1134518582 J * Cru ~mindwarp@instructor.e.de.wahlich.com
1134518594 M * Cru morning
1134518858 J * anarcat ~anarcat@H144.C72.B0.tor.eicat.ca
1134518865 M * anarcat hey micah 
1134518891 M * anarcat i'm starting a deb.koumbit.net that'll build amd64 versions of your backport
1134519368 J * ddlp ~danieldlp@shell2.sea5.speakeasy.net
1134519625 M * anarcat hmmm
1134519647 M * anarcat micah, there's no tarball of util-vserver on deb.riseup.net... i guess that means you don't provide a 'deb-src' repo?
1134519712 Q * micah Quit: leaving
1134519751 J * micah ~micah@69.90.134.205
1134519764 M * derjohn anarcat, micah's packages are in sid .. apt-get source from there?
1134519779 N * Bertl_zZ Bertl
1134519790 M * Bertl ahh, that was good :)
1134519798 M * derjohn anarcat, ah, util-vserver ?
1134519818 M * derjohn Bertl, hm?
1134519842 M * Bertl derjohn: don't look at me, regarding debian packages :)
1134519873 M * daniel_hozac derjohn: didn't micah push a util-vserver recently too?
1134519912 M * derjohn daniel_hozac, i didn't follow if he's co-maintainer by now. But there was a talk about that ongoing.
1134519935 M * derjohn Bertl, I meant "ahh, that was good :)" .. I thought you meant /me 
1134519984 M * daniel_hozac 2005-12-08 02:57:55 <micah>     I just made a util-vserver-0.30-209 backport to sarge
1134520053 M * Bertl http://www.testing.osdl.org/plm-cgi/plm?module=patch_info&patch_id=4877
1134520063 M * Bertl I wonder what went wrong this time ...
1134520090 M * Cru I am running a vserver x build -m yum for fc4 on a CentOS4 host - yum now claims about the Fedora RPM GPG key missing, importing it into the hosts rpmdb does not help - where shall I place it to make yum find it?
1134520127 M * daniel_hozac /etc/vservers/.distributions/fc4/pubkeys/, IIRC.
1134520133 M * daniel_hozac check the flower page for the exact path.
1134520176 M * Cru heh, okay - already had a look at it but did not find it - will look closer ;)
1134520179 P * stefani I'm Parting (the water)
1134520214 M * daniel_hozac "#  /etc/vservers/.distributions/dist/pubkeys
1134520214 M * daniel_hozac Directory with GPG pubkeys which are used to sign the packages of this distribution."
1134520215 M * daniel_hozac ;)
1134520225 M * Cru just found it, thanks ;)
1134520356 M * Cru is there a ChangeLog vs2.0.1rc5 -> vs2.01?
1134520367 M * derjohn Bertl, looks promising ;) 2089 warnings, 3910 errors  ... what is this test good for? 
1134520385 M * derjohn Cru, linux-vserver.org/Changelog26
1134520397 M * Cru thanks
1134520447 M * Cru should be linked at www.13thfloor.at/vserver/s_rel26/*/ ;)
1134520469 M * daniel_hozac the 2.0.1 release page has a brief list of fixes...
1134520481 M * micah anarcat: I dont have source packages available there yet
1134520507 M * micah daniel_hozac: I made a util-vserver package for sarge in my own personal repository, not in debian
1134520518 M * micah daniel_hozac: not yet... I'm still working to get co-maintainership with Ola
1134520520 M * daniel_hozac ah, ok. i misunderstood that then.
1134520524 M * micah but it is getting clsoe
1134520558 M * micah in fact, he has agreed, but I am waiting for him to allow me to import util-vserver into our project tree
1134520583 M * daniel_hozac Bertl: those logs aren't pretty...
1134520608 M * Bertl daniel_hozac: yes, but I'm pretty sure it's an OSDL/PLM fault 
1134520618 M * daniel_hozac indeed.
1134520629 M * Bertl let me upload the report/delta
1134520713 M * daniel_hozac ppc64 seems like it might be valid, but doesn't look vserver related at all.
1134520829 M * Bertl http://vserver.13thfloor.at/Experimental/PLM/cross_report_4877.txt
1134520830 M * Cru daniel_hozac: I know, but I think it would also be nice for pre-release users to know whether upgrading is needed ;)
1134520844 M * Bertl Cru: which pre release?
1134520879 M * daniel_hozac Cru: there's a 2.0.1-rc5 to 2.0.1 delta in Experimental.
1134520906 M * Bertl Cru: are you referring to some changelog for 2.0.1-rc* releases?
1134520906 M * Cru Bertl: rc5
1134520929 M * daniel_hozac Cru: in general, the typical "look in FOR-X for deltas with recent timestamps" works.
1134520948 M * Bertl daniel_hozac: yes, except that I broke that yesterday :/
1134520961 M * Bertl but I can recover the timestamps
1134520964 M * daniel_hozac i noticed that, that's why i wanted the delta ;)
1134520979 M * Bertl yeah, was some kind of accident
1134520984 M * Cru daniel_hozac: well, I suppose, reading the source is not everyone's preference
1134521000 M * Bertl Cru: what filesystem do you use?
1134521008 M * Cru all vservers are on Reiser
1134521010 M * Cru 3
1134521012 M * daniel_hozac Cru: the filenames usually tell you a lot.
1134521021 M * Bertl then it's a simple answer for you, upgrade!
1134521025 M * Cru daniel_hozac: just talking in general ;)
1134521033 M * Cru Bertl: heh, ok
1134521054 M * Bertl the only change besides cleanups is a reiserfs change which was forgotten
1134521089 M * Bertl (well, fixing disk limits on reiser)
1134521101 M * Bertl so if you don't use those, you're probably fine too
1134521175 M * Cru heh, okay, then I will wait until a regular reboot is scheduled
1134521393 Q * Aiken Quit: Leaving
1134521650 M * Bertl daniel_hozac, Cru, all: time stamps should be fixed now :)
1134521674 M * Cru heh
1134521756 J * Aiken ~james@tooax6-122.dialup.optusnet.com.au
1134522108 M * anarcat micah, could you give me a hand with dput?
1134522131 M * anarcat i'm trying to give it a pgp signature, but i can't figure out how exactly
1134522138 M * anarcat what do I sign anyways? the .changes?
1134522297 M * anarcat is it possible to mak dput do an ssh to sign the .debs?
1134522651 M * Bertl daniel_hozac: the ia64 compile report looks interesting .. that might be a bug, but I can't tell without cross compiling (which I will do soon)
1134522816 M * daniel_hozac yeah, annoying how it won't show the error(s) that caused the build to fail.
1134522844 M * micah anarcat: how are you building your final packages? you should be signing them then
1134522862 M * micah anarcat: i wonder if it makes sense to have multiple repositories for these backports?
1134522903 M * anarcat micah, i can upload them to your place if you want...
1134522908 M * anarcat i'm using apt-src to build the packages
1134522953 M * micah anarcat: i've never used apt-src
1134522960 M * anarcat what do you use?
1134522961 M * daniel_hozac ah, i found the errors! heh.
1134522963 M * micah anarcat: does it call dpkg-buildpackage, or debuild or something?
1134522969 M * anarcat yep, dpkg-buildpackage
1134522979 M * anarcat my main problem is that my private key is on another machine
1134522995 M * micah anarcat: ok, thats fine, you can just use debsign to sign things afterwards
1134523008 M * anarcat debsign eh?
1134523016 M * anarcat micah, can I upload to your server then?
1134523016 M * micah yeah, it will sign the .dsc and .changes file
1134523035 M * micah anarcat: sure! let me get it setup so you can
1134523042 M * anarcat excellent!
1134523045 M * anarcat glad i can do that
1134523120 M * daniel_hozac Bertl: heh, looks like you had the same problem back in may ;) (/me just googled the error)
1134523467 M * Bertl excellent, what was the solution?
1134523478 M * Bertl (if there was any :)
1134523487 M * daniel_hozac i couldn't see one ;)
1134523501 M * Bertl ah, that's probably why it is still there :)
1134523551 M * daniel_hozac (http://ds9a.nl/klogbot/?year=2005&month=5&day=31&hour=13.5)
1134523614 M * daniel_hozac http://gcc.gnu.org/ml/gcc-bugs/2005-05/msg03980.html
1134523634 M * Bertl lol, tx for that memory (which I had already lost :)
1134523655 M * daniel_hozac hehe.
1134523665 M * Bertl yes, but we fixed that one, actually
1134523681 M * Bertl the __label__ should not be there ... checking now
1134523743 M * daniel_hozac history code.
1134523845 M * Bertl hmm, yeah, maybe I just changed it in the devel release
1134523863 M * daniel_hozac heh.
1134523937 M * Bertl ah, aha ...
1134523947 M * Bertl current_text_addr() there seems to be a macro for this
1134523999 M * Bertl yeah, which is defined differently (hacked) for ia64
1134524067 M * daniel_hozac heh.
1134524091 M * Bertl that quite simplifies the stuff .. great!
1134524816 M * Bertl interesting that I didn't find that earlier ...
1134524850 Q * comfrey Ping timeout: 480 seconds
1134524889 M * daniel_hozac well, why make things simple when you can make them difficult? :)
1134524944 M * Bertl hrm ...
1134525387 M * anarcat micah, uploading now
1134526094 J * mep__ mep@p5091BF0E.dip0.t-ipconnect.de
1134526534 Q * mep_ Ping timeout: 480 seconds
1134527266 Q * FireEgl Read error: Connection reset by peer
1134527741 J * FireEgl Atlantica@Atlantica.US
1134527751 M * Cru bbl
1134527755 Q * Cru Quit: use Unices; $live->free() || die;
1134527841 Q * sebi Ping timeout: 480 seconds
1134528381 J * sebi ~sebi@Fd66c.f.strato-dslnet.de
1134528411 M * Bertl http://www.testing.osdl.org/plm-cgi/plm?module=patch_info&patch_id=4878
1134528418 M * Bertl (let's see if it helps :)
1134529457 M * Bertl okay, I'm off for now .. maybe back later ...
1134529462 N * Bertl Bertl_oO
1134529613 M * anarcat oookay
1134529624 M * anarcat so i have the latest util-vserver here now, on debian sarge
1134529652 M * anarcat and it doesn't start (on boot) my server i just created with vserver build -m debootstrap...
1134529702 M * anarcat micah, you told me to upgrade to the latest, here i am ;)
1134529729 M * micah do you have a mark file for that vserver?
1134529738 M * anarcat i don't think so
1134529748 M * anarcat i don't know what a mark file is :/
1134529749 J * cemil ~cemil@defiant.wavecon.de
1134529762 M * anarcat i'm used to the old /etc/vservers/server.conf stuff
1134529773 M * anarcat i really don't understand zlit about the new layout...
1134529830 Q * cemil_ Read error: Connection reset by peer
1134529947 M * anarcat wow
1134529950 M * anarcat ok, i get it
1134529953 M * anarcat sorry about that brain fart
1134530302 M * micah got it?
1134530369 M * anarcat yeah
1134530374 M * anarcat it's confusing, that thing
1134530710 M * anarcat wow
1134530724 M * anarcat this thing has an obscure way of figuring out the "DISTRIBUTION" we're running here
1134530735 M * anarcat vserver-build.functions:            DISTRIBUTION=`debconf-get-selections 2> /dev/null | grep ^base-config | grep "mirror/suite" | awk '{ print $4 }'`
1134530745 M * anarcat so i'm running "testing"
1134530746 M * anarcat sucks
1134530943 M * anarcat i guess i don't have a better way though
1134531318 M * micah anarcat: you can pass which distro as an argument to debootstrap
1134531421 M * anarcat yep, i do that
1134531440 M * anarcat but i'm still stuck with my problem of filling /etc/apt automagically
1134531441 M * anarcat #  /etc/vservers/.distributions/dist/apt
1134531441 M * anarcat Default content of the /etc/apt/ directory.
1134531447 M * anarcat that just doesn't work
1134531448 M * micah ahh
1134531452 M * anarcat or i can't figure out what "dist" is
1134531458 M * micah what happened to the skeleton/clone approach?
1134531458 M * anarcat ..supposed to be
1134531464 M * anarcat i scrapped it
1134531474 M * anarcat i liked better the debootstrap approach
1134531476 M * anarcat ... which i know
1134531497 M * micah right, well, if you figure this out, I'm curious to know what the answer is!
1134531506 M * anarcat hehehe
1134531529 M * anarcat so far, i think it might be that "DISTRIBUTION" gets set to "testing" using the above insane grep/awk:
1134531532 M * anarcat 22:25 < anarcat> vserver-build.functions:            DISTRIBUTION=`debconf-get-selections 2> /dev/null | grep ^base-config | grep "mirror/suite" | awk '{ 
1134531536 M * anarcat                  print $4 }'`
1134531538 M * anarcat oups
1134531558 M * anarcat or it might be set to the -d option of the debootstrap method
1134531564 M * anarcat although that didn't work the first time round
1134531676 M * micah that is pretty weird
1134531691 M * anarcat yep
1134531700 M * anarcat the scripts are quite obscure at time
1134531708 M * anarcat and they all use libraries
1134531713 M * anarcat so it's quite hard to debug...
1134531737 M * anarcat although libraries are better than copy paste, i wonder if all this could have been implemented in something better than "sh"... :)
1134531745 M * anarcat but who am i to complain eh? :)
1134531764 M * micah or more sane bash?
1134531789 M * anarcat sane.. bash... you gotta be kidding
1134531790 M * anarcat :)
1134531796 M * anarcat trying something:
1134531798 M * anarcat for d in sarge defaults testing 
1134531798 M * anarcat > do
1134531798 M * anarcat > mkdir -p $d/apt
1134531798 M * anarcat > echo $d > $d/apt/sources.list
1134531801 M * anarcat > done
1134531806 M * anarcat now i'll now :)
1134531854 M * anarcat know
1134532247 M * anarcat wow
1134532252 M * anarcat it does *nothin*
1134532260 M * anarcat this just doesn't work
1134532296 M * anarcat so cool
1134532944 M * anarcat wow
1134532954 M * anarcat each 10 minutes, i wow on something new :)
1134532961 M * anarcat apt.initFilesystem
1134532969 M * anarcat doesn't get executed when using debootstrap
1134532975 M * anarcat only when using apt-rm
1134532976 M * anarcat rpm
1134533033 A * anarcat considers patching
1134533399 M * anarcat micah, how is the sarge util-vserver package broken?
1134533405 M * anarcat ie. why the backport, after all?
1134533579 M * micah anarcat: there are many things broken in the debian package overall, but the backport is to get .209 available in sarge
1134533593 M * micah as there have been many fixes from .208 and previous (which are only available in sarge)
1134533805 M * anarcat is there a changelog somewhere?
1134533925 M * micah well, this issue sounds like an upstream util-vserver issue, not a debian package issue
1134533950 M * micah most of the debian packaging issues have to do with missing syscalls for odd architectures and some other random things
1134534116 M * anarcat sure
1134534161 M * anarcat this is insane
1134534177 M * anarcat there's so m any global variables in there it's frightning
1134534362 M * anarcat found a bug!
1134534365 M * anarcat a real one now
1134534366 M * anarcat         colorize echo -n "$@" >&2
1134534371 M * anarcat ## Usage: colorize <style> <command>
1134534371 M * anarcat function colorize
1134535186 J * dos000 ~dos000@CPE0013109a586e-CM00137186e53a.cpe.net.cable.rogers.com
1134535720 M * anarcat i guess i should report this in the bugtracker eh
1134535947 M * micah yeah
1134535952 M * micah that would be useful!
1134536000 M * anarcat boy, this is a mess
1134536003 M * anarcat i give up, again
1134536022 M * anarcat no way that the apt directory can be auto-filled with the debootstrap method
1134536030 M * anarcat not even sure it actually works in the other methods
1134536047 M * anarcat from what i can tell, only .pkg directories get filled, nothing actually useful
1134536056 M * anarcat completely absurd, from here
1134536060 M * anarcat <sigh>
1134536383 M * anarcat phew
1134536390 M * anarcat sent the bug report
1134536479 Q * dos000 Quit: Leaving
1134536800 M * anarcat my head hurst
1134536801 M * anarcat ts
1134536840 M * anarcat http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=343277
1134537585 M * anarcat http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=343278
1134537877 M * micah anarcat: good, thanks for reporting those!
1134537930 M * anarcat yeah
1134537935 M * anarcat really frustrating experience
1134537938 M * anarcat if you ask me
1134537940 M * anarcat but anyways
1134537966 M * anarcat i'm also considering downgrading to the regular sarge util-vserver since the backport doesn't really seem to solve my initial problems...
1134537971 M * anarcat and i don't know *what* they fix
1134538092 M * anarcat http://cvs.koumbit.net/cgi-bin/cvsweb.cgi/~checkout~/koumbit-maint/bin/vcreate?rev=HEAD;content-type=text%2Fplain
1134538099 M * anarcat is what i use to create the vservers these days
1134538180 M * anarcat i'll add some features like detecting already existing ip addresses at some point...
1134538207 M * anarcat my eyes are itchy
1134538210 M * anarcat i'm done :)
1134538211 M * anarcat good night
1134538226 M * anarcat good holidays, micah :)
1134538237 Q * anarcat Quit: time to sheep
1134539155 Q * matti Ping timeout: 480 seconds
1134539658 J * balbir ~balbir@59.145.136.1
1134541475 J * Smutje_ ~Smutje@xdsl-84-44-186-28.netcologne.de
1134541580 Q * Smutje Ping timeout: 480 seconds
1134542734 J * matti matti@linux.gentoo.pl
1134543752 Q * cemil Ping timeout: 480 seconds
1134544123 J * cemil ~cemil@defiant.wavecon.de
1134545345 J * infowolfe infowolfe@66-230-115-37-cdsl-rb1.nwc.acsalaska.net
1134549885 N * Smutje_ Smutje
1134550105 Q * brcc Quit: No windows for this server
1134550548 J * Cru ~mindwarp@bastardrouterfromhell.e.de.wahlich.com
1134550552 M * Cru morning
1134551146 M * matti LOL
1134551152 M * matti Router from hell?
1134551157 M * matti :)
1134551218 M * Cru bastard router! ;)
1134551264 M * Cru I had some problems with OpenWRT when I had to give it a hostname ;)
1134551280 M * Cru s/had to give/gave/
1134551486 J * Doener doener@i5387EEBF.versanet.de
1134551508 M * Cru hmm, lecker
1134551523 M * Doener *sigh*
1134551583 M * Cru heh
1134551591 Q * Aiken Quit: Leaving
1134551684 Q * ddlp Remote host closed the connection
1134551713 J * ddlp ~danieldlp@69.17.116.3
1134554062 Q * infowolfe Quit: Leaving
1134554345 M * Cru bbl
1134554353 Q * Cru Quit: use Unices; $live->free() || die;
1134555160 J * infowolfe infowolfe@66-230-115-37-cdsl-rb1.nwc.acsalaska.net
1134556467 J * JViz Anomaly@adsl-065-013-131-023.sip.int.bellsouth.net
1134558132 J * infowolfe_ infowolfe@66-230-115-37-cdsl-rb1.nwc.acsalaska.net
1134558154 J * infowolfe__ infowolfe@66-230-115-37-cdsl-rb1.nwc.acsalaska.net
1134558159 Q * infowolfe__ Read error: Connection reset by peer
1134558585 Q * infowolfe Ping timeout: 480 seconds
1134558616 Q * infowolfe_ Ping timeout: 480 seconds
1134558873 J * infowolfe jthm@66-230-115-37-cdsl-rb1.nwc.acsalaska.net
1134560736 M * _Roey hi all
1134560740 N * _Roey Roey
1134563471 M * Roey hello
1134563473 M * Roey anyone here?
1134563489 M * Roey I want to make it so that all my vservers share the same /usr, /lib and /etc....
1134563510 M * aba here are always people. But you usually don't get an answer w/o asking a real question :)
1134563512 M * Roey (just that they'd have different /var's or that they could may only write to a certain directory under /var)
1134563517 M * Roey ah
1134563521 M * Roey aba:  so, how would I do that?
1134563527 M * Roey aba:  your nick means father in hebrew
1134563534 A * aba doesn't know the answer to that
1134563538 M * Roey ok
1134563540 M * aba ah, interessting :)
1134563543 M * Roey thanks anyway though :)
1134563558 M * Roey aba:  well yeah.. it's one of the first things out of a baby's mouth ;)
1134563559 M * aba but with sharing /etc, I'm not too sure if that's really wise
1134563569 M * Roey I was thinking about that.
1134563578 M * aba however, IIRC there is a script that joins "just" the binaries
1134563597 M * Roey aba:  the thing is, I don't want to duplicate the same stuff and have to apt-get dist-upgrade for each vserver.
1134563600 M * Roey aba:  ok
1134563634 M * Roey aba:  where at?
1134563694 M * daniel_hozac unification.
1134563709 M * aba sorry, I forgot all about that. And I dist-upgrade each vserver seperate
1134563711 M * daniel_hozac would still mean that you'd have to apt-get in all the vservers though.
1134563770 M * daniel_hozac but you wouldn't lose any disk space for doing so.
1134564244 M * Roey how so?
1134564341 Q * balbir Quit: Leaving
1134564465 M * Roey Bertl_oO:  I guess you're asleep still ;)
1134564944 M * Doener Roey: if you really want to share the _same_ /usr in all vservers, bind mount extensions will be your friend, as they allow read-only bind mounts
1134564956 M * Doener not sure if it really is a good idea though
1134565000 M * Doener packages updates may involve more than just the stuff you share between the vservers and you might end up with really broken setups
1134565086 M * Doener (just because you don't do the update in each vserver, so the problem is tied to your idea)
1134565107 M * Doener the package management itself would have to support that kind of stuff i guess
1134565438 J * Ben81 ~Ben81@tipi0e.lri.fr
1134565835 J * mep_ mep@p509182EA.dip0.t-ipconnect.de
1134565851 M * Roey Doener:  hmm
1134565860 M * Roey Doener:  well I intend to apt-get update from the host
1134565866 M * Roey not the vserver guests.
1134566243 Q * mep__ Ping timeout: 480 seconds
1134567755 Q * Doener Ping timeout: 480 seconds
1134567801 J * Doener doener@i5387C4AD.versanet.de
1134568588 Q * infowolfe Quit: Leaving
1134571288 N * Bertl_oO Bertl
1134571293 M * Bertl morning folks!
1134571388 M * Doener morning Bertl!
1134571519 M * Roey hey Bertl!
1134571792 M * JViz does anyone know where i can find a howto or example of setting up a network of NAT'd vservers?
1134571821 M * Roey you don't do nat'd vservers :)
1134571833 M * JViz no?
1134571833 M * Roey they're just separate partitions of the same system
1134571836 M * JViz right
1134571837 M * Roey with the same hardware
1134571863 M * Roey so 'eth0' in one guest is the same 'eth0' on other guests is the same 'eth0' on the host
1134571866 M * JViz but i want to have a set of fake IP's that the vservers use
1134571879 M * Roey JViz:  that's just standard Linux networking then
1134571899 M * Roey you can have multiple IPs bound to the same device (I think... Bertl, is that right?)
1134571972 M * JViz yeah
1134571983 M * JViz i know you can have multihomed adapters
1134571988 M * Roey yeah
1134571995 M * Roey but just know that the devices are all shared
1134572064 M * JViz yeah
1134572066 M * JViz that's just it
1134572137 M * JViz if i have a private IP and a public IP on the same adapter, and someone else changes their address to match on the network, they will be on the same subnet, or there could even be a collision
1134572149 M * Roey ok
1134572163 M * JViz well
1134572164 M * Roey note though that Doener or Bertl know more than I -- I am just starting
1134572270 M * Bertl JViz: don't know your setup, but what you probably want is to SNAT your private IPs to the public one
1134572276 M * Bertl something like this:
1134572288 M * JViz yeah
1134572300 M * Bertl iptables -t nat -I POSTROUTING -s <private>/24 -j SNAT --to <public-ip>
1134572307 M * JViz ah
1134572555 M * JViz so even if the public ip is eth0 and the private ip is eth0:1, if someone else has a second IP in the same subnet as my private IP, it won't be able to talk directly to my private IP?
1134572602 M * Bertl well, no, a network on the private range will be able to talk to the guest
1134572620 M * Bertl (without the SNAT)
1134572642 M * Bertl of course, you can restrict that in arbitrary ways (with iptables)
1134572646 M * JViz ah, so i need to have a different (fake)physical device?
1134572662 M * JViz oh
1134572673 M * Bertl JViz: the important point with networking is: there is no virtual/whatever networking except for the host networking
1134572695 M * Bertl the guests use IP isolation 
1134572714 M * Bertl i.e. you assign a subset of the host IPs to the guest, which can then bind ports to them
1134572742 M * JViz right, but i can still make fake adapters with bridging and use those, right?
1134572771 M * Bertl you can do whatever you do on a normal linux box (on the host)
1134572792 M * Bertl vlans, bridging, aliases, tunnels, whatever
1134572986 M * JViz what do you suggest using for having more vservers than IP's?
1134573038 M * Bertl assign them private IPs (e.g. 10.0.0.2, 10.0.0.3 ...) and do portmapping for incomming conenctions and SNAT for outgoing (as above)
1134573247 M * Bertl one border case is the situation with a single public IP and N guests
1134573263 Q * mnemoc Read error: Connection reset by peer
1134573264 M * Bertl if they, for example, should be reachable via ssh
1134573315 M * JViz it's on a LAN with other machines that could verywell be using the same private IP or subnet thereof, so I should restrict that with arbitrary ways (with iptables) you previously mentioned?
1134573321 J * mnemoc ~amery@200.75.27.40
1134573396 M * Bertl JViz: hmm, probably, maybe also put them on dummy0 and disable arp proxying?
1134573413 M * Bertl (so that they are not advertized outside of the host)
1134573447 M * JViz ok
1134573483 M * Bertl okay, moving out now .. back later ...
1134573489 N * Bertl Bertl_oO
1134573492 M * JViz thank you
1134573502 M * JViz one last question?
1134573556 M * JViz how does this compare to bridging?
1134573565 M * JViz more efficient?
1134573663 J * jeeves ~jeeves@c-24-11-171-10.hsd1.mi.comcast.net
1134575174 M * jeeves Hey guys!.   I got it installed successfully.  I followed the FC4 instructions.  Easy as cake.
1134575198 M * jeeves or pie.  I like pie more.
1134576442 M * JViz piece of pie
1134576448 M * JViz or cake, i like cake more.
1134576950 M * daniel_hozac that's the way it should be ;)
1134577162 M * daniel_hozac RPM or source, btw?
1134577573 N * Bertl_oO Bertl
1134577682 M * Bertl daniel_hozac: btw, I found another 'issue" yesterday ... it's kind of unusual/weird but we should think about a solution for the OSDL testing
1134577715 M * daniel_hozac hmm, details?
1134577722 M * daniel_hozac (or timestamp for the logs :))
1134577749 M * Bertl the vserver network stuff uses *dev_put(), which in turn calles some inet* function, which might not be in the kernel if the entire ipv4 is compiled as module :)
1134577790 M * daniel_hozac ah, is that the reason for the linking failures?
1134577810 M * Bertl I'd say so ...
1134577848 M * Bertl I'm not sure if we should try to 'solve' this dependacy or if we should simply force the ipv4 into the kernel
1134577887 M * daniel_hozac well, i guess it only makes sense to force it.
1134577921 M * Bertl yeah, probably ...
1134577956 M * Bertl OTOH, we could move the vserver related code into the ipv4 module
1134577964 M * Bertl (at least at link level)
1134578075 M * daniel_hozac hmm, can IPv4 really be built as a module?
1134578082 M * daniel_hozac CONFIG_INET is a bool here.
1134578336 M * Bertl hmm, maybe it's the allno config ... not 100% sure, the PLM logs always confuse the hell out of me ...
1134578364 M * Bertl have to do some cross compiling myself ...
1134578394 M * daniel_hozac yeah, allnoconfig looks like it gets errors on bzImage.
1134578438 M * daniel_hozac i can't access the x86 regression test results... they don't seem to be present.
1134578456 M * Bertl yes, same here ... already sent a bug report ...
1134578538 M * jeeves Sorry, I was afk.  I used the rpms
1134578630 M * daniel_hozac the ppc 40x and ppc64 errors don't appear to be vserver related. ppc regression results don't have the error files present...
1134578752 J * comfrey ~comfrey@peoples.coop
1134578777 M * daniel_hozac jeeves: new RPMs will be showing up any minute now...
1134578824 M * daniel_hozac (i think Dave Jones hates me, i built a new set yesterday night, and an errata kernel was released this morning...)
1134578874 M * jeeves doh
1134578876 M * jeeves lol
1134578899 M * Bertl welcome comfrey!
1134578907 M * jeeves these are working ok for now.  I just need to get the guest servers to surf the net
1134578922 M * Bertl jeeves: private IPs?
1134578943 M * jeeves yea, behind a router.
1134578945 M * Bertl iptables -t nat -I POSTROUTING -s <private>/24 -j SNAT --to <public-ip>
1134578960 M * Bertl I should probably bind that to some key :)
1134578961 M * jeeves in the host or gues
1134578967 M * jeeves lol
1134578972 M * Bertl on the host, always on the host
1134578987 M * daniel_hozac jeeves: which ones? 2.6.14-1.1651_FC4.vs2.0.1.1? 2.6.14-1.1650_FC4.vs2.0.1.0.rc5.2?
1134578988 M * jeeves makes sense.  fogive my ignorance.  I am a noob to this
1134579000 M * jeeves vs2
1134579004 M * Bertl jeeves: np, you are forgiven :)
1134579040 M * daniel_hozac (first one hasn't been QA'd yet, busy building the next ones, heh)
1134579294 M * comfrey morning Bertl!
1134579597 M * Hollow Bertl: fyi, will see to get 2.0.1 stable next week, including new stage releases with a fixed baselayout, and will promote it on the weekly news then
1134579599 J * stefani ~stefani@superquan.apl.washington.edu
1134579617 Q * shedi Quit: Leaving
1134579621 M * Bertl Hollow: excellent!
1134579625 M * Bertl welcome stefani!
1134579658 N * jeeves jeeves_2HourLunch
1134579660 M * Hollow does also mean stable openvz support in gentoo, so i'd like to do the comparision thing next week too
1134579667 M * stefani hello Bertl 
1134579931 J * shedi ~siggi@inferno.lhi.is
1134579948 M * FaUl europe sucks :-(
1134579956 M * Bertl hmm?
1134579962 M * Bertl welcome shedi!
1134579963 M * mnemoc chile sucks more
1134579978 M * shedi thank you Bertl
1134580005 M * FaUl Bertl: they finally got this datamiming-stuff through the european parlament 
1134580046 Q * Ben81 Quit: Leaving
1134580162 M * Bertl FaUl: well, we can still hope that they do the mining with M$ apps :)
1134580170 M * FaUl ehe
1134581620 Q * comfrey Remote host closed the connection
1134582523 N * jeeves_2HourLunch jeeves
1134582716 J * jeeves_ ~jeeves@c-24-11-171-10.hsd1.mi.comcast.net
1134582716 Q * jeeves Read error: Connection reset by peer
1134583138 M * jeeves_ That was weird
1134585438 J * Smutje_ ~Smutje@xdsl-84-44-187-2.netcologne.de
1134585540 Q * Smutje Ping timeout: 480 seconds
1134586011 N * Smutje_ Smutje
1134588661 M * Bertl Hollow: ping!
1134588764 M * baggins he bertl
1134588784 M * baggins there seems to be a problem with token bucket on SMP
1134588899 M * baggins Bertl: for some unknown reason the bucket is filled so fast that nothing gets stopped
1134589007 M * Bertl hmm?, could you provide your scheduler config?
1134589034 M * Bertl basically /proc/virtual/<xid>/sched
1134589091 M * baggins Bertl: http://pastebin.com/464125
1134589201 Q * sebi Ping timeout: 480 seconds
1134589211 M * baggins it worked up to a point, and then suddenly it just stopped
1134589249 M * Bertl cool, 8cpus :)
1134589293 M * Bertl well, no idea _what_ happens, never got a chance to test it on more than 4 cpu systems
1134589329 M * Bertl but, the result of one of the discussions I had here at princeton is, that we will do some 'interesting' modifications to the scheduler ...
1134589363 M * Bertl and the first modification will be to make the token bucket per cpu
1134589400 M * Bertl so .. asumed that you are interested in testing here, this might perfectly solve your issues ...
1134589536 M * baggins if this machine will be available for testing at that time...
1134589560 M * baggins but it has to go for production at the end of january
1134589572 M * baggins BTW, after a reboot it works again
1134589583 M * Bertl well, we can probably start next week (or maybe at the weekend)
1134589640 M * baggins nice :) cpuset satisfy my needs but I can test scheduler too
1134589652 A * baggins gotta go
1134589662 M * Bertl k, cya!
1134590535 Q * alexx Ping timeout: 480 seconds
1134591101 J * prae ~benjamin@sherpadown.net
1134591111 M * Bertl welcome prae!
1134591116 M * prae hi Bertl 
1134591224 M * jeeves_ noob question:  names are not resolving.  I can ping yahoo's IP, but not the name.
1134591233 M * jeeves_ from within the gues
1134591234 M * jeeves_ t
1134591241 M * Bertl check the /etc/resolv.conf inside the guest
1134591253 M * jeeves_ I put two nameservers in here
1134591278 M * Bertl can you reach them (via ping) from inside?
1134591286 M * jeeves_ ip only
1134591294 M * Bertl the name servers
1134591301 M * jeeves_ yea
1134591321 M * Bertl then you have to check _why_ the nameservers do not answer your guest requests
1134591330 M * jeeves_ one is my little router, and one belongs to msu.edu
1134591357 M * jeeves_ hmmm.
1134591362 M * Bertl so have a look at your router (log files)
1134591374 M * jeeves_ good idea.  Looking....
1134591652 J * sebi ~sebi@Fcf9d.f.strato-dslnet.de
1134591839 M * Bertl welcome sebi!
1134592044 M * jeeves_ ok, my little router doesn't log DNS lookups.
1134592072 M * Bertl what about putting a tcpdump there?
1134592280 M * jeeves_ is that something I need to install on the guest?
1134592294 M * Bertl ahem, no, I thought more about the router here
1134592309 M * Bertl but maybe tcpdump on the host might also give you valuable info
1134592346 A * FaUl would tcpdump on the host, too :-)
1134592382 M * jeeves_ yea, I am looking at it.  It is huge b/c I am VNCing from work right now.
1134592397 M * Bertl limit it to port 53?
1134592447 M * FaUl tcpdump host <routerip> ip proto udp port 53 :-)  
1134592459 M * Bertl might be tcp too
1134592473 M * FaUl yes, but it is very unlikely
1134592492 M * FaUl in fact my authorative-nameserver don't even listen on 53 tcp :-)
1134592545 M * Bertl hmm, interesting, might cause some 'stranger' issues
1134592649 M * FaUl but i experienced some advanced-routing (source-based-routing) issues with udp and masquerading on 2.4 kernels (logfile said: "MASQUERADE: no route: Rustys brain broke!"), are you using source-based-routing and masquerading?
1134592693 M * jeeves_ I remember that
1134592703 M * FaUl what?
1134592715 M * jeeves_ Rusty's little error messages from 2.4
1134592733 M * FaUl never got such a message before :-)
1134592742 M * jeeves_ or the earlier versions of iptables
1134592936 M * FaUl jeeves_: do you remember when this issues where fixed? 
1134592944 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1134592969 M * FaUl i personally run 2.4.29 because i never got my router to reboot since then
1134592998 M * FaUl are this issues fixed in 2.4.32?
1134593002 M * jeeves_ um, the problem I had was user (my) error configuring iptables.  Or ipchains
1134593018 M * FaUl ok, then this problem was not related to mine :-)
1134593039 M * jeeves_ I doubt it, but Rusty's messages are interesting
1134593103 M * jeeves_ I did: tcpdump host 192.168.10.1 ip proto udp port 53
1134593111 M * jeeves_ got syntax error
1134593144 M * jeeves_ reading man page for tcpdump
1134593192 M * FaUl try without ip proto udp
1134593196 M * FaUl aeh
1134593198 M * FaUl npe
1134593211 M * FaUl try tcpdump host 192.168.10.1 and ip proto udp and port 53 ;-)
1134593372 M * Bertl welcom cdrx!
1134593631 M * jeeves_ and?
1134593640 M * FaUl yes
1134593679 M * cdrx hi everybody, i've been reading the list a while, it was time to join the list
1134593740 Q * monrad Quit: leaving
1134593842 M * Bertl cdrx: great!
1134593853 M * Bertl will be back later ... have to get something to eat ...
1134593860 N * Bertl Bertl_oO
1134593875 M * jeeves_ ok my test vserver is called test.  if I do:  tcpdump -i eth0:test0 > dump.out      It should only get the packets from the vserver?
1134593908 J * monrad ~mikkel@213083190131.sonofon.dk
1134594239 J * jag ~jag@199.232.75.2
1134594264 M * TheSeer afaik, you cannot tcpdump -i on alias interfaces
1134594404 M * jeeves_ that is what I am finding.  I get everything from eth0
1134594425 M * jeeves_ It lists it as an interface.  tcpdump -D
1134594623 J * hallyn ~xa@adsl-68-72-170-201.dsl.chcgil.ameritech.net
1134594675 M * jeeves_ ok good stuff here.  I did: tcpdump port 53.  I then went into the guest and pinged yahoo.  It of course says unknown host www.yahoo.com.  I stop the dump and it didn't collect any packets.
1134594815 J * baughj ~baughj@199.232.75.2
1134595116 M * jeeves_ ok, got it.  Are you ready for the answer?
1134595127 M * jeeves_ I had to start named on the host.
1134595139 M * jeeves_ vserver rocks.
1134595197 M * Eyck  yupp
1134595236 M * mnemoc jeeves_: bind needs to be built with "--disable-linux-caps" to run without extra capabilities on vserver
1134595268 M * daniel_hozac or with a special patch.
1134595281 M * daniel_hozac (to still allow for chroot mode with a non-root UID)
1134595328 M * daniel_hozac s/special//
1134595343 M * jeeves_ what happens if I just use the default rpm?
1134595359 M * daniel_hozac it won't work on the vserver.
1134595370 M * daniel_hozac it'll fail with capset: Operation not permitted.
1134595382 J * Aiken ~james@tooax6-148.dialup.optusnet.com.au
1134595385 M * jeeves_ oh, ok I get it
1134595401 M * jeeves_ yea,  I started it on the host.  
1134595420 M * jeeves_ I will recompile if I put named on a guest.
1134595508 M * jeeves_ Does it sound like I am getting this, or am I still out in left field?
1134595696 Q * prae Quit: Pwet
1134596067 M * baughj Has anyone noticed that the 2.6.8.1 bind mount extension patches don't compile cleanly...?
1134596104 M * baughj It looks like fs/ext(2|3)/acl.c just needed an #include <linux/namei.h>.
1134596132 M * daniel_hozac i don't think there are any 2.6.14 patches for BME.
1134596144 M * daniel_hozac except for those integrated into the devel series.
1134596196 N * Bertl_oO Bertl
1134596199 M * Bertl back now ...
1134596270 M * Hollow Bertl: pong
1134596272 M * Bertl baughj: what kernel do you use?
1134596322 M * Bertl Hollow: hmm, hmm, ah, yes I remember now :)
1134596329 M * Hollow :)
1134596330 M * baughj I'm using the ubuntu warty sources (2.6.8.1-16.24)
1134596352 M * Bertl Hollow: do you need anything which isn't in 2.1.0-rc10 yet for your tools (to be in 2.1.0)?
1134596376 M * Bertl baughj: that is probably more a 2.6.12-14 kernel than a 2.6.18.1, right?
1134596409 M * Hollow hm, the only issue i can think of would be the double entries in /proc/virtnet
1134596417 M * Hollow but feature-wise i guess no
1134596439 M * Hollow at least it works in the current state
1134596440 M * Bertl hmm, k, that will probably be fixed with the sequencer stuff (not sure when we do that)
1134596491 M * baughj bertl: IIRC warty is 2.6.8.1, hoary is 2.6.10, and breezy is 2.6.12..?
1134596520 M * Bertl that's what they call it :)
1134596525 M * baughj heh
1134596539 M * Bertl baughj: no, actually the kernels include patches from later releases too
1134596545 M * baughj ahhh....okay.
1134596548 M * baughj makes sense.
1134596561 M * Bertl but you should look at the bme patches for 2.6.12 or so
1134596570 M * baughj hmm, okay
1134596572 M * Bertl let me see if I can find them ...
1134596585 M * baughj the 2.6.8.1 diffs applied cleanly, once I added the include line it compiled cleanly as well
1134596599 M * Bertl http://vserver.13thfloor.at/Experimental/BME/split-2.6.11-rc5-bme0.06.1/
1134596634 M * Bertl baughj: you should verify that they work as expected too
1134596641 M * Hollow Bertl: could you explain what checkpointing is?
1134596642 M * baughj we intend on it :)
1134596750 M * Bertl baughj: soo .. you might be interested in getting a few test scripts, you could then clean up and make available for the public, yes?
1134596838 M * baughj sure, we could do that
1134596964 M * Bertl so, interested? (i.e. do we have a deal there :)
1134596988 M * baughj yeah, just email me the stuff (baughj@gnu.org) and i will see what i can do
1134597051 M * daniel_hozac speaking of testing scripts, any plans for disk limit testing in testfs? would've caught that reiserfs bug...
1134597064 M * Bertl baughj: will upload it shortly ...
1134597194 M * daniel_hozac (i wouldn't mind writing it, just wondering if it's in progress and i'd be duplicating effort)
1134597217 M * Bertl daniel_hozac: plans a lot, but nothing in progress
1134597225 M * Bertl daniel_hozac: so go crazy there!
1134597229 M * daniel_hozac ok.
1134597423 M * jeeves_ ok, time for dinner.  Thank you for your help today.  C U tom.
1134597439 Q * jeeves_ Quit: Leaving
1134597739 Q * baughj Quit: Leaving
1134598110 M * Bertl for the record: http://vserver.13thfloor.at/Experimental/BME/fstest-0.03.tar.bz2
1134598119 M * Bertl now going to send email :)
1134598123 Q * jag Quit: Leaving
1134598586 M * Bertl daniel_hozac: btw, it would be great if you could consider the (per context) quota too ... (not urgent)
1134598651 M * daniel_hozac well, i have never used quota, but i can try it.
1134598679 M * Bertl it's probably quite similar from the check POV
1134598703 M * Bertl i.e. zero, known usage, limit
1134598729 M * Bertl (but I'm more than happy with dlimits for a start, just consider it :)
1134599158 M * Bertl okay, off for a rest ... back later I guess
1134599165 N * Bertl Bertl_zZ
1134599179 M * daniel_hozac good night!
1134599607 Q * hallyn Quit: leaving
1134601411 M * micah util-vserver-0.30.209-1 debian package just migrated to testing
1134601418 M * micah and I'm about to migrate to a plane
1134602320 J * alexx ~alexx@proxy.ikse.net