1393203768 Q * undefined Ping timeout: 480 seconds
1393203792 J * undefined ~undefined@00011a48.user.oftc.net
1393203802 Q * undefined 
1393204947 J * robmc ~robmc@198.164.160.214
1393211770 N * Bertl_zZ Bertl
1393211776 M * Bertl back now ...
1393218832 J * undefined ~undefined@00011a48.user.oftc.net
1393228274 J * VirMan ~oftc-webi@91.218.79.125
1393228312 M * VirMan hello
1393228336 M * Bertl hey
1393228462 M * VirMan quick question - when i try to mount /vservers on ext4 with xattrs i got wrong option
1393228477 Q * Guy- Read error: No route to host
1393228480 M * VirMan i mean i can;t use the same flags as on reiserfs (old server)
1393228502 J * Guy- ~korn@elan.rulez.org
1393228505 M * VirMan and i have every option (except debugging) for ext4 marked
1393228542 M * VirMan when using tag option i got werid messages when setting disk lmits
1393228546 M * Bertl reiserfs and ext4 are very different, so I'm not surprised the same options don't work
1393228595 M * VirMan is it possible to use xid limits on ext4 or i need to use reiserfs ?
1393228649 M * Bertl xid limits as in Linux-VServer disk limits?
1393228657 M * VirMan yes
1393228667 J * Ghislain ~aqueos@adsl1.aqueos.com
1393228669 M * VirMan dlimits/root/space_total
1393228670 M * Bertl depends on the kernel and Linux-VServer patch
1393228701 M * VirMan 3.9.10-vs2.3.6.5  / 0.30.216-pre3038
1393228828 M * Bertl what are your mount options and what error do you get?
1393228836 M * Bertl what is in dmesg afterwards?
1393228870 M * VirMan fstab : /dev/sda2<-----><------>/vservers<----->ext4<--><------>defaults,noatime,tag<--><------>0 2
1393228908 M * VirMan vdlimit: vc_set_dlimit(): Invalid argument
1393228920 M * VirMan when executing vserver start
1393228931 M * Bertl so /vservers did get mounted?
1393228948 M * VirMan only works with tag option
1393228980 M * Bertl well, that's fine, no?
1393229004 M * Bertl check with 'grep vservers /proc/mounts'
1393229006 M * VirMan well i cant start vserver ;)
1393229021 M * VirMan with disk limits
1393229064 M * VirMan output : /dev/sda2 /vservers ext4 rw,tag,noatime,data=ordered 0 0
1393229084 M * Bertl so far so good, what tagging scheme do you use?
1393229133 M * VirMan ok - you got me.. - i just added tag to kernel options and i'm trying to use disk limits, putting in dlimits figures i desire.
1393229138 M * VirMan so i have no idea
1393229159 M * Bertl okay, do you have a kernel .config?
1393229165 M * VirMan yes
1393229174 M * Bertl then let's upload that to a pastebin
1393229174 M * VirMan pastebin ?
1393229182 Q * undefined Quit: Closing object
1393229284 M * VirMan http://pastebin.com/7jUf1DwR
1393229290 M * VirMan um.. not complete
1393229376 M * VirMan http://pastebin.com/z1xUXvFv
1393229382 M * VirMan hould be better
1393229479 M * Bertl okay, looks good so far, you are using the ID24 tagging scheme
1393229492 M * VirMan default nothing touched
1393229499 M * VirMan so should be ok
1393229506 M * Bertl I also just checked that ext4 works fine with 3.9.11-vs2.3.6.5
1393229559 M * Bertl do you have a device which is currently unused on that server?
1393229568 M * VirMan oficial util-vserver is bit newer than mine, but there is no newer ebuild on gentoo 
1393229571 M * Bertl i.e. which doesn't matter if it gets reformatted
1393229593 M * VirMan well.. that is a problem ;) - i did not assumed that i'd have such case
1393229605 M * Bertl okay, no problem, let's create one
1393229612 M * VirMan i can try to add a drive and configure raid0 on it
1393229630 M * VirMan system on file ?
1393229674 M * VirMan btw - is there any method of creating file without writing all data to it ? like just allocating file ?
1393229710 M * Bertl dd if=/dev/zero of=/path/to/test.img bs=1M seek=1024 count=1
1393229726 N * Guest755 JvdK
1393229735 M * Bertl then use losetup /dev/loop0 /path/to/test.img
1393229768 M * VirMan yes - btw - it writes those zeros anyway :)
1393229783 M * Bertl it only writes the last block
1393229799 M * Bertl unless you have sparse files disabled
1393229812 M * VirMan hmmm - that i do not know.
1393229852 M * VirMan ok now what - format this with ext4 ?
1393229887 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testfs.sh
1393229895 M * Bertl grab this, and run it like this:
1393229949 M * Bertl testfs.sh -n -xyz -D /dev/loop0 -F ext4
1393229960 M * Bertl then upload the output to a pastebin
1393230057 M * VirMan http://pastebin.com/FckwTmvs
1393230350 M * Bertl so that looks fine as well
1393230380 M * Bertl I'd suggest to manually update util-vserver just to check that it isn't a problem with the ebuild
1393230534 M * VirMan even though there are failed tests ?
1393230562 M * VirMan wher to get newer vserver-tools ?
1393230581 M * Bertl yes, the failed tests are deviations from the accounted values which are normal for ext4
1393230589 M * Bertl as ext4 does some pre-allocations
1393230602 M * Bertl as long as they do not deviate too much it's okay
1393230613 M * Bertl (you see both values on the right_
1393230614 M * Bertl )
1393230687 J * undefined ~undefined@66-190-97-211.dhcp.unas.tx.charter.com
1393230708 M * fback morning Bertl :)
1393230716 M * VirMan ok.
1393230817 M * fback have you noticed privmsg I sent you last thursday?
1393230930 M * Bertl yes I read it, but I didn't fully understand it :)
1393230944 M * Bertl basically it's about the change from gz to xz, yes?
1393231067 M * VirMan sorry to bother - i downloaded latest util-vserver, compiled and "make install"' ed but it still shows old version :(
1393231090 M * VirMan should i replace old files - where are new ones?
1393231091 M * Bertl most likely you didn't configure them to your path setup
1393231106 M * Bertl so they probably endet up in /usr/local or so
1393231130 M * Bertl firt run uninstall, then change the pathes, then uninstall the old tools and reinstall the new ones
1393231742 M * VirMan ok i edited pathconfig.h but after reinstalling nothing changed
1393231755 M * VirMan i checked where original files were and installed it.
1393231841 M * VirMan ahh sory - when i uninstalled old one, and tried install new - it is missing files..
1393231876 M * daniel_hozac pathconfig.h is not used for basically anything
1393231884 M * daniel_hozac set your paths with ./configure
1393231894 M * VirMan ok..
1393231967 M * VirMan how ?
1393231980 M * VirMan it isrunning on automatic
1393232504 M * hijacker VirMan,  perhaps ./configure --help might give you some hints?
1393232570 M * VirMan ok... then a lot of options.. 
1393232579 M * hijacker aye
1393232587 M * hijacker you might be interested in those few... /configure --prefix=/usr/local --sysconfdir=/etc --with-apis=NOLEGACY --with-vrootdir=/data/vservers --with-initscripts=sysv --enable-dietlibc
1393232608 M * hijacker just tweak the variables according to your needs
1393232611 M * Bertl where --prefix probably will be /usr
1393232648 M * Bertl (and --with-vrootdir=/vservers at least that's the default)
1393234016 M * VirMan ok trying different
1393236253 M * VirMan why NOLEGACY ?
1393236266 M * daniel_hozac do you have a 2.4 kernel?
1393236273 M * VirMan nope
1393236280 M * VirMan 3.9 or sth
1393236301 M * VirMan and what gives --with-initscripts=sysv
1393236328 M * daniel_hozac then you don't need the legacy APIs.
1393236353 M * VirMan --sysconfigdir=/ ? it adds etc/ at end ?
1393236361 M * VirMan ok.
1393236503 M * VirMan ./configure --bindir=/usr --sbindir=/usr --libexecdir=/usr --sysconfigdir=/ --withlocalstatedir=/ --libdir=/usr --with-apis=NOLEGACY --with-vrootdir=/vservers
1393236503 M * VirMan ?
1393236551 M * daniel_hozac you want all of your files right in /usr?
1393236565 M * daniel_hozac and config in /?
1393236570 M * VirMan nope - but in manual there is stated :
1393236581 M * VirMan --bindir=DIR            user executables [EPREFIX/bin]
1393236589 M * VirMan so it added bin/ ???
1393236604 M * daniel_hozac to your configured prefix.
1393236605 M * VirMan and it goes /usr/bin
1393236617 M * daniel_hozac so if you set --prefix=/usr, you get /usr/bin
1393236651 M * VirMan hmm i guess ?
1393236665 M * VirMan thi is how it is written in --help
1393236689 M * VirMan or... not ?
1393236714 M * VirMan i mean - one needed to set this - so author should know- whoever it is :)
1393236715 M * distemper you want to do a "./configure --prefix=/usr" skip the rest ;)
1393236769 M * daniel_hozac no, you wan at least --localstatedir=/var and --with-apis=NOLEGACY as well.
1393236779 M * daniel_hozac if you are on Gentoo, you probably should have --with-initscripts=gentoo too.
1393236847 M * VirMan yes i use gentoo
1393236990 M * VirMan ok i guess documentation is bad - it requires full paths it does not add those suffixes mentioned in --help
1393237026 M * daniel_hozac nowhere does it say that it adds suffixes.
1393237042 M * daniel_hozac it says the default is a suffix after prefix.
1393237057 M * VirMan so how would you understand this line :  --sbindir=DIR           system admin executables [EPREFIX/sbin]
1393237100 M * VirMan that through --sbindir i add DIR variable which is placed in EPREFIX  giving DIR/sbin
1393237329 M * VirMan ok i made : ./configure --bindir=/usr/bin --prefix=/usr --sbindir=/usr/sbin --libexecdir=/usr/libexec --sysconfdir=/etc --localstatedir=/var --libdir=/usr/lib --with-apis=NOLEGACY --with-vrootdir=/vservers --with-initscripts=gentoo
1393237344 M * distemper err, sorry my bad, i meant: don't mess with bindir/sbindir and so on, just configure the prefix... btw, there's nothing wrong with the current util-vserver ebuild on gentoo, imho
1393237367 M * VirMan i got this : util-vserver-0.30.216-pre3054 # vserver vs6018 start Can not find a vserver-setup at '/usr/local/etc/vservers/vs6018/'.
1393237390 M * VirMan where it got this /usr/local ???
1393237412 M * VirMan well i got problem starting guest with dlimits
1393237421 M * VirMan no messages in dmesg
1393237423 M * VirMan using ext4
1393237432 M * VirMan so i was suggested by bertl to get newest
1393237497 M * Bertl just a hint, here (with the default setup) all I need to specify for ./configure is --prefix=/usr
1393237507 M * Guy- hi
1393237515 M * Bertl you also get a list of pathes once configure completed to check
1393237538 M * Guy- anyone tried 3.13.5 with vs2.3.6.10 yet, or will I be breaking new ground if I do? :)
1393237546 M * distemper if your a bit familiar with gentoo, you can rename the ebuild to the latest version and put it in a local portage dir
1393237596 M * Guy- the incremental patches seem to apply pretty cleanly (with only minimal fuzz) over 3.13.1+vs2.3.6.10
1393237616 M * VirMan i once did that... we'll see
1393237652 M * VirMan anyway it is strange that i specified all paths including --prefix and it still plays crazy
1393237687 M * distemper Guy-: i'm running it, works fine... i also grabbed this patch from the mailing list (http://list.linux-vserver.org/archive?mss:6511:201402:filjmhgamabjmadaafcm), because i hit that, too
1393237709 M * distemper Bertl: any comment on that patch?
1393237773 M * Bertl hmm, looks good, will double check shortly and update
1393237780 M * Guy- distemper: oh, thanks
1393237782 M * distemper ok, thanks :)
1393237850 M * Bertl was that patch sent to the Linux-VServer ML? did I just miss it?
1393237868 M * distemper it was, i also received it via email
1393237996 M * VirMan vserver2 etc # vserver vs6018 start The following problem(s) were encountered while verifying vshelper functionality: * The configured vshelper '/usr/libexec/util-vserver/vshelper' does not match the 'vshelper'   script of the util-vserver package. Maybe you have two versions installed?
1393238199 J * renihs ~arf@83-65-34-34.arsenal.xdsl-line.inode.at
1393238529 M * Bertl distemper: okay, many thanks for pointing it out, I somehow missed it
1393238589 M * Guy- http://list.linux-vserver.org/archive?sss:6511:201402:filjmhgamabjmadaafcm#b gives an internal server error
1393238594 M * VirMan vdlimit: vc_set_dlimit(): Invalid argument
1393238597 M * VirMan ok guys
1393238609 M * VirMan i'll try to solve it somehow
1393238621 M * Bertl is that now with the recent tools?
1393238642 M * VirMan i was not able to build recent tools so they work....
1393238651 M * Bertl really?
1393238726 M * Bertl daniel_hozac: any ideas what could cause that except for the kernel interface and why it seems to work fine with testfs.sh?
1393238732 M * VirMan well - i might not be an expert, but there were all the time some werid directories popping, or mysterious messages - i tried symlink directories to usr/local/share/etc and so on... then i his vshelper problem
1393238771 M * Bertl the helper message just states that it wasn't updated
1393238783 M * Bertl i.e. you probably didn't re-run the init scripts
1393238799 M * VirMan i restarted util-vservers
1393238813 M * VirMan ok - i'll try to do it again
1393238838 M * Bertl as I suggested, first, try to uninstall everything you installed by now
1393238853 M * Bertl double check for left overs, i.e. search the filesystem
1393238854 M * daniel_hozac Bertl: that's only the kernel interface.
1393238963 M * VirMan ok after uninstalling and installing newer version i run into vshelper then run echo "/usr/local/libexec/util-vserver/vshelper" >/proc/sys/kernel/vshelper and it stated (with rnetling message but it was probably minor ip leftover)
1393238971 M * VirMan now i'll try with dlimits
1393238991 M * daniel_hozac Bertl: perhaps the limits themselves are invalid.
1393239007 M * Bertl hmm, good point
1393239023 M * daniel_hozac e.g. used > total
1393239030 M * Bertl daniel_hozac: can we get the commands with --debug?
1393239032 M * VirMan util-vserver: 0.30.216-pre3054; Feb 24 2014, 10:54:32
1393239041 M * daniel_hozac yeah
1393239044 M * VirMan sure
1393239047 M * Bertl ah, looks good!
1393239069 M * VirMan vdlimit: vc_set_dlimit(): Invalid argument
1393239078 M * VirMan the same error with new utils
1393239101 M * daniel_hozac please run it with --debug
1393239102 M * VirMan vserver vs6047 start --debug
1393239104 M * VirMan ?
1393239110 M * daniel_hozac no, vserver --debug vs6047 start
1393239113 Q * JvdK Quit: Ex-Chat
1393239220 M * VirMan http://pastebin.com/hexW8i4v
1393239369 M * daniel_hozac you're reserving 5000% for the root user?
1393239380 M * VirMan 5000% ????
1393239386 M * VirMan or 5000 inodes ?
1393239394 M * daniel_hozac it's percent.
1393239401 M * VirMan facepalm....
1393239407 M * Bertl 5000 percent seems a little much :)
1393239424 M * daniel_hozac that would be a nifty way to get more space though.
1393239427 M * VirMan ok next time please include facepalm virman message :D
1393239441 M * renihs :)
1393239455 M * VirMan one line of code saves hours of frustration :P
1393239480 M * VirMan this is oficial path request :D
1393239485 M * Bertl you know, enough rope to shoot yourself in the foot :)
1393239499 M * VirMan hehe
1393239544 M * Bertl but yes, I'm pretty sure if somebody whips up a patch to check for such a case, daniel_hozac will gladly add it
1393239571 M * Bertl (there is a chance he will add a check without patch :)
1393239609 M * VirMan some less merged with vservers would be grateful ;)
1393239649 M * VirMan anyway - i would probably wait for a year for newer ebuilds ;)
1393239665 M * VirMan now all works
1393239666 M * VirMan thanks guys
1393239703 Q * undefined Quit: Closing object
1393240715 M * CcxCZ hmm, can someone give me rundown of differences of LXC-1.0 and current vserver?
1393240903 M * Bertl LXC is lacking a number of security relevant limits
1393240920 M * ard I think context vs namespace, a lot of restrictions based on context
1393240922 M * Bertl it is also missing a number of virtualizations
1393240936 M * Bertl like uptime, fake pid 1, etc
1393240958 M * Bertl and it doesn't provide lightweight IP isolation
1393240983 M * Bertl and there are different tools to configure each
1393240986 M * Bertl other than that, it is basically identical
1393240991 A * CcxCZ is reading https://www.stgraber.org/2014/01/01/lxc-1-0-security-features/ atm
1393241002 M * CcxCZ kthnx
1393241052 M * CcxCZ well, with 1.0 they seem to have something called "unprivileged containers"
1393241086 M * arekm vserver disables that, no support for user namespace
1393241127 M * Bertl you could simply enable it and test the implications btw
1393241156 M * Bertl (where simply means removing the depends clause)
1393241202 M * Bertl CcxCZ: does it help against fork bombs for example?
1393241276 M * CcxCZ dunno. I have the hunch some cgroup limits are supposed to handle that
1393241279 M * arekm don't resource limits still apply for user_ns?
1393241304 M * Bertl but I'm confident LXC is going into the right direction, after 12 years they should better be :)
1393241405 M * Bertl which reminds me ...
1393241432 M * CcxCZ well, it'd be nice to get closer to vanilla kernel w/ vserver. integrating different patchsets is surely tedious
1393241433 M * Bertl daniel_hozac: what do you think, can we remove the device mapping feature and utilize the cgroup for that?
1393241467 M * Bertl (not the mapping, which is rarely used, but the access control)
1393241496 M * daniel_hozac yeah, i don't see why not.
1393241544 Q * robmc Ping timeout: 480 seconds
1393241635 M * Bertl okay, just checking, not that we hit any major issues where util-vserver goes crazy or so :)
1393241677 M * daniel_hozac the utils should probably do it automatically, so the vdevmap configuration keeps working.
1393241743 M * daniel_hozac do you remember how it works? if you set one, it goes into default-deny, right?
1393241754 M * arekm CcxCZ: other vserver thins: no ipv6-loopback in isolation, no user_ns and recent /proc problem are tiny issues with vserver, rest works well here
1393242021 M * Bertl I do not remember, I'm old and senile :)
1393242053 M * Bertl but I will see when I start removing the interfaces
1393242085 M * Bertl do we know, is the device mapping feature used often?
1393242113 M * Bertl it might be worth adding it to the cgroup
1393242120 M * Bertl shouldn't be too hard I guess
1393242134 M * ard Anyway, it is usually easy to use lxc features in vserver, while having vserver security in lxc is impossible without just running vserver 8-D
1393242163 M * ard Bertl : I never knew how and why to use device mapping features :-)
1393242178 M * daniel_hozac build roots
1393242204 M * Bertl the idea was to have e.g. /dev/console (which has a specific major/minor) to a different tty for example
1393242233 M * Bertl later it became a necessity for chroot/buildroot enviroments which insist on creating devices
1393242234 M * ard Ah.. yes... like ubuntu upstart *needs* a working tty as /dev/console
1393242237 M * daniel_hozac i thought the most useful case was with some admin GUI
1393242248 M * daniel_hozac and quotatools
1393242260 M * daniel_hozac where the device had to match the filesystem
1393242261 M * ard I worked around that by actually giving it a real tty :-(
1393242262 M * daniel_hozac and vroot doesn't
1393242270 M * Bertl well, yes and no, the quotatools where happy with the fake mtab :)
1393242295 M * daniel_hozac still? i thought that changed in the past year or two
1393242312 M * ard So I would be able to build a network boot image with device mapping?
1393242312 M * Bertl yes, it changed, everything gets better :)
1393242361 M * ard Hmmm, I don't think so, the most important one is the ethernet....
1393242385 M * Bertl you can't remap network devices :)
1393242405 M * Bertl we are talking device nodes in the filesystem
1393242749 M * ard well, I could create veth devices and such :-)
1393242758 M * ard and use network namespaces...
1393242761 M * Bertl yes, that will work
1393242783 M * VirMan ok another issue with new host :)
1393242784 M * VirMan vlimit -c 6047 --rss 131072 vc_set_rlimit(): Invalid argument
1393242785 M * ard But I need something that I can change and then cpio to a ramdisk image
1393242805 M * Bertl VirMan: memory limits are done via cgroups nowadays
1393242839 M * Bertl i.e. the memory limit subsystem was removed when the cgroups provided similar features
1393242843 M * VirMan ahh - true
1393242871 M * VirMan and i guess thre is not such nice command now .. ? 
1393242890 M * Bertl you can 'simply' echo the correct values to the config dir
1393242908 M * Bertl (which will be copied to the cgroup)
1393242936 M * ard The current place is /dev/cgroup/<vservername>
1393242976 M * ard so you can change it live
1393242979 M * daniel_hozac or /sys/fs/cgroup/memory/<vservername> or wherever else you might have configured it.
1393242991 M * ard is there a standard?
1393243013 M * daniel_hozac /sys/fs/cgroup/<subsystem> is the standard.
1393243024 M * ard oh , right :-(
1393243046 M * ard Then I need to fix my munin before I upgrade
1393243073 M * hijacker fellows, there is a dead link on vserver website, the link to the latest kernel is broken
1393243077 M * CcxCZ daniel_hozac, Guy-: there are now several initiatives to runit-ify OpenRC so there's proper daemon supervisor support. just fyi :)
1393243105 M * hijacker there is no .bz2 kernel on kernel.org, just .xz and .gz
1393243176 M * VirMan whoaa .. now i see a lot of options :P - any hins how to use it ?
1393243180 M * Bertl hijacker: are you sure?
1393243216 M * hijacker Bertl, just browsed the kernel.org repo
1393243222 M * CcxCZ Bertl: pretty sure, I also reported it earlier
1393243225 M * hijacker https://www.kernel.org/pub/linux/kernel/v3.x/
1393243239 M * CcxCZ but no one seemed to react :]
1393243247 M * hijacker CcxCZ, sorry, I am not actively reading in here
1393243257 M * hijacker thus I doubled ;-(
1393243257 M * Bertl we have been working on it, it should be fixed by now
1393243275 M * distemper https://www.kernel.org/happy-new-year-and-good-bye-bzip2.html
1393243279 M * Bertl i.e. reload the page, check if it works
1393243288 M * hijacker great!
1393243301 M * Bertl kudos go to fback .. thanks!
1393243481 M * CcxCZ daniel_hozac, Guy-: also I'm writing this http://wpr.cz/ccx/loggerhead/zsv/view/head:/README.rst (should update TODO, most of the hooks are in now)
1393243630 Q * ircuser-1 Read error: Operation timed out
1393243729 M * Guy- CcxCZ: oh wow; otoh, I frankly don't get the point of svtab
1393243788 M * CcxCZ Guy-: well, the main point of the project is to put the init/teardown code in reusable parametrisable shell functions
1393243854 M * CcxCZ svtab is about the easiest and most readable way I came up with on the spot, I'll probably make it more modular laters
1393243927 M * CcxCZ the point is I can make things like wpa_supplicant and dhcpcd start/stop on external conditions, like cable plugged in / rfkill disabled, etc..
1393244000 M * renihs yay, another init :)
1393244016 M * CcxCZ well, this one has been around for a while ;-)
1393244128 M * CcxCZ it's more of a core which you can script around. unlike that other one which tries to do everything by itself and then some
1393244172 M * renihs i see :)
1393244179 M * renihs CcxCZ, you seem offline in freenode :)
1393244195 M * renihs ahno, i am just blind
1393244327 M * CcxCZ if you are familiar with DJB's daemontools it's basically that plus some essential things that make it saner
1393244351 M * Bertl I hope it doesn't require a CcxCZ universe to work :)
1393244407 M * renihs just a small galaxy :)
1393244724 M * CcxCZ Bertl: I try to keep my stuff self-contained. I do prototype in zsh though. rewrite to posix-sh is usually feasible, but here I don't care much
1393244889 M * Bertl was just a kick towards DJB, don't take it personal
1393245772 M * Guy- CcxCZ: I'm a heavy runit user too
1393245793 M * Guy- CcxCZ: my approach has been to make the run scripts configurable via /etc/defaults and similar
1393245845 Q * Aiken Remote host closed the connection
1393245851 M * Guy- CcxCZ: I must admit I still don't quite see how that svtab thing is supposed to work, but I'm interested
1393245892 M * CcxCZ Guy-: did that before, when I was on daemontools-based setup
1393245893 M * Guy- CcxCZ: also I have run scripts for 100+ services in case you're interested (their quality varies)
1393245932 M * Guy- CcxCZ: how did you find me with this, btw? based on my contribution to the linux-vserver wiki?
1393245972 M * CcxCZ I remembered talking to you before because of the runit page on the wiki, yeah. so I looked you up there again :)
1393246095 M * Guy- ah, indeed, that was like more than a year ago :) I didn't remember
1393246206 Q * VirMan Remote host closed the connection
1393246450 J * ircuser-1 ~ircuser-1@35.222-62-69.ftth.swbr.surewest.net
1393246570 M * CcxCZ Guy-: I'm certainly interested, will poke you later if I may
1393246798 M * Guy- CcxCZ: sure
1393246869 J * undefined ~undefined@66-190-97-211.dhcp.unas.tx.charter.com
1393247476 N * l0kit Guest1083
1393247481 J * l0kit ~1oxT@0001b54e.user.oftc.net
1393247881 Q * Guest1083 Ping timeout: 480 seconds
1393248205 J * beng_ ~BenG@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net
1393251467 J * alpha_one_x86 ~kvirc@190.171.235.197
1393251790 P * undefined 
1393251790 J * undefined ~undefined@00011a48.user.oftc.net
1393252473 Q * undefined Quit: Closing object
1393258756 Q * renihs Quit: narf
1393259009 J * virman ~oftc-webi@91.218.79.125
1393261473 J * zerick ~eocrospom@190.187.21.53
1393261968 Q * zerick Quit: Saliendo
1393262084 Q * rawplayer Remote host closed the connection
1393262084 J * zerick ~eocrospom@190.187.21.53
1393262454 J * rawplayer ~xyzzy@shell.students.os3.nl
1393262750 J * bonbons ~bonbons@ppp-157-178.adsl.restena.lu
1393263260 Q * beng_ Ping timeout: 480 seconds
1393264093 J * thierryp ~thierry@lns-bzn-47f-62-147-212-202.adsl.proxad.net
1393264115 Q * thierryp 
1393264117 J * thierryp ~thierry@lns-bzn-47f-62-147-212-202.adsl.proxad.net
1393264890 Q * thierryp Quit: ciao folks
1393266378 P * alpha_one_x86 No matter how dark the night, somehow the Sun rises once again
1393268897 M * shaggy63 If I modify a hashified file from the host instead of the guest will the hardlink still be broken or will it actually update the .hash/ file?
1393272156 M * daniel_hozac yes
1393272166 M * daniel_hozac the copy on write doesn't care about who breaks it.
1393272573 J * beng_ ~BenG@cpc29-aztw22-2-0-cust128.18-1.cable.virginm.net
1393272732 M * Bertl off for a nap ... bbl
1393272746 N * Bertl Bertl_zZ
1393274949 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1393275426 M * CcxCZ hmm, I should read up on hashification
1393275464 M * CcxCZ is it tied to vserver contexts, or could I use it as generic COW solution?
1393276114 J * bzed_ ~bzed@bzed.netrep.oftc.net
1393276280 Q * jrklein_ Ping timeout: 480 seconds
1393276425 Q * bzed Ping timeout: 480 seconds
1393276430 N * bzed_ bzed
1393276858 M * daniel_hozac it's generic.
1393277658 Q * bonbons Quit: Leaving
1393277697 Q * beng_ Quit: I Leave
1393277884 J * jrklein ~osx@proxy.dnihost.net
1393278063 M * CcxCZ neat
1393278143 M * CcxCZ any good doc on it aside from http://linux-vserver.org/util-vserver:Vhashify ?
1393278277 M * daniel_hozac that's only about hashify.
1393278285 M * daniel_hozac not the generic COW links.
1393278363 M * daniel_hozac http://linux-vserver.org/Paper#Unification has more about the concept.
1393278637 M * CcxCZ hm, no info on cow there either as far as my searches go
1393280034 M * daniel_hozac it's the same, just with COW link breaking added.
1393283107 M * CcxCZ what kind of marking does it use? some special FS support needed?
1393283164 M * daniel_hozac yes, the vserver iattrs are used for it. same ones as in the paper, and recently (two years ago maybe?), an additional COW iattr.
1393283207 Q * Ghislain Quit: Leaving.
1393284223 M * virman echo "536870912" > memory.kmem.limit_in_bytes ===>  bash: echo: write error: Device or resource busy
1393284432 M * virman when i set memory.limit_in_bytes to 512m  guest sees all available memory...
1393284456 M * virman is there any way to make it look as in old vserver so that guest would see only his memory ?
1393284507 J * undefined ~undefined@66-190-97-211.dhcp.unas.tx.charter.com
1393284537 P * undefined 
1393284540 J * undefined ~undefined@00011a48.user.oftc.net
1393285967 Q * virman Remote host closed the connection
1393285977 M * Jb_boin virman, put "virt_mem" on /etc/vserver/VSERVERNAME/flags
1393286075 M * Jb_boin too late :)
1393286103 M * Jb_boin ill paste the "answer" anyway http://paste.debian.net/hidden/6707ffcf/