kalou Is it right or wrong ?
1100642054 M * kalou lsattr gives i and u .. Trying to -iu gives operation not permitted
1100642074 M * mugwump u != unify :)
1100642088 M * Doener ah, hi mugwump!
1100642096 M * mugwump hi doener, how's tricks?
1100642115 A * Doener still lacks knowledge about unifications
1100642145 M * kalou with i+u, I should be able to: remove a file inside a vserver
1100642153 M * kalou not able to change it
1100642158 M * Doener hacked alpha tools work fine except for the fact, that mounts appear in host and vserver namespace
1100642172 M * mugwump you can't set that `u' with the standard `chattr'
1100642193 M * Doener kalou: try +it ... IIRC...
1100642215 M * kalou okay ... lol ... what a mess :-) 
1100642215 M * mugwump ok, +it used to be the right thing to do, when immutable was the same bit as tail merge
1100642233 M * mugwump it may still work on reiserfs, as the bits are IIRC the same there still
1100642246 M * Doener mugwump: means that f.e. /vservers/xxx/proc etc. show up in the host namespace as well and won't go away until you umount it yourself
1100642257 J * rs rs@imhotep.rhapsodyk.net
1100642271 M * Doener hm, he's on vs1.28, that the stuff really change there?
1100642273 M * rs re
1100642277 M * Doener wb rs!
1100642291 M * mugwump no, vs1.28, use +it, should be fine
1100642312 M * kalou what does IIRC mean, sorry ?
1100642322 M * Doener if i remember/recall correctly
1100642413 M * kalou lol okay !
1100642433 M * mugwump kalou: you could try http://utsl.gen.nz/unify-dirs.vs1.2x, but test it works first like this:
1100642450 M * mugwump cd /vservers/
1100642452 M * mugwump mkdir test
1100642452 M * mugwump cd test
1100642454 M * mugwump mkdir a b
1100642458 M * mugwump cp /etc/passwd a
1100642459 M * mugwump cp /etc/passwd b
1100642470 M * mugwump unify-dirs.vs1.2x -vil a b
1100642514 M * kalou ok will try this just after a little coffee ..
1100642542 A * mugwump has been told off before for using /etc/passwd as a temporary scratch data file
1100642577 M * Doener http://doener.homeip.net/doener/vserver/util-vserver-0.30.196-clean-namespace-test5.diff
1100642581 M * Doener if you dare ;)
1100642591 M * Doener probably breaks some setups
1100642616 M * Doener should work when migrate is available, and namespaces are used
1100642645 M * Doener everything except start/stop/suexec/exec/enter is untested
1100642653 M * mugwump Will that work against 1.9.3 doener ?
1100642663 M * Doener that's against 0.30.196 ;)
1100642723 M * Doener does not need any kernel changes
1100642770 M * Doener and "of course" it breaks adding mounts to the vserver namespace
1100642770 M * mugwump ok I'll give it a crack later on.  So, what works?  Everything except outside admin of the namespace I presume
1100643846 Q * infowolfe Ping timeout: 480 seconds
1100643886 N * Bertl_oO Bertl
1100643931 M * Bertl okay, got later than expected ... so I'm off for today ...
1100643938 M * Bertl have a nice one everyone!
1100643943 N * Bertl Bertl_zZ
1100644052 M * Doener mugwump: start/stop/enter etc. works, all flags etc. are set as usual, usual fstab mounting etc. happens, you just don't have the host tools at hand, if you enter the vserver namespace
1100644446 M * Hollow i setup a svn repos with my guide and ebuilds in it @ http://home.xnull.de/svn/vserver/ URL, ChangeLogs are in the guide too
1100644453 J * infowolfe infowolfe@ns30831.ovh.net
1100644464 M * Hollow hi infowolfe 
1100644712 M * kalou mugwump ... still online ? why is unify-dirs asking for Digest::SHA1 ?
1100644846 M * mugwump oh, it uses that to check that two files are the same
1100644906 M * kalou oki .. Makes me think of a perfect daemon that would do the following:
1100644941 M * kalou automatic unification while running ... compare checksums of files, and unify them without any user interraction..
1100644953 M * mugwump just run it out of cron :)
1100645009 M * kalou yep indeed.
1100645011 M * mugwump also, it's probably a good idea to only use unify-dirs on certain directories, like /usr, /lib, /sbin, /bin
1100645041 M * mugwump lots of programs expect to be able to open files in /etc and /var for writing :)
1100645159 M * kalou ok ... In fact unify would be perfect to take place at "ln " stage
1100645189 M * kalou why should we link files and then unify where we just could link ?
1100645390 M * mugwump linking is unifying, in a sense
1100645414 M * mugwump the only extra thing is the permissions that make the file immutable (but still removable)
1100645578 M * kalou hmmm. cp is using a mask when a file is copied
1100645607 M * kalou maybe this could be used to get unification while cp -ral when desired ?
1100645618 M * mugwump I've got a script for that too :)
1100645668 M * mugwump utsl.gen.nz/immucp.vs1.2x
1100645674 M * mugwump use it like this:
1100645683 M * mugwump immucp -il directory1 directory2
1100645700 A * mugwump should really put them on CPAN or something
1100645704 M * kalou hey hey .. nice repository ... everything seems to be hidden at this time .. do you need someone to do the documentation stuff ?
1100645749 M * kalou hmm I *love* immucp
1100645829 M * kalou or maybe simply ask for them to be included in utils-vserver ?
1100645853 M * mugwump the util-vserver is C and shell only
1100645875 M * mugwump I think the best form to distribute them would probably be a CPAN module
1100645997 M * mugwump but, I need to add code to detect which bit is in use for immulink...
1100646119 M * kalou Hmmm. I'm getting some Immutable, No Tail_merging dirs, some not
1100646233 M * kalou ok .. Files are like this, and directories are normal
1100646458 M * mugwump arse.  that's probably a bug.
1100646509 M * kalou Hmm. I have to say that I did not dig into the immutable iunlink directories questions yet :)
1100646532 M * mugwump ok.  I did have a bug like that at one point, but that version should be fixed
1100646547 M * kalou ok ready to help you debug
1100646628 M * mugwump ok, so the bug is: immucp is making +it directories ?
1100646645 M * Hollow i'm off to bed, cu!
1100646691 M * kalou cu Hollow
1100646722 M * kalou nope, immucp -il is making -------- directories
1100646760 M * kalou (---)
1100646847 M * mugwump according to lsattr?  that's expected
1100646877 M * kalou ok so no bug. We probably mis understood earlier.
1100646896 M * mugwump probably.
1100646906 M * kalou Just reported that files are put as +it and directories without no flag.
1100646949 M * kalou Ok very fine so.. Thank you very much for this great tool. I'll dream happy tonight !
1100646950 M * mugwump without _any flags_ :-)
1100646976 A * kalou is going to bed ! 
1100646981 M * mugwump a bientôt