1177373050 M * fromport since i got it working under debian on an amd sempron I certainly started to doubt if vserver actually works on a via C3 cpu. I googled around extensively but couldn't find an conclusive answer
1177373208 M * ruskie don't see why not
1177373215 M * ruskie it's not dependant on cpu last I checked
1177373241 M * ruskie the way I understand it is it's a glorified chroot with some nice features...
1177373338 M * fromport older via c3 doesn't have same instructions, i dont know if vserver use them
1177373371 M * ruskie you mean it's not x86 compatible?
1177373430 M * fromport i just checked that i actually had compiled the 2.6.20.7 kernel with the 2.3.0.12 code. i just recompiled with 2.2.0, let's see if that works ;-)
1177373501 M * fromport Whilst the CPU is 686 class, it lacks the cmov extension which gcc
1177373528 Q * bored2sleep Server closed connection
1177373589 J * bored2sleep ~bored2sle@66.111.53.150
1177374250 M * fromport still no go:
1177374253 M * fromport # vserver-info
1177374253 M * fromport Versions:
1177374253 M * fromport                    Kernel: 2.6.20.7
1177374253 M * fromport                    VS-API: 0x00020200
1177374253 M * fromport              util-vserver: 0.30.212; Dec  9 2006, 12:26:51
1177374296 M * fromport will try a new install of a vserver next
1177374460 Q * eyck Server closed connection
1177374472 J * eyck ~eyck@nat.nowanet.pl
1177374504 M * fromport this would a valid install command line right ?
1177374507 M * fromport newvserver --vsroot /var/lib/vservers/ --hostname test1 --domain int.bitsource.net --ip 10.1.10.99/24 --dist etch --mirror http://mirrors.kernel.org/debian/ --interface eth1
1177374524 M * ruskie no clue I use vserver build
1177374541 M * fromport oki, this seems to be debian specific (i guess)
1177374698 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net
1177374803 Q * _jthm_ Read error: No route to host
1177375176 M * fromport still gives me the same problems:
1177375183 M * fromport during install i get:
1177375187 M * fromport I: Configuring tasksel...
1177375187 M * fromport I: Base system installed successfully.
1177375187 M * fromport ncontext: vc_net_create(): Invalid argument
1177375187 M * fromport An error occured while executing the vserver startup sequence; when
1177375187 M * fromport there are no other messages, it is very likely that the init-script
1177375189 M * fromport (/etc/init.d/rc 3) failed.
1177375368 Q * kwowt Ping timeout: 480 seconds
1177375385 Q * fromport Read error: Connection reset by peer
1177375424 J * fromport ~dth@cpe-75-82-131-136.socal.res.rr.com
1177375561 M * fromport it works on 2 other machines (with stock debian kernel) , not on my compiled kernel. on the via c3 i cannot run a vserver enabled stock kernel (no full 686 instruction set)
1177375570 M * fromport thanks for the help/feedback
1177376083 Q * Solver Server closed connection
1177376100 J * Solver ~robert@CPE00a0c96b79ba-CM000039aeec61.cpe.net.cable.rogers.com
1177377100 Q * johnny Remote host closed the connection
1177377698 Q * bzed Quit: Leaving
1177378439 Q * Adrinael Server closed connection
1177378442 J * Adrinael adrinael@rid7.kyla.fi
1177378507 J * johnny ~urr@ip70-160-117-211.hr.hr.cox.net
1177378514 M * johnny hi folks
1177378517 M * johnny i was here http://linux-vserver.org/Frequently_Asked_Questions#How_do_I_assign_a_static_context_to_an_existing_vserver.3F
1177378536 M * johnny and it says 1 isn't a valid static context. is that true?
1177378567 M * johnny oh oops.. thought my context was 1.. :)
1177378580 M * johnny i was 95% sure i picked 1, but i actually picked 2
1177379066 M * johnny what does the SINGLE_IP flag do?
1177379249 M * bored2sleep anyone ever have any problems getting guest ldconfig to pick up settings in /etc/ld.so.conf? (setting is also in the host ld.so.conf)
1177379371 M * bored2sleep hmm, actually it seems that with vserver ... enter, it works right.  with vserver ... exec it doesn't work
1177379552 M * lyli1 johnny: xid 1 -> http://www.mail-archive.com/vserver@list.linux-vserver.org/msg10518.html
1177379596 M * johnny aha nice
1177379597 M * johnny thanks
1177379611 M * johnny i just recently started using 2.2
1177379616 M * johnny err 2.2.0
1177379662 M * lyli1 which is how: vcontext --migrate --xid 1 htop
1177379689 M * lyli1 works w/ accounting all process across vservers :)
1177379703 M * lyli1 nice... 2.2.0 works well; fixed some lingering issues
1177379896 M * johnny my distro didn't have nice vserver init scripts when i set this one up
1177379900 M * johnny back during 1.9 series
1177379909 M * johnny so i set it up manually
1177379920 M * johnny so now i've had to twidde some things to keep it workin
1177380047 Q * opuk Server closed connection
1177380059 J * opuk ~kupo@c213-100-138-228.swipnet.se
1177380987 Q * toidinamai Ping timeout: 480 seconds
1177381129 J * toidinamai ~frank@i59F72623.versanet.de
1177382109 Q * ensc Ping timeout: 480 seconds
1177383499 J * ensc ~irc-ensc@p54b4ed46.dip.t-dialin.net
1177383774 Q * infowolfe Read error: Connection reset by peer
1177383796 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net
1177384336 J * infowolfe_ ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net
1177384372 Q * arachnist Quit: brb
1177384579 Q * mountie Server closed connection
1177384594 J * mountie ~mountie@trb229.travel-net.com
1177384596 J * arachnist ~arachnist@088156189068.who.vectranet.pl
1177384793 Q * infowolfe Ping timeout: 480 seconds
1177385769 Q * mnemoc Server closed connection
1177385781 J * mnemoc ~amery@kilo105.server4you.de
1177385983 Q * Radiance Server closed connection
1177386003 J * Radiance 9cea374280@halt.1984world.eu
1177386839 Q * TrueBrain Server closed connection
1177386847 J * TrueBrain truelight@openttd.org
1177386958 Q * ntrs Server closed connection
1177386973 J * ntrs ntrs@68-188-55-120.dhcp.stls.mo.charter.com
1177387798 Q * Hunger Server closed connection
1177387843 J * Hunger Hunger.hu@Hunger.hu
1177388318 Q * infowolfe_ Ping timeout: 480 seconds
1177388418 Q * bronson Ping timeout: 480 seconds
1177389105 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net
1177389198 Q * ray6 Server closed connection
1177389695 M * fromport Q: my machine has 2 ethernet cards: 1 outside, 1 inside (10.1.10.1) if i want a vserver on the inside with ip 10.1.10.9 is this the right start command ?
1177389703 M * fromport  vserver lala build -n lala --hostname lala --interface eth1:10.1.10.9/24 -m debootstrap --force -- -d etch -m http://mirrors.kernel.org/debian/
1177389780 M * fromport the installation ends normally without a hitch but: 
1177389781 M * fromport # vserver lala start
1177389781 M * fromport ncontext: vc_net_create(): Invalid argument
1177390038 Q * Wonka Server closed connection
1177390043 J * Wonka produziert@chaos.in-kiel.de
1177391349 M * daniel_hozac fromport: you're missing the --context option.
1177392014 Q * FaUl Server closed connection
1177392015 J * FaUl immo@shell.chaostreff-dortmund.de
1177392145 Q * nox Server closed connection
1177392157 J * nox ~nox@static.88-198-17-175.clients.your-server.de
1177392207 Q * pusling Server closed connection
1177392215 J * pusling pusling@195.215.29.124
1177392257 M * fromport daniel_hozac: ok will try and read something about it in the documentation
1177392743 Q * softi42 Ping timeout: 480 seconds
1177392881 M * fromport daniel: according to the man page:
1177392883 M * fromport cfg-options are:
1177392883 M * fromport     --context   ...  the static context of the vserver [default: none; a dynamic
1177392886 M * fromport                      context will be assumed]
1177392924 M * fromport I read that it is optional.. but i will try with...
1177392945 J * bronson ~bronson@adsl-76-202-197-80.dsl.pltn13.sbcglobal.net
1177393095 M * daniel_hozac fromport: well, your kernel doesn't have dynamic contexts, so you'll have to specify one.
1177393227 M * fromport ah ! thanks
1177393313 N * DoberMann[ZZZzzz] DoberMann
1177393322 M * fromport daniel: asked here before but no answer yet: does vserver run on a via c3 processor ? "Whilst the CPU is 686 class, it lacks the cmov extension which gcc"
1177393333 M * daniel_hozac doesn't matter.
1177393344 M * daniel_hozac as long as you don't compile your kernel for 686...
1177393357 J * softi42 ~softi@p549D7FDD.dip.t-dialin.net
1177393413 M * fromport nope, but since i cant use a standard (debian) kernel since it's 686 only that is vserver enabled, i compiled my own. and that's the one with the problems ;-) and the debian newvserver gives me the same problem.
1177393463 M * daniel_hozac well, specify a static context and all of your problems will be gone...
1177393582 M * fromport Enable dynamic context IDs (NEW) (found the kernel option) *MANY* thanks
1177393599 M * daniel_hozac you really want to use static ids.
1177393668 M * fromport so each vserver runs with a static context ?
1177393789 M * daniel_hozac yes
1177394614 M * fromport daniel: it works ! thanks!
1177394641 M * fromport i've being trying for almost 2 days now! i now finally can sleep again ;-)
1177394655 M * daniel_hozac you're welcome!
1177395134 Q * bronson Quit: Ex-Chat
1177395974 Q * Bertl_vV Server closed connection
1177395976 J * Bertl_vV herbert@IRC.13thfloor.at
1177395995 M * fromport daniel: i'm currenlty running 2.2.0 patch. I'm using encrypted filesystem which i want only to be availble in one ofthe guests. When i use dmcrypt now i get
1177395998 M * fromport Command failed: Incompatible libdevmapper 1.02.08 (2006-07-17)(compat) and kernel driver
1177396013 J * oliwel ~chatzilla@ppp-82-135-74-227.dynamic.mnet-online.de
1177396023 M * fromport do you think i have more luck with the developement version ?
1177396062 A * oliwel waves a friendly good morning to the world
1177396085 A * fromport waves back from a place where it is 23:30 ;-)
1177396106 M * oliwel fromport: ;) - well - for some people this is close to noon
1177396154 Q * johnny Quit: Ex-Chat
1177396536 N * DoberMann DoberMann[PullA]
1177396713 J * dna ~naucki@255-216-dsl.kielnet.net
1177396746 Q * fromport Quit: Zzzzz
1177397297 J * sharkjaw ~gab@158.36.45.236
1177397325 Q * badari Server closed connection
1177397356 J * badari ~badari@bi01p1.co.us.ibm.com
1177397540 Q * kaner Server closed connection
1177397542 J * kaner kaner@strace.org
1177398280 J * borgfish ~bla@141.12.67.238
1177398386 Q * oliwel Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.3/2007030916]
1177399141 J * zLinux ~zLinux@88.213.13.130
1177399340 Q * arachnist Read error: Operation timed out
1177399679 J * oliwel ~chatzilla@ppp-82-135-74-227.dynamic.mnet-online.de
1177399699 M * oliwel knockknock
1177399703 M * oliwel daniel_hozac: doener: present ?
1177401492 J * derjohn2 ~aj@80.69.41.3
1177401824 Q * tzafrir_laptop Ping timeout: 480 seconds
1177403080 Q * weasel Quit: Reconnecting
1177403082 J * weasel weasel@asteria.debian.or.at
1177403479 Q * FloodServ charon.oftc.net services.oftc.net
1177403737 J * FloodServ services@services.oftc.net
1177403869 Q * FloodServ cation.oftc.net charon.oftc.net
1177403872 J * FloodServ services@services.oftc.net
1177404056 J * fs fs@213.178.77.98
1177404183 J * thessy H2Jb4755@nat-1.rz.uni-karlsruhe.de
1177404288 J * dothebart ~willi@xdsl-84-44-228-174.netcologne.de
1177404292 Q * tudenbart Read error: Connection reset by peer
1177405395 Q * thessy Quit: using sirc version 2.211+KSIRC/1.3.12
1177405531 J * meandtheshell ~markus@85-124-206-194.dynamic.xdsl-line.inode.at
1177405705 Q * kir Quit: Leaving
1177407677 J * chand ~chand@212.99.51.254
1177407823 J * johan ~johan@mx1.ttesa.net
1177407842 M * johan Hello
1177408283 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at
1177408351 Q * oliwel Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.3/2007030916]
1177408474 M * johan Has anyone successfully setup courier-imap inside a vserver?
1177408501 M * johan I have an issue it always wants to connect to 127.0.0.1 111, which takes 3 minutes to time-out...
1177408541 M * johan It seems 127.0.0.1 is hardcoded as I don't use this address in any config files.
1177408587 M * johan Basically I don't know how to proceed to fix it. (And I don't want to change the source an recompile.)
1177408677 M * PowerKe imap is port 143 btw
1177408764 M * PowerKe doesn't seem to be binding to any other IP then what I tell it to
1177408784 M * PowerKe check /etc/courier-imap/imapd for ADDRESS=0 and replace with your IP
1177408854 Q * sharkjaw Quit: Leaving
1177408855 M * johan if it would connecto to 127.0.0.43 111 (.43 is the localhost inside the vserver) it would get a "connection refused" too.
1177408864 M * johan but it wants to connect to 127.0.0.1
1177408897 M * johan Imap is bound in the config to the external IP address. This is done correctly.
1177408948 M * johan So, it is not that imap binds to the wrong IP, it is just that it tries to access 127.0.0.1 during mailbox access.
1177409089 M * johan strace:
1177409092 M * johan bind(3, {sa_family=AF_INET, sin_port=htons(943), sin_addr=inet_addr("0.0.0.0")}, 16) = -1 EACCES (Permission denied)
1177409092 M * johan connect(3, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("127.0.0.1")}, 16
1177409103 M * johan And then it takes 3 minutes to timeout...
1177409113 M * johan This is inside a vserver.
1177409154 M * johan inside the host it goes like this:
1177409156 M * johan bind(3, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
1177409156 M * johan connect(3, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("127.0.0.1")}, 16) = -1 ECONNREFUSED (Connection refused)
1177409156 M * johan close(3)                                = 0
1177409163 M * johan Connection refused, no timeout.
1177409187 M * johan Is there anyway to reject connections to 127.0.0.1 inside a guest?
1177409393 M * PowerKe seems to be fam/portmap related, maybe try compiling without fam support?
1177409411 M * PowerKe (I wouldn't even know what that is, but it turns up on google and I seem to have it compiled without)
1177409540 M * sid3windr file alteation monitor :)
1177409708 M * PowerKe yes, just googled that. Does courier-imap use it to notify clients when new email arrives?
1177409840 M * sid3windr think so.
1177410764 M * johan I don't use FAM, I tried setting IMAP_ENHANCEDIDLE to 1, but that doesn't make any difference.
1177410791 M * johan I use the stable debian packages for imapd
1177411800 M * johan I googled for a bit, but I don't think this is the FAM problem. I don't get any errormessaged. Basicaly there are no errors. It wants to connect to 127.0.0.1 and that needs to timeout...
1177412538 M * PowerKe I think it want to connect to 127.0.0.1 to ask portmap if fam is running
1177412586 M * johan But how do I stop it? ;-)
1177412633 M * PowerKe I think by telling it not to use fam
1177412680 M * PowerKe Maybe you could make a copy of the guest and put a self-compiled imapd (without fam) in there to see if that works?
1177412737 M * PowerKe Does it timeout when starting imapd, or when a client connects?
1177412747 M * johan OK, I have a friend also running imapd inside a vserver, and he gets connection refused when trying to connect to 127.0.0.1 from inside the guest. (127.0.0.1 is not the localhost of the guest.)
1177412775 M * johan I timeout everytime the mailbox is accessed.
1177413588 J * onox ~onox@kalfjeslab.demon.nl
1177414004 Q * Aiken Quit: Leaving
1177414699 J * LarsW ~wayden@p5080E0C5.dip.t-dialin.net
1177415223 M * johan OK, I solved it.
1177415265 M * johan I allowed access from the guest interfaces to the host lo0 interface.
1177415288 M * johan It's more a workaround instead of a solution. But that's OK for now.
1177415575 M * derjohn2 i run courier in a vserver guest . you might simply apt-get remove fam and try again. i dint know if i enabled fam in that gueszt.
1177415693 M * johan I don't have fam
1177415729 M * johan It seems imap has hardcoded 127.0.0.1 somewhere in stead of using localhost from the hostfile.
1177417018 M * brcc_ good morning
1177419180 J * knopkopp ~knoppix@dialin-145-254-194-033.pools.arcor-ip.net
1177419384 P * knopkopp Verlassend
1177419524 J * arachnist ~arachnist@88.156.189.68
1177419884 Q * glen_ Read error: Connection reset by peer
1177419896 J * glen_ ~glen@elves.delfi.ee
1177420250 J * Fire_Egl ~FireEgl@adsl-220-216-111.bhm.bellsouth.net
1177420545 M * ex anyone knows which means 'vxW: !!! limit: caab104c[SHMEM,19] = 1024 on exit.
1177420623 M * ex in vserver 2.2.0 while restarting fms (flash media server in guest)
1177420638 Q * FireEgl Ping timeout: 480 seconds
1177420940 M * cruser Hi. I had change selinux "permissive" to "disabled" on a vserver host to be able to install (yum) bzip2 on a guest vserver :( That did not seem right.
1177421989 Q * Fire_Egl Remote host closed the connection
1177422215 P * johan Leaving
1177422281 J * ema ~ema@rtfm.galliera.it
1177422486 J * djpavlik ~djpavlik@195.160.234.1
1177422628 M * djpavlik hi all,
1177422628 M * djpavlik did anyone tried to use PAE Vserver kernel on fedora 6?
1177423024 Q * SadMan Quit: .
1177423319 M * daniel_hozac djpavlik: i take it it doesn't work?
1177423484 M * daniel_hozac cruser: well, umm, have you setup SELinux properly to cope with guests?
1177423504 M * daniel_hozac ex: it means the guest had 1024 <unit> of shared memory when the guest was stopped.
1177423659 M * ex daniel_hozac, looks harmless, thx
1177423693 M * cruser daniel_hozac: I understood that "permissive" mode was the same as "disabled" only logging is activated in permissive. If you ever want to active selinux it is best done from a previous "permissive" configuration as opposed to "disabled".
1177423753 M * cruser So I am thinking there is no selinux setup when you are in the permissive mode.
1177423779 M * daniel_hozac there is.
1177423804 M * daniel_hozac AFAIK, anything explicitly disallowed is still disallowed.
1177424816 J * FireEgl ~FireEgl@Sebastian.Atlantica.US
1177424990 M * cruser daniel_hozac: thanks. I will read up.
1177425019 J * Piet hiddenserv@tor.noreply.org
1177425479 Q * daniel_hozac Ping timeout: 480 seconds
1177425793 Q * sladen Server closed connection
1177425801 J * sladen paul@starsky.19inch.net
1177425886 M * djpavlik daniel_hozac: the system is being automatically restarted each time I'm selecting the kernel to boot
1177425999 Q * click Remote host closed the connection
1177426015 Q * starcode Quit: starcode
1177426092 M * djpavlik official PAE kernel (means the one released by Red Hat) works normally but with vs2.2.0.1 patch crashes at startup
1177426927 Q * DavidS Quit: Leaving.
1177427084 J * hardwire ~bip@rdbck-3091.wasilla.mtaonline.net
1177427805 M * Bertl_vV greetings folks!
1177427906 J * stefani ~stefani@flute.radonc.washington.edu
1177428309 M * Bertl_vV good morning stefani!
1177428319 M * stefani hola Bertl_vV 
1177428391 J * bzed ~bzed@dslb-084-059-115-096.pools.arcor-ip.net
1177428438 M * Bertl_vV welcome bzed!
1177428499 M * bzed thanks Bertl_vV  :)
1177428701 M * djpavlik rpm.hozac.com is down?
1177430738 Q * derjohn2 Ping timeout: 480 seconds
1177430985 J * bonbons ~bonbons@83.222.38.145
1177431021 Q * toidinamai Quit: Leaving
1177431023 M * Bertl_vV djpavlik: looks like ...
1177431068 M * djpavlik Bertl_vV: today's day is not mine :(
1177431128 M * Bertl_vV what do you need from there?
1177431250 M * djpavlik I have a problem with PAE kernel version 2.6.20-1.2936.fc6.vs2.2.0.1. Isaw there's some other minor version VS kernel with PAE patch. Probable this once will work. That's my last hope, otherwise my vservers will see only 4 Gb of RAM :(
1177431484 M * Bertl_vV why PAE? what machine are you using?
1177431614 M * djpavlik I'm using Dell Power Edge 1850 with 6Gb of DDR2 ECC
1177431758 M * Bertl_vV could you upload the output of /proc/cpuinfo to some pastebin please?
1177431787 J * marcfiu ~mef@aegis.CS.Princeton.EDU
1177431801 M * Bertl_vV hey marcfiu!
1177431821 M * djpavlik yes, moment...
1177432014 J * DoberMann_ ~james@AToulouse-156-1-178-109.w90-38.abo.wanadoo.fr
1177432123 Q * DoberMann[PullA] Ping timeout: 480 seconds
1177432213 M * djpavlik Bertl_vV: see here http://bgf-mirc.pastebin.ca/455987
1177432280 M * Bertl_vV yup, your machine is 64bit capable, thus I'd strongly advise to install a 64bit host system
1177432350 M * Bertl_vV which will not only give you better performance, but also make the memory issue a non-issue
1177432367 M * djpavlik thanks :)
1177432367 M * djpavlik I doubts about x64 support as I do not know how to determine it via cpuinfo
1177432384 M * Bertl_vV it's the 'lm' flag
1177432387 M * djpavlik I mean I had doubts...
1177432426 M * djpavlik thanks for advise, but the next question is: can I install x32 guests on x64 host OS?
1177432437 M * Bertl_vV yes, of course :)
1177432448 M * djpavlik great :)
1177432484 M * arachnist i actually prefer x86_64 kernel + i686 system, but that requires one to have additional toolchains
1177432605 M * djpavlik arachnist: what do you mean saying additional toolchains?
1177432615 Q * AndrewLee Server closed connection
1177432626 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1177432663 M * arachnist djpavlik: to compile x86_64 kernel on a i686 system you need a crosstoolchain
1177432696 M * Bertl_vV yes, but kernel cross toolchains are fairly easy to build
1177432729 M * Bertl_vV djpavlik: you do not have the problem if your host system is 64bit
1177432741 M * Bertl_vV djpavlik: the 64bit toolchain can build 32bit stuff too
1177432771 M * djpavlik I can install Fedora from official x64 distro and then install just an x64 vs2.2.0 kernel RPM, right?
1177432787 M * Bertl_vV yep
1177432839 M * djpavlik okay, folks, now I'm going to play with Fedora x64
1177432839 M * djpavlik thanks for help
1177432855 M * djpavlik aha, one more questions...
1177432859 M * Bertl_vV you're welcome!
1177432909 M * djpavlik will the guest x32 processes see all 6Gb of RAM in case the host kernel is x64?
1177432960 M * Bertl_vV no
1177432974 M * Bertl_vV it will be limited to 3GB or something
1177432995 M * Bertl_vV but different processes will be able to use different portions
1177433008 M * djpavlik okay, that's sounds good
1177433031 M * djpavlik thanks
1177433086 M * Bertl_vV feel free to hang around btw, you might get some en-passant information ...
1177433433 M * lyli1 Bertl_vV: g/a
1177433484 M * lyli1 chattin w/ daniel_hozac last night and he pointed us to the mainline network namespace patches from biederman
1177433487 M * lyli1 looks interesting
1177433581 J * derjohn2 ~aj@80.69.41.3
1177433614 M * Bertl_vV lyli1: yep, we actually have that running here together with Linux-VServer, but don't tell :=
1177433707 M * lyli1 O_O
1177433762 M * sid3windr :p
1177433775 A * sid3windr perks his ears
1177433994 Q * chand Quit: chand
1177434000 M * bzed does anybody have an idea where this comes from: grsec: From 123.123.123.123: denied resource overstep by requesting 69419390976000 for RLIMIT_STACK against limit 8388608 for /usr/lib/util-vserver/secure-mount[secure-mount:31088] uid/euid:0/0 gid/egid:0/0, parent /usr/sbin/vserver[vserver:31039] uid/euid:0/0 gid/egid:0/0
1177434045 M * bzed it happened after setting up a new guest, not sure when exactly, though
1177434182 M * lyli1 Bertl_vV: how long would you gauge until mainline acceptance?
1177434226 M * onox djpavlik: have you tried Gentoo?
1177434242 M * djpavlik no, i didn't
1177434272 J * daniel_hozac ~daniel@c-081472d5.010-230-73746f22.cust.bredbandsbolaget.se
1177434281 M * bzed not sure how it works on other linux flavours, but setting up a complete guest is done in a minute in debian
1177434374 M * djpavlik onox: I have to use an RPM-based distro because my custom application is being installed from RPMs
1177434412 M * onox which custom app?
1177434497 M * djpavlik onox: a set of WEB-applications written in Java and running on Tomcat. I'm packing them into RPMs and then deploy to the vserver(s)
1177434526 M * onox ah
1177434537 J * svbb55980 kad@c-956070d5.01-142-6b736810.cust.bredbandsbolaget.se
1177434559 M * onox well, you can have a distro A running guests of type B
1177434624 Q * daniel_hozac Read error: Connection reset by peer
1177434643 M * djpavlik onox: yes, but my managed hosting provider provides only RH distros :)
1177434665 M * sid3windr ew
1177434703 M * onox managed hosting provider: does that mean you run your own guest server?
1177434783 J * daniel_hozac ~daniel@c-081472d5.010-230-73746f22.cust.bredbandsbolaget.se
1177434906 M * djpavlik djpavlik: I'm running whole cluster with a set of physical host machines running vserver kernel and vservers. Mostly, managed hosting is just called managed as we do a lot of work ourselves. But in case of non-RH distro I'm gonna lose "live hands" at all, because - "your distro - you do support" :)
1177434909 Q * phedny Server closed connection
1177434926 J * phedny ~mark@ip56538143.direct-adsl.nl
1177435024 M * onox lol
1177435026 M * arachnist /etc/vserver/name.conf are old-style configs, right?
1177435032 M * Bertl_vV yep
1177435034 M * svbb55980 ZEnet IRC Network - irc.zenet.org - www.zenet.org
1177435038 M * svbb55980 ZEnet IRC Network - irc.zenet.org - www.zenet.org
1177435046 M * Bertl_vV arachnist: ancient to be precise :)
1177435053 M * Bertl_vV svbb55980: hmm?
1177435064 M * svbb55980 ZEnet IRC Network - irc.zenet.org - www.zenet.org
1177435066 M * svbb55980 ZEnet IRC Network - irc.zenet.org - www.zenet.org
1177435069 M * arachnist Bertl_vV: and what if vserver-copy pretends to need it?
1177435070 M * svbb55980 ZEnet IRC Network - irc.zenet.org - www.zenet.org
1177435073 M * onox Bertl_vV: why sits #vserver on irc.oftc.net, why not on irc.freenode.org?
1177435076 Q * svbb55980 
1177435096 M * Bertl_vV onox: because we (the community) decided so :)
1177435126 M * arachnist global msg/notice annoyance? :>
1177435135 M * sid3windr etc
1177435144 Q * derjohn2 Ping timeout: 480 seconds
1177435148 M * sid3windr oftc >> freenode :]
1177435179 M * arachnist the only difference i notice are hostmask and i=/n= for ident on freenode
1177435189 M * arachnist hostmasks*
1177435196 M * Bertl_vV but till his tragic accident, even lilo was on this channel on a regular basis :)
1177435220 M * Bertl_vV and recent/modern irc clients handle multiple servers quite well, no?
1177435397 M * daniel_hozac djpavlik: rpm.hozac.com should be up again.
1177435400 M * daniel_hozac stupid power outage...
1177435416 M * arachnist huh
1177435435 M * Bertl_vV daniel_hozac: ah, good, tx :)
1177435438 M * arachnist there used to be at least one/month here, until recently
1177435482 M * djpavlik daniel_hozac: I'm getting 502 Bad Gateway
1177435502 M * onox Bertl_vV: yes, but i am too lazy to find out how irssi saves my current channels and connects to it when started
1177435520 M * sid3windr :p
1177435546 M * daniel_hozac djpavlik: really? that's odd.
1177435557 M * djpavlik daniel_hozac: yeep, now rpm.hozac.com it is back!
1177435565 M * Bertl_vV onox: save (layout)?
1177435594 M * djpavlik I mean error 502 disappeared
1177435663 M * arachnist what to do if vserver-copy asks for /etc/vserver/name.conf?
1177435680 M * daniel_hozac arachnist: realize that it's legacy and use vserver ... build -m rsync/clone?
1177435681 M * Bertl_vV arachnist: upgrade the tools?
1177435699 Q * djpavlik Quit: Be free
1177435818 J * click click@ti511110a080-2856.bb.online.no
1177436014 J * TheSasch ~TheSasch@p57af6a20.dip.t-dialin.net
1177436059 M * TheSasch Hello There
1177436064 M * daniel_hozac hi
1177436089 M * TheSasch i have a question i use 2.6.20.4 with stable 2.2 on a debian etch 
1177436117 M * TheSasch few hours ago i added 2 harddisks to my system 
1177436136 M * TheSasch and since then i got the following errors when i start my guests:
1177436149 M * TheSasch secure-mount: mount(): Device or resource busy 
1177436175 Q * ema Quit: leaving
1177436189 M * daniel_hozac and where did you mount it?
1177436194 M * daniel_hozac (the harddrives)
1177436210 M * TheSasch  /dev/hdg1 and /dev/hdh1
1177436230 M * daniel_hozac i mean the mountpoint.
1177436245 M * TheSasch  /mnt/hdg1 and /mnt/hdh1 
1177436270 M * TheSasch fstab from atlas ist default:
1177436271 M * TheSasch none    /proc           proc    defaults                0 0 
1177436271 M * TheSasch none    /tmp            tmpfs   size=16m,mode=1777      0 0 
1177436271 M * TheSasch none    /dev/pts        devpts  gid=5,mode=620          0 0
1177436306 M * daniel_hozac well, that's interesting.
1177436334 M * Bertl_vV TheSasch: what filesystem do they use?
1177436340 M * TheSasch xfs
1177436364 M * Bertl_vV could you reformat them for a test? or do they contain data?
1177436388 M * TheSasch they contain data. i used them 2 years or more at a 1.xx vserver machine
1177436418 M * TheSasch is there a limit to moutpoints?
1177436427 M * Bertl_vV no, not that I know of :)
1177436429 M * TheSasch mount|wc -l gives me 33
1177436449 M * Bertl_vV /proc/mounts should give you even more, I guess
1177436479 M * TheSasch one problem i could have is that i use standard debian vserver-utils (0.30.212-1 --> debian versionid)
1177436488 M * Bertl_vV very bad :)
1177436502 M * daniel_hozac hmm?
1177436507 M * Bertl_vV never use out of the box debian tools :)
1177436519 M * daniel_hozac etch should be good for now.
1177436535 M * daniel_hozac (for now == another day or two ;))
1177436540 M * Bertl_vV wasn't there a -2 with a lot of fixes?
1177436565 M * daniel_hozac not that i know of.
1177436573 M * Bertl_vV not speaking of 0.30.213-rc which actually can use debootstrap again :)
1177436614 M * Bertl_vV daniel_hozac: you're the expert on that ... TheSasch: so please listen to daniel_hozac ...
1177436627 M * onox Bertl_vV: what's is wrong with debian tools?
1177436629 M * Hollow it works quite well actually, i've been using debian packages for the last weeks...
1177436649 M * Hollow morning btw, or whatever :)
1177436650 M * daniel_hozac 0.30.212-1 is the latest version.
1177436662 M * Bertl_vV onox: usually they are outdated and broken ...
1177436673 M * daniel_hozac (see sarge)
1177436783 M * Bertl_vV onox: if you guest building starts with 'newvserver' it's already terribly wrong for example ...
1177436815 M * TheSasch hmmm i used newvserver to get an examle :-))
1177436847 M * TheSasch i had to migrate the 1.2 configs to new config format. but there are no other errors at starttime
1177436855 M * Bertl_vV daniel_hozac: btw, the dynamic id works quite nicely ...
1177436867 M * daniel_hozac as long as you don't already use that range :)
1177436884 M * onox Bertl_vV: I understand 'outdated' :p, but not broken, with debian I would have thought people say 'tried and true' or something like that :p
1177436914 M * Bertl_vV onox: the problem is, most reported issues are not fixed in a timely fashion
1177436931 M * Bertl_vV onox: thus the tools are not only outdated but also broken in some regard
1177436960 M * daniel_hozac "some" :)
1177436966 M * Bertl_vV onox: but that is the debian way ...
1177437092 M * onox k, thx
1177437143 M * onox i'll paste those lines into a file, so I can cure more people who are afraid of compiling
1177437182 M * onox Bertl_vV: have you tried Gentoo? ;)
1177437208 M * Bertl_vV onox: I've tried almost any distro out there (over time)
1177437255 M * TheSasch is that a problem for my vservers to get this error messages from mount? they are still come up and seem to run ok
1177437283 J * oliwel ~chatzilla@ppp-82-135-82-123.dynamic.mnet-online.de
1177437570 M * oliwel daniel_hozac: ??
1177437711 M * onox Bertl_vV: :|
1177437717 M * onox it was supposed to be a joke
1177437722 M * onox you should say LOL then :p
1177437742 M * oliwel onox: hi ;)
1177437769 Q * rob-84x^ Server closed connection
1177437772 J * rob-84x^ rob@submarine.ath.cx
1177437818 M * onox oliwel: hi
1177437858 M * oliwel might I help you ?
1177437876 M * onox with what?
1177437882 M * oliwel or are you just wating your time here ;)
1177437889 M * Bertl_vV onox: okay, LOL! :)
1177437891 M * oliwel dont know - look like you want to contact bertl...
1177437899 M * oliwel Haaa Bertl is back ;)
1177437914 M * oliwel Welcome Bertl
1177437914 M * onox oliwel: you can help me with saying LOL
1177437920 M * Bertl_vV oliwel: well, still on vacation (or something like that :)
1177437922 M * oliwel LOL 
1177437932 M * onox LOl, he said LOL
1177437944 M * oliwel Bertl_vV: I thought that helping vserver guys is always like vacation;)
1177437974 M * Bertl_vV yeah, but less the 'sight-seeing' type of vacation
1177437986 M * oliwel were are you ?
1177437994 M * oliwel where.....
1177438000 M * Bertl_vV currently in Princeton, NJ
1177438016 J * waldi ~waldi@bblank.thinkmo.de
1177438025 M * oliwel Not bad
1177438041 M * oliwel really for vacation or are you working there 
1177438051 M * waldi Bertl_vV: ETA for vserver stacking?
1177438084 M * Bertl_vV well, I'm in the US for vacation (really), but I'm currently helping the princeton folks with Linux-VServer (today)
1177438103 M * oliwel ;)
1177438103 M * Bertl_vV waldi: you mean guests inside guests?
1177438119 M * oliwel Might I interruot your vacation with a question ?
1177438127 M * Bertl_vV sure .. go ahead
1177438131 M * onox Bertl_vV: what are the princeton folks using Linux-VServer for?
1177438140 M * Bertl_vV onox: planetlab
1177438159 M * Bertl_vV http://www.planet-lab.org/
1177438232 M * waldi Bertl_vV: jup
1177438267 M * yang hey Bertl_vV, how is the USA for you :)
1177438347 M * Bertl_vV yang: quite nice, thanks ...
1177438378 M * Bertl_vV waldi: not going to happen anytime soon, as it would require priviledges a guest is not supposed to have
1177438413 M * waldi Bertl_vV: bad
1177438425 M * Bertl_vV waldi: besides that, there is little point in making them hierarchical .. maybe pseudo hierarchical at some point in the future
1177438547 M * yang Bertl_vV: If you are going to NY, don't forget that you can use www.nycwireless.net :)
1177438910 M * Bertl_vV oliwel: I assume you had another one, yes?
1177439113 M * onox idea of planetlab looks pretty cool
1177439229 M * oliwel Bertl_vV: jep actually
1177439239 M * oliwel Bertl_vV: baaaad kernel oopses with DRBD
1177439252 M * oliwel I played around with daneil yesterday but didnt find a solution
1177439273 M * oliwel but I wont distrurb your vacation ;)
1177439344 M * Bertl_vV oliwel: url?
1177439360 M * oliwel http://paste.linux-vserver.org/1546
1177439387 M * oliwel happens when I try to send a drbd device into secondary state on shutdown of the vserver
1177439399 M * oliwel is definitly related to some context / namespace stuff
1177439412 M * oliwel if I run the same command after the stop script finishes it wrks like a charm
1177439438 M * Bertl_vV the kernel version is actually vs2.2.0 yes?
1177439449 M * oliwel yes
1177439452 M * Bertl_vV and drbd was patched with version ...?
1177439457 M * oliwel gentoo - didnt try vanilla
1177439463 M * oliwel 0.7.23
1177439469 M * oliwel its the only one that works
1177439480 M * oliwel earlier one dont patch on prior 2.6.19
1177439494 M * Bertl_vV did you get any rejects or something like that?
1177439497 M * oliwel no
1177439511 M * Bertl_vV okay, you have the kernel tree at hand?
1177439512 Q * lilalinux_ Remote host closed the connection
1177439514 M * oliwel the guys from drbd guessed that some ifdefs or so are not fitting in correctly
1177439517 M * oliwel sure
1177439556 M * Bertl_vV okay, let's try to addr2line the EIP  
1177439582 M * oliwel what ever you tell me ,=)
1177439612 M * oliwel but I am not a kernel hacker - so please helo me a bit ,=)
1177439621 M * Bertl_vV addr2line -e vmlinux c01184b1
1177439635 M * Bertl_vV (inside the kernel source tree)
1177439637 M * oliwel ok
1177439682 M * oliwel hm - guess the tree is not matching the running kernel... mom
1177439800 M * oliwel I must first reprocude the error with teh current kernel - the snap is a bit older and I recompiled meanwhile
1177439854 M * Bertl_vV okay
1177439869 M * Bertl_vV it might also help to enable a few things in the kernel
1177439882 M * Bertl_vV i.e. debug info/ framepointers/unwind stuff
1177439888 M * oliwel just tell me your wishes
1177439935 M * oliwel where to find these ?
1177440039 M * oliwel hm ,kernel hacking framepointer is here
1177440045 M * oliwel but where to find the rest
1177440256 P * TheSasch 
1177440321 M * Bertl_vV CONFIG_DEBUG_INFO=y
1177440321 M * Bertl_vV CONFIG_UNWIND_INFO=y
1177440337 M * Bertl_vV you can use / to search in 'make menuconfig'
1177440342 M * oliwel great....you have magic hands - the error is gone.....grml
1177440350 M * Bertl_vV excellent :)
1177440401 M * oliwel Basically - there is still another issue
1177440413 M * oliwel the oops only occurs after reverting changes in the utils....
1177440439 M * oliwel with the newer utils I cant send the device to secondary because I get a "still mounted"
1177440454 M * oliwel Tried to work artound with daniel and then had these oopses.....
1177440482 M * Bertl_vV okay, so let's try to recreate it with whatever 'broken' tools you can find
1177440487 M * oliwel the lines in charge are where the stop script changes the namespave (/usr/sbin/vserver, line 204)
1177440509 M * oliwel I can break tthe tools by commenting out this line ,)
1177440579 M * Bertl_vV go ahead then ... :)
1177440593 M * oliwel tring hard to break it....
1177440595 M * oliwel wont
1177440686 M * oliwel heheh I got him 
1177440731 M * oliwel hmm, addr2line shows "??:0"
1177440748 M * oliwel Might be that the recompile that is running affects this...
1177440758 M * Bertl_vV which means that you do not have the info stuff enabled :)
1177440769 M * oliwel no - this kernel is still compiling
1177440780 M * Bertl_vV okay, so let's wait for that one ...
1177440785 M * oliwel I cant find unwind in the config
1177440794 M * Bertl_vV you might also want to prepare a 2.6.19 kernel, just to make sure
1177440826 M * Bertl_vV oliwel: 'make menuconfig' - press '/' - enter UNWIND
1177440846 M * oliwel did so
1177440852 M * oliwel also grepped through config
1177440857 M * oliwel no matches found
1177440863 M * Bertl_vV shows me two lines here...
1177440868 M * oliwel hm
1177440876 M * oliwel 2.6.20  gentoo....
1177440882 M * oliwel v2.6.20-vs2.2.0-gentoo
1177440951 M * Bertl_vV ah, got changed in 2.6.20
1177440957 M * oliwel ;)
1177440993 M * oliwel so - I will boot the one with debug_info and framepointer
1177441001 M * Bertl_vV yep, should work
1177441198 M * oliwel ok - system crashed successful ;)
1177441221 M * oliwel kernel/signal.c:222
1177441247 M * Bertl_vV okay, what does that line contain in your source tree?
1177441306 M * oliwel fastcall void recalc_sigpending_tsk(struct task_struct *t)
1177441308 M * oliwel {
1177441309 M * oliwel         if (t->signal->group_stop_count > 0 ||
1177441315 M * oliwel guess this is drbd stuff
1177441341 M * Bertl_vV nope
1177441351 M * oliwel vserver ?
1177441355 M * Bertl_vV nope
1177441359 M * oliwel ;)
1177441364 M * oliwel upstream..... :)
1177441371 M * oliwel any ideas
1177441384 M * Bertl_vV yep, but it's not immediately clear where it crashes there
1177441398 M * Bertl_vV could you upload the actual oops please?
1177441406 M * oliwel mom
1177441442 M * oliwel http://paste.linux-vserver.org/1560
1177441607 P * cruser 
1177441615 M * hardwire mom?
1177441635 M * oliwel Bertl_vV: must change places - back in a minute
1177441650 P * oliwel 
1177441666 M * Bertl_vV hardwire: like in sec :)
1177441689 M * hardwire moment!
1177441694 M * hardwire not mommy!
1177441782 M * Bertl_vV that is what I assumed, yes :)
1177441811 M * hardwire Uno Minuto!
1177442169 J * oliwel ~mail-at-o@ppp-82-135-82-123.dynamic.mnet-online.de
1177442184 M * oliwel Bertl_vV: - so back again
1177442215 Q * Piet Quit: Piet
1177442223 M * Bertl_vV oliwel: looks like a drbd issue, we should try with a kernel without Linux-VServer patched in
1177442232 M * Bertl_vV oliwel: and with namespaces
1177442237 M * oliwel ok
1177442246 M * oliwel what should  I do
1177442253 M * oliwel 2.6.20 vanilla ?
1177442271 M * Bertl_vV well, starts with building such a kernel .. i.e. 2.4.20.4 or so with drbd patches
1177442284 M * oliwel 2.4 ???
1177442289 M * oliwel 2.6...
1177442297 M * Bertl_vV yep, then you probably have to recreate the situation with vnamespace
1177442310 M * Bertl_vV maybe daniel_hozac can be of some assistance there
1177442324 M * oliwel is namespace in the vanilla ?
1177442332 M * Bertl_vV it looks like the ext3 is simply disposed because mounted inside a namespace
1177442336 M * Bertl_vV oliwel: yes
1177442345 M * oliwel ok
1177442363 M * oliwel hmm I can slightly remember we had similar issues last year
1177442522 M * oliwel sys-kernel/vanilla-sources-2.6.20.7 ?
1177442525 M * oliwel is this fine ?
1177442567 M * Bertl_vV I'd take the same version you had with Linux-VServer
1177442583 M * bXi anyone fluent in php to give me a simple yes or no?
1177442596 M * Bertl_vV return 1;
1177442603 M * bXi alomst :P
1177442618 M * bXi i've been coding for too long today so the question is quite newbish
1177442620 M * bXi elseif($data->rank+3 <= $verdediger->rank && !in_array($verdediger->login, $nokill))
1177442621 M * bXi will this trigger if $data->rank = 3 and $verdediger->rank = 15 and the $verdediger->login is in the $nokill array
1177442623 M * oliwel bXi: try...
1177442633 M * bXi oliwel: i cant
1177442640 M * bXi need to have this correctly on a live server
1177442651 M * bXi not a chance to test this easily
1177442692 M * Bertl_vV bXi: no, won't trigger
1177442709 M * bXi so any other name will trigger
1177442716 M * Bertl_vV maybe
1177442756 M * bXi thanks
1177442769 M * oliwel Bertl_vV: an idea where I find the orginal kernel version in the tree ?
1177442774 M * bXi i hate it when you've been coding for 14 hours straight that you cant even debug something simple like this
1177442784 Q * FireEgl Quit: ...
1177442825 M * oliwel seems to be a 7
1177443103 M * Bertl_vV oliwel: well, what does uname -a say?
1177443148 M * oliwel gentoo version
1177443179 M * oliwel I guess it is a #6 or #7 - 
1177443243 M * Bertl_vV well, you should figure that or 'just' revert the Linux-VServer patch then
1177443261 M * oliwel hm 
1177443290 M * oliwel then you get a gentoo-kernel
1177443374 J * boci^ ~boci@pool-6582.adsl.interware.hu
1177443433 M * Bertl_vV oliwel: I'm fine with that :)
1177443449 M * oliwel too late - vanilla is almost compiled ;)
1177443456 M * Bertl_vV oliwel: we want to recreate the issue without Linux-VServer to figure who we have to blame :)
1177443461 M * oliwel okok....
1177443475 M * Bertl_vV but chances are good that vanilla has the issue too
1177443475 M * oliwel than I need the patches from hollow
1177443482 M * oliwel lets try 
1177443486 M * oliwel I am almost done
1177443499 M * Bertl_vV try first with that, but don't spend ages on it
1177443505 M * oliwel no
1177443513 M * Bertl_vV you still need to figure how to trigger it with namespaces only
1177443519 M * oliwel hm
1177443523 M * oliwel any ideas already ?
1177443539 M * Bertl_vV I'd start with mounting it inside a namespace
1177443549 M * Bertl_vV something like:
1177443552 M * oliwel sounds like an idea
1177443565 M * oliwel ok I will try and ping you if I get any results
1177443573 M * Bertl_vV vnamespace -n -- mount /dev/something /mnt/somewhere
1177443603 M * oliwel ok
1177443923 M * oliwel looks like the show is over....the machine does not boot
1177443998 M * Bertl_vV well,go back to a previous kernel
1177444018 M * oliwel Had luck - is back
1177444147 M * oliwel how can I see the namespaces ?
1177444159 M * oliwel the mount was successful...but I cant see it
1177444196 M * daniel_hozac no oops?
1177444201 M * oliwel hi daniel_hozac 
1177444219 Q * boci^ Ping timeout: 480 seconds
1177444225 M * oliwel I get oopses on unmonuting . which I cant do actually 
1177444239 M * Bertl_vV ah, hmm, where do you do the unmount?
1177444241 M * oliwel I create a new namespace that I cant find 
1177444243 M * Bertl_vV on the host?
1177444258 M * oliwel does not work on the hoist
1177444260 M * Bertl_vV no, the namespace disappears immediately after the mount
1177444267 M * oliwel so how can I see them 
1177444275 M * Bertl_vV causing an implicit unmount
1177444289 M * oliwel so the device should not be mounted
1177444297 M * Bertl_vV what is the procedure you have to trigger it with Linux-VServer?
1177444308 M * oliwel sorry... ?
1177444315 M * oliwel ok the mount is gone
1177444324 M * Bertl_vV what are the steps you normally do to trigger the oops
1177444326 M * oliwel I can send it to secondary again
1177444343 M * oliwel I have a prestart script that turns the device to primary state
1177444350 M * oliwel I mount the device in fstab
1177444361 J * derjohn2 ~aj@213.55.131.22
1177444376 M * oliwel than I unmount in fstab and run a postpost-script that turns the device to secondary state
1177444395 M * oliwel the "old" utils DONT change namespaces and cause the oopses
1177444422 M * oliwel the new ones fail to send the device to secondary state
1177444431 M * oliwel because the device is not mounted
1177444436 M * Bertl_vV how do you unmount in the fstab?
1177444450 M * oliwel I just have a line that mounts the device
1177444466 M * oliwel  - /dev/drbd/test /data dev 0 0
1177444477 M * Bertl_vV so lets get the mount/umount steps right in your setup
1177444485 M * oliwel ok
1177444493 M * Bertl_vV 1) primary state
1177444498 M * oliwel did so
1177444499 M * Bertl_vV 2) mount in fstab
1177444503 M * oliwel ?
1177444507 M * oliwel in the host ?
1177444509 M * oliwel or what ?
1177444510 M * Bertl_vV yep
1177444534 M * Bertl_vV 3) ?? secondary?
1177444540 M * oliwel mom
1177444556 M * Bertl_vV you did say something about unmounting I didn't understand
1177444565 M * Bertl_vV 21:52 < oliwel> than I unmount in fstab and run a postpost-script that turns 
1177444579 M * oliwel Actually I dont unmount
1177444581 M * oliwel utils do....
1177444598 M * Bertl_vV hmm, no they shouldn't unmount either
1177444602 M * oliwel ??
1177444604 M * oliwel NOT ?
1177444628 M * Bertl_vV the context will disappear, and that might cause (if using namespaces) the guest mounts to become unmounted
1177444633 M * oliwel ok
1177444639 M * oliwel that clearifies a lot
1177444648 M * Bertl_vV we can emulate that with the vnamespace thingy
1177444754 M * oliwel so it seems like the drbd stuff is not happy riunning within a changed context
1177444767 M * Bertl_vV not confirmed yet
1177444785 M * Bertl_vV let's try the following scenario
1177444794 M * oliwel affirmative ;)
1177444796 M * Bertl_vV 1) primary
1177444812 M * Bertl_vV 2) on second login, vnamespace -n -- bash
1177444823 M * Bertl_vV 3) mount device somewhere
1177444833 M * Bertl_vV 4) on first login, turn secondary
1177444845 M * Bertl_vV 5) leave second login (exit)
1177444858 M * Bertl_vV make sure to do the mount with -n 
1177444866 M * Bertl_vV so that it doesn't add an entry to the mtab
1177444873 M * oliwel 4) ->  ioctl(,SET_STATE,) failed: Device or resource busy
1177444908 M * Bertl_vV that actually sounds sane ...
1177444909 M * oliwel to late....already in mtab :(
1177444911 M * oliwel ok
1177444925 M * oliwel I cant unmount it even in the second bash
1177444925 M * Bertl_vV but try again with the -n just to make sure
1177444944 M * Bertl_vV you just need to leave the second logon
1177444950 M * oliwel ok it disappeared from mtab
1177444951 M * Bertl_vV it will get auto-unmounted
1177444951 M * oliwel ok
1177444953 M * oliwel mom
1177444970 M * oliwel works and no oopes
1177444979 M * oliwel but wait
1177444990 M * oliwel I guess I must do some changes to the drive so it gets synces
1177445047 J * Aiken ~james@121.45.222.137
1177445053 J * ema ~ema@rtfm.galliera.it
1177445057 M * oliwel still no oopses
1177445059 M * oliwel looks sane
1177445151 M * Bertl_vV it would be a good idea to run the guest startup with --debug, and verify that we have all the relevant steps
1177445168 M * Bertl_vV then try to recreate it on the Linux-VServer patched kernel from the host
1177445185 M * Bertl_vV (preferably without creating a context)
1177445188 M * oliwel you mean I should snapshot the vserver startup ?
1177445197 M * oliwel ok - will boot the old kernel
1177445415 M * oliwel so machine is up
1177445499 M * oliwel hm - runs up and down withput oops
1177445562 M * Bertl_vV so maybe on the second run?
1177445568 M * oliwel yes...
1177445569 M * Bertl_vV did you capture the --debug stuff?
1177445598 M * oliwel will upload it
1177445600 M * oliwel mom
1177445727 Q * nox Ping timeout: 480 seconds
1177445895 J * nox ~nox@static.88-198-17-175.clients.your-server.de
1177445909 J * boci^ ~boci@pool-6582.adsl.interware.hu
1177445979 M * Bertl_vV please could everyone who cannot reach list.linux-vserver.org via web, send an email to martin at list-petersen dot dk ?
1177446003 M * Bertl_vV that might inspire him to take a look at the issue :)
1177446027 M * sid3windr lol =)
1177446033 M * sid3windr everyone who... == everyone
1177446033 M * sid3windr :p
1177446048 M * sid3windr the entire intarweb will send him an email!
1177446086 Q * oliwel Read error: No route to host
1177446144 Q * derjohn2 Ping timeout: 480 seconds
1177446171 J * oliwel ~mail-at-o@ppp-82-135-82-123.dynamic.mnet-online.de
1177446217 Q * oliwel Read error: Connection reset by peer
1177446277 M * daniel_hozac Bertl_vV: i doubt he reads that email.
1177446365 M * sid3windr why?
1177446369 M * sid3windr I actually think he does
1177446381 M * daniel_hozac he just doesn't care?
1177446384 M * sid3windr you wouldn't read daniel@zakrisson.se ? :p
1177446415 M * daniel_hozac well, Bertl_vV emailed him a while back, i sent another a couple of weeks ago...
1177446424 M * daniel_hozac still nothing has happened.
1177446431 M * sid3windr myup
1177446437 M * sid3windr but it's still his email address
1177446441 M * sid3windr so he does read it
1177446444 M * sid3windr he just doesn't care
1177446448 M * sid3windr that's not what you said ;)
1177446466 M * daniel_hozac i like to think he cares.
1177446467 P * stefani I'm Parting (the water)
1177446475 M * daniel_hozac and that something is keeping him from reading it.
1177446491 M * daniel_hozac because if he's just ignoring it, well, that would be really bad.
1177446517 M * sid3windr :-)
1177446523 M * sid3windr sorry ;)
1177446633 M * Bertl_vV well, martin tends to priorize stuff based on urgency and/or demand
1177446651 M * Bertl_vV he sent me a reply stating that he will look into it soon ...
1177446664 M * Bertl_vV (that was more than a week ago:)
1177446696 M * Bertl_vV actually it was april 2nd, and he wrote:
1177446706 M * Bertl_vV I'll look into that today.
1177446708 M * sid3windr yea, you pasted it
1177446763 J * oliwel ~chatzilla@ppp-82-135-82-123.dynamic.mnet-online.de
1177446885 M * oliwel this is actually not my day..now my notebook fetched the cold
1177447070 J * derjohn2 ~aj@213.55.131.23
1177447110 M * sid3windr gesundheit
1177447121 M * oliwel thx 
1177447133 M * sid3windr hehe
1177447136 M * sid3windr this isn't my month
1177447153 M * sid3windr I just realized both disks in my workstation RAID1 are broken
1177447163 M * sid3windr either works, but not quite
1177447170 M * sid3windr 3ware controller is complaining, loudly
1177447175 M * daniel_hozac d'oh!
1177447189 M * sid3windr last week, both disks in RAID1 in one of my servers went dead
1177447197 M * sid3windr PLUS one in a RAID5 in another server
1177447217 M * sid3windr software raid 5; I replaced the disk and while rebuilding the raid the thing PANIC'd twice
1177447226 M * sid3windr I hate computers.
1177447271 M * daniel_hozac haha
1177447275 M * daniel_hozac i know that feeling...
1177447325 M * sid3windr I need to downsize
1177447332 M * sid3windr too many disks means too many replacements
1177447354 M * sid3windr and this is all my personal stuff, not my work :p
1177447455 M * daniel_hozac hehe.
1177447474 M * daniel_hozac i'm thinking of doing a 3-drive RAID-1 for my next server...
1177447477 M * onox now your porn collection is gone :D
1177447502 M * sid3windr onox: no, I have backups :p
1177447511 M * onox ha ha :D
1177447518 M * sid3windr on a 13-disk raid5+raid5+raid5
1177447522 M * sid3windr AND
1177447529 M * onox :p
1177447529 M * sid3windr I just heard from someone who has the same controller as me
1177447536 M * sid3windr that it doesn't have any error recovery
1177447545 M * sid3windr so if one of your drive fucks up, up they all go
1177447564 M * sid3windr _that_ was the (EXPERIMENTAL) in menuconfig ;)
1177447564 M * onox sid3windr: what controller do you have?
1177447566 M * sid3windr *knock on wood*
1177447568 M * sid3windr marvell
1177447579 M * sid3windr 01:08.0 SCSI storage controller: Marvell Technology Group Ltd. MV88SX6081 8-port SATA II PCI-X Controller (rev 09)
1177447582 M * sid3windr sata_mv module
1177447681 M * daniel_hozac sid3windr: you mean in the RAID-array, or you'll-have-to-reboot?
1177447726 M * sid3windr rmmod sata_nv; modprobe sata_nv
1177447752 M * sid3windr the entire controller system currently locks when one of your disk is fucking up its bus
1177447758 M * daniel_hozac ouch
1177447763 M * sid3windr I have 2 of 'em
1177447767 M * onox is the controller onboard?
1177447768 M * sid3windr and yes, my root fs is also on one of them
1177447769 M * sid3windr no :)
1177447795 M * sid3windr it's a cheap 8-port sata controller (100 eur), and it works well
1177447800 M * sid3windr apart from that little omission in the driver
1177447802 M * sid3windr heh
1177447807 M * sid3windr but I haven't been struck by it yet
1177447809 Q * bonbons Ping timeout: 480 seconds
1177447809 M * sid3windr and I hope I won't be
1177447818 M * sid3windr because that would mean just more and more dead disks :(
1177447884 M * onox I only heard of areca, adaptec and promise
1177447912 J * tudenbart ~willi@xdsl-213-196-253-169.netcologne.de
1177447972 M * daniel_hozac sid3windr: so it's a driver problem? it should be fixable then, no?
1177448009 M * sid3windr yes, it's a driver problem
1177448018 M * sid3windr it's just that *I* don't have any clue to fix it
1177448025 M * sid3windr :)
1177448028 M * onox i'm going to sleep
1177448034 M * sid3windr onox: this isn't a raid controller, it's just a dumb sata controller
1177448048 M * sid3windr areca is da bomb, I'd gladly put in a 16port pci-x card like I have many at work but hm
1177448056 M * sid3windr 800 eur isn't quite teh cheap for a home server
1177448059 M * onox :p
1177448089 M * onox my home server (if I would have one) would be a p2 :p
1177448095 M * sid3windr :-)
1177448099 M * sid3windr this is a dual xeon 2.4ghz ;)
1177448151 M * onox that's a little more powerful :p
1177448210 M * sid3windr :)
1177448217 M * onox anyway, bye everyone
1177448224 M * onox vserver #vservers left
1177448241 M * sid3windr =)
1177448241 M * sid3windr cya
1177448243 Q * onox Quit: baselayout-2 \o/
1177448354 Q * dothebart Ping timeout: 480 seconds
1177448364 M * oliwel Bertl_vV: here is debug from the startsequence
1177448366 M * oliwel http://paste.linux-vserver.org/1562
1177448478 M * Bertl_vV okay, let'stry to filter out the mount/umount operations and in what context/space they happen
1177448488 M * oliwel wait
1177448491 M * oliwel drbd was not up
1177448536 Q * nou Server closed connection
1177448540 M * oliwel http://paste.linux-vserver.org/1563
1177448540 J * nou Chaton@causse.larzac.fr.eu.org
1177448572 M * sid3windr poor Bertl_vV 
1177448578 M * sid3windr working on vacation
1177448835 M * Bertl_vV oliwel: same goes for that one :)
1177448882 M * oliwel sorry - its late....and all it is going crazy - notebook is down, sitting on a blody server console
1177448923 M * Bertl_vV hmm .. maybe we delay it then?
1177448989 M * oliwel hmm - so I pstet the right one, but you cant see anything bout the mounts
1177448991 M * oliwel http://paste.linux-vserver.org/1564
1177449242 M * Bertl_vV hmm, no, I haven't looked through it yet :)
1177449265 M * Bertl_vV what I meant was, _you_ might have a look and filter out the relevant data (i.e. mounts and such)
1177449297 M * oliwel yes - there is nothing
1177449361 M * Hollow http://paste.linux-vserver.org/1565 :)
1177449373 M * oliwel I guess this does the trick 
1177449375 M * oliwel ++ /usr/lib/util-vserver/secure-mount -a --chroot --fstab /etc/vservers/test/fstab --rootfs no
1177449390 M * Bertl_vV Hollow: hehe
1177449391 M * oliwel Hi Hollow
1177449524 Q * boci^ Quit: Távozom
1177449857 M * Bertl_vV oliwel: with that you can recreate it without a context?
1177449873 M * oliwel recreate what ?
1177449884 M * oliwel the oops ?
1177449889 M * Bertl_vV yes
1177449900 M * oliwel havent found a real solution
1177449913 M * oliwel What daniel sugested was
1177449916 M * Bertl_vV ah, it sounded like you did
1177449924 M * oliwel yesno
1177449953 M * oliwel If I see everything correctly this happens:
1177449977 M * oliwel The current util-vserver (212) changes into the guests namespace to run the vserver.stop script
1177449998 M * oliwel inside this namespace the physical drbd device is still mounted
1177450018 M * oliwel the postpost script that calls "drbd secondary" runs inside this namespace
1177450030 M * oliwel as the mount is still active, the command fails
1177450041 M * oliwel this was changed between 210 and 212 
1177450059 M * oliwel the behaviour seems clear to me after ypur explantion that utils DONT unmount
1177450081 M * oliwel Now: daniel suggested to hack the vserver stop script to NOT enter the namespace 
1177450087 M * oliwel I did so and get the oopses
1177450092 M * oliwel so for me it looks like:
1177450123 M * Bertl_vV the question is: why do you get the proper/expected behaviour with 'just' the namespaces
1177450127 M * oliwel Switching drbd into secondary with recent utils fails, because the mount is active inside the namespace where we try to shutdown the device
1177450157 M * oliwel Switching drbd into secondary in root context while the shutdown is runnig causes an oops
1177450177 M * oliwel but - why dont I get an oops if I try to shutdown the device while the guest is running
1177450182 M * oliwel so I guess somewhat interfered
1177450191 M * oliwel yes thats a good pint
1177450264 M * oliwel and I dont have a single idea how to get beyond
1177450299 M * Bertl_vV well, we should try to simplify the steps to trigger that
1177450323 M * Bertl_vV first, we should make sure that we can reliably recreate it
1177450441 M * oliwel hm ok
1177450449 M * oliwel Idea.
1177450455 M * oliwel if I create a namespace by hand
1177450485 M * daniel_hozac that's what you've been doing with vnamespace -n.
1177450488 M * oliwel mount the device inside and than try to go to secondary outside - this should be comparable to the issue we see ?
1177450505 M * oliwel hm yep
1177450507 M * oliwel right
1177450519 M * oliwel but we didnt get oopese
1177450521 M * oliwel oopses
1177450524 M * daniel_hozac 2007-04-24T14:56:56 <@gamla_kossan> ok. då är det skickat.
1177450527 M * daniel_hozac d'oh.
1177450531 M * Bertl_vV :)
1177450533 M * daniel_hozac stupid mouse with too many buttons ;)
1177450534 Q * derjohn2 Ping timeout: 480 seconds
1177450551 M * oliwel you should buy a MAC
1177450551 M * daniel_hozac anyway, you did try to set it to secondary, right?
1177450557 M * Bertl_vV what about other namespaces in that context?
1177450565 M * oliwel yes I did
1177450570 M * Bertl_vV daniel_hozac: could those affect the result?
1177450571 M * oliwel other namespaces ?
1177450582 M * daniel_hozac Bertl_vV: shouldn't the effect be the same?
1177450597 M * Bertl_vV I'm more thinking about unshare vs. old setup
1177450609 M * daniel_hozac ah.
1177450612 M * Bertl_vV ushare/clone
1177450629 Q * ema Quit: leaving
1177450630 M * daniel_hozac well, is there anything really using unshare already?
1177450644 M * oliwel Bertl_vV: daniel_hozac  - a simply other idea - if we can figure out how to unmount the device during shutdown I guess I am lucky
1177450653 M * Bertl_vV daniel_hozac: yes, vcmd of course :)
1177450666 M * oliwel anyway it would be nice to fix the oopese
1177450669 M * daniel_hozac Bertl_vV: i meant inside the guest ;)
1177450678 M * daniel_hozac or did i misunderstand what you were saying?
1177450681 M * Bertl_vV no, probably not, and that's not what I meant
1177450702 M * Bertl_vV oliwel: you said it came with newer tools + workaround, yes?
1177450712 M * oliwel Bertl_vV: yes
1177450720 M * daniel_hozac oliwel: you can't really unmount it before all of the processes have exited, in which case you've already lost...
1177450798 M * oliwel daniel_hozac: what is the last chance to catch the originail conext - might it possible to unmount the device from here
1177450814 M * oliwel I am considering to monut the device from init instead fsta
1177450816 M * oliwel b
1177450829 M * daniel_hozac oliwel: it's the root filesystem, right?
1177450836 M * oliwel argh - wont work, the device contains /var
1177450849 M * oliwel no, var and etc bind mounted
1177450910 Q * duckx Remote host closed the connection
1177451005 M * oliwel I guess I got a workaround....
1177451018 M * oliwel the oops seems to happen only if the two peers are connected
1177451111 Q * dna Quit: Verlassend
1177451180 M * oliwel hm - calling a disconnect also oopses
1177451201 M * oliwel so I guess we can exclude fs related dtuff
1177451203 M * Bertl_vV could you upload that one too?
1177451212 M * oliwel sure
1177451242 M * oliwel its the same signal
1177451271 M * oliwel http://paste.linux-vserver.org/1568
1177451322 M * oliwel hmm
1177451323 M * oliwel static inline int test_tsk_thread_flag(struct task_struct *tsk, int flag)
1177451325 M * oliwel {
1177451326 M * oliwel         return test_ti_thread_flag(task_thread_info(tsk), flag);
1177451328 M * oliwel }
1177451347 M * oliwel from  include/linux/sched.h
1177451618 M * Bertl_vV okay, leaving now .. will be back in half an hour ...
1177451637 M * oliwel Bertl_vV: guess I will be nin bed - close to midnight here
1177451647 M * oliwel any ideas before you leave 
1177451787 M * sid3windr yes, send all your hardware to sid3windr 
1177451792 M * sid3windr he will make good use of it
1177451797 M * sid3windr especially harddisks, he loves those
1177451822 M * oliwel if you care bout my customers and I can charge the money - no problem
1177451836 M * sid3windr hm
1177451840 M * sid3windr that's taking it too far ;)
1177452029 P * marcfiu 
1177452331 Q * mattzerah_ Quit: mattzerah_
1177452564 J * _mcp ~hightower@wolk-project.de
1177452605 Q * mcp Read error: Connection reset by peer
1177453306 J * derjohn2 ~aj@213.55.131.22
1177453376 Q * DoberMann_ Remote host closed the connection
1177453390 J * DoberMann[PullA] ~james@AToulouse-156-1-178-109.w90-38.abo.wanadoo.fr
1177455403 M * Bertl_vV back now ...
1177455735 M * oliwel Hi Bertl_vV 
1177455741 M * oliwel still playing games here....
1177455756 M * oliwel looks like DRBD is not to blame
1177455775 M * oliwel 2.6.15 (my old kernel) works fine even with 0.7.23 drbd patchset
1177455801 M * oliwel just this second treid 2.6.20.7 vanilla with current vserver patch and 0.7.23 drbd -> crashes
1177455803 M * Bertl_vV well, that could have changed with 2.6.19+
1177455816 M * oliwel should i try 19 or 18 ?
1177455829 M * Bertl_vV I would definitely appreciate a check with 2.6.19
1177455846 M * Bertl_vV and if that fails too, with 2.6.18 
1177455862 M * oliwel ok - wil take 20 mins to fetch and compile...
1177455871 M * Bertl_vV okay, np here ...
1177455881 M * oliwel vanilla or gentoo flavoured ?
1177455893 M * Bertl_vV preferablevanilla
1177455939 M * oliwel 2.6.19 is on the way
1177456109 M * Bertl_vV quite a lot changed between 2.6.18 and 2.6.20, so I wouldn't be surprised if something got broken for out of mainline stuff like drbd
1177456506 Q * derjohn2 Ping timeout: 480 seconds
1177456986 J * ballen ~ballen@nssa-cluster.rit.edu
1177457139 M * ballen In ubuntu 6.10, has anyone had the issue inside of there vserver with updating udev, and getting this error: Setting up udev (093-0ubuntu18.0edgy2) ...
1177457140 M * ballen mknod: `/lib/udev/devices/ppp': Operation not permitted
1177457140 M * ballen dpkg: error processing udev (--configure):
1177457249 M * Bertl_vV no, but it is not unecpected
1177457259 M * Bertl_vV *unexpected
1177457272 M * Bertl_vV i.e. you do not install udev in a vserver guest
1177457285 M * Bertl_vV (mainly because there is no point in doing so)
1177457287 M * ballen yea I figured as much, but its stopping apt-get from working and udev is a dev of a number of things
1177457294 M * ballen like ubuntu-minimal
1177457307 M * ballen dep*
1177457345 M * Bertl_vV I would replace the scripts with a link to /bin/true or so
1177457591 M * ballen hmm the vserver seems to run without udev, and ubuntu-minimal packages
1177457597 M * ballen problem solved
1177457611 Q * rob-84x^ Ping timeout: 480 seconds
1177457624 M * trippeh Isn't ubuntu-minimal just a meta-package anyway?
1177457637 M * ballen no idea, sounded somewhat important
1177457640 M * trippeh Eg doesn't do anything but depending on a standard package set.
1177457649 M * Bertl_vV yes, I think so
1177457688 M * ballen yea it is
1177457710 M * ballen in a std install it provides boot files, etc
1177457721 M * oliwel Bertl_vV: 2.6.19 fails also - same line as before in recalc_sigpending_tsk
1177457827 M * Bertl_vV hmm, okay, 2.6.18?
1177457843 M * oliwel last chance - its late ;)
1177457884 M * Bertl_vV okay :)
1177457898 M * oliwel But still the more interessring question should be -. how to shutdown the device in a reular manner...
1177458053 M * oliwel Bertl_vV: three is no stable against 2.6. 18  ?
1177458075 M * Bertl_vV no but there should be a release candidate
1177458121 M * oliwel patch-2.6.18.5-vs2.2.0-pre5.diff
1177458146 M * Bertl_vV yes, let's take this one
1177458206 M * tanjix good evening
1177458224 M * Bertl_vV and a good 1 2u2!
1177458225 M * tanjix hmm good morning sounds better :)
1177458240 M * Bertl_vV well, it'sactually evening here :)
1177458251 M * tanjix :)
1177458258 M * tanjix Bertl, you have a minute?
1177458266 M * Bertl_vV sure
1177458286 M * tanjix i am reading the FAQ reagrding the thing to limit the ram of a vserver
1177458308 M * tanjix it says "Check the size of memory pages. On x86 and x86_64 is usually 4 KB per page."
1177458317 M * tanjix how di i check this?
1177458347 M * Bertl_vV IIRC, there is a tool/source for that on the same page
1177458444 M * tanjix hmm i dont see this... :(
1177458557 N * DoberMann[PullA] DoberMann[ZZZzzz]
1177458579 M * tanjix but lets say that the 4KB value is correct
1177458584 Q * bzed Quit: Leaving
1177458605 M * tanjix my host has 2 GB RAM and I want ot limit each guest to 128 MB
1177458647 M * tanjix what must i calculate then? don't really understand the one in the FAQ :(
1177458660 M * Bertl_vV http://linux-vserver.org/Resource_Limits
1177458674 M * Bertl_vV the code is there (just verified that)
1177458696 M * Bertl_vV oliwel: I have to get something to eat, please just leave me a message with the results ... TIA
1177458713 M * Bertl_vV okay, back in a few
1177458721 M * oliwel Bertl_vV: enjoy - I will post