1126483433 J * nayco_laptop ~nayco@lns-bzn-8-nan-82-250-240-246.adsl.proxad.net
1126483445 M * nayco_laptop http://pastebin.com/361100
1126483499 M * nayco_laptop ohh dear, i'm sure i've compiled with XFS quota support :(
1126483527 M * nayco_laptop It looks like no... Ok, I check into the kernel conf
1126483531 M * Bertl yes? and what about 'normal' quota support?
1126483645 M * nayco_laptop Hum.... er... Ok, I'm going to check. (I thought it was default on my config...)
1126483696 M * nayco_laptop   [*] Quota support                                                        ? ?
1126483696 M * nayco_laptop   ? ?           <M>   Old quota format support                                           ? ?
1126483696 M * nayco_laptop   ? ?           <*>   Quota format v2 support    
1126483706 M * nayco_laptop seems ok...
1126483849 M * Bertl so both enabled?
1126483856 M * Bertl xfs and normal quota?
1126483865 M * nayco_laptop yep
1126483890 M * Bertl in the kernel now running?
1126484173 M * nayco_laptop Well, how can I tell it without mistake ?... Yes, uname -a : 
1126484192 M * nayco_laptop Linux nayport 2.6.13.1-vs2.1.0-pre7 #2 SMP Mon Sep 12 01:32:31 CEST 2005 i686 Pentium III (Coppermine) unknown GNU/Linux
1126484257 M * nayco_laptop and I've only got one vserver kernel on this machine, I deleted the old (yesterday's ;)) one
1126484260 M * Bertl you should enable the configuration in procfs
1126484281 M * nayco_laptop config.gz ? It's done already, you want it ?
1126484294 M * Bertl no, but you can zgrep there
1126484365 M * nayco_laptop what do you need ?
1126484390 M * nayco_laptop [root@nayport ~]# zgrep -i quota /proc/config.gz
1126484391 M * nayco_laptop CONFIG_XFS_QUOTA=y
1126484391 M * nayco_laptop CONFIG_QUOTA=y
1126484391 M * nayco_laptop CONFIG_QUOTACTL=y
1126484422 M * Bertl zgrep Q..T /proc/config.gz (gets it all)
1126484434 M * Bertl but that is already sufficient
1126484476 M * Bertl check devices and quota capability and (if you use vroot) the vroot setup
1126484480 M * nayco_laptop [root@nayport ~]# zgrep Q..T /proc/config.gz
1126484480 M * nayco_laptop CONFIG_CPU_FREQ_STAT=m
1126484480 M * nayco_laptop # CONFIG_CPU_FREQ_STAT_DETAILS is not set
1126484480 M * nayco_laptop CONFIG_XFS_QUOTA=y
1126484480 M * nayco_laptop CONFIG_QUOTA=y
1126484481 M * nayco_laptop CONFIG_QFMT_V1=m
1126484483 M * nayco_laptop CONFIG_QFMT_V2=y
1126484485 M * nayco_laptop CONFIG_QUOTACTL=y
1126484521 M * nayco_laptop well, the pastebin I've done above tells this, no ?
1126484538 M * Bertl ah, I have to update my zgrep to _Q..T :)
1126484571 M * nayco_laptop like this : 
1126484575 M * nayco_laptop # zgrep _Q..T /proc/config.gz
1126484575 M * nayco_laptop CONFIG_XFS_QUOTA=y
1126484575 M * nayco_laptop CONFIG_QUOTA=y
1126484575 M * nayco_laptop CONFIG_QFMT_V1=m
1126484575 M * nayco_laptop CONFIG_QFMT_V2=y
1126484577 M * nayco_laptop CONFIG_QUOTACTL=y
1126484578 M * nayco_laptop ?
1126484627 M * nayco_laptop er... Do I use vroot ?
1126484711 P * kusznir Leaving
1126484736 M * Bertl IIRC, we agreed to check with the / first, then go for the /home ...
1126484759 M * Bertl nayco_laptop: so let's check a) if quota on /vservers works on the host
1126484774 M * Bertl and b) if that works, what happens inside the guest for /
1126484804 M * Bertl but the ENOSYS doesn't make me happy ...
1126484834 M * nayco_laptop well, the (set|rep)quota commands seem not to wrok on the host.... Checking again
1126484849 M * Bertl yes, that's what I suspect
1126484929 M * nayco_laptop http://pastebin.com/361118 :(
1126484971 M * Bertl hmm, did you mount the fs with quota support?
1126484995 M * nayco_laptop see my first pastebin, I think so... : 
1126485007 M * Bertl check in /proc/mounts ...
1126485058 M * nayco_laptop http://pastebin.com/361120
1126485061 M * nayco_laptop wait...
1126485064 M * Bertl hmm ... maybe the quota hashes interfere somehow ...
1126485109 M * nayco_laptop http://pastebin.com/361121
1126485138 M * nayco_laptop seems like the quota redirector fails to direct to the xfs quota functions.
1126485154 M * nayco_laptop Is it called "redirector" ?
1126485248 M * Bertl you mean the syscall or what?
1126485386 M * nayco_laptop yes... I haven't got the kernel language ;) => If I understand well, a userland program make a quotactl syscall, inside the kernel it passes through various functions to get the fs type, and finally the right function from xfs_qm_syscall.c is executed, right ?
1126485407 M * nayco_laptop not "passes", but "travels"
1126485419 M * nayco_laptop french language ;)
1126485447 M * Bertl well, I think the problem is my modification to the xqm_quotactl_valid()
1126485459 M * nayco_laptop in that file
1126485460 M * nayco_laptop ?
1126485469 M * Bertl in fs/quota.c
1126485497 M * Bertl let's add some debugging there, you have a timestamped version of your kernel by now?
1126485531 M * Bertl i.e. a make will take a few seconds instead of hours?
1126485605 M * nayco_laptop in france, we are used to make a "yellow laugh", in that case :D !
1126485610 M * nayco_laptop yes, I think so ;)
1126485623 M * nayco_laptop ok, go on.
1126485735 M * Bertl hmm, I guess I already found the issue :)
1126485746 M * nayco_laptop a typo ;) ?
1126485752 M * Bertl fs/xfs/linux-2.6/xfs_super.c line 771
1126485759 M * Bertl sb->s_qcop = &linvfs_qops;
1126485767 M * Bertl below that, add the following:
1126485781 M * Bertl sb->s_dqh->dqh_qcop = &linvfs_qops;
1126485830 M * Bertl then let's recompile and retry
1126485871 M * nayco_laptop pffff, I'll never understand anything to the kernel, no hope. This is chinese to me, not because it is C, but because one nneds to known the whole kernel to do thinks like you :-| ... 
1126485920 M * Bertl heh, it only needs time an patience ...
1126485925 M * Bertl *and
1126485939 M * Bertl btw, do you need any special quota tools for xfs quota?
1126486001 N * lilo_ lilo
1126486017 M * nayco_laptop er.... Well, I don't know... What do you mean ? A special userland tool other than setquota and the fstab "quota" keyword ?
1126486073 M * nayco_laptop How long has it been since you started studying and coding the kernel ?
1126486127 M * Bertl ad tool, figured that my quota tools should suffice ...
1126486144 M * Bertl ad time: well, linux kernel since about 2-3 years now ...
1126486165 M * nayco_laptop :O !
1126486181 M * nayco_laptop er... well, you spent nights on it, so :P !
1126486233 M * Bertl ah, yes that fixes the host issues :)
1126486293 M * nayco_laptop ok, for the tools, I dont know : I can say that I usually use setquota, repquota, and that's all. I even do not need quotaon|off, because XFS quota are on from the boot. Sadly, it requires often a reboot because mount -o remount (-a|<mountpoint>) does not always does it.
1126486324 M * nayco_laptop ok, I'm rebooting (Nice noatime ;))
1126486336 M * Bertl yes, remount is not supposed to work with quota stuff
1126486340 M * nayco_laptop brb
1126486349 Q * nayco_laptop Remote host closed the connection
1126486358 M * nayco <worm hole>
1126486398 M * nayco before linux, you used to program/study an other kernel ?
1126486418 M * nayco kernel panic : attempted to kill the idle task 
1126486433 M * Bertl yes, mach and micro kernels ...
1126486441 M * nayco atkbd.c : too many keys pressed
1126486449 M * Bertl nayco: hmm, how so?
1126486459 M * nayco mach ? And why did you switch to linux ?
1126486470 A * Bertl .oO( cat on the keyboard? )
1126486494 M * Bertl nayco: got interested in this fancy linux-vserver project .)
1126486550 M * nayco not cat :D !!! No, it happened during the shutdown process. Ok, booting goes well, pfewww !
1126486583 M * nayco Well, 3 years of linux is not long to reach your hacking level, good job ;-)
1126486600 M * Bertl thanks! I love challenges ...
1126486633 J * nayco_laptop ~nayco@lns-bzn-8-nan-82-250-240-246.adsl.proxad.net
1126486637 M * nayco ah
1126486654 M * nayco_laptop back ;)
1126486717 M * nayco_laptop ohh dear, same error messages (Or no messages at all)
1126486720 M * nayco_laptop =(
1126486742 M * Bertl it works here on the host .. but I guess you're talking about the guest
1126486749 M * nayco_laptop I guess this @]=^Г kernel isn't recompiled.
1126486768 M * nayco_laptop no, the host ;(
1126486798 M * nayco_laptop but : [root@nayport ~]# uname -a
1126486798 M * nayco_laptop Linux nayport 2.6.13.1-vs2.1.0-pre7 #2 SMP Mon Sep 12 01:32:31 CEST 2005 i686 Pentium III (Coppermine) unknown GNU/Linux
1126486799 M * nayco_laptop .
1126486800 M * nayco_laptop ?*
1126486824 M * Bertl 02:16 < nayco_laptop> Linux nayport 2.6.13.1-vs2.1.0-pre7 #2 SMP Mon Sep 12 
1126486830 M * Bertl same version :)
1126486840 M * nayco_laptop not same date !
1126486852 M * nayco_laptop oh, sorry
1126486865 M * nayco_laptop I did not see the time :O
1126486885 M * Bertl the #2 means the build version
1126486900 M * Bertl so it wasn't rebuilt/reinstalled
1126486903 M * nayco_laptop I got to sleep, wrok tommorow... Anyway, is there some ".o" file I can deleted to force the building of the needed part ?
1126486925 M * Bertl btw, quota in the guest work here too ...
1126486943 M * nayco_laptop ok, I was asking myslef what was this #2 ;) => Told you I would learn thing here ;)
1126487054 M * nayco_laptop e.g. : .tmp_vmlinux1 ?
1126487059 M * nayco_laptop to delete
1126487083 M * Bertl really depends .. maybe you should avoid the distcc for the rebuilds?
1126487106 M * Bertl there _is_ a certain path from the edited file to the kernel
1126487129 M * nayco_laptop well, If I do that, I think I saw that it will recompile the whole kernel and modules :|
1126487149 M * Bertl let me figure the relevant file to remove for you
1126487154 M * nayco_laptop thx.
1126487182 M * nayco_laptop I guessed that, but to track it down is not in my abilities
1126487184 M * nayco_laptop ;)
1126487250 M * Bertl rm -f fs/xfs/linux-2.6/xfs_super.o fs/xfs/xfs.o fs/filesystems.o vmlinux
1126487287 M * Bertl you edited the fs/xfs/linux-2.6/xfs_super.c, and saved it back, yes?
1126487371 M * nayco_laptop er... lemme look
1126487402 M * nayco_laptop #endif
1126487402 M * nayco_laptop         sb->s_qcop = &linvfs_qops;
1126487402 M * nayco_laptop         sb->s_dqh->dqh_qcop = &linvfs_qops;
1126487402 M * nayco_laptop         sb->s_op = &linvfs_sops;
1126487409 M * Bertl good!
1126487481 M * nayco_laptop compiles...
1126488348 M * nayco_laptop ok, it is #3, yes !!!! (thanks strings) => reboot
1126488354 Q * nayco_laptop Quit: Leaving
1126488396 M * nayco ...
1126488461 M * nayco the kernel printed a message talking about dqhash when booting... With strange characters
1126488466 M * nayco but it boots ;)
1126488483 M * Bertl yep, that's fine :)
1126488605 M * nayco hmmm.... er... crap !
1126488613 M * nayco doesn't work :(
1126488627 M * Bertl _what_ doesn't work?
1126488692 M * nayco ok, here is what i'm going to do : gonna go to sleep, and tomorrow i'll build it from the start at work (my distcc farm is 6 powerful machines). so tommorow night, i'll have the actual kernel built !
1126488715 M * nayco uname -a, like I saw with strings, 
1126488716 M * Bertl k, good night then! :)
1126488728 M * nayco gives #3 for the kernel rev...
1126488747 M * nayco I think there is something wrong with distcc.
1126488760 M * Bertl could be ...
1126488769 M * nayco I tried setquota and repquota on the host, and strace qives
1126488772 M * nayco gives
1126488788 M * nayco "function not implemented" for the  syscall
1126488795 M * nayco sh**
1126488797 M * nayco ;)
1126488807 M * Bertl okay, as this works here, I'm pretty sure it is a compile issue
1126488837 M * nayco ok, I will correct that tomorrow... Good night, and thanks :)
1126488845 M * Bertl will upload a patch tomorrow, which you can use for your kernel, so best start a compile with the 2.1.0-pre7 (without modifications an distcc)
1126488857 M * Bertl nayco: sweet dreams!
1126488883 M * nayco ok, nice coding^Wsleep ;Гp !!!
1126488948 Q * nayco Quit: bloody distcc, crappy laptop !!!!
1126489166 M * Bertl k, folks, I'm off to be now too .. have a good whatever ... and cya tomorrow!
1126489177 N * Bertl Bertl_zZ
1126494077 M * micah hrm, using 2.6.12, vserver 2.0 and vapt-get updating some packages, and am getting:
1126494087 M * micah Unpacking replacement apache ...
1126494087 M * micah dpkg: error processing /var/cache/apt/archives/apache_1.3.33-6sarge1_i386.deb (--unpack):
1126494090 M * micah  failed to rmdir/unlink `//usr/lib/apache/suexec.disabled.dpkg-tmp': Operation not permitted
1126494095 M * micah just on that particular package
1126494111 M * micah then when I try to reissue the upgrade of apache, it completes fine
1126494228 M * micah in another vserverr:
1126494234 M * micah Preparing to replace login 1:4.0.3-30.8 (using .../login_1%3a4.0.3-31sarge5_i386.deb) ...
1126494237 M * micah Unpacking replacement login ...
1126494240 M * micah dpkg: error processing /var/cache/apt/archives/login_1%3a4.0.3-31sarge5_i386.deb (--unpack):
1126494243 M * micah  failed to rmdir/unlink `//usr/bin/newgrp.dpkg-tmp': Operation not permitted
1126494258 J * Aiken_ ~james@tooax6-043.dialup.optusnet.com.au
1126494583 Q * Aiken Ping timeout: 480 seconds
1126494984 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126495387 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net
1126495590 Q * dddd44 Ping timeout: 480 seconds
1126496000 Q * kas_3 Ping timeout: 480 seconds
1126496298 J * dddd44 U2FsdGVkX1@tor-irc.dnsbl.oftc.net
1126496400 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il
1126496795 Q * dddd44 Ping timeout: 480 seconds
1126496961 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126500253 Q * obi Ping timeout: 480 seconds
1126500275 J * obi ~obi@asus.saftware.de
1126500875 Q * dddd44 Ping timeout: 480 seconds
1126500924 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126501452 J * kas_3 U2FsdGVkX1@tor-irc.dnsbl.oftc.net
1126501790 Q * dddd44 Ping timeout: 480 seconds
1126506170 Q * ag- Quit: BRB
1126507529 Q * fluor Ping timeout: 480 seconds
1126507780 Q * tchan Ping timeout: 480 seconds
1126509363 J * tchan ~tchan@c-67-174-18-204.hsd1.il.comcast.net
1126511272 J * ag- ag@caladan.roxor.cx
1126511300 Q * kas_3 Ping timeout: 480 seconds
1126511333 Q * Aiken_ Ping timeout: 480 seconds
1126511953 J * Aiken ~james@tooax8-043.dialup.optusnet.com.au
1126513881 Q * Hollow Remote host closed the connection
1126514070 J * wam ~wigwam@proxy2.msh.de
1126514164 M * wam Hi! When I enter a vdlimit (see below), I recieve an error message: "vdlimit: vc_xidopt2xid("id"): No such file or directory". What could be wrong? Did I miss a kernel patch?
1126514164 M * wam vdlimit -xid 50 --set space_total=500000 --set space_used=`du -s /vservers/vhost0/ | awk '{print $1}'` --set inodes_total=500000 --set inodes_used=`ls -1aRi /vservers/vhost0/ | awk '/^[0-9]+ / { print $1 }' | sort -u | wc -l` --set reserved=5 /vservers/vhost0/
1126514285 M * wam here's my testme.sh - output: http://pastebin.com/361307
1126515163 M * wam buaaaahaaahaa - sorry. just had to add another "-" before "xid"... Damn
1126515333 M * Greek0 hehe
1126515388 M * SiD3WiNDR damn, such a complicated commandline and already in the first parameter it breaks ;)
1126515565 M * wam hmm - but I can use lots more diskspace in the vserver. "df -k" tells me, I have ~300MB free. But I can create a 600MB file... Maybe I really missed a kernel-patch?
1126515978 J * lilo_ ~lilo@lilo.usercloak.oftc.net
1126516055 Q * lilo Ping timeout: 480 seconds
1126516243 J * erwan_taf ~erwan@81.80.43.77
1126516753 P * erwan_taf Leaving
1126516803 J * Nicoli ask@208.53.159.170
1126517728 J * Blogmeister ~Blogmeist@site.lycos.de
1126517936 P * Blogmeister 
1126521080 J * Hollow ~Hollow@home.xnull.de
1126521096 M * Hollow Greek0: you around?
1126522992 M * wam I want to limit disk-usage with vdlimit as explained here: http://linux-vserver.org/Disk+Limits The command executes fine and "df" in the vserver shows the correct values. But I can create files which are much bigger than the (simulated) diskspace. Would I have to install another kernel-patch?
1126523285 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126523394 N * BobR_oO BobR
1126523438 J * fluor ~fluor@tanneries.squat.net
1126523535 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net
1126523830 Q * dddd44 Ping timeout: 480 seconds
1126523885 M * wam ok ;) The other way round: Are the patches qh, cx, cq, d required for disk-limits or are they already included in the latest 2.0-patches (for kernel 2.6)?
1126523900 M * wam (patches which are found here: http://www.13thfloor.at/vserver/s_addons/overview/	)
1126524671 M * Greek0 Hollow: yep
1126524688 M * Hollow Greek0: what do you think about libConfuse ?
1126524707 M * Greek0 I'm confused. what is libConfuse?
1126524715 M * Hollow *g* a config file parser
1126524803 M * Greek0 where? in the utils?
1126524815 M * Greek0 I've lost that url _again_ :-/
1126524852 M * Hollow http://www.nongnu.org/confuse/
1126524916 M * Greek0 didn't know it yet
1126524930 M * Hollow i did neither, just stumbled across it today
1126524947 M * Greek0 the syntax doesn't make me want to puke immediately, which is a good sign ;)
1126524971 M * Hollow heh
1126525040 M * Greek0 do you want a global one or one-config-per-vserver?
1126525064 M * Hollow i'd prefer one file per vserver
1126525076 M * Hollow imagine you have 200 vservers, what a bloat if you want to start only 1
1126525106 M * Hollow and even per-vserver config files can get quite large
1126525109 M * Greek0 yep, sounds reasonable
1126525134 M * Greek0 what about supporting the old config scheme?
1126525146 M * Hollow uh.. *g*
1126525158 M * Hollow i'd rather write a converter
1126525224 M * Greek0 hmm, well.. a converter woud be ok probably. however if it's easy to get support integrated, why not?
1126525272 M * Greek0 would at least provide a smooth starting point, both, for us, and for new users
1126525274 M * Hollow imo it'll not be worth the trouble ;)
1126525342 M * Hollow but writing some bash script which reads the old tree and forms the new config is probably a lot easier
1126525362 M * Hollow but i'm not all against it if it's possible
1126525414 M * Greek0 well, I think it wouldn't be so hard to parse the old config, I'm just thinking of how to represent all the info in memory
1126525455 M * Hollow structs?
1126525517 M * Hollow well, i also like the xml approach, but xml is quite difficult to parse..
1126525555 M * Greek0 something like query("vserver-name", "apps.init.mark"); would be nice, but I think it's not doable because of stuff like interfaces/
1126525582 M * Greek0 I don't think xml is the best thing for config files..
1126525642 M * Hollow depends... imo you can describe the context hierarchy quite nice, but you have to do much overhead xml work..
1126525644 M * Greek0 well, we could have query("vserver-name", "interfaces") return a list of interfaces.. 
1126525672 M * Hollow libconfuse supports multiple sections...
1126525683 M * Greek0 I find it a bit hard to read. IMO simple plain-text stuff (like libconfuse) would rock
1126525690 M * Hollow yep
1126525700 J * liQuid ~liQuid@dsl-082-082-116-220.arcor-ip.net
1126525707 M * Hollow you can even define custom functions for config files, so your quite flexible with confuse
1126525728 M * Hollow and we don't have to care about char parsing *narf* ;)
1126525729 N * Bertl_zZ Bertl
1126525733 M * Bertl morning folks!
1126525733 M * Hollow hola Bertl 
1126525745 M * liQuid hello, does this vserver software need much load?
1126525754 M * Bertl *short visit*
1126525772 M * Bertl liQuid: what software do you mean? and what do you mean by load?
1126525783 M * liQuid the vserver thing here
1126525789 M * Bertl Hollow, Greek0: take a look at property lists!
1126525817 M * liQuid i mean how much vservers u can do with this
1126525853 M * Bertl ah, well, we have reports from fantastic 200 guests on a dual Xeon machine ...
1126525872 M * Bertl in a typical setup I would aim for about 30-60 though
1126525908 M * Greek0 Bertl: uh, property lists? -v?
1126525935 M * liQuid i want to use a amd xp 2400+ with 512 memory
1126525938 M * liQuid @bert
1126525946 M * Bertl liQuid: the linux-vserver stuff has neglectible overhead, and given that the guests can share some stuff (libraries, executables, ...) it can save a lot of resources
1126525980 M * Bertl liQuid: this, in contrary to Xen/UML approaches allows to use the resources more efficiently
1126526010 M * Bertl Greek0: google for it, basically you have: Value, Array, Dictionary
1126526023 M * liQuid well it looks like uml
1126526037 M * Bertl liQuid: yes, but we have a single kernel for all guests
1126526079 M * Bertl Greek0, Hollow: I can contribute a free/GPL rock-solid hash table implementation for the dictionary if you want to code the propertylists yourself ...
1126526095 M * wam Sorry, again my problem - I couldn't solve it. I read *every* doc in the wiki... I'm trying to limit disk-usage per context and used vdlimit for this. Inside the vhost, "df" shows the values I entered with vdlimit. But I can create files that are much larger than the (simulated) available disk. The vhost runs on a ext3-filesystem, which should support xid-tagging per inode, as described in the wiki. But somehow the values in "df" don't change when I create file
1126526117 M * Hollow hm, i'm too dumb to use google today..
1126526136 M * Bertl wam: did you try a non-debian version yet?
1126526154 M * wam Bertl: no, I compiled the latest sid-version of the util-vserver.
1126526163 M * Bertl wam: also, did you mount the partition with tagxid?
1126526172 M * wam Bertl: do I have to? How?
1126526174 M * Bertl (not remount, mount!)
1126526187 M * wam Bertl: I didn't see this in the wiki. Maybe I overread this... :/
1126526200 M * Bertl wam: mount -o tagxid,rw /dev/whatever /wherever
1126526217 M * wam aah ;) That could be it...
1126526230 M * Bertl if it is missing in the wiki, please add it ...
1126526287 M * wam is it possible to "retag" existing files in a mounted partition?
1126526301 M * wam because they all were created without tagging....
1126526315 M * Bertl wam: yes, that is done with chxid
1126526345 M * Bertl wam: so no problem there ...
1126526359 M * wam coool ;) thanks! I'll try and update the wiki!
1126526373 M * Bertl TIA!
1126526394 M * Bertl okay, off for now, back in an hour or so ...
1126526399 N * Bertl Bertl_oO
1126526590 M * Hollow well, seems like property lists are a bunch of arrays values and hashes... that's what libconfuse does anyway
1126526735 M * Greek0 Hollow: link?
1126526758 M * Hollow http://www.absoluteastronomy.com/encyclopedia/a/as/associative_array.htm
1126526782 M * Hollow it's not a definition of property lists
1126526809 M * Hollow i just guessed
1126526816 M * Hollow after googling without much success ;)
1126526918 M * Greek0 thanks.
1126526941 M * Greek0 I think it's funny as bertl drops by, spews some enlightment all over the place, and hurls away again ;)
1126526963 M * Hollow heh
1126527241 M * Hollow i'm bit off watching tv, just ping me..
1126527248 M * Greek0 mm
1126527325 Q * Doener Ping timeout: 480 seconds
1126527329 J * Doener ~doener@p5487682D.dip.t-dialin.net
1126528564 Q * liQuid Quit: ( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )
1126528611 Q * Hunger arion.oftc.net neutron.oftc.net
1126528611 Q * virtuoso arion.oftc.net neutron.oftc.net
1126528648 J * Hunger Hunger.hu@Hunger.hu
1126528648 J * virtuoso ~s0t0na@shisha.spb.ru
1126530448 Q * litage Ping timeout: 480 seconds
1126530899 J * Mac_Cloud ~julien@nat1.cri74.org
1126531015 M * Mac_Cloud hi all I am looking for a good doc to understand the new rlimits mecanics :)
1126531067 M * Mac_Cloud at least a kernel memory managment doc
1126531152 J * litage ~nick@203.201.96.81
1126531283 Q * Aiken Quit: Leaving
1126532151 M * Greek0 Mac_Cloud: uhm. I've got a book here that describes linux kernel memory management.. 748 pages..
1126532181 M * Greek0 somewhere the author said that he hopes that his book can bring down the time needed to understand linux kernel MM from 6 weeks to 3-4.. hopefully
1126532607 M * SiD3WiNDR =)
1126532770 Q * derbien oxygen.oftc.net nova.oftc.net
1126532770 Q * michal oxygen.oftc.net nova.oftc.net
1126532770 Q * ag-2 oxygen.oftc.net nova.oftc.net
1126532770 Q * ntrs oxygen.oftc.net nova.oftc.net
1126532770 Q * Vudumen oxygen.oftc.net nova.oftc.net
1126532770 Q * eyck oxygen.oftc.net nova.oftc.net
1126532770 Q * Medivh oxygen.oftc.net nova.oftc.net
1126532770 Q * ray6 oxygen.oftc.net nova.oftc.net
1126532770 Q * SiD3WiNDR oxygen.oftc.net nova.oftc.net
1126532770 Q * lonewolff oxygen.oftc.net nova.oftc.net
1126532803 J * derbien ~derbien@whiterabbit.nbmc.de
1126532803 J * michal ~michal@michal.usercloak.oftc.net
1126532803 J * ag-2 ag@muaddib.roxor.cx
1126532803 J * ntrs ~ntrs@Dardeene-68.188.50.87.charter-stl.com
1126532803 J * Vudumen vudumen@perverz.hu
1126532803 J * eyck eyck@81.219.64.71
1126532803 J * Medivh ck@paradise.by.the.dashboardlight.de
1126532803 J * ray6 ~ray@klon3.gcsc2.ray.net
1126532803 J * SiD3WiNDR luser@bastard-operator.from-hell.be
1126532803 J * lonewolff ~lonewolff@host86-128-17-74.range86-128.btcentralplus.com
1126532925 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126533295 Q * kas_3 Ping timeout: 480 seconds
1126533312 N * Bertl_oO Bertl
1126533320 M * Bertl evening folks!
1126533391 M * Bertl Mac_Cloud: what do you mean by 'new rlimit mechanics'?
1126533989 M * wam what is preferred for virtual servers, which are opened for clients: sched_hard or sched_prio?
1126534459 T * * http://linux-vserver.org/ | latest stable 2.0, 2.0.1-pre2, 1.2.10, 1.2.11-rc1, devel 2.1.0-pre7 | 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 ;) 
1126534459 T * Bertl -
1126534485 M * Mac_Cloud Bertl what are the things under all the rlimits flags
1126534514 M * Bertl cryo: wb say!
1126534527 M * Bertl Mac_Cloud: you are referring to the rmilits page?
1126534546 M * Mac_Cloud yes this one http://linux-vserver.org/Resource+Limits
1126534564 M * Bertl okay, and you want to know what the code means?
1126534580 M * Bertl or what resource limits are in general?
1126534616 M * Mac_Cloud I want to know what I can do exactly with this
1126534634 M * Mac_Cloud limit memory usage per vserver or cpu usage etc ...
1126534635 M * Bertl ah, okay ... you know the bash 'ulimit' command?
1126534647 M * Mac_Cloud yes
1126534661 M * Bertl okay, those limits are per user and/or per process
1126534667 M * Mac_Cloud oki
1126534683 M * Bertl they are listed with the 'U' code
1126534707 M * Bertl some of those limits make sense to be extended to 'per context' (i.e. per vps/guest)
1126534708 M * wam Bertl: Thanks!
1126534720 M * Bertl wam: you're welcome!
1126534735 M * Mac_Cloud R limits are the same but for securyity context
1126534735 M * cryo Bertl: thnx :)
1126534811 M * Bertl cryo: how is FreeVPS? 
1126534841 M * Mac_Cloud and if i want to knows in details what is   NPROC or  RTPRIO I can find it in umilits docs ?
1126534841 M * Bertl Mac_Cloud: yep, so they limit the entire context ..
1126534854 M * Mac_Cloud Bertl thanks
1126534857 M * Bertl most of them, except for the last 4
1126534872 M * Mac_Cloud ok
1126534878 M * Bertl (which were adde, because they do not exist in ulimits)
1126534924 M * Mac_Cloud Bertl ok thanks a lot !
1126534944 M * Mac_Cloud I ll read it and comme back for details ! :)
1126534945 M * Bertl Mac_Cloud: my pleasure!
1126534955 M * Bertl Mac_Cloud: feel free to idle around ...
1126535011 M * wam aah - that's the reason I can't see the rlimits inside the vhost. Is there a way to hide the real memory from the user, which she could for example view via top?
1126535049 M * Bertl yes, this is done via the virt_* and hide_* flags
1126535061 M * wam ok, I'll read.
1126535075 M * Bertl FlowerPage -> bcaps, ccaps, flags
1126535160 M * Hollow Bertl: could you elaborate on "property lists"?
1126535207 M * Bertl sure, the (very old) concept of property lists was developed to store/represent arbitrary data in a generic way
1126535211 M * cryo Bertl: Actually, I have not already returned to freevps, because now I'm working on next release of NOCMonkey with FreeBSD support. I will be back to FreeVPS next week, may be :)
1126535268 M * Bertl cryo: ah, i.c. well, good to that you found your way to us ... would be nice to have a chat now and then ... no?
1126535410 M * wam Bertl: so if I --echo "virt_mem" >> flags--, it takes rss*4 for mem and as*3,55 for swap?
1126535498 M * Bertl the 4 I udnerstand, the 3,55 escapes my comprehension :)
1126535515 M * Bertl actually limits are in pages, so it depends on the arch
1126535525 M * wam ah ok ;)
1126535643 M * Bertl and the VM/AS is taken as total display amount (which is wrong but looks nice)
1126535672 M * Bertl the RSS is used as RAM (in the display) and both are cropped to the actually available amount
1126535686 M * Bertl (because you can and usually will overbook them)
1126535854 M * wam that rocks
1126535871 M * Bertl yeah! :=
1126536080 Q * ray6 oxygen.oftc.net nova.oftc.net
1126536080 Q * Medivh oxygen.oftc.net nova.oftc.net
1126536080 Q * eyck oxygen.oftc.net nova.oftc.net
1126536080 Q * Vudumen oxygen.oftc.net nova.oftc.net
1126536080 Q * ntrs oxygen.oftc.net nova.oftc.net
1126536080 Q * derbien oxygen.oftc.net nova.oftc.net
1126536080 Q * lonewolff oxygen.oftc.net nova.oftc.net
1126536080 Q * SiD3WiNDR oxygen.oftc.net nova.oftc.net
1126536080 Q * michal oxygen.oftc.net nova.oftc.net
1126536080 Q * ag-2 oxygen.oftc.net nova.oftc.net
1126536107 J * derbien ~derbien@whiterabbit.nbmc.de
1126536107 J * michal ~michal@michal.usercloak.oftc.net
1126536107 J * ag-2 ag@muaddib.roxor.cx
1126536107 J * ntrs ~ntrs@Dardeene-68.188.50.87.charter-stl.com
1126536107 J * Vudumen vudumen@perverz.hu
1126536107 J * eyck eyck@81.219.64.71
1126536107 J * Medivh ck@paradise.by.the.dashboardlight.de
1126536107 J * ray6 ~ray@klon3.gcsc2.ray.net
1126536107 J * SiD3WiNDR luser@bastard-operator.from-hell.be
1126536107 J * lonewolff ~lonewolff@host86-128-17-74.range86-128.btcentralplus.com
1126536174 M * Bertl Hollow: the basic idea is, to store arbitrary data, you need a) soemthing to store values, b) a way to store indexed data a[i], and c) another container to store 'hashed' entries b["color"]
1126536215 M * Hollow in an abstract way i can do that quite fine with libConfuse
1126536315 M * Bertl I hope so, propertly lists are the 'basic' building blocks
1126536335 M * Bertl but IMHO they are also sufficient from the 'container' perspective
1126536627 M * cryo Bertl: not now, sorry, we have no time for next release, so works very hard.
1126536657 M * Bertl cryo: np, we have time :)
1126536705 Q * cryo Quit: Покидаю
1126536714 J * cryo ~say@gw.psoft.od.ua
1126536953 N * BobR BobR_afk
1126537917 M * wam ah, btw: I extended the page a little: http://linux-vserver.org/Disk+Limits  -- just in case someone wants to look over it, to make sure I didn't write complete crap ;)
1126537989 M * Bertl This only works on ext2 and ext3 at the moment. should work on xfs and reiser too, even jfs, IIRC ...
1126538002 M * wam oh - I read that on another page...
1126538011 M * Bertl (maybe check it?)
1126538105 M * Bertl "has to be run after every boot of the host-system" .. well, some folks (lycos and rosehosting, IIRC) developed a smart way to 'keep' the limits over a reboot, by storing them at guest shutdown, and using the stored values on startup .. you might want to mention taht option?
1126538139 M * Bertl wam: inodes are tagged, files are 'jsut imagination' :)
1126538143 M * wam ok.
1126538192 M * Bertl regarding ext2/3 only, I guess you read that because the 'internal' xid tagging is limited to 'certain' filesystems
1126538208 M * wam i think so. don't remember...
1126538214 M * wam I'll just delete that.
1126538224 M * Bertl but the usual UID24/GID24 and friends are available on most 32bit capable filesystems
1126538248 M * Bertl wam: but thanks for adding this stuff ...
1126538269 M * Bertl okay, off for dinner now .. back soon ...
1126538275 N * Bertl Bertl_oO
1126538572 J * stefani ~stefani@superquan.apl.washington.edu
1126538916 P * stefani I'm Parting (the water)
1126538983 Q * Mac_Cloud Quit: Leaving
1126538997 M * wam ok , stuff corrected: http://linux-vserver.org/Disk+Limits -- I have to go now. Please feel free to correct whatever you feel is wrong. Even my english ;)
1126539031 P * wam 
1126539493 N * Bertl_oO Bertl
1126539506 M * Bertl back now ...
1126540845 N * fluor fluor-
1126541167 N * BobR_afk BobR
1126541734 J * micah_ micah@micha.hampshire.edu
1126542126 N * BobR BobR_oO
1126542153 Q * micah Ping timeout: 480 seconds
1126542196 Q * eyck Quit: leaving
1126542488 J * eyck eyck@81.219.64.71
1126547980 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net
1126548020 J * wam ~wigwam@p549CD3A9.dip.t-dialin.net
1126548350 Q * dddd44 Ping timeout: 480 seconds
1126548873 Q * wam Ping timeout: 480 seconds
1126549227 N * BobR_oO BobR
1126549388 Q * BobR Quit: leaving
1126549509 N * micah_ micah
1126549560 J * wa1 ~wigwam@p549CFB1D.dip.t-dialin.net
1126549580 M * Bertl hey micah! wb wa1!
1126549608 M * micah hey Bertl!
1126549631 M * Bertl had a few reports/complaints about 'mising' backports for debian?
1126549707 M * micah hmm... missing backports?
1126549732 M * Bertl don't ask me, but I think folks want to use recent utilities (0.30.208+) on woody or sarge :)
1126549774 M * micah ah, ok... i'll look into backporting that
1126549791 M * micah 208 should be in sarge soon, although ola still hasn't updated the one in unstable to the new patch level ;p
1126549800 M * wa1 I did that today. at least for i386
1126549801 M * Bertl iirc. there was also one issue with dietlibc and/or libbeecrypt ...
1126549815 M * micah yes, but that has been resolved
1126549815 M * wa1 exactly. They are required in a recent version
1126549876 M * micah sep 3 Ola said he would upload a new version with the patches and build against the fixed libbeecrypt
1126549911 M * Bertl which year? (just kidding :)
1126549919 M * micah heheh
1126549950 M * mnemoc *G*
1126550124 J * spd1snd ~psingh@68-232-133-13.chvlva.adelphia.net
1126550301 M * wa1 maybe those users would be satisfied with that instructions: http://linux-vserver.org/util-vserver+on+sarge (I hope, the wiki is open for such articles / workingdrafts?)
1126550355 M * Bertl np, but maybe check with the distro specific pages 
1126550370 M * Bertl http://linux-vserver.org/Documentation 
1126551022 M * wa1 how do you mean?
1126551050 M * Bertl we have a section ther, called 'Distro specific Howtos'
1126551055 M * Bertl *there
1126551085 M * Bertl you might want to look at those, and if it suits, add this info there (either on the existing pages or as link)
1126551104 M * wa1 ah ;)
1126551511 M * spd1snd anyone running CPanel inside individual vservers? 
1126552118 M * spd1snd I've got a Gentoo machine setup to run VServers... can I install Fedora into a vserver at all? haven't been able to find much info through Google.
1126552363 M * Bertl spd1snd: you can basically isntall any distro of your liking inside a guest
1126552387 M * Bertl sometimes it requires a few modifications to the startup/shutdown scripts, but usually it's painless
1126552413 M * spd1snd Bertl: oh ok, i'll give that a shot then... thank you
1126552441 M * Bertl you're welcome!
1126553092 Q * monrad Quit: Leaving
1126553368 Q * alexx Ping timeout: 480 seconds
1126553736 Q * litage Read error: Connection reset by peer
1126553765 J * menomc ~amery@200.75.27.22
1126553863 Q * mnemoc Ping timeout: 480 seconds
1126553864 N * menomc mnemoc
1126554056 J * nayco ~nayco@lns-bzn-8-nan-82-250-229-159.adsl.proxad.net
1126554087 M * nayco Hello :) !!!
1126554410 M * Bertl hey nayco!
1126554569 M * nayco nice job ;)
1126554577 M * Bertl tx
1126554580 M * nayco oh, brb, meal ;)
1126554748 J * litage ~nick@203.201.96.19
1126554990 M * Bertl welcome litage!
1126555144 M * spd1snd anyone running Fedora as a guest on a Gentoo machine? if so, how did you go about installing fedora in the guest vserver?
1126555172 M * Bertl well, the apt-rpm should work as well as simply old copy ...
1126555195 M * spd1snd hmm, can you elaborate a bit?
1126555214 M * Bertl I assume you can compile the apt-get for rpm on gentoo, no?
1126555225 M * spd1snd well im emerging rpm on there now
1126555256 M * spd1snd ill try creating the vserver again, if i get another error... ill look for apt-rpm as well
1126555267 M * Bertl rpm itself might work too, probably apt-rpm is easier ...
1126555275 M * Bertl IIRC, you could also try/use yum
1126555634 P * wa1 
1126555955 M * spd1snd vserver servername build -m apt-rpm --hostname=servername --interface eth0:1.2.3.4/25 -- -d fc3
1126555964 M * spd1snd is that the right syntax to install fc3?
1126555989 M * Bertl looks good
1126556011 M * spd1snd ok, well gentoo does not have any packages for apt-rpm
1126556019 M * spd1snd can i replace that with rpm and give it a shot?
1126556041 M * Bertl ahem, no, as rpm does not support downloading from a repository
1126556047 M * spd1snd ah
1126556055 M * Bertl well, you can download the packages yourself, and then use the rpm emthod
1126556089 M * spd1snd ah ok, well it looks like there is no package to emerge yum with
1126556097 M * spd1snd i guess ill need to download the packages
1126556120 M * spd1snd so if i download all of the packages from a repository, what dir should i be putting them in?
1126556307 M * Bertl anywhere, you specify that with the rpm build method (see vserver - build --help)
1126557180 J * _maharaja maharaja@ip52.ipax.at
1126558054 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126558400 Q * kas_3 Ping timeout: 480 seconds
1126560201 J * Aiken ~james@tooax6-241.dialup.optusnet.com.au
1126560595 Q * dddd44 Ping timeout: 480 seconds
1126561141 M * nayco Bertl: What about the vroot devices ? How does it work ?
1126561160 M * Bertl morning Aiken!
1126561168 M * Bertl nayco: it works like a proxy ...
1126561184 M * Bertl basically just passing on the quota commands, but blocking normal I/O
1126561341 M * Aiken cold morning :(
1126561353 M * Aiken someone forgot to tell the weather this is spring :(
1126561701 M * nayco Bertl: Ok. What are the steps needed  to configure a vserver to do quota with vroot ? Is the wiki's documentation up to date ? And more, do you think there is a way to automate the setup for quotas with vroot in the utils, based on parameters in the config directory ?
1126561726 M * Bertl you now have it working with hda8/9 copied into the guest, right?
1126561735 M * nayco Aiken: Well, be happy, fall begin here :( ... ;)
1126561744 M * nayco Bertl: Yep.
1126561749 M * Bertl did you try with the /home partition yet?
1126561755 M * nayco yes
1126561763 M * Bertl works fine?
1126561852 M * nayco fine, yes, my poor test user (me) had only a 10Mb quota (I am a true bastard), and creating two files with "dd" (9Mb and 1Mb) showed the limits ;)
1126561880 M * Bertl excellent!
1126561906 M * Bertl okay, the vroot device is a replacement for hda8/9 inside the guest
1126561913 M * nayco yes..
1126561931 M * Bertl so you configure for example vroot1 for hda8, and then copy the vroot1 into the guest, instead of the hda8
1126561933 Q * spd1snd Quit: spd1snd
1126561949 M * nayco so the docs are up to date ;)
1126561981 M * Bertl good to know :)
1126562008 M * nayco that is this procedure that i'd like to be automated by the tools in the future... But right now, i'm gonna test it manually, to learn.
1126562018 M * Bertl k
1126562088 M * nayco http://linux-vserver.org/Standard+non-shared+quota <= is it this one ? (24 july)
1126562089 M * Hollow going to dream of future tools too ;) night everyone
1126562098 M * nayco 'night ;)
1126562290 M * nayco Mmmm, some points are not clear to me in this doc...
1126562361 M * nayco First (Oh, this one is clear enough, but sad) : "you should have a directory /dev/vroot/ with 8 vroot devices (0-7)" => What if I need 10 or 20 quota-enabled partitions ?
1126562363 M * Bertl night Hollow!
1126562388 M * Bertl nayco: then you need to raise the max_vroot up to 255
1126562396 M * Bertl (which is currently the upper limit)
1126562402 M * nayco at kernel config type ?
1126562406 M * nayco sorry
1126562413 M * nayco at kernel config time ?
1126562426 M * Bertl either kernel boot command line or at module load time, if compiled as module
1126562468 M * nayco ...mmm... so, not in .config ?
1126562488 M * Bertl no, currently not, do you consider this a feature?
1126562680 M * nayco Well, dunno... If eight is the default, it will certainly be not enough for large file vserver hosts, but, to speak frankly,  I currently don't nee more than one or two quota-enabled partitions on my vservers, and have only one file vserver per host. But with bigger servers, who knows... And I'm not the only one to use/need these features, I guess ;)
1126562770 M * nayco And I think giving options to the kernel on the command line can get heavy if there are to many. But, hey, I'm not very experienced with kernels, and surely this method is the good one !
1126562923 M * nayco or it would be more elegant to pass the option to the module when loading it, in /etc/modules.???, if possible (dunno how it works)...  But well, built-in vroot seems to be the default in the kernel conf, so...
1126562981 M * nayco ...Er... Is vroot device a standard kernel feature, or a vserver-related one made by you ?
1126563017 M * Bertl a vserver related one :) but it is very similar to the loop device, so I chose similar defaults
1126563048 M * nayco Strange, it is not under vserver section in menuconfig !!
1126563065 M * Bertl yes, it's under block devices
1126563109 M * nayco ...Is making the default number bigger is ressource-consuming for the kernel ?
1126563143 M * Bertl yes, but not very much ... very similar to the loop device
1126563299 M * nayco well, maybe making it 16 would be cool. But, well, I got no experience with it, so... When I have a multi-processor-hyperthreading-multicore server, with 10 teraBytes fibre-channel storage to host tenth of samba vserver for 5000 users, well, that day, I me be accurate :D :D !
1126563721 M * nayco oh, talking about that, I stressed-test a few hosts and vservers today: Remember I have 4 distcc vservers (on 4 hosts), plus one non-vserver distcc, including PIII (866,1133,1400) and 2*PIV HT 2800 ? Well, I installed BOINC with seti@home in two of these vservers, and launched the compilation of my newest host's vserver kernel with distcc (I sure know how to party) : The load on the BOINC+distcc vservers grown up to ~3.00, and 
1126563722 M * nayco the kernel+modules were correctly built (Well, no magic here) in about 30 minutes. No errors nor instabilities. Well, if that helps ;)
1126563888 M * nayco And, again, the distcc-cleint machine is definitely the bottleneck : With a similar setup my laptop took about 1 hour to achieve its task, while the PIV HT 2.8 GHz used today only ~30 minutes. This is sad...
1126564133 M * Bertl it is also advantageous to compile with 'make -j N' N being a number between #cpus*2 and #cpus*3
1126564163 Q * maharaja Ping timeout: 480 seconds
1126564169 N * _maharaja maharaja
1126564216 M * nayco well, I used (2*cpu)+1, and because It gave 9, I rounded to 10.
1126564273 M * [MUPPETS]Gonzo don't give a Number, it will use all avaiable cpu :)
1126564289 M * nayco well, I'm testing vroot on my laptop.
1126564325 M * nayco [MUPPETS]Gonzo: thanks, i'll try that ;) ... Er, do I need to tell "-j" in this case ?
1126564333 M * nayco for "make"
1126564413 M * nayco My mistake : I used cpu+1, not (2*cpu)+1 !
1126564560 M * [MUPPETS]Gonzo cpu * 2 is a good value most times
1126564590 M * nayco Oh, Bertl : Another thing, I _always_ compile my vs kernels with preemptible and SMP. My PIV HT are recognized as 2 cpus (Well, after activating it in the bios, because they were delivered with HT to "off" =( !!!)
1126564635 M * nayco [MUPPETS]Gonzo: Thanks. That mean an HT proc "gives" +4 to N ?
1126564638 M * nayco +s
1126564670 M * [MUPPETS]Gonzo I would use -j 4    on HT P4
1126564675 M * nayco k
1126564680 M * [MUPPETS]Gonzo or if nothing else runs on the server, use -j
1126564685 M * [MUPPETS]Gonzo without any number
1126564711 M * nayco so make hogs the cpu to the value it finds best ?
1126564716 M * Aiken do a bit of benchmarking to see which N with -j N is faster
1126564749 M * Bertl *nap attack* probably off for today ...
1126564755 M * nayco is it the same thing when compiling on a simple host ?
1126564791 M * Bertl nayco: I'm confident you will master the vroot setup/testing ... let me know how it went ...
1126564807 M * Bertl night everyone ... cya tomorrow!
1126564809 M * nayco :D !!! Ok, nice sleep 
1126564818 N * Bertl Bertl_zZ
1126565451 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net
1126566465 M * nayco Does not fully work... Mountpoint resolution failures: "repquota /home" fails, but "repquota -a" shows /home quotas :O !!! ok, 'night all !
1126566522 Q * nayco Quit: `sleep 30000`
1126566600 Q * Loki|muh_ Read error: Connection reset by peer
1126566626 J * Loki|muh loki@satanix.de
1126568075 J * lilo ~lilo@lilo.usercloak.oftc.net
1126568203 Q * lilo_ Ping timeout: 480 seconds