1219364484 Q * dowdle_ Remote host closed the connection
1219365082 J * doener ~doener@i577B8611.versanet.de
1219365188 Q * doener_ Ping timeout: 480 seconds
1219367931 N * Bertl_zZ Bertl
1219367939 M * Bertl back now .. couldn't sleep
1219367983 M * hparker :P
1219369790 Q * tam Quit: Lost terminal
1219369804 J * tam ~tam@gw.nettam.com
1219372284 Q * loddafni1 Remote host closed the connection
1219373904 Q * hparker Quit: Read error: 104 (Peer reset by connection)
1219376256 Q * derjohn_mob Ping timeout: 480 seconds
1219378412 Q * ag- Ping timeout: 480 seconds
1219378504 J * ag- ~ag@fedaykin.roxor.cx
1219379644 Q * AndrewLee Quit: leaving
1219380551 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1219380625 Q * AndrewLee 
1219380686 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1219384159 Q * balbir Read error: Connection reset by peer
1219384485 J * derjohn_mob ~aj@e180216006.adsl.alicedsl.de
1219385071 J * balbir ~balbir@122.167.242.7
1219387052 Q * larsivi Quit: Konversation terminated!
1219387817 Q * DoberMann[ZZZzzz] Ping timeout: 480 seconds
1219388102 Q * derjohn_mob Ping timeout: 480 seconds
1219388916 J * derjohn_mob ~aj@80.69.42.51
1219390001 J * larsivi ~larsivi@85.221.53.194
1219390304 J * dna ~dna@209-241-dsl.kielnet.net
1219390718 Q * john Read error: Connection reset by peer
1219390963 J * john ~john@glou.nurvnet.org
1219391366 Q * balbir Ping timeout: 480 seconds
1219392120 M * squat danke
1219392129 M * squat mal sehen also der path applied schonmal nicht..
1219392150 J * balbir ~balbir@122.171.21.62
1219394699 M * Bertl squat: probably wrong window :)
1219394704 J * loddafni1 ~mike@193.170.48.107
1219395502 J * DoberMann[ZZZzzz] ~james@cap31-6-88-180-72-76.fbx.proxad.net
1219395584 N * DoberMann[ZZZzzz] DoberMann
1219396153 J * ktwilight__ ~ktwilight@87.66.203.64
1219396381 Q * balbir Ping timeout: 480 seconds
1219396427 Q * ktwilight_ Ping timeout: 480 seconds
1219397171 J * yarihm ~yarihm@whitehead2.nine.ch
1219401554 M * Bertl nap attack ... bbl
1219401559 N * Bertl Bertl_zZ
1219404195 J * infowolfe ~infowolfe@c-67-160-149-42.hsd1.or.comcast.net
1219406403 Q * Aiken Remote host closed the connection
1219406707 Q * cehteh Ping timeout: 480 seconds
1219408531 Q * duckx Remote host closed the connection
1219409066 J * balbir ~balbir@122.171.6.234
1219412912 Q * larsivi Quit: Konversation terminated!
1219414126 Q * nkukard Remote host closed the connection
1219416433 J * dowdle ~dowdle@scott.coe.montana.edu
1219416627 Q * dna Quit: Verlassend
1219417328 J * dna ~dna@137-213-dsl.kielnet.net
1219417858 N * Bertl_zZ Bertl
1219417861 M * Bertl back now ...
1219418952 N * DoberMann DoberMann[PullA]
1219419126 Q * infowolfe Ping timeout: 480 seconds
1219419625 J * pmenier ~pmenier@ACaen-152-1-63-64.w83-115.abo.wanadoo.fr
1219419917 Q * kir Quit: Leaving.
1219420859 J * larsivi ~larsivi@169.80-202-217.nextgentel.com
1219421247 J * ViRUS ~mp@p579B41CD.dip.t-dialin.net
1219421346 Q * yarihm Quit: Leaving
1219422559 J * nkukard ~nkukard@196.212.73.74
1219422945 J * mfv ~mfv@host69-22-dynamic.2-87-r.retail.telecomitalia.it
1219422978 J * dcdave ~dcdave@f054228127.adsl.alicedsl.de
1219423123 Q * mfv 
1219424215 J * Tooom ~toom@rejna.docisland.org
1219424256 M * Tooom hi
1219424326 M * Tooom Bertl: using vanilla util-verser 0.30.215 solve my problem (the bug is in debian package)
1219424352 Q * dcdave Quit: Ex-Chat
1219424619 M * Bertl Tooom: that was the context issue, yes?
1219424621 M * Tooom I try to identify where is the bug
1219424655 M * Bertl IIRC, you used the 0.30.215 from debian, yes?
1219424663 Q * derjohn_mob Ping timeout: 480 seconds
1219424677 M * Tooom vc_migrate_context(): Function not implemented
1219424770 M * Tooom I have to leave, thanks and good bye
1219424771 M * Bertl yep, the interesting part is, 0.30.215 should not cause that (and as you verified, mainline 0.30.215 doesn't), but we know that you can get that with 0.30.216 pre releases (they will be fixed shortly, maybe they are already)
1219424801 M * Bertl so I suspect that your 0.30.215 is actually a 0.30.216pre or so
1219424812 M * Bertl (the one from debian, that is)
1219424872 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d
1219424938 M * Tooom I used the debian package : util-vserver-0.30.216~r2750
1219424950 M * Tooom bbl
1219424954 M * Bertl ah, well, that explains it then ...
1219424956 M * Bertl cya
1219425877 J * hparker ~hparker@linux.homershut.net
1219426437 Q * pmenier Quit: Konversation terminated!
1219426730 M * Bertl daniel_hozac: feature request: could you add the vcmd info (category, command, version) to the error reporting function?
1219427333 J * ViRUS_ ~mp@p579B448F.dip.t-dialin.net
1219427511 M * daniel_hozac not easily.
1219427536 M * Bertl I thought the syscall/error stuff is wrapped?
1219427544 M * daniel_hozac error reporting happens two to three layers above from the function that knows such info.
1219427572 M * Bertl hmm, bad, maybe some kind of errno variable?
1219427589 M * Bertl vcmdno?
1219427615 Q * dowdle Remote host closed the connection
1219427674 M * daniel_hozac i guess that would be possible... will take some magic to do the right thing wrt. threads though.
1219427693 M * Bertl as does errno, IIRC
1219427696 M * daniel_hozac right.
1219427750 M * Bertl as I said, only a feature request, but I think it could be quite helpful in debugging issues
1219427763 Q * ViRUS Ping timeout: 480 seconds
1219428589 J * ktwilight_ ~ktwilight@187.101-66-87.adsl-dyn.isp.belgacom.be
1219428589 Q * ktwilight__ Read error: Connection reset by peer
1219428628 Q * esa Ping timeout: 480 seconds
1219429173 Q * ktwilight_ Read error: Connection reset by peer
1219429391 Q * dna Ping timeout: 480 seconds
1219432192 N * DoberMann[PullA] DoberMann
1219433753 M * waldi Bertl: http://bugs.debian.org/495919
1219433800 Q * pisco_ Remote host closed the connection
1219433806 M * daniel_hozac i don't see anything Linux-VServer related?
1219433821 M * waldi vserver changes the memory management
1219433824 M * Bertl nah, looks like tg3 page allocation issue
1219433832 M * daniel_hozac uh, no we don't.
1219433890 M * Bertl what kernel version is that?
1219433913 M * Bertl I mean, what mainline/vserver is related to this version?
1219433928 M * waldi 2.6.26.2 2.3.0.35
1219433973 M * Bertl waldi: google for 'tg3_alloc_rx_skb page allocation failure'
1219434072 M * waldi hmm
1219434122 M * Bertl but of course, Linux-VServer (or better the increased process/socket activity) will trigger it easier
1219434181 M * Bertl also note that this is not a crash or so, just a 'report' that the allocation failed
1219434181 M * waldi sure
1219434187 J * pisco_ ~pisco@tor.noreply.org
1219434193 M * Bertl (which actually means that a packet got dropped :)
1219434207 M * waldi i had a bug which produced such notes in a loop
1219434490 M * Bertl it would probably help this guy to disable highmem
1219434522 M * waldi or use a 64 bit kernel
1219434596 M * Bertl yes, of course :)
1219434948 J * Beuc ~yo@82.238.35.175
1219435086 M * Beuc Hi. I'm trying to use mock in a vserver, as a non-root user part of group 'mock', and I get "ERROR: [Errno 1] Operation not permitted". I added "SYS_ADMIN" to bcapabilities and "secure_mount" in ccapabilities just in case, but this doesn't help. Do you have an idea where the problem is?
1219435125 M * Beuc This is Fedora 9 (built with rinse) and mock 0.9.9
1219435146 J * derjohn_mob ~aj@e180216006.adsl.alicedsl.de
1219435156 M * Bertl what does strace -fF say?
1219435240 M * daniel_hozac CAP_MKNOD
1219435250 M * daniel_hozac and depending on the kernel, you'll need to setup the device map.
1219435276 M * Beuc Thanks.
1219435982 Q * NaioN_ Ping timeout: 480 seconds
1219436098 M * Beuc Something else: I retried building fedora hosts using 'vserver build -m yum' today, and I keep getting "rpm-fake-resolver: vc_ctx_migrate(): No such process" errors;
1219436105 M * Beuc I added a wrapper to 'rpm-fake-resolver' to trace the parameters (-F 4 -C 2147483648 -r . -c 40058) and Evc_ctx_migrate is trying to join a context that doesn't exist yet (not shown in vserver-stat or vps).
1219436115 J * NaioN ~stefan@misc.mordor.unilogicnetworks.net
1219436158 M * daniel_hozac it exists.
1219436191 M * daniel_hozac rpm-fake.so checks if the context exists, if it does, it tells rpm-fake-resolver to migrate to it.
1219436201 M * daniel_hozac if it doesn't, rpm-fake-resolver is told to create it.
1219436205 M * daniel_hozac thus, the race.
1219436263 M * Beuc What can I do to debug this?
1219436291 M * daniel_hozac i think i've tried pretty much everything, short of adding printk's to do_exit...
1219436406 M * Beuc Am I right to say that, for now, I have essentially no way to make it work?
1219436426 M * daniel_hozac just run it till it works.
1219436428 M * daniel_hozac it will, eventually.
1219436475 M * Beuc Is there a way to recode the 'yum' method to avoid this race?
1219436515 M * daniel_hozac no.
1219436527 M * Beuc That's an interesting issue.
1219436565 M * Beuc Strangely enough, I had better luck with the v216 pre-release version (that is packaged in Debian).
1219436585 M * Bertl couldn't the 'create' part be moved out of the resolver libraries completely?
1219436607 M * daniel_hozac you still need to use proper uids.
1219436827 Q * hparker Quit: off to get the kid
1219436917 M * Beuc May I ask what is the point of the resolver, in fact? What is the difference with building a system with 'yum' independently and only then start a context with it?
1219436943 M * daniel_hozac we might work around it by creating a persistent context, running vyum, and killing it off... but i'd like to get this fixed right, as it also affects things like vserver ... pkgmgmt internalize and subsequent vyums.
1219436966 M * daniel_hozac Beuc: it resolves uids/gids in the guest, and runs scriptlets in it.
1219436976 J * Aiken ~Aiken@ppp118-208-65-150.lns1.bne4.internode.on.net
1219436979 M * Bertl Beuc: if you _can_ build a guest with yum in a directory (not affecting host stuff), that would be really helpful
1219436994 M * daniel_hozac Bertl: that's not a problem at all.
1219436998 M * daniel_hozac that's what util-vserver's doing.
1219437007 M * Bertl what about scripts?
1219437091 M * daniel_hozac they're run in a chroot.
1219437109 M * Beuc Bertl: well I did a "yum --installroot=/vservers/yum-f9 install initscripts" - in what way does it affects the host?
1219437119 M * Bertl and that's enough? think apache going down/up?
1219437152 M * daniel_hozac Bertl: that's exactly my point.
1219437164 M * daniel_hozac rpm-fake is needed to put things in the guest.
1219437165 M * Bertl IMHO _all_ side effects need to happen inside a complete guest
1219437190 M * Bertl OTOH, if you can make yum have no side effects at all, that would be fine too
1219437206 M * Bertl (no guest context would be required)
1219437226 M * daniel_hozac scripts need to be run. that's why they're there, after all :-)
1219437252 M * Bertl note: my suggestion to Beuc was not a practical one :)
1219437313 M * Bertl but it contains at least some theoretical aspects, e.g. it would be better to delay all those scripts till guest startup .. I fully realize that this is not something which can be done easily
1219437334 A * Beuc has been running 'while true; do vserver test build -m yum...; done' for the past 10 minutes. The race condition winner is going Schumarer
1219437554 M * Beuc The system I made with 'yum --installroot...' starts fine now (after a manual 'initpost'), but I have problem with entering the vserver: "vlogin: openpty(): No such file or directory".
1219437602 M * daniel_hozac mkdir /dev/pts
1219437604 M * daniel_hozac restart
1219437605 M * Beuc I made sure /dev/pts is created, and also MAKEDEV std console pty fd
1219437610 M * daniel_hozac ptmx
1219437669 M * Beuc yay!
1219437770 M * Beuc Does util-vserver manage to suppress "initctl: Unable to send message: Connection refused" startup messages?
1219437779 M * Beuc Apparently they are harmless, but I'd like to know :)
1219437884 M * daniel_hozac yes.
1219437930 M * Beuc Do you remember what part of 'initpost' is involved?
1219437958 M * daniel_hozac http://svn.linux-vserver.org/projects/util-vserver/changeset/2719
1219438124 M * daniel_hozac Bertl: do you remember when we added support for persistent contexts?
1219438177 M * daniel_hozac looks like it was vs2.0.2.
1219438317 M * Bertl a long time ago, hollow 'required' it
1219438342 M * daniel_hozac yeah.
1219438356 J * ktwilight ~ktwilight@186.78-66-87.adsl-dyn.isp.belgacom.be
1219439038 M * Beuc I bootstrapped my fedora vserver using "yum --installroot=/vservers/yum-f9 install initscripts" directly without fake-rpm, what problems may I run into?
1219439177 M * daniel_hozac for such a minimal install, i don't think you should have any problems.
1219439191 M * daniel_hozac but check with rpm -Va that all the files have the right owner.
1219439272 M * Beuc Ok, nothing to report, except for the altered /etc/rc.d/init.d/halt
1219439349 M * daniel_hozac you can reproduce this issue 100% of the time, right?
1219439414 M * Beuc The "no such process" one? Apparently yes.
1219439544 M * Beuc It can work with the SVN version though (r2750 apparently).
1219439713 Q * john Ping timeout: 480 seconds
1219439744 J * john ~john@glou.nurvnet.org
1219441377 M * Bertl okay, off to bed now ... have a good one everyone!
1219441382 N * Bertl Bertl_zZ
1219441512 M * daniel_hozac Beuc: could you try http://people.linux-vserver.org/~dhozac/t/uv-testing/util-vserver-0.30.216-pre2765.tar.bz2 and see if you can make it fail?
1219441777 M * Beuc Let's see
1219442348 M * Beuc 'vserver monvserver build -m yum --force -- -d f7' says nothing.
1219442360 M * Beuc 'vserver monvserver build -m yum --force -- -d f7' again says "save_ctxinfo: open("/usr/local/var/run/vservers/monvserver"): File exists"
1219442419 M * daniel_hozac don't use --force... use vserver ... delete.
1219442464 M * daniel_hozac and FYI, f7 is no longer supported by the Fedora project.
1219442523 M * Beuc Same with "vserver vyum-f8 build -m yum -- -d f8"
1219442559 M * daniel_hozac what?
1219442601 M * Beuc It doesn't look like it's doing much: http://www.pastepine.com/142
1219442623 M * daniel_hozac -fF
1219442627 M * daniel_hozac but strace won't help.
1219442662 M * daniel_hozac sounds like you've either got empty yum.repos.d directories in /etc/vservers, or the mirrorlist specified isn't available for you
1219442739 M * Beuc yeah, my /etc/vservers/.distributions stops at f7
1219442791 M * Beuc I'll link it to lib/util-vservers/distributions
1219442806 M * daniel_hozac don't.
1219442848 M * Beuc That didn't help anyway
1219442865 M * daniel_hozac that would make it worse.
1219442882 M * daniel_hozac run it with --debug
1219442973 M * Beuc There: http://www.pastepine.com/143
1219443120 M * daniel_hozac looks like vrpm --initdb is failing...
1219443315 Q * john 
1219443331 M * daniel_hozac what kind of host system is that?
1219443374 M * Beuc Debian testing (updated this morning)
1219443391 M * daniel_hozac and you have /etc/rpm?
1219443429 M * Beuc Just a sec, yum --installroot is complaining now
1219443447 N * DoberMann DoberMann[ZZZzzz]
1219443458 Q * loddafni1 Remote host closed the connection
1219443565 M * Beuc Weird, it worked a while ago. Maybe I messed /etc/yum.conf. I get "Error: Failure finding best provider of libc.so.6(GLIBC_2.4) for initscripts-8.60-1.i386, exceeded maximum loop length"
1219443738 M * daniel_hozac never seen that before.
1219443741 M * daniel_hozac try f9?
1219443777 M * Beuc ah yeah, naturally I don't have sources for f8
1219443813 M * Beuc well I do actually..
1219443832 M * Beuc yum.conf: http://www.pastepine.com/145
1219443905 M * daniel_hozac yum.conf isn't relevant for vserver...
1219443922 M * Beuc Same behavior with f9
1219443928 M * Beuc I do have an empty /etc/rpm
1219443955 Q * nebuchadnezzar Quit: ERC Version 5.3 (IRC client for Emacs)
1219444070 J * nebuchadnezzar ~dad@zion.asgardr.info
1219444094 M * daniel_hozac well, i have never seen --initdb fail, and without any output it's hard to guess why...
1219444145 M * daniel_hozac try an strace -fF.
1219444223 M * Beuc apparently my yum fails to get dependencies right with f8
1219444271 M * Beuc yet, it works with f9, but not with util-vserver -- on to strace
1219444392 Q * ViRUS_ Quit: Leaving
1219444470 M * Beuc It behaves differently with strace -fF - it stalls
1219444489 M * daniel_hozac well, that's expected.
1219444509 Q * bonbons Quit: Leaving
1219444525 M * daniel_hozac did you see the rpm --initdb call?
1219444561 M * Beuc Nope
1219444578 M * daniel_hozac so where did it stop?
1219444594 M * Beuc "[pid 10742] waitpid(-1, Process 10742 suspended"
1219444602 M * Beuc I was trying to pastebin it, but it's too large
1219444650 M * Beuc http://www.pastepine.com/146 beginning - http://www.pastepine.com/147 last few lines
1219444663 M * daniel_hozac so that's the entire strace?
1219444684 M * Beuc something is probably missing
1219444734 M * daniel_hozac that looks pretty good to me.
1219444785 M * daniel_hozac but it looks like strace messes stuff up too much.
1219444790 M * Beuc http://beuc.net/trace.gz if needed
1219444842 M * Beuc I'll have to stop here, I'm a tad tired, in this kind of state where you usually make easy mistakes :)
1219444911 M * daniel_hozac okay, we can continue some other time.
1219445023 M * Beuc nice, see you
1219445593 Q * Beuc Ping timeout: 480 seconds