1182730626 M * slack101 back to setting up Vserver
1182730630 M * slack101 going to get this donetoday 
1182730804 M * slack101 kerrnel processes dont show in vserver do they ?
1182730898 M * Bertl no, usually not, but if yes, then it is a bug :)
1182730987 M * slack101 just chekcing Bertl
1182730989 Q * [BiG^BrotheR] Quit: ][DreaM-ScripT][
1182731297 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1182732253 J * DoberMann_ ~james@AToulouse-156-1-46-224.w90-16.abo.wanadoo.fr
1182732355 Q * DoberMann[ZZZzzz] Ping timeout: 480 seconds
1182733836 Q * Piet Quit: Piet
1182735991 Q * slack101 Quit: Leaving.
1182736926 Q * ||Cobra|| Ping timeout: 480 seconds
1182737416 J * slack101 ~Administr@cpe-71-74-77-84.insight.res.rr.com
1182738808 M * slack101 anyone here ever use debootstrap
1182738928 M * Bertl not directly, just via the tools 
1182738989 M * slack101 Bertl: yea im using it directly 
1182739002 M * Bertl why?
1182739014 M * slack101 getting a base default install for my host system
1182739050 M * Bertl ah, interesting ...
1182739062 M * Bertl always installed debian via network install
1182739088 M * Bertl I mean, via some install floppy/cd/netboot image
1182739406 M * slack101 Bertl: netboot image ?
1182739414 M * slack101 i think debian is the best distro lol
1182739421 M * slack101 fast small and easy
1182739433 M * slack101 all the reasons i like slackware and it has good package management
1182739665 M * Bertl well, when I installed it (so far) I used a small installer image which could be booted over network
1182739681 M * Bertl (as most machines I installed it on, didn't have a floppy or cdrom drive)
1182739996 M * slack101 this is a remote server 
1182740003 M * slack101 a installer iamge ?
1182740009 M * slack101 you got any links of anyting about this ?
1182740020 M * Bertl shouldn't be hard to find .. sec
1182740091 M * Bertl http://ftp.debian.org/debian/dists/sarge/main/installer-sparc/current/images/sparc64/netboot/
1182740095 M * Bertl (just one example)
1182740430 M * slack101 i wonder if thats better then the 
1182740438 M * slack101 deboostrapp
1182740464 Q * Johnnie Ping timeout: 480 seconds
1182740486 M * Bertl well, it will use debootstrap too :)
1182740505 M * slack101 ooo?
1182740530 M * Bertl but wrapped in a nice installer :)
1182740596 M * slack101 so
1182740609 M * slack101 i can jus put that on my server andf it will automatically install it ?
1182740619 M * slack101 i would need a remote console for that though corect ?
1182740646 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1182740683 M * Bertl yep
1182740698 M * Bertl but a serial one is sufficient for debian
1182741011 J * Johnnie ~jdlewis@c-67-163-246-136.hsd1.pa.comcast.net
1182741514 Q * Johnnie Ping timeout: 480 seconds
1182742062 J * Johnnie ~jdlewis@c-67-163-246-136.hsd1.pa.comcast.net
1182742201 Q * Wonka Remote host closed the connection
1182742502 J * Wonka produziert@chaos.in-kiel.de
1182742906 M * Bertl okay, off to bed now .. have a good one everyone!
1182742912 N * Bertl Bertl_zZ
1182746517 J * meandtheshell ~markus@85.127.102.3
1182748186 J * lylix ~eric@dynamic-acs-24-154-33-109.zoominternet.net
1182748616 M * slack101 is /dev the same on all guest ?
1182748634 M * slack101 like is it set in the kerrnel exaclty which stuff they have access too ?
1182748823 M * coderanger  /dev isn't managed by the kernel 
1182748891 M * slack101 sorry
1182748895 M * slack101 wrong use of words
1182748908 M * lylix what you see is what they get...
1182748924 M * slack101 i was jus curious 
1182748935 M * slack101 how hte /dev is made 
1182748948 M * slack101 i know how its made but who / what decides what hardware to include 
1182748963 M * lylix there is a minimalist tarball around that is generally all that is needed...
1182748982 M * slack101 i use skeleton mode 
1182748988 M * slack101 so it gets created
1182748997 M * slack101 so it only has access to certain parts of the kerrnel but yea 
1182749007 M * lylix http://mirrors.sandino.net/vserver/images/dev.tar.gz
1182749036 M * lylix idk, dont use skeleton, but it prob sets the mins
1182749042 M * slack101 well it was a stupid question
1182749050 M * slack101 why not use skeleton mode ?
1182749068 M * lylix i dont use util-vserver ;)
1182749104 M * slack101 oooooo
1182749116 M * slack101 so you dont have vserver-stat or nothing ?
1182749124 M * lylix nope
1182749641 M * slack101 lylix: how do you keep track and know and see what vservers you are running ?
1182749893 M * lylix there are other tools w/ the software we're using
1182750113 M * lylix but i usually write shell wrappers to get the info i want, ie. vls = vserver-stat
1182750156 M * lylix w/ ability to sort wach column, ie. by rss usage, xis, et. al
1182750165 M * lylix xis=>xid
1182750367 M * slack101 lylix: why not jsut use vserver-stat ?
1182750448 M * lylix it doesn't exist... ~~~
1182751022 Q * Johnnie Ping timeout: 480 seconds
1182751175 J * DavidS ~david@p54810930.dip0.t-ipconnect.de
1182751217 N * DavidS DavidS|Vechta
1182751586 J * Johnnie ~jdlewis@c-67-163-246-136.hsd1.pa.comcast.net
1182751650 M * slack101 lylix: its in util vserver i would think
1182751684 M * slack101 also a question i never asked is how is iptables managed in linux vsserver i see it uses isolation
1182751701 M * lylix from the host
1182751706 J * phedny ~mark@ip56538143.direct-adsl.nl
1182752553 M * slack101 lylix: but how are things seperated ? by ip's ?
1182753707 M * lylix yep
1182753808 M * lylix if you want iptables control within the vserver, thats a different beast
1182753839 M * lylix there is a php daemon & shell script client set floating around that handles this
1182753897 M * lylix and supposedly some other kernel level patchset in the works (?)
1182754395 J * ktwilight ~ktwilight@165.197-66-87.adsl-static.isp.belgacom.be
1182754718 J * dna ~naucki@189-235-dsl.kielnet.net
1182754745 Q * ktwilight_ Ping timeout: 480 seconds
1182754757 J * ktwilight_ ~ktwilight@221.68-66-87.adsl-dyn.isp.belgacom.be
1182754809 N * DoberMann_ DoberMann[PullA]
1182755016 Q * ktwilight Ping timeout: 480 seconds
1182756969 Q * ||Cobra|| Remote host closed the connection
1182756975 J * andrew ~andrew@linux3.cc.ntu.edu.tw
1182757027 Q * AndrewLee Quit: leaving
1182757429 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1182758246 J * bzed ~bzed@10-205-116-85.dsl.manitu.net
1182758419 M * awk hmm, vserver-copy doesn't work anymore?
1182758436 M * awk it states that /etc/vservers/name.conf doesn't exsist?
1182758938 M * DavidS|Vechta the latter is (very) old legacy stuff ... don't know about -copy though
1182759049 M * awk well i have done the copy by hand
1182759053 M * awk where do I set the context though ?
1182759059 M * awk cant find it in /etc/vservers....
1182759064 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1182759079 Q * andrew Quit: leaving
1182759106 J * AndrewLe1 ~andrew@flat.iis.sinica.edu.tw
1182759110 Q * AndrewLe1 
1182759161 M * awk never mind, all working
1182759215 M * DavidS|Vechta echo $UNIQUE_RANDOM > /etc/vservers/$NAME/context   # IIRC
1182759525 M * awk 1 last question if you don't mind.
1182759546 M * awk what capabilities needs to be added for within a vserver to have a certain device?
1182759550 M * awk I need /dev/zap in my vserver
1182759566 M * awk it's allready loaded in the host
1182760213 M * awk root@pbx:/# zttool 
1182760213 M * awk Unable to open /dev/zap/ctl: No such file or directory
1182760236 M * awk as you can see I can view the modules with lsmod but I need that /dev/zap/* to be passed to the vserver
1182760430 M * DavidS|Vechta awk: just copy them in / mknod them
1182760434 M * DavidS|Vechta (from outside
1182762983 Q * bzed Quit: Leaving
1182763001 J * HeinMueck ~Miranda@host-88-217-199-211.customer.m-online.net
1182763890 M * awk DavidS|Vechta: so add something say to rc.local on the host? so as to have them in on each reboot
1182764184 M * awk DavidS|Vechta: how would a symbolic work?
1182764189 J * pmenier ~pmenier@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr
1182764300 M * awk root@pbx:/# zttool 
1182764301 M * awk Unable to open /dev/zap/ctl: Too many levels of symbolic links
1182764304 M * awk that doesn't work...
1182764421 M * lylix symbolic wouldnt work, in the chroot it would just point to itself :)
1182764433 M * awk ye I see
1182764440 M * DavidS|Vechta you can add hook scripts into the vserver config ... read the flower page for details
1182764462 M * lylix do as DavidS says, make the device nodes
1182764480 N * DavidS|Vechta Simo1
1182764491 N * Simo1 DavidS|Vechta
1182764492 M * DavidS|Vechta ;)
1182764562 M * awk mkdir -p /dev/zap && mknod /dev/zap/ctl c 196 0 && mknod /dev/zap/timer c 196 253 &&mknod /dev/zap/channel c 196 254 && mknod /dev/zap/pseudo c 196 255
1182764565 M * awk that seems to do the job
1182764567 M * lylix not sure why you want to mess w/ init/script/hooks... ince they're there, they are persistent
1182764575 M * lylix ince=once
1182764578 M * awk so i'll just add that to an rc script from within the vserver
1182764615 M * awk least now I can read the zap channels, etc
1182764641 M * lylix wont work in an rc script
1182764680 M * lylix mknod does work unless you give the vserver the right caps... and as mentioned, no need... unless you are purposely wiping the nodes on halt
1182764689 M * lylix does=doesnt... its late :/
1182764707 M * awk lylix I have allready gave it mkdir caps
1182764716 M * awk err mknod, as I stated earlier its all working.. thanks
1182764722 M * lylix still dont see the point :/
1182764736 M * lylix your doing alot of extra that isnt required, heh
1182764753 M * awk lylix so how do I get the devices in /dev/zap in your opinion? 
1182764785 M * awk everything is scripted now, so its no extra work anymore.
1182764790 M * lylix add them to the /vserver/$NAME/dev/zap/ directory from the host... 
1182764800 M * lylix do it once and its done, no need to redo
1182764808 M * awk lylix oh
1182764809 M * lylix trust me, ive done it 1000X
1182764810 M * awk let me try
1182764820 M * lylix ++
1182764824 M * awk but if done from within the vserver its allready there
1182764831 M * awk so I can just now remove the caps mknod ?
1182764835 M * lylix why give it the caps though... 
1182764837 J * chand ~chand@212.99.51.254
1182764885 M * awk you miss what I said,I said so I can remove the caps now?
1182764893 M * lylix unless you have complete control of the vserver, thats dangerous
1182764896 M * lylix yes
1182764926 M * awk great, let me remove it
1182765099 M * lylix same concept works for other apps... ie hylafax, et. al.
1182765138 M * DavidS|Vechta yeah, malicious vservers with CAP_MKNOD can break your system
1182765166 M * DavidS|Vechta e.g. mknod hda && cat /dev/zero > hda
1182765176 M * DavidS|Vechta or read /dev/kmem
1182765244 M * awk what do you mean by malicious vservers?
1182765269 M * DavidS|Vechta awk: clients, crackers, script kiddies
1182765291 M * DavidS|Vechta not necessarily in that order
1182765309 M * awk ahh but nobody has access to the vserver except me
1182765327 M * awk but I guess 1 less piece of control over the system is all that much better if it does get ./0dayed
1182765454 M * DavidS|Vechta you got the right there
1182766793 J * cedric ~cedric@80.70.39.67
1182766894 M * eSa| do some has util-vserver0.30.213 compiled for etch i386?
1182767075 Q * AndrewLee Ping timeout: 480 seconds
1182767187 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1182768047 M * eSa| I've found one here which seems up-to-date
1182768049 M * eSa| http://www.zbla.net/debian/
1182768643 Q * eSa| Ping timeout: 480 seconds
1182768874 N * pmenier pmenier_off
1182770109 J * ktwilight ~ktwilight@36.220-66-87.adsl-static.isp.belgacom.be
1182770171 J * bonsaikitten dreeevil@dev.gentooexperimental.org
1182770179 M * bonsaikitten hey people ... it's me again :-)
1182770188 M * bonsaikitten got a really weird issue now
1182770213 N * Bertl_zZ Bertl
1182770215 M * bonsaikitten within the vserver sshd accepts one connection, then falls into a coma. Also sudo hangs
1182770217 M * Bertl morning folks!
1182770240 M * bonsaikitten I'm not sure, but I assume it's a lack of usable /dev/pty devices
1182770245 M * bonsaikitten Bertl: good morning!
1182770248 M * Bertl bonsaikitten: probably exhausted entropy pool :)
1182770288 M * bonsaikitten no, it connects, but then hangs with "Entering interactive session"
1182770348 M * Bertl well, do you have a problem to start screen on the host?
1182770356 M * bonsaikitten the host works well
1182770371 M * bonsaikitten I can even enter the vservers from the host without problems
1182770396 M * Bertl with enter, but not ssh I presume?
1182770401 M * bonsaikitten screen as root within the vserver works also
1182770413 M * bonsaikitten yes, vserver foo enter, that works
1182770415 M * Bertl then it cannot be the ptys
1182770425 M * Bertl because screen will allocate one
1182770444 M * bonsaikitten ok, so why do ssh and sudo fail? :-)
1182770447 M * Bertl check 'cat /proc/sys/kernel/random/entropy_avail' on the host
1182770462 M * bonsaikitten 604
1182770469 M * Bertl not very much :)
1182770492 M * bonsaikitten should be enough for one ssh session!? ;-)
1182770495 M * Bertl wait until it reaches 1200+
1182770516 Q * ktwilight_ Ping timeout: 480 seconds
1182770548 J * lilalinux__ ~plasma@dslb-084-059-028-208.pools.arcor-ip.net
1182770549 M * bonsaikitten that might take some time
1182770558 M * bonsaikitten seems to be oscillating between 400 and 700
1182770579 M * Bertl see, something is using up your entropy
1182770584 J * ktwilight_ ~ktwilight@246.220-66-87.adsl-static.isp.belgacom.be
1182770596 M * Bertl probably an sshd key generation or so
1182770622 M * Bertl get a good entropy source to add ...
1182770630 M * bonsaikitten also weird ... load seems to go up continuously
1182770650 M * Bertl because more and more processes get stuck waiting for entropy
1182770667 M * Bertl did you create a number of guests recently?
1182770674 M * bonsaikitten I added two or three
1182770680 M * Bertl debian?
1182770683 M * bonsaikitten ah, found the bugger ... cron goes crazy
1182770685 M * bonsaikitten no, all gentoo
1182770711 M * Bertl with sshd each, I presume?
1182770753 Q * ktwilight Ping timeout: 480 seconds
1182770753 M * bonsaikitten yes
1182770754 M * sid3windr hehe
1182770769 M * Bertl you might consider increasing the pool size before you do entropy intensive stuff
1182770770 M * sid3windr the idea of using the motion sensors in thinkpads is really cool
1182770779 M * sid3windr "please shake your laptop to generate your ssh key"
1182770791 M * Bertl (especially on servers which have no good entropy source)
1182770792 M * bonsaikitten sid3windr: that might be difficult with a rackmounted server
1182770798 M * sid3windr etch-a-sketch style ;)
1182770807 M * bonsaikitten "please trigger an earthquake for key generation"
1182770808 M * sid3windr "please kick over the rack"
1182770819 M * sid3windr "keep removing floortiles until collapse"
1182770842 M * Bertl what works quite nicely for servers is to attach a microphone :)
1182770851 M * bonsaikitten hmmm, entropy gets depleted quite fast
1182770853 J * Piet hiddenserv@tor.noreply.org
1182770863 M * Bertl the noise in rack rooms are a good entropy source :)
1182770877 M * Bertl s/are/is/
1182770889 M * Bertl bonsaikitten: vps | grep key
1182770906 J * lilalinux ~plasma@dslb-084-058-245-221.pools.arcor-ip.net
1182770915 M * bonsaikitten Bertl: nothing
1182770921 Q * lilalinux_ Ping timeout: 480 seconds
1182770943 M * Bertl bonsaikitten: hmm, maybe you have something else draining entropy?
1182770958 M * Bertl bonsaikitten: like a copy/dd from /dev/random?
1182770970 M * bonsaikitten unlikely
1182770995 M * Bertl ipsec or vpn?
1182771010 M * bonsaikitten possibly vpn
1182771028 M * sid3windr Bertl: extra entropy when a raid array is failing! beep.. beep.. beep.. :)
1182771063 M * Bertl actually the beep has less entropy than the white noise :)
1182771071 M * sid3windr yesh
1182771080 M * sid3windr but beep+varying noise !
1182771081 M * sid3windr =)
1182771082 M * sid3windr hehe
1182771111 M * Bertl :)
1182771112 M * DavidS|Vechta good point with the microphone .. I always wondered, why all those server machines have on board sound :)
1182771127 M * Bertl yeah, it's aprefect match :)
1182771129 M * DavidS|Vechta i guess one then needs egd or so?
1182771153 M * Bertl for most sound systems, yes
1182771162 M * DavidS|Vechta that'd be a nifty gadget: microphone integrated into the plug :)
1182771197 M * Bertl not hard to do, costs you about 2 EUR :)
1182771201 J * rgl ~Rui@84.90.10.107
1182771203 M * rgl hello
1182771208 M * Bertl welcome rgl!
1182771225 Q * lilalinux__ Ping timeout: 480 seconds
1182771225 M * rgl hi :)
1182771246 P * DavidS|Vechta 
1182771257 M * rgl have you guys built the server using ubuntu/debian make-kpkg to produce .deb files?
1182771262 J * DavidS|Vechta ~david@p54810930.dip0.t-ipconnect.de
1182771272 M * DavidS|Vechta damn ^W
1182771279 M * Bertl rgl: hmm?
1182771295 M * bonsaikitten Bertl: hmmm, I wonder ... I can ssh to the host, but not the vserver. Thus unlikely that it's only entropy related ...
1182771296 M * rgl errr, s,server,kernel
1182771367 M * Bertl bonsaikitten: try to chroot into a guest and start the sshd there (for the guest) then see if it works better ...
1182771378 M * bonsaikitten Bertl: no, that does not help
1182771396 M * rgl Bertl, I'm fuzzing with having the root FS (/) on LVM2 stripped in two disks, but when I build the vanilla kernel, it fails to boot unless I supply a initrd, which seems to do some vudu to make it work.  so, I'll want to also create an initrd for this vanilla+vserver kernel :D
1182771400 M * Bertl bonsaikitten: then it is not Linux-VServer related either :)
1182771408 M * bonsaikitten :-(
1182771413 M * bonsaikitten well, thanks for the support
1182771422 M * Bertl bonsaikitten: you're welcome!
1182771426 M * bonsaikitten I'll have to figure out what's wrong on my own :-)
1182771466 M * Bertl rgl: doesn't your distro provide 'mkinitrd' or 'mkinitramfs'?
1182771491 M * Bertl bonsaikitten: I would definitely check where the entropy is bein used up .)
1182771517 M * sid3windr DavidS|Vechta: you have server machines with onboard sound?
1182771518 M * rgl Bertl, *shrug*  but oh, there is http://linux-vserver.org/Installation_on_Ubuntu  which is using mkinitramfs :)
1182771556 M * sid3windr hmm, yesh - usb plug with sound2usb with built-in microphone, 2x1x0.5cm entropy generator 
1182771826 M * Bertl well, a really good source ca be built with a fire detector and a web cam :)
1182771834 M * Bertl *can
1182771837 M * rgl Bertl, oh so sweet, it worked.   thx for putting up with my lameness ;-)
1182771847 M * Bertl rgl: np
1182771865 M * sid3windr a fire detector?
1182771940 M * Bertl yeah, well, US fire detectors have a small radiation source
1182772003 M * DavidS|Vechta sid3windr: 00:1b.0 Audio device: Intel Corporation 82801H (ICH8 Family) HD Audio Controller (rev 02)
1182772025 M * DavidS|Vechta which, i have to admit, is a custom built "server"
1182772058 M * DavidS|Vechta all my mini-ITX i use as firewalls also have on-board sound too, though i'd hardly call _them_ servers ;)
1182772063 M * sid3windr :)
1182772069 M * sid3windr mini itx are rather all-purpose ;)
1182772084 M * DavidS|Vechta but, missing hdd activity, they really need more entrop y:)
1182772109 M * sid3windr hmm
1182772116 M * sid3windr mine don't seem to have trouble with that
1182772123 M * sid3windr don't they get entropy from network traffic?
1182772148 M * DavidS|Vechta i checked (didn't look for a long time) and after adding the hw-rng it looks quite good: http://www.edv-bus.at/munin/edv-bus.at/fw-schmidg.edv-bus.at-entropy.html
1182772179 M * DavidS|Vechta it does depend on the machine though: http://www.edv-bus.at/munin/mariatreu.at/fw-maria.mariatreu.at-entropy.html
1182772192 M * DavidS|Vechta (both the same hw and basic config)
1182772202 M * rgl humm, whats the advised user program do use?  util-vserver or vserver-utils? :D
1182772230 J * ema ~ema@rtfm.galliera.it
1182772409 Q * ema 
1182772438 M * bonsaikitten aha, I think I have narrowed it down
1182772446 M * DavidS|Vechta rgl: util-vserver
1182772469 M * bonsaikitten looks like /dev/pts has bad permissions, can run screen as root, but not as normal user
1182772471 M * rgl DavidS|Vechta, thx.  you known if there is a ubuntu package for it?
1182772541 M * Loki|muh wasn't there a command line switch to enable debugging output to vserver xy start?
1182772545 J * dna_ ~naucki@86-218-dsl.kielnet.net
1182772556 M * DavidS|Vechta rgl: no idea
1182772556 M * Bertl Loki|muh: --debug ?
1182772561 M * rgl DavidS|Vechta, oh there is.  nm :)
1182772751 M * Loki|muh Bertl: oops, sorry, thank you
1182772784 M * Bertl Loki|muh: np
1182772953 Q * dna Ping timeout: 480 seconds
1182774316 M * rgl gag.. util-vserver segfaults :(
1182774384 J * gerrit gerrit@host661461325d.prpl.res.tor.fcibroadband.com
1182774441 J * ema ~ema@rtfm.galliera.it
1182775090 Q * gerrit Ping timeout: 480 seconds
1182775195 M * rgl btb. later guys.
1182775196 P * rgl Leaving
1182776166 N * dna_ dna
1182776566 Q * HeinMueck Ping timeout: 480 seconds
1182776820 M * bonsaikitten hah, I fix0r!
1182776832 M * bonsaikitten it was a hanging metalog that blocked a few things
1182776839 M * bonsaikitten using syslog-ng ssh works again
1182776848 A * bonsaikitten headdesks
1182776863 M * Bertl m4s70r fix0r0r! :)
1182776869 M * bonsaikitten y0
1182776880 M * bonsaikitten well, you learn something new every day ...
1182776888 M * Bertl yep, my words ...
1182776900 J * gerrit ~gerrit@fw54.torolab.ibm.com
1182776948 M * bonsaikitten silly power failure has cost me ~8h to fix everything
1182776955 M * bonsaikitten that'll teach me to test more :-)
1182777010 M * Bertl derjohn: ping?
1182777126 M * derjohn Bertl, hello !
1182777183 N * pmenier_off pmenier
1182777218 M * Bertl derjohn: hey, I have a patch for the diskio accounting (almost) ready
1182777268 M * derjohn Bertl, an own one? or adapted the atop !
1182777288 M * Bertl well, a vserver patch based on the aatop accounting
1182777368 Q * dna Ping timeout: 480 seconds
1182777405 M * bonsaikitten that sounds wicked
1182777483 M * Bertl derjohn: http://vserver.13thfloor.at/Experimental/delta-ioacc-feat01.diff
1182777508 M * Bertl (needs some decent testing on a real world scenario)
1182777509 M * derjohn fine, I had a look into atop itself and saw that it doesnt apply to a vserver prepatched kernel. ( In sched etc.), did you simply make that fit? Or did you change more? Can I use atop userspace ? where can I find that patch? 
1182777515 M * derjohn race ...
1182777529 M * derjohn ah, you called it a new vserver feature :)
1182777530 M * Bertl you don't need any userspace
1182777539 M * derjohn even better.
1182777550 M * Bertl i.e. the values (per guest) will show up in proc for now
1182777559 M * derjohn it should work an all $arch? I'll try on amd64 first.
1182777574 M * Bertl yep, as usual, we support all archs :)
1182777578 J * dna ~naucki@86-218-dsl.kielnet.net
1182777592 M * DavidS|Vechta "support"? *g*
1182777627 M * Bertl well, "support" hear means: we code for all archs and with all archs in mind :)
1182777632 M * Bertl *here
1182777651 M * DavidS|Vechta good :)
1182777675 M * Bertl contrary to other projects (Xen, OVZ, ...) which only "support" a few archs :)
1182778321 Q * bonsaikitten Quit: thanks and bye-bye!
1182778540 M * meandtheshell Bertl: and that is good because
1182778540 M * meandtheshell Specialization is for Insects.
1182778540 M * meandtheshell       -- Robert A. Heinlein
1182778545 M * meandtheshell ;-]
1182778675 M * Bertl lol
1182779166 Q * Aiken Quit: Leaving
1182779853 Q * sladen Ping timeout: 480 seconds
1182779886 J * sladen paul@starsky.19inch.net
1182780223 J * HeinMueck ~Miranda@host-88-217-199-211.customer.m-online.net
1182781469 M * Bertl nap attack ... back later ...
1182781478 N * Bertl Bertl_zZ
1182781543 Q * HeinMueck Quit: Aah!
1182781552 J * HeinMueck ~Miranda@host-88-217-199-211.customer.m-online.net
1182782211 Q * arachnist Quit: bbl
1182783462 M * jkl good morning everyone
1182783500 M * jkl does anyone know how to increase the size of /tmp within a vserver w/o restarting?
1182783536 M * daniel_hozac vnamespace -e guest mount -n -o remount,size=32m /vservers/<guest>/tmp or something like that.
1182783759 A * ard has 2 versions:
1182783769 Q * pmenier Read error: Connection reset by peer
1182783770 M * ard either: vnamespace --enter vactest mount -t tmpfs -o remount,size=512m,mode=1777 none tmp
1182783776 M * jkl mount: can't find /tmp in /etc/fstab or /etc/mtab :(
1182783784 M * ard morfoh: vnamespace --enter vacacc chroot ./ mount -o remount,size=512m  /tmp
1182783799 M * ard both in /var/lib/vservsers/<yourvserver>
1182783825 M * jkl ah, thanks let me try those.
1182783827 M * ard the latter wil use the /etc/mtab of the vserver
1182783836 M * ard the first will assume you know everything
1182783865 J * ninou ~sylvain_f@bredele.imag.fr
1182783872 M * ard so, it's what daniel_hozac said+ "-t tmpfs" and a bogus device "none"
1182783873 M * ninou hi
1182783879 M * ard hi\
1182783936 M * jkl ard: first one completed w/o error, but tmp is still mounted as having only 32m
1182783945 M * ninou why have i to put the host + the guest in /etc/exports for NFS when i just want to mount in the guest only ?
1182783963 M * daniel_hozac jkl: df says that?
1182783971 M * daniel_hozac ninou: are you using fstab.remote?
1182783976 M * ninou yes
1182784003 M * jkl ard: ha, the second command you gave me now shows this in 'mount'
1182784005 M * jkl none on /tmp type tmpfs (rw,size=32m,mode=1777,size=128m)
1182784028 M * jkl daniel_hozac: I'm checking mount
1182784031 M * ninou but if i not put host in exports, when starting the vserver, it returns '/etc/vservers/scm-timc/fstab.remote:5:1: failed to mount fstab-entry'
1182784171 J * DavidS ~david@p548103A2.dip0.t-ipconnect.de
1182784353 M * daniel_hozac jkl: mount lies.
1182784382 M * daniel_hozac jkl: use df or cat /proc/mounts to get a view of the real world...
1182784598 Q * DavidS|Vechta Ping timeout: 480 seconds
1182784894 J * jesusch ~jesusch@bjoern.schiebtsich.net
1182785467 J * stefani ~stefani@flute.radonc.washington.edu
1182785514 J * tanjix ~tanjix@office.star-hosting.de
1182785527 M * tanjix vps023:~# chcontext --xid 49152 kill -9 2929
1182785528 M * tanjix vcontext: vc_ctx_migrate(): Invalid argument
1182785536 M * tanjix what is wrong here`
1182786224 M * jkl daniel_hozac: yep, you're right. df reports the size correctly. thanks!
1182786329 Q * ninou Quit: Chatzilla 0.9.67+ [Iceape 1.0.9/2007051000]
1182787165 J * arachnist ~arachnist@088156185052.who.vectranet.pl
1182787356 Q * ensc Ping timeout: 480 seconds
1182787710 Q * HeinMueck Quit: Aah!
1182787990 J * bonbons ~bonbons@ppp-110-252.adsl.restena.lu
1182788072 J * bzed ~bzed@dslb-084-059-098-045.pools.arcor-ip.net
1182788438 N * DavidS DavidS|Vechta
1182788448 J * ensc ~irc-ensc@p54B4D56E.dip.t-dialin.net
1182788960 J * click_ click@ti511110a080-0461.bb.online.no
1182789060 Q * click Ping timeout: 480 seconds
1182789710 J * click click@ti511110a080-5534.bb.online.no
1182789815 Q * click_ Ping timeout: 480 seconds
1182790318 M * jesusch bonbons: hi
1182790337 M * bonbons jesusch: hi
1182790338 M * jesusch I just have a general question
1182790344 M * bonbons go ahead
1182790361 M * jesusch why aren't your IPv6 patches in the main-branch of vserver?
1182790511 Q * chand Ping timeout: 480 seconds
1182790544 M * bonbons there are plans to get them into 2.3 series of patches, but nobody pushed enough to make Bertl think about including them earlier
1182790633 M * jesusch :/
1182790642 M * jesusch bonbons: thnx for the info
1182790690 M * bonbons those using it don't care that much about one patch more or less, there are so many patches for their kernels anyhow
1182790731 M * jesusch bonbons: I'm using debian stock kernel and I will stick to it
1182790779 M * bonbons isn't the ipv6 patch included in the debian vserver kernel? (at least the one that derjohn maintains)
1182790802 M * jesusch bonbons: but it would have been great if that patch would have been included before 2.6.18 :(
1182790821 M * jesusch bonbons: to be honest I don't know
1182790846 M * jesusch haven't tried it (as I thought I would need a extra kernel)
1182790857 M * jesusch is there any way to check?
1182790907 M * derjohn bonbons, I am in no way an official DD. The -vserver kernel flavor is maintained by waldi. It does not include the vp patch. basically ist a 2.2.0rc9 what is in the etch kernel.
1182790929 Q * cedric Quit: cedric
1182790964 M * bonbons hm, that's already old ;)
1182790981 M * derjohn jesusch, on linux-vserver.derjohn.de you find a debian repo with my kernels. check blog.derjohn.de for further infos if interested. I worte some snips about the kernel.
1182790988 M * bonbons but I don't look that much at distro kernels, building vanilla+patches anyhow
1182791028 M * derjohn bonbons, yes I build from vanilla, too. Debian lacks e.g. several blob firmware (Tigen cards etc.) for copyright doubts.
1182791052 M * bonbons always optimize the kernel config for my setup (not compiling all those unneeded drivers, building in what I need to mount the / filesystem as I dislike initrds)
1182791079 M * derjohn jesusch, no, there is no ipv6 patch for vserver in debian. Also no capabilty masking. Take my kernel as a drop-in replacement for debian etch.
1182791127 M * jesusch derjohn: thnx for the info - I'm starting to read :)
1182791138 M * derjohn bonbons, funny, I build as much modular as possible, so i dont have to compile the kernel more than once. at least after you manage some dozem machines, you'll begin to like it ;)
1182791274 M * daniel_hozac jesusch: Debian is using 2.0 anyway. it would never have been merged there.
1182791308 M * derjohn jesusch, ok, thx. give my blog a link, if you run a blog or a site ;) YOu can also grab the sources and build the kernel of even the dpatch set an patch the vanilla tree yourself.
1182791309 M * bonbons that depends, for the number of kernel I compile that's not an issue anyhow, ontime config that gets adjusted at each kernel upgrade with make oldconfig
1182791313 Q * ema Quit: leaving
1182791325 M * bonbons and for similar/identical machines the configs are reused
1182791349 M * derjohn daniel_hozac, did you tell me debian uses 2.2.0rc9 ? I didnt check changelogs myself (as I only rarely use the etch kernels)
1182791361 M * daniel_hozac derjohn: 2.0.2.2-rc9.
1182791384 M * derjohn ah ;(, because above you wrote: 2.0 , not 2.2.0 ....
1182791397 M * derjohn oh
1182791407 A * derjohn rubs his eyes ... 
1182791428 M * derjohn ey, 2.0.2 .. thats long ago ...
1182791855 J * click_ click@ti511110a080-5717.bb.online.no
1182791970 Q * click Ping timeout: 480 seconds
1182792338 N * Bertl_zZ Bertl
1182792342 M * Bertl back now ...
1182792366 M * daniel_hozac morning Bertl!
1182792418 J * marcfiu ~mef@aegis.CS.Princeton.EDU
1182792426 M * Bertl wb marcfiu!
1182792561 M * derjohn jesusch, you might also check the PDF/article i wrote about vserver if you want to know about capabilties (I think you can read german language)
1182792831 M * Bertl derjohn: modularity can also cost you quite some performance
1182792870 M * Bertl i.e. a 30-50% speedup (kernel side) is not that unusual, if you compare a mostly monolithic and optimized kernel to a distro kernel :)
1182792875 M * derjohn Bertl, example? (except maybe things like "number of cpu")
1182792896 M * Bertl ah, not just CPU stuff, think bridge and raw tables for example
1182792913 M * Bertl if you don't use them, the hooks are just slowing you down
1182792929 M * derjohn really? incredible! I thought a loaded module vs a statically compiled in should not make much difference after it's loaded.
1182792931 M * Bertl (not to speak of cache misses for much bigger structures)
1182792954 M * Bertl no, it's like this:
1182792960 M * derjohn why should there be more cache misses with a module that is loaded ?
1182792963 M * Bertl modular: A -> B
1182792982 M * Bertl is actually A->stub->B
1182793000 M * Bertl while monolithic makes it a direct call
1182793023 M * Bertl the stub often has cache indirections, where directly compiled in code is often inlined
1182793037 M * derjohn that means, I can accelerate by boxes by 30% on system level (e.g. nertwork io, maybe disk io ...) if I build monolitic kernels ?
1182793041 M * Bertl that aside, if you don't use the code, the stub is just dead code
1182793081 M * Bertl yep, kernel side operations (not the disk or network I/O itself, but the kernel processing required) can gain 30-50%
1182793107 M * Bertl and regarding the cache misses, it's quite simple:
1182793128 M * Bertl just think struct wossname
1182793134 M * derjohn well, the disk example is not very good as the controllers itself are the bottleneck ...
1182793155 M * Bertl if it needs to hold data for ipv4, ipv6, etc ... you need more space
1182793174 M * Bertl OTOH, if you say, ipv4 it is. period. only that space will be used up
1182793199 M * Bertl note: the space in struct wossname will have the ipv6 data even if the ipv6 module is not loaded :)
1182793299 M * derjohn Bertl, ok, thx. I havo leave now, see you l8ter !
1182793313 M * Bertl cya
1182793362 N * DoberMann[PullA] DoberMann
1182793568 Q * click_ Ping timeout: 480 seconds
1182793932 Q * slack101 Read error: Connection reset by peer
1182793963 J * slack101 ~Administr@cpe-71-74-77-84.insight.res.rr.com
1182793972 M * slack101 ey
1182793975 M * slack101 anyone up :P
1182794001 M * Bertl nope, everybody is sleeping ...
1182794008 J * HeinMueck ~Miranda@dslb-088-065-042-032.pools.arcor-ip.net
1182794011 M * slack101 i wonder if its possible to ever let customers see the actual start up
1182794042 M * slack101 or is that kind of pointl;ess hmmm 
1182794043 M * Bertl if you use sysv init stryle, sure, why not
1182794067 M * Bertl even with playin init style you could do that, but it's a little trickier
1182794068 M * slack101 im saying some how give them a remote console ;)
1182794073 M * slack101 so they could change distros 
1182794082 M * daniel_hozac uh, in the guest?
1182794087 M * Bertl that is simpler done via a web frontend
1182794087 M * slack101 yes
1182794095 M * slack101 yea i know 
1182794105 M * slack101 but having that would give the real features of a real server kinda 
1182794114 M * daniel_hozac you realize you can't use _any_ installers inside a guest, right?
1182794127 M * Bertl slack101: if you want a real server, you buy a real one :)
1182794179 M * Bertl slack101: we could, for example, delay a guest startup by 60 seconds or so, so that it looks more real :)
1182794201 M * slack101 well no its just gicing the customer more full control
1182794221 M * slack101 maybe give them ftp access to that directory 
1182794221 M * Bertl no, a read only console log gives no real control
1182794222 M * slack101 ;)
1182794241 M * slack101 so they could change anything they wanted too
1182794241 M * Bertl ftp access is something worth considering
1182794261 M * slack101 so when the vserver is off they could change anytning 
1182794265 M * Bertl even allowing to 'upload' your own guest template could be interesting
1182794288 M * slack101 hmm i will do this 
1182794291 M * slack101 sftp of course
1182794300 M * slack101 or FTPS
1182794318 M * slack101 also is iptales rules per guest os supported yet ?
1182794335 M * Bertl if you configure them on the host, yes
1182794361 M * Bertl inside the guest, you are not allowed to manipulate the tables
1182794457 M * slack101 i mena is there ever gonna be a way to change iptables inside of the guest Bertl
1182794501 M * Bertl we had a few attempts on that, but it seems it fell into the low interest category like the quota stuff
1182794512 M * slack101 quota ?
1182794526 M * slack101 quota's for what ?
1182794527 M * Bertl shared user/group quota
1182794564 M * slack101 theres still individual quota for each vserver correct ?
1182794594 M * Bertl disk limits on shared filesystems and user/group quota on separate filesystems
1182794632 M * slack101 ah
1182794634 M * slack101 well
1182794639 M * slack101 whats high on the  list ?
1182794653 M * Bertl ipv6
1182794661 M * slack101 is that done yet :P ?
1182794668 M * Bertl yep, basically
1182794681 M * slack101 i understand ipv6 is just an add on to the regular ip system now correct ?
1182794703 M * Bertl well, no, ipv6 is the second? generation ip after ipv4 :)
1182794709 M * slack101 btw what is the mac address of each vserver ?
1182794713 M * daniel_hozac sixth generation, obviously. :)
1182794721 M * slack101 ah'
1182794725 M * Bertl daniel_hozac: ah, no, actually not :)
1182794733 M * slack101 does each vserver have a different mac or no ?
1182794740 M * tanjix vps023:~# chcontext --xid 49152 kill -9 2929
1182794741 M * tanjix vcontext: vc_ctx_migrate(): Invalid argument
1182794742 M * Bertl daniel_hozac: ipv5 was already taken for something else, IIRC :)
1182794743 M * tanjix what is wrong here`
1182794755 M * daniel_hozac tanjix: is 49152 running?
1182794757 M * Bertl tanjix: the dynamic xid
1182794765 M * tanjix daniel_hozac: yes,it is
1182794776 M * Bertl slack101: no, only the host has MACs
1182794800 M * slack101 so everyone uses the same MAC ?
1182794803 M * tanjix i have some processes under this context
1182794808 M * tanjix hoe can i kill them?
1182794903 M * daniel_hozac tanjix: vkill?
1182795005 M * Bertl slack101: no, hopefully other machines use different MACs :)
1182795009 M * slack101 how could that work if every ip uses the same MAC address 
1182795030 M * Bertl MAC is layer 2, IP is layer 3
1182795039 M * slack101 what if one ip gets banned via mac address from some service now everyone is blocked ?
1182795049 M * slack101 on that mac 
1182795064 M * Bertl such a banning would not make sense
1182795085 M * Bertl slack101: note on a router, every IP passing through has the same (the router's) mac :)
1182795108 M * Bertl slack101: and the MAC changes every hop too :)
1182795131 M * slack101 every hop ?
1182795142 M * Bertl yep :)
1182795147 M * slack101 so whats the purpose of a mac address then just for the ISP ?
1182795165 M * Bertl the mac address is the 'physical' identifier for your network card
1182795203 M * Bertl so that switches and router can send packets to you
1182795237 M * slack101 but lets say a service blocks the mac address connected to it 
1182795239 M * Bertl in linux, you can change the MAC quite easily
1182795243 M * slack101 every vserver would become affected ?
1182795251 M * slack101 of every ip ?
1182795259 M * slack101 or just per neetwork card 
1182795265 M * Bertl yes, if you lose layer 2 connectivity, all your servers are unreachable
1182795282 M * Bertl very similar to the case when your upstream router stops working :)
1182795306 M * Bertl (or when you pull the ethernet cable :)
1182795378 M * slack101 lol
1182795526 M * slack101 so its impossible for each user to have a different mac 
1182795540 M * Bertl no, just a little more complicated
1182795578 M * Bertl one option is to have a network card per guest :)
1182795625 M * slack101 Bertl: your kidding :P
1182795633 M * slack101 is there any other way ?
1182795640 M * slack101 it would jus make it more like a true vserver
1182795649 M * Bertl true vserver?
1182795839 M * slack101 Bertl: jus play along
1182795842 M * slack101 :P
1182795861 M * slack101 how exactly could each person have thier own Mac without having diff network  cards 
1182796044 M * Bertl if you add another layer of indirection, making the networking slower, you can put a virtual MAC (like Xen or VMWare do it) on the virtual network card
1182796059 M * Bertl then you can bridge that (adding even more overhead) to your router
1182796162 M * slack101 Bertl: no other options ? :P
1182796175 M * Bertl not really :)
1182796250 M * slack101 so i would have to viertulize it Bertl ?
1182796261 M * bonbons slack101: but what's the advantage of individual mac addresses for each guest? Are the guests LAN-only, as soon as you leave your local LAN (cross a router) the guest's own MAC address is replaced by the one of the last router the packet(s) crossed
1182796563 Q * Hunger Ping timeout: 480 seconds
1182797670 J * Hunger Hunger@Hunger.hu
1182798550 J * _markus__ ~chatzilla@chello213047089232.17.14.vie.surfer.at
1182798555 M * _markus__ hi
1182798636 M * _markus__ Back when I was using chroots inside my server, I had the fine option to easily share files (given that I take care of uid/gids) between my chroots by mounting directories accross the chroots. I'm having trouble doing this with vservers. I usually to mount --bind /source /var/lib/vservers/vserv01/test but when I enter vserv01 and look into /test/ it appears still empty
1182798738 M * Bertl works there too
1182798770 M * Bertl just put it in the guest's fstab, and it will be mounted on startup, or if you want temporary mounts, use vnamespace
1182798780 M * bonbons _markus__: you need to bind-mount before starting the guest (or in the guest's config fstab)
1182798787 M * _markus__ ah .. ok, mom
1182798797 M * Bertl bonbons: won't work with recent tools, they do cleanups :)
1182798841 M * bonbons do they clean everything that is below target root and not in fstab in guest's config?
1182798863 M * Bertl daniel_hozac has the details for what is cleaned up
1182798881 M * _markus__ bonbons: great, works!
1182798970 M * daniel_hozac anything below the guest's root is left alone.
1182798989 M * _markus__ I'm confused now with your comments, Bertl. How can I put a directory to be mounted inside my guests fstab when it is physically outside ... ?
1182798992 M * bonbons that makes sense
1182799022 M * bonbons _markus__: it's the vserver utils that do the mounting, not the guest's init scripts
1182799033 M * _markus__ ah ;)
1182799035 M * Bertl _markus__: the config fstab has different source than destination pathes
1182799173 M * _markus__ Bertl:  achso, you mean /etc/vservers/vserv01/fstab ?
1182799179 M * _markus__ (for example)
1182799183 M * daniel_hozac yes.
1182799199 M * _markus__ ok, I now got that. Thanks
1182799287 M * _markus__ I guess when we've the new infrastructure in our company running it's time to donate some money to that ueber-cool project
1182799324 M * _markus__ vservers help me making me more important by drawing even more important network plans with even more vservers ;-)
1182799396 Q * lilalinux Remote host closed the connection
1182799553 M * Bertl _markus__: feel free to do so :)
1182799694 M * _markus__ :)
1182799707 M * _markus__ good night and, as always, many thanks to the quick responses!
1182799713 M * Bertl you're welcome! 
1182799826 Q * _markus__ Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.4/2007051502]
1182799953 Q * Johnnie Quit: G'bye!
1182799962 M * slack101 Bertl: that ftp access thing is a good idea 
1182799974 J * Johnnie ~jdlewis@c-67-163-246-136.hsd1.pa.comcast.net
1182800012 M * slack101 a good idea for recovery
1182800050 J * rgl ~Rui@84.90.10.107
1182800053 M * rgl hi
1182800073 A * rgl has is new box on his hands *G*
1182800110 M * slack101 Bertl: so whats the best way to have each vserver be able to setup thier own iptables ?
1182800445 M * slack101 is a web frontend the only way Bertl ?
1182800486 M * daniel_hozac only? no. the PHP scripts lylix mentioned to you earlier should work as well.
1182800536 M * slack101 i didnt see what he said
1182800559 M * daniel_hozac you might want to start reading the responses to your questions then...
1182800572 M * slack101 well
1182800579 M * slack101 php script could clasify as a web frontend
1182800595 Q * meandtheshell Quit: Leaving.
1182800608 M * daniel_hozac that's only because you haven't looked at it, at all.
1182800627 M * slack101 lol
1182800638 M * slack101 not in a good mood huh ? :P
1182800681 M * daniel_hozac well, it's getting rather annoying when you ask the same things, over and over...
1182800691 M * slack101 well i admit i have asked that a few times
1182800692 M * slack101 but
1182800710 M * slack101 i have a bad habit of asking and leaving and not checking 
1182800720 M * daniel_hozac so?
1182800725 M * daniel_hozac start checking.
1182800728 M * jkl slack101: how's it going ?
1182800735 M * slack101 still means you shouldnt trsat me like a dog ;)
1182800750 M * slack101 jkl: pretty well ....... 
1182800769 M * slack101 i will just finish integrating my iptables front end 
1182800783 M * slack101 i am suprised no one has never made a patch for iptables 
1182800802 M * daniel_hozac that does what, exactly? virtualizes it?
1182800830 M * slack101 well no
1182800834 M * jkl slack101: yeah, I was surprised to hear you asking about having a separate firewall per vserver
1182800853 M * slack101 huh ?
1182800854 M * slack101 well
1182800867 M * jkl slack101: it would be tedious to maintain that :) my single firewall on the host does a great job
1182800877 M * slack101 i dont think it should have a seperate firewall but something to manage your ip/chain within the guest would be good 
1182800888 M * daniel_hozac ... which is what exists.
1182800896 M * slack101 where ?
1182800904 M * daniel_hozac 21:41 < daniel_hozac> only? no. the PHP scripts lylix mentioned to you earlier should work as well.
1182800912 M * daniel_hozac seriously.
1182800914 M * slack101 im sorry i dont have logs atm ;)
1182800922 M * daniel_hozac that's... 7 minutes ago.
1182800925 M * bXi thats about 20 lines up
1182800929 M * slack101 no no
1182800934 M * slack101 what lylix said 
1182800948 M * slack101 i heard you about php scripts 
1182800950 M * daniel_hozac besides, all the logs are at http://irc.13thfloor.at/LOG/
1182800954 M * slack101 but where are they is what i was asking ?
1182800989 M * daniel_hozac google virtuatables.
1182801424 M * slack101 hmmm
1182801432 M * slack101 not having very good luck finding them except for logs 
1182801453 M * slack101 you would think this would have been implemented into the kerrnel but yea 
1182801464 M * daniel_hozac why? it's a userspace feature.
1182801592 M * slack101 you said that in the logs not me ;)
1182801598 M * slack101 i was just agree'ing with you
1182801607 M * slack101 or did your mind change ?
1182801613 M * jkl hehe
1182801655 J * haxier ~haxier@eu85-84-166-67.clientes.euskaltel.es
1182801666 M * Bertl welcome haxier!
1182801667 M * jkl hello haxier!
1182801671 M * haxier hi!
1182801737 M * slack101 also it seems this thing uses some iffy way of doing it 
1182801807 M * slack101 change the variable IPTABLES_SERVER_IP to match the IP of your host server. 
1182801807 M * slack101 change IPTABLES_SERVER_PORT to match localport defined in virtuatables.php. 
1182801833 M * slack101 looks if you put a different ip in the guest one that is not the guest true ip it will still change the iptables 
1182801868 M * jkl well that's a neat feature
1182801895 M * slack101 per say your ip is x.x.x.x.5 and the person puts x.x.x.x.6 it will change the other persons iptables ;)
1182801923 M * slack101 seems like my web front end of  iptables kills that idea ;) 
1182801944 M * jkl slack101: personally, I shudder at the thought of having php scripts mess with iptables ... *especially* if the php script is form based and allows arbitrary user input
1182802002 M * jkl but I'm sure it is perfectly safe ;)
1182802071 M * slack101 it allows the guest to put their own ip in it ;)
1182802079 M * slack101 what was this person thinking 
1182802085 M * daniel_hozac i highly doubt that.
1182802101 M * slack101 http://www.virtuaserver.com.br/forum/viewtopic.php?p=215
1182802114 M * slack101 maybe the instructions of the author is lying
1182802142 M * Bertl what do you think you see there?
1182802164 M * slack101 download it 
1182802207 M * Bertl file not found
1182802211 M * slack101 i see it says server ip so thier not deciding thier ip but i am more owrried about the daemon
1182802282 M * slack101 it just using netcat to send commands 
1182802350 M * slack101 really sounds like an iffy thing 
1182802427 M * slack101 would prefer it to be a compiled app also not a shell script where they can clesrly see whats going on 
1182802475 M * daniel_hozac why is that? objdump is perfectly capable of showing you what a binary does...
1182802583 M * Bertl security by obfuscation (pending micro$oft patent :)
1182802741 M * slack101 well
1182802749 M * slack101 makes it much much harder then just using a shell script 
1182802850 M * daniel_hozac really? i don't think so.
1182802874 M * daniel_hozac even if you assume your users don't know asm, the source is going to be available online.
1182803139 M * Bertl slack101: you can write a small bash obfuscator, which converts any script into binary :)
1182803272 M * slack101 Bertl: good point 
1182804583 J * chand ~chand@m167.net81-64-156.noos.fr
1182804662 J * tuxmania ~bonbons@2001:5c0:85e2:0:20b:5dff:fec7:6b33
1182804978 Q * bonbons Ping timeout: 480 seconds
1182805404 J * coderanger_ ~laptop@wireless-118.media.mit.edu
1182805452 M * Bertl wb coderanger_!
1182805507 M * coderanger_ Yo
1182805536 M * coderanger_ Bertl: Something michael was mentioning this morning, is rate-limiting and other such disk magic implemented for jffs2?
1182805586 M * Bertl partially, I still need to have a logner chat with dwmw2
1182805708 M * Bertl but the dlimit interface will be used for disk limits
1182805724 M * Bertl and for the bandwidth limits we still ahve to define an interface
1182805748 M * daniel_hozac network? or disk?
1182805764 M * Bertl actually, nand reflash cycles :)
1182805777 M * daniel_hozac ah, interesting.
1182805807 M * Bertl but it might boil down to disk ops, as jffs2 cannot really separate the nand write per guest
1182805851 M * Bertl that is why I still need to chat with David
1182805869 J * Aiken ~james@ppp121-45-220-241.lns2.bne1.internode.on.net
1182805883 M * Bertl greetings Aiken!
1182805991 M * Aiken hello
1182806017 M * coderanger_ Bertl: Okay, we'll just try to ignore it for now
1182806269 M * Bertl coderanger_: you can test and implement the dlimits with ext3
1182806301 M * slack101 Bertl: how do you control bandwidth ?
1182806345 M * Bertl usually with token buckets
1182806521 M * coderanger_ Bertl: Not really
1182806534 M * coderanger_ Bertl: We are doing basically all of this on an XO
1182806548 M * coderanger_ Bertl: But dealing with dlimits is a fair bit down the path
1182806559 M * Bertl coderanger_: a simple 2GB sdcard will do nicely
1182806590 M * daniel_hozac i usually run my tests on a 100 MiB loopback mount.
1182806688 Q * tuxmania Quit: Leaving
1182807007 M * Bertl dilinger: btw, how is the new git branch called?
1182807059 M * Bertl dilinger: (or which branch contains the vserver patches?)
1182807733 M * neuralis Bertl: http://dev.laptop.org/git?p=olpc-2.6;a=shortlog;h=vserver
1182807744 Q * DavidS|Vechta Quit: Leaving.
1182807795 M * Bertl neuralis: hmm, how do I check that out?
1182807865 M * neuralis git clone git://dev.laptop.org/olpc-2.6 vserver; cd vserver; git checkout origin/vserver; git checkout -b vserver
1182807890 N * DoberMann DoberMann[ZZZzzz]
1182807898 M * neuralis (you'll want to be using git 1.5)
1182807929 M * Bertl grml
1182807942 M * Bertl is 1.4.2 not goot enough :)
1182807947 M * Bertl *good
1182807959 M * Bertl what is the origin/vserver ?
1182807970 M * neuralis so 1.4 pulls in remote branches by default, 1.5 doesn't
1182807982 M * slack101 Bertl: you say you have used debootstrap right or something ?
1182808002 M * neuralis if you're using 1.4, clone and then 'git checkout vserver' should just work
1182808004 M * Bertl neuralis: I have the olpc-2.6 branch here
1182808012 M * slack101 o nevermind you use netboot
1182808027 M * neuralis olpc-2.6 is not a branch, it's a tree
1182808030 M * neuralis vserver is a branch in that tree
1182808034 M * neuralis slack101: what do you need?
1182808035 M * Bertl sorry, tree
1182808035 M * slack101 i am wondering if i make a debootstrap distro on one system will it be ok for another different system 
1182808051 M * Bertl neuralis: the problem is, it says it doesn't know a vserver branch?
1182808087 M * neuralis is this after cloning fresh or with an existing tree?
1182808105 M * Bertl existing tree, do I have to clone it from new every time?
1182808114 M * neuralis no, you just need to pull in the new branch
1182808135 M * Bertl isn't 'git fetch' supposed to do that?
1182808145 M * slack101 i have a remote console a network booted distro ....i have access to the server ....but if i do a deboostrap then i think it will make the dev and stuff for this recovery system 
1182808158 M * slack101 maybe if i chroot then make it using debootstrap
1182808166 M * neuralis Bertl: it won't pull in new branches by default, afaik
1182808239 M * neuralis try 'git-fetch origin vserver:vserver'
1182808260 A * neuralis profoundly hates whoever designed git's CLI
1182808267 M * Bertl second that!
1182808278 M * Bertl Cannot fetch into the current branch.
1182808291 M * neuralis ugh
1182808302 M * neuralis that exceeds my git expertise, then. dilinger might know more.
1182808327 M * Bertl ah, what the hack, I check it out from scratch
1182808337 M * Bertl will take a few hours though ...
1182808372 M * neuralis hm
1182808381 M * neuralis try 'git-fetch -f origin vserver:vserver' first
1182808421 M * neuralis oh
1182808426 M * neuralis actually, git-fetch might be the wrong way to do it altogether
1182808436 M * neuralis according to the (convoluted) docs
1182808457 M * Bertl forget it, don't spend any time on it ... this part of git really sucks :)
1182808461 M * neuralis yeah
1182808487 M * neuralis there are three poeple in the world who understand git branching. one is linus, and another two are tibetan monks that haven't been seen in over 16 years.
1182808498 M * Bertl lol
1182808535 M * Bertl I would be fine with pasky in my left pocket :)
1182808546 M * neuralis heh, true
1182808854 Q * gerrit Ping timeout: 480 seconds
1182809229 M * slack101 so when you start a vserver it just starts the init correct ?
1182809240 M * slack101 since all the kerrnel stuff is already started 
1182809243 M * daniel_hozac depends on the initstyle.
1182809247 M * Bertl if you use the plain init style, yes
1182809255 M * slack101 how does it do it other ways 
1182809256 M * slack101 ?
1182809276 M * daniel_hozac it executes the rc script directl.
1182809277 M * daniel_hozac +y
1182809294 M * slack101 i seee
1182809306 M * slack101 so really theres no point for a remote console 
1182809313 M * slack101 for guest
1182809329 M * Bertl nope, not really :)
1182809346 M * slack101 and if i offer ftp access as a recovery system i think that will be sweet ?
1182809359 M * slack101 so they can see the errors if it doesnt start
1182809360 M * Bertl could be nice
1182809371 M * slack101 and they can still change a whole distro to thier own custom distro etc 
1182809376 P * marcfiu 
1182809379 M * slack101 oooo wooo im excited 
1182809418 M * slack101 my little recovery ftp thingy + my web interface with iptables + i will use openVCP == i will own the VPS market 
1182809445 M * Bertl definitely :)
1182809464 M * slack101 that dude that makes 120 k a month i know his shit sucks compared to this 
1182809478 M * slack101 and he got slow connection 
1182809479 M * Bertl watch your language :)
1182809481 M * slack101 i got a 100mbit 
1182809485 M * slack101 sorry
1182809486 M * slack101 :P
1182809800 Q * dna Quit: Verlassend
1182809976 Q * chand Read error: Connection reset by peer
1182810905 M * slack101 Bertl: might be getting alot of donations soon :P
1182810912 M * Bertl great!
1182810932 M * slack101 i told a bunch of vserver people to donate
1182810951 M * slack101 and i will donate too :P
1182811051 J * _mcp ~hightower@wolk-project.de
1182811055 M * Bertl neuralis: regarding the defconfig, you might want to change to internal tagging, otherwise jffs2 will not me tagged
1182811080 M * Bertl neuralis: and I'm not sure the guest privacy is of any interest to you
1182811150 M * Bertl CONFIG_TAG_NFSD is obviously a leftover ... i.e. we will remove that as it is not present in this branch
1182811259 Q * mcp Read error: Connection reset by peer
1182811509 Q * HeinMueck Quit: Aah!
1182811538 M * rgl omg.  I've installed ubuntu, but when it tries to boot from HD nothing happens... can you guys help me out troubleshoot this? 
1182811561 M * Bertl in a guest?
1182811570 M * rgl in the host
1182811585 M * Bertl then probably #ubuntu is a better place, no?
1182811599 M * rgl no one answers there :-(
1182811612 M * daniel_hozac maybe you chose the wrong distro? :)
1182811617 M * Bertl rgl: shows how good ubuntu support is :)
1182811644 M * rgl :-(
1182811663 M * Bertl rgl: I can help you with Mandriva :)
1182813226 M * Bertl okay, off to bed now .. have fun!
1182813237 N * Bertl Bertl_zZ
1182813294 M * rgl Bertl_zZ, bye.
1182813299 M * rgl no fun at all... :/
1182814169 M * slack101 mandriva hasa server distro lol ?
1182814173 M * slack101 as*