1219709544 J * mrfree ~mrfree@host132-183-dynamic.13-79-r.retail.telecomitalia.it
1219709693 Q * mrfree 
1219710690 J * doener ~doener@i577BAFAF.versanet.de
1219710793 Q * doener_ Ping timeout: 480 seconds
1219710983 Q * ViRUS_ Quit: Leaving
1219714478 J * SlackLnX ~Lee@bl7-139-214.dsl.telepac.pt
1219717346 Q * SlackLnX Quit: I'll Be back
1219717729 Q * kaner Remote host closed the connection
1219718221 J * kaner ~kaner@zzz.strace.org
1219718457 J * derjohn_foo ~aj@p5B23D4C4.dip.t-dialin.net
1219718887 Q * derjohn_mob Ping timeout: 480 seconds
1219722574 J * balbir ~balbir@122.171.6.234
1219723382 Q * hparker Quit: Read error: 104 (Peer reset by connection)
1219723861 M * daniel_hozac Bertl: http://paste.linux-vserver.org/12386
1219723940 M * daniel_hozac happened at vserver ... stop after a find /vservers/lenny -type f -print0 | xargs -0 setattr --iunlink; vserver lenny2 build -m clone -- --source lenny
1219723956 M * daniel_hozac (lenny was stopped at the time)
1219723989 M * daniel_hozac trying to reproduce now.
1219724191 M * daniel_hozac yeah, it's reproducible.
1219724289 M * Bertl interesting ...
1219724359 M * Bertl can you do the addr2line for me please?
1219724388 M * daniel_hozac which one?
1219724425 M * Bertl preferably all of them, but the nsproxy, mntput* are most interesting
1219724580 M * daniel_hozac http://paste.linux-vserver.org/12387
1219724583 M * daniel_hozac all of them, in order.
1219724840 M * Bertl seems I somehow messed up with the link breaking writers count
1219724958 M * daniel_hozac i don't even see that we mess with it directly at all...
1219725257 M * Bertl well, that's the problem, we should
1219725311 M * daniel_hozac ah, so you think the count is in the negatives?
1219725491 M * Bertl could you test that?
1219725823 M * daniel_hozac it's 1.
1219725830 M * Bertl hmm ...
1219725848 M * Bertl and I assume we had a bunch of link breaks, yes?
1219725878 M * daniel_hozac 17.
1219725929 M * daniel_hozac i'll make it output the sb too, so we can make sure it's at all COW-related...
1219725958 M * Bertl okay, good idea
1219726464 M * daniel_hozac hmm.... / is the mount point.
1219726522 M * Bertl hmm, that would look like an rbind issue then?
1219726557 M * daniel_hozac i guess so.
1219726589 M * Bertl but you need to do link breaking to trigger it?
1219726728 M * Bertl do you use/try with a bind mount - remount atm?
1219726773 M * Bertl it could be, that we are holding a reference in the context to the rootfs?
1219726804 M * daniel_hozac well, we can't get rid of the rootfs.
1219726817 M * daniel_hozac this guest's fstab is clean, no bind mounts.
1219726858 M * Bertl what does the namespace counter say, are we leaking anything?
1219726880 M * daniel_hozac hmm. it does seem like link breaking is required. if i reboot and don't recreate the guest, it works fine.
1219726893 M * daniel_hozac no, it seems correct.
1219726968 M * Bertl so, we have one more writers than we should have, when we dispose of the guest mounts, and this situation is triggered by the cow link breaking ...
1219727115 M * Bertl I guess I have it ... sec
1219727143 M * Bertl yep
1219727164 M * Bertl fs/namei.c goto line 1962
1219727187 M * Bertl in the immediately following lines, we do
1219727191 M * Bertl error = mnt_want_write(nd.path.mnt);
1219727197 M * daniel_hozac ah.
1219727205 M * Bertl and in line 1974
1219727213 M * Bertl we leave out with exit
1219727257 M * Bertl where we should jump to line 1985 instead
1219727295 M * Bertl so I propose to make that goto exit_cow
1219727314 M * Bertl and put a label inside, but at the end of the #ifdef block
1219727345 M * daniel_hozac we should drop the write in the goto restart case too.
1219727388 M * Bertl nah, that should be fine as long as we do not change the dentry
1219727415 M * Bertl and probably it's perfectly fine after that too, I doubt it uses anything of the inode/Dentry
1219727426 M * Bertl and we are still on the same filesystem (have to be)
1219727452 M * Bertl ah, wait, restart, yes
1219727575 M * daniel_hozac testing the fix now.
1219727669 M * daniel_hozac yeah, works fine now. thanks.
1219727679 M * Bertl excellent! thanks for spotting :)
1219727694 M * Bertl will upload and update that shortly
1219727999 M * Bertl http://vserver.13thfloor.at/Experimental/delta-cowbl-fix02.diff
1219728006 M * Bertl (does that match your changes?)
1219728058 J * cryptronic ~oli@p54A3B1B6.dip0.t-ipconnect.de
1219728159 M * daniel_hozac should mnt_drop_write be before path_put?
1219728181 M * Bertl probably, yes
1219728191 M * Bertl let's move it there
1219728192 Q * derjohn_foo Ping timeout: 480 seconds
1219728208 M * daniel_hozac then it's a match :-)
1219728292 Q * balbir Ping timeout: 480 seconds
1219728293 M * Bertl as a matter of fact, we should even do it before the release_open_intent(&nd);
1219729308 Q * cryptronic Quit: Leaving.
1219729794 M * daniel_hozac Bertl: http://people.linux-vserver.org/~dhozac/p/k/delta-vspid-revert02.diff http://people.linux-vserver.org/~dhozac/p/k/delta-pidspace-feat05.diff
1219729808 M * daniel_hozac the latter being the most interesting.
1219729828 M * daniel_hozac it's not perfect yet, but init doesn't complain.
1219729839 M * daniel_hozac ps doesn't show it though...
1219729861 M * Bertl incorrect order/sequence?
1219729886 M * Bertl but nice, if init doesn't complain, we are at least half there
1219729895 M * daniel_hozac hmm?
1219729915 M * Bertl processes are listed in ascending sequence
1219729933 M * Bertl it might be that this process is at the wrong location in the list
1219729974 M * Bertl is it shown in /proc ?
1219729981 M * daniel_hozac yeah.
1219729982 M * Bertl (but unreachable)?
1219729985 M * daniel_hozac i can read the files too.
1219729988 M * daniel_hozac top shows it fine.
1219730004 M * Bertl aha, nice, strace -fF of ps?
1219730048 M * Bertl hmm, what are the many vs_base removals in this patch?
1219730062 M * daniel_hozac accidental cleanup.
1219730075 M * Bertl okay, can we break that out into a separate one?
1219730077 M * daniel_hozac (i was reading through the patch to find the vx_checks)
1219730106 M * Bertl I mean all the non pid related ones :)
1219730385 M * Bertl hmm, I think we might need a replacement for the exit_vx_info_early() no?
1219730404 M * Bertl or how is that handled now?
1219730649 J * sharkjaw ~gab@149-66-195.231210.adsl.tele2.no
1219730657 M * daniel_hozac yeah.
1219730683 M * daniel_hozac it's partially handled by exit_child_reaper.
1219730773 M * Bertl which I do not see in those patches?
1219730842 M * daniel_hozac no, that's in http://people.linux-vserver.org/~dhozac/p/k/delta-pidns-fix01.diff
1219730882 M * Bertl at least partially, I see :)
1219731003 M * Bertl so we now _assume_ that child reaper and init are identical, yes?
1219731020 M * daniel_hozac do we?
1219731036 M * Bertl vx_initpid()
1219731042 M * Bertl   return task_tgid_vnr(vxi->vx_nsproxy->pid_ns->child_reaper);
1219731062 M * daniel_hozac ah, yes.
1219731132 M * daniel_hozac that should probably be a pid_vnr(find_pid_ns(1, vxi->vx_nsproxy->pid_ns)) instead.
1219731288 J * ntrs ~ntrs@77.29.69.237
1219731807 Q * larsivi Quit: Konversation terminated!
1219732014 M * daniel_hozac Bertl: what about the rest of the patch? in particular, the __alloc_vx_info and vx_migrate_task portions...
1219732281 M * Bertl definitely needs a little more error handling
1219732314 M * Bertl what is the purpose if the set_reaper checks?
1219732332 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr
1219732377 M * daniel_hozac checking if the task is in the guest's pid namespace?
1219732408 M * daniel_hozac i guess the checks can be removed now, when the kernel's doing all the setup.
1219732475 M * Bertl checks are fine, we can make them WARN_ON()s
1219732498 M * Bertl what about setting the reaper now?
1219732515 M * daniel_hozac what about it?
1219732565 M * Bertl I mean, via the vc_* interface
1219732602 M * daniel_hozac should still be supported?
1219732622 M * daniel_hozac i don't think i've changed any of that.
1219732642 M * Bertl will that work? for that we would need the checks, I think
1219732672 M * daniel_hozac i don't see why it wouldn't.
1219732742 J * ntrs_ ~ntrs@77.29.76.53
1219732751 M * daniel_hozac interesting... ps aux works, ps faux does not.
1219733171 Q * ntrs Ping timeout: 480 seconds
1219734362 Q * doener Ping timeout: 480 seconds
1219734517 Q * pmenier Quit: Konversation terminated!
1219734558 J * doener ~doener@i577B88FB.versanet.de
1219734960 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr
1219735132 Q * pisco_ Ping timeout: 480 seconds
1219735216 J * balbir ~balbir@59.145.136.1
1219735241 J * DavidS ~DavidS@p5B2A2DA1.dip0.t-ipconnect.de
1219735247 M * DavidS hi all
1219735259 M * DavidS Bertl: have you seen http://paste.linux-vserver.org/12383 yesterday?
1219735376 M * Bertl nope
1219735418 M * Bertl what kernel is that?
1219735454 M * DavidS that's on a brand new sid 2.6.26-1-vserver-amd64 which includes the 2.3.0.35 patch
1219735481 M * Bertl do you have a build tree for that kernel?
1219735490 J * larsivi ~larsivi@85.221.53.194
1219735491 M * Bertl i.e. can you convert the addresses to lines?
1219735506 M * DavidS i can get the debian kernel tree, so i should be able to
1219735521 M * Bertl that would definitely help
1219735526 M * DavidS or you can get it yourself, if you have a amd64 box at hand
1219735533 M * DavidS with debian
1219735563 M * Bertl as you can imagine, I don't :)
1219735943 J * pisco ~pisco@tor.noreply.org
1219736677 J * DavidS1 ~DavidS@vpn.uni-ak.ac.at
1219737048 Q * DavidS Ping timeout: 480 seconds
1219737144 N * DavidS1 DavidS
1219737847 M * DavidS darn, ksymoops tells me: "Error (regular_file): read_ksyms stat /proc/ksyms failed"
1219737867 M * Bertl try addr2line -e vmlinux
1219737960 M * DavidS probably i'm doing something wrong here: addr2line: /boot/vmlinuz-2.6.26-1-vserver-amd64: File format not recognized
1219737981 M * daniel_hozac not vmlinuz, vmlinux.
1219737982 M * Bertl yes, you are using the compressed and installed kernel
1219737996 M * Bertl while you want to use the re-built vmlinux in the kernel tree
1219738006 M * DavidS debian doesn't seem to ship much debug infos for their kernel :-/
1219738021 M * DavidS or can i just build the same kernel/config again and get usable results?
1219738025 M * Bertl that's all stable and tested, why use debug info :)
1219738041 M * Bertl yes, that is probably the best way to get there
1219738088 M * DavidS ok, that'll take a while
1219738111 M * Bertl looks like an x86_64 to me, should be quite fast there
1219738146 M * daniel_hozac well, it's a kitchen-sink build... ;)
1219738162 M * Bertl all the good things :)
1219738202 M * Bertl the 000000ff000000ff looks suspiciously like some kind of marker
1219738208 M * Bertl (or poison)
1219738291 M * DavidS Bertl: since the host-2.6.26 is unstable i'm running a 2.6.22-vserver within kvm :->  which doesn't help performance either
1219738316 M * daniel_hozac no virtio?
1219738329 M * Bertl http://lkml.org/lkml/2008/4/20/182
1219738418 M * Bertl DavidS: host is unstable, even without guests active?
1219738454 M * DavidS the oopses happen on the ext3 where i have my vservers
1219738465 M * DavidS even if i access them from ctx 0
1219738480 M * Bertl intersting, the filesystem was checked?
1219738486 M * DavidS several times
1219738503 M * Bertl you can trigger them by what exactly?
1219738523 M * DavidS rsyncing a vserver to another fs always fails at the exact same file
1219738529 M * DavidS and trigges the oops
1219738554 M * Bertl okay, can you examine this file with e.g. 'file'?
1219738554 M * daniel_hozac cool.
1219738570 M * Bertl does that give you that specific oops too?
1219738594 M * Bertl DavidS: btw, make sure to keep that filesystem around!
1219738608 M * DavidS Bertl: i still have it
1219738642 M * Bertl would be great to have a (shrinked down) version of that for further testing
1219738650 M * Bertl but I guess that won't be easy
1219738713 M * Bertl sys32_lstat64 <- maybe that's some clue?
1219738793 M * DavidS that's all the recent BUGs: http://paste.linux-vserver.org/12388
1219738807 M * DavidS I'm running 32bit userspace
1219738839 M * Bertl can you get us a detailed upload of a few of them
1219738850 M * Bertl (looks like they are quite similar)
1219738863 M * Bertl and they are all triggered by the very same file?
1219738917 M * Bertl is that even a file?
1219738917 M * DavidS I've got ~20 vservers on this FS and like 5 or so of them i couldn't completely rsync
1219738936 M * DavidS one of the things i looked into was part of a git repo
1219738943 M * Bertl what do the 'files' where the rsync stops have in common?
1219738945 M * DavidS and it rsyncs fine on the older kernel
1219738959 M * DavidS i use hashify
1219738968 M * Bertl ah, wait a moment, you are switching between older and newer kernels?
1219738992 M * DavidS i just checked the logs, it's always the same do_lookup+0x2a/0x34d; permission,__link_path_walk stack
1219739006 M * DavidS the FS was done with a older kernel as well as the hashify
1219739019 M * Bertl the filesystem flags changed recently, we had to adjust for newer flags .. that might cause the stranges issues if you do not remap them properly
1219739046 M * Bertl what are the two kernel versions in question?
1219739047 M * DavidS how can i remap them?
1219739054 M * daniel_hozac shouldn't cause BUGs though...
1219739058 M * Bertl one is 2.6.26-vs2.3.0.35, yes?
1219739072 M * DavidS yes, i'll check the other
1219739080 M * Bertl and the filesystem is ext3
1219739106 M * DavidS 2.6.22+vs2.2.0.5
1219739110 M * DavidS ext3, yes
1219739210 Q * ntrs_ Ping timeout: 480 seconds
1219739234 M * Bertl in this case, please also do an lsattr and showattr on one of those 'evil' files (besides the 'file')
1219739278 M * DavidS lsattr showed ----i---- (no idea how many dashes)
1219739311 M * DavidS i just started the kernel build
1219739494 M * daniel_hozac what about showattr?
1219739547 M * DavidS i don't have that handy .. i'm testing now 
1219739559 M * DavidS let's hope the kernel doesn't go bye bye again
1219740295 J * kir ~kir@swsoft-msk-nat.sw.ru
1219740492 Q * DavidS Ping timeout: 480 seconds
1219740584 J * DavidS1 ~DavidS@p5B2A2DA1.dip0.t-ipconnect.de
1219740641 P * kir Leaving.
1219740795 N * DavidS1 DavidS
1219741310 M * DavidS ffs, the same rsync today didn't trigger the oops .. I'll try the other ones
1219741330 M * DavidS could it be the destination FS ?
1219741340 M * DavidS which is now another one
1219741346 M * DavidS another=different
1219741366 M * Bertl sure, everything is possible
1219741384 M * DavidS :-(
1219741431 M * Bertl if you have a suspect (file/dir/etc), try all three of the tools (file,lsattr,showattr) on it, and upload the output
1219741444 M * DavidS will do
1219741778 Q * DavidS Quit: Leaving.
1219741905 J * DavidS ~DavidS@vpn.uni-ak.ac.at
1219742474 J * fosco_ fosco@212.85.148.86
1219742490 J * Alteisen_ alteisen@shell.chaostreff-dortmund.de
1219742504 J * blathijs_ ~matthijs@drsnuggles.stderr.nl
1219742512 J * morrigan_ morrigan@IRC.13thfloor.at
1219742517 J * Bertl_ herbert@IRC.13thfloor.at
1219742521 J * bXi_ bluepunk@irssi.co.uk
1219742528 Q * Mr_Smoke synthon.oftc.net kilo.oftc.net
1219742528 Q * Alteisen synthon.oftc.net kilo.oftc.net
1219742528 Q * bXi synthon.oftc.net kilo.oftc.net
1219742528 Q * nox synthon.oftc.net kilo.oftc.net
1219742528 Q * karasz synthon.oftc.net kilo.oftc.net
1219742528 Q * mEDI_S synthon.oftc.net kilo.oftc.net
1219742528 Q * eyck_ synthon.oftc.net kilo.oftc.net
1219742528 Q * waldi synthon.oftc.net kilo.oftc.net
1219742528 Q * morrigan synthon.oftc.net kilo.oftc.net
1219742528 Q * Bertl synthon.oftc.net kilo.oftc.net
1219742528 Q * fosco synthon.oftc.net kilo.oftc.net
1219742528 Q * blathijs synthon.oftc.net kilo.oftc.net
1219742528 J * nox ~nox@static.88-198-17-175.clients.your-server.de
1219742529 J * Mr_Smoke ~smokey@layla.lecoyote.org
1219742534 J * mEDI_S ~medi@snipah.com
1219742534 J * karasz ~karasz@yoda.expert-erp.net
1219742534 J * waldi ~waldi@bblank.thinkmo.de
1219742534 J * blathijs ~matthijs@drsnuggles.stderr.nl
1219742534 J * morrigan morrigan@IRC.13thfloor.at
1219742534 J * fosco fosco@marx.wirefull.org
1219742534 J * Alteisen alteisen@shell.chaostreff-dortmund.de
1219742534 J * bXi bluepunk@irssi.co.uk
1219742534 J * eyck UwZxKTwV@nat05.nowanet.pl
1219742539 Q * bXi Read error: Connection reset by peer
1219742547 N * Bertl_ Bertl
1219742581 Q * karasz Read error: Connection reset by peer
1219742593 Q * fosco Ping timeout: 481 seconds
1219742608 Q * Alteisen Ping timeout: 481 seconds
1219742626 Q * blathijs Read error: Connection reset by peer
1219742628 Q * morrigan Ping timeout: 481 seconds
1219742638 Q * mEDI_S Ping timeout: 481 seconds
1219742704 J * mEDI_S ~medi@snipah.com
1219742786 J * karasz ~karasz@yoda.expert-erp.net
1219742826 J * fanto666 fantomas@fantomas.fantomas.sk
1219742837 M * fanto666 Hello, do I need special privileges to be able to strace in vserver?
1219742852 M * Bertl no, not really
1219742853 M * DavidS got it: http://paste.linux-vserver.org/12389   
1219742885 M * DavidS there is the dyning rsync as well as lsattrs from the region where it is killed
1219742997 M * DavidS in this case it seems to be a directory that cannot be scanned
1219742997 Q * xdr_ Read error: Connection reset by peer
1219743008 J * lilalinux ~plasma@80.69.41.3
1219743152 M * Bertl DavidS: do we have addr2line info?
1219743237 M * DavidS still compiling :-/
1219743267 M * Bertl ah, I thought you already bootet that kernel
1219743642 N * morrigan_ morrigan
1219743658 M * DavidS find of this directory dies on "fstatat64(AT_FDCWD, "perfmon",  <unfinished ...>"
1219743694 M * Bertl what is the dir? 'perfmon'?
1219743713 M * DavidS it's $vserver/var/lib/samba/perfmon
1219743755 M * Bertl can you try to start a find right above that?
1219743762 M * Bertl and also provide the flags for that dir
1219743771 M * Bertl (i.e. lsattr/showattr) 
1219743858 M * DavidS http://paste.linux-vserver.org/12390 is the relevant part of the find
1219743868 M * DavidS *attr dies before doing any output
1219743904 M * Bertl nice, do you get the same kernel trace for them?
1219743910 M * Bertl or do they slightly vary?
1219743923 N * blathijs_ blathijs
1219743964 M * DavidS the last three lines of the stack are the same: do_lookup,permission,__link_path_walk
1219743981 M * DavidS but __link_path_walk seems to be called from many different places
1219744052 M * Bertl okay, let's wait for the addresses
1219744101 M * DavidS :-/ is building the kernel enough or do i really have to boot into it an get a fresh oops?
1219744159 M * Bertl depends, if the config and build (gcc) are the same, it should be fine
1219744186 N * DoberMann[ZZZzzz] DoberMann
1219744672 M * DavidS if rebuilding isn't enough i can create another kvm guest and test there .. that should hve less risk than playing on the host kernel
1219744681 M * DavidS have*
1219744743 Q * pmenier Quit: Konversation terminated!
1219744938 J * xdr ~xdr@gote2.62.cust.blixtvik.net
1219745235 J * friendly ~friendly@ppp59-167-65-187.lns1.mel6.internode.on.net
1219746245 N * bXi_ bXi
1219746278 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr
1219746513 Q * derjohn Ping timeout: 480 seconds
1219746543 J * derjohn ~derjohn@80.69.41.3
1219746752 J * yarihm ~yarihm@84-74-147-84.dclient.hispeed.ch
1219746999 J * Zigara ~LSD@CPE00179a51d1ee-CM0018c0d8b4d0.cpe.net.cable.rogers.com
1219748838 Q * doener Ping timeout: 480 seconds
1219748922 J * dna ~dna@224-203-dsl.kielnet.net
1219749032 J * doener ~doener@i577B9D1D.versanet.de
1219749907 Q * Aiken Quit: Leaving
1219749951 Q * friendly Ping timeout: 480 seconds
1219750819 M * DavidS yahoo :-) the compile finished 
1219751043 M * DavidS but it seems there is some mismatch between the built kernel and what I'm running: http://paste.linux-vserver.org/12391
1219751092 M * Bertl try to boot with your new kernel
1219751124 M * DavidS on my way
1219751154 M * Zigara does vserver need an special hardware to run a vm? i just want to play around it it on an old amd box to test it out
1219751167 M * DavidS Zigara: no
1219751179 M * Zigara it doesnt require that your cpu has that VT thing?
1219751183 M * DavidS Zigara: no
1219751186 M * Zigara oh great.
1219751188 M * Zigara thanks :]
1219751267 M * DavidS Zigara: but your question leads me to believe that you should read up on what VT and vserver does: e.g. starting with http://linux-vserver.org/Frequently_Asked_Questions#What_kind_of_Operating_System_.28OS.29_can_I_run_as_guest.3F
1219751280 M * Zigara okay.
1219751727 Q * AndrewLee Ping timeout: 480 seconds
1219751998 M * Zigara hrhrhr
1219751999 M * Zigara i like this ;]
1219752517 M * Bertl okay, nap attack ... bbl
1219752523 N * Bertl Bertl_zZ
1219752642 M * Zigara hmm
1219752652 M * Zigara where can I get this "vprocunhide" script?
1219752671 M * daniel_hozac install util-vserver.
1219752688 M * Zigara i think the debian version is outdated or something
1219752889 M * Zigara odd.. the vserver still wont start
1219752895 M * Zigara vc_migrate_context(): Function not implemented
1219752918 M * DavidS Zigara: use older util-vserver;  .215 works
1219752939 M * DavidS there is a mismatch between lenny's utils and current kernels/patches
1219752963 M * Zigara ah
1219752967 M * Zigara okay, thanks
1219752988 M * Zigara 0.30.215 ?
1219753008 M * DavidS yeah
1219753013 M * Zigara okie dokie
1219753026 M * DavidS get it from snapshot.debian.net for example
1219753054 M * DavidS it installed and worked fine for me 
1219753055 M * Zigara hmm
1219753056 M * daniel_hozac better to get 0.30.216~r2772.
1219753063 M * Zigara vserver --version
1219753063 M * Zigara vserver 0.30.215 -- manages the state of vservers
1219753139 M * DavidS daniel_hozac: hmm .. reality seems to have -- again -- out-run my knowledge :-)
1219753783 M * micah 0.30.216~r2772 has been uploaded to unstable and will be in lenny in 10 days
1219754049 M * Zigara rawr:/# localedef -v -c -i en_US -f UTF-8 en_US.UTF-8
1219754049 M * Zigara character map file `UTF-8' not found: No such file or directory
1219754049 M * Zigara cannot read character map directory `/usr/share/i18n/charmaps': No such file or directory
1219754053 A * Zigara cries
1219754059 M * Zigara (sorry for the spam)
1219754643 Q * yarihm Quit: This computer has gone to sleep
1219755959 J * hijacker_ ~hijacker@213.91.163.5
1219755959 Q * hijacker Read error: Connection reset by peer
1219756414 Q * sharkjaw Remote host closed the connection
1219757186 J * cryptronic ~oli@p54A3B1B6.dip0.t-ipconnect.de
1219757757 M * fanto666 anyone workd with gentoo vserver? I'm unable start it because scripts are tryibng to mount /proc etc, which 1. is mounted by vserver scripts 2. permission denied 
1219757801 M * fanto666 hmmm
1219758247 J * kir1 ~kir@swsoft-msk-nat.sw.ru
1219758416 J * ntrs ~ntrs@77.29.79.67
1219758588 J * kir ~kir@swsoft-msk-nat.sw.ru
1219758967 Q * kir1 Ping timeout: 480 seconds
1219759292 Q * larsivi Ping timeout: 480 seconds
1219759427 M * PowerKe fanto666: vserver build should fix the startup scripts after you created it
1219759553 M * fanto666 PowerKe: should i re-build the vserver?
1219759555 M * PowerKe If you updated the initscripts afterwards, you can run the fix script again by executing: /usr/lib/util-vserver/distributions/gentoo/initpost /etc/vservers/<vservername> /usr/lib/util-vserver/util-vserver-vars
1219759567 M * fanto666 aha!
1219759772 M * fanto666 chroot-sh: link("/etc/init.d/net.vserver"): File exists
1219759774 M * fanto666 hmmm
1219759865 Q * mcp Remote host closed the connection
1219759912 J * mcp ~mcp@wolk-project.de
1219760055 M * PowerKe Can you start it now?
1219760068 M * fanto666 no
1219760102 M * fanto666 well, I can provide root password and comment everything in /sbin/rc that tries to mount those filesystems (dev, proc, sys, devpts)
1219760132 M * fanto666 I have installed stage4 I found on webpage (~1 year old) and upgraded 
1219760148 M * PowerKe baselayout 2 ?
1219760152 M * fanto666 downgraded the baselayout...
1219760169 M * PowerKe hmm, downgrading it might not be supported
1219760185 M * PowerKe In that case the fix scripts probably are not enough
1219760248 M * fanto666 hmmm should I install baselayout-vserver?
1219760270 M * PowerKe I'm just using baselayout 2 on my guests
1219760323 M * fanto666 was that the problem? there's 2.0rc4 in the stage, 2.0.0 in portage, depending on openrc that is masked...
1219760335 M * fanto666 do you use openrc? which version?
1219760453 M * PowerKe Hmm, just checking, seems I also have 1.13.0_alpha12 in one of the guests as well
1219760481 M * PowerKe 2.0.0_rc6-r1 in another (no openrc yet on that one)
1219760507 M * fanto666 well, only 2.0.0 is in portage now...
1219760532 Q * ruskie Ping timeout: 480 seconds
1219760575 M * PowerKe I have one on that version as well
1219760588 M * PowerKe openrc-0.2.5
1219760599 M * fanto666 that was broken according to some bug
1219760619 M * fanto666 there's -9999 version in portage because of that bug
1219760662 M * fanto666 (openrc)
1219760781 M * PowerKe hmm, changelog mentions: 'always update dependency tree, fix net.lo and do timezone update before moving clock conf.d', doesn't seem to be something really bad in there, mostly annoying if it goes wrong (for a guest)
1219760820 M * fanto666 http://bugs.gentoo.org/show_bug.cgi?id=224171
1219760829 M * fanto666 and thus http://packages.gentoo.org/package/openrc
1219760883 M * PowerKe I'm not running in parallel for guests
1219761012 M * PowerKe Which baselayout did you install now?
1219761074 M * PowerKe Hmm, 1.13 versions have been removed
1219761088 M * fanto666 I installd 2.0.0
1219761184 M * PowerKe You just changed it now? Since you said you downgraded earlier.
1219761210 M * fanto666 yes
1219761229 M * PowerKe any luck getting that to work (after running the fix script again)?
1219761262 M * fanto666 damn
1219761272 M * fanto666 /lib/rc/sh/init-vserver.sh: line 28: /sbin/rc: No such file or directory
1219761301 M * fanto666 it got lost
1219761325 M * fanto666 I think I'll try to start from the scratch again
1219761802 Q * balbir Ping timeout: 480 seconds
1219761802 Q * xdr Read error: Connection reset by peer
1219763510 J * xdr ~xdr@gote2.62.cust.blixtvik.net
1219763769 M * fanto666 works!@
1219763801 M * DavidS fanto666: yeah .. scrapping the whole vserver is always a good recipe to repair wedged installs ;-D
1219763901 M * fanto666 DavidS: well, I've had to install 2 masked packages ...  
1219763917 M * fanto666 I haven't foud that mentioned in gentoo veerver docsb
1219764014 Q * xdr Quit: leaving
1219764020 J * xdr ~xdr@gote2.62.cust.blixtvik.net
1219764738 Q * DavidS Quit: Leaving.
1219764877 Q * xdr Ping timeout: 480 seconds
1219765082 J * dowdle ~dowdle@scott.coe.montana.edu
1219765988 J * larsivi ~larsivi@169.80-202-217.nextgentel.com
1219766672 J * balbir ~balbir@122.167.245.59
1219766921 Q * dna Quit: Verlassend
1219767155 J * xdr ~xdr@gote2.62.cust.blixtvik.net
1219767267 J * ntrs_ ~ntrs@77.29.74.202
1219767469 J * yarihm ~yarihm@84-74-147-84.dclient.hispeed.ch
1219767677 Q * ntrs Ping timeout: 480 seconds
1219767681 J * ruskie ruskie@ruskie.user.oftc.net
1219767933 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d
1219768486 J * ntrs__ ~ntrs@77.29.74.202
1219768574 Q * ntrs_ Read error: Connection reset by peer
1219768655 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1219768926 N * DoberMann DoberMann[PullA]
1219769324 J * FireEgl FireEgl@adsl-220-221-184.bhm.bellsouth.net
1219769952 N * pmenier pmenier_off
1219770132 Q * pmenier_off Remote host closed the connection
1219770543 M * quote iis it possible to run freebsd as guest vserver?
1219770548 Q * nkukard Ping timeout: 480 seconds
1219770866 M * PowerKe fanto666: check /etc/portage/package.keywords and you'll find masked packages there as well
1219770956 M * PowerKe quote: everything runs under the same kernel, you can't run another kernel for a guest
1219771081 M * quote right..
1219771982 N * Bertl_zZ Bertl
1219772114 M * Bertl back now ..
1219772255 J * nkukard ~nkukard@196.212.73.74
1219773930 Q * lilalinux Remote host closed the connection
1219775052 Q * derjohn Ping timeout: 480 seconds
1219775217 J * loddafnir ~mike@193.170.138.233
1219775422 J * derjohn ~derjohn@dslb-084-058-243-121.pools.arcor-ip.net
1219775514 J * LSD_ ~LSD@CPE00179a51d1ee-CM0018c0d8b4d0.cpe.net.cable.rogers.com
1219775514 Q * Zigara Read error: Connection reset by peer
1219775741 J * hparker ~hparker@linux.homershut.net
1219775742 J * olgagirl ~olgagirl@ANantes-257-1-135-233.w90-32.abo.wanadoo.fr
1219775853 Q * LSD_ Quit: This is D-Block
1219775857 Q * olgagirl 
1219778143 Q * derjohn Ping timeout: 480 seconds
1219778167 J * derjohn ~derjohn@80.69.41.3
1219779633 Q * pisco Ping timeout: 480 seconds
1219780525 M * snooze hey Bertl you there?
1219780559 M * Bertl yup
1219780621 M * snooze we were talking about traceroute and all in vservers
1219780632 M * snooze tracepath acts weirdly as well
1219780638 M * Bertl how so?
1219780647 M * snooze no respone from the final destination
1219780655 M * snooze but from all nodes before
1219780670 M * Bertl and you get one from the host?
1219780675 M * snooze yea
1219780680 M * snooze on the host system it responds
1219780713 M * Bertl make a tcpdump -vvnei ... of both tries (tcpdump on the host) and upload somewhere
1219780728 M * snooze alright
1219780822 M * snooze hm, lots of packets. busy system
1219780834 M * Bertl narrow it down via the target ip
1219780844 M * snooze ah ye
1219780848 M * Bertl (choose an unused target)
1219781241 M * snooze http://paste.linux-vserver.org/12393 host system
1219781254 M * snooze http://paste.linux-vserver.org/12394 guest (i aborted tracepath after couple of timeouts)
1219781416 M * Bertl you are using different addresses
1219781447 M * snooze well yeah
1219781466 M * snooze host system got a different ip to the internet than the guests (which are runs behind nat of that ip)
1219781467 M * Bertl 1.0.0.2 is assigned to the guest, yes?
1219781495 M * snooze 1.0.0.2 would be the internet ip for the guests, the guests themself have local 172.16.* ip's
1219781512 M * Bertl what is assinged in the config?
1219781556 M * snooze the guest has the ip 172.16.0.12
1219781558 M * Bertl or even better, get me the /proc/virtnet/<nid>/* contents uploaded
1219781578 M * Bertl and you use some NAT to make that 1.0.0.2?
1219781597 M * snooze yes
1219781607 M * Bertl then you need to handle the icmp replies too
1219781619 M * snooze SNAT       all  --  172.16.0.0/21        anywhere            to:1.0.0.2
1219781650 M * snooze i can ping inside the guest tho?
1219781672 M * Bertl check the tcpdumps, compare them
1219781719 M * Bertl you want a DNAT entry for the public guest IP, at least for the icmp
1219781839 M * Bertl something like iptables -t nat -A PREROUTING -d 1.0.0.2 -p icmp DNAT --to-destination 172.16.0.12
1219781853 M * Bertl (probably with an -i ethx too
1219781902 M * Bertl you should get away by limiting it to the proper icmp-type(s) though
1219781978 M * Bertl btw, probably the best package to do tracing is tcptraceroute
1219782357 M * snooze iptables -t nat -A PREROUTING -d 1.0.0.2 -p icmp -j DNAT --to-destination 172.16.0.12 didnt help anyway
1219782610 M * snooze tcptraceroute looks good btw, i'll test that some time :>
1219783016 M * Bertl you need to make sure that the public ip is nat-ed back .. then it should work
1219783033 M * Bertl okay, off to bed now .. have a good one everyone!
1219783041 N * Bertl Bertl_zZ
1219783124 J * cybergirl ~cybergirl@ANantes-257-1-135-233.w90-32.abo.wanadoo.fr
1219783261 Q * cybergirl 
1219783986 Q * ntrs__ Ping timeout: 480 seconds
1219784144 Q * bonbons Quit: Leaving
1219784345 J * rooijan ~rrosso@cuba.sonosite.com
1219784386 M * rooijan I have a really weird behaviour after a recent/ubuntu update and reboot.  Can someone help?
1219784494 M * rooijan Basically my backup client EMC Legato now fails because it is saying I am talking from one of my vserver guests and not the host.
1219784563 M * rooijan So a tcpdump shows that my host is talking back as one of the guests.  Which obviously confuses the backup server...I have to run to a meeting but can provide any information necessary later when I get back.
1219784659 J * derjohn_mob ~aj@e180210209.adsl.alicedsl.de
1219784760 J * Aiken ~Aiken@ppp118-208-65-150.lns1.bne4.internode.on.net
1219785085 M * dustybin im using this ip route setup to seperate my HOSTS traffic on eth0 and my vserver guest traffic on eth1:
1219785088 M * dustybin http://paste.linux-noob.com/index.php?query=2703
1219785107 M * dustybin if i pull the cable out of the vserver guests NIC, i can no longer ping anything apart from the host
1219785123 M * dustybin anybody know a simple iptables rule i can setup to stop the vserver guest being able to ping the host
1219785721 Q * cryptronic Quit: Leaving.
1219785746 J * pisco ~pisco@tor.noreply.org
1219786453 J * androsch ~androsch@dslb-084-060-202-196.pools.arcor-ip.net
1219786458 M * dustybin "And as I'm sure you vserver experts know, inter-vserver communications occur over the loopback interface.  Once I move my rule(s) to disable communication between vserver clients above the loopback rule everything worked as expected."
1219786463 M * dustybin right............
1219786466 M * dustybin thats the PROBLEM
1219787879 Q * loddafnir Remote host closed the connection
1219787938 M * dustybin is it possible to assign a vserver guest its own local loopback interface?
1219788009 N * DoberMann[PullA] DoberMann[ZZZzzz]
1219789734 M * rooijan Someone on here know why my host will suddenly after an update/reboot be confused on what the primary host interface is?
1219789806 M * rooijan everything is working except my backup software client.  when talking to the server it is confused and thinks its one of the guests talking back as opposed to the host
1219790706 Q * yarihm Quit: Leaving
1219791309 J * ViRUS ~mp@p579B4048.dip.t-dialin.net
1219792999 Q * nou Ping timeout: 480 seconds
1219793247 J * nou Chaton@causse.larzac.fr.eu.org
1219793597 M * hparker http://paste.linux-vserver.org/12395   Ideas?
1219793694 M * daniel_hozac rooijan: paste ip a; ip r output at paste.linux-vserver.org, please.
1219793735 M * daniel_hozac hparker: missing space in front of --, missing - in front of d lenny.
1219793757 M * hparker ugh.. thanks daniel_hozac 
1219793764 A * hparker spanks fingers
1219793796 M * daniel_hozac dustybin: no, since Linux-VServer is an IP-based isolation technique.
1219793958 M * rooijan daniel output at http://paste.linux-vserver.org/12396
1219793984 M * hparker ugh... chroot-sh: open("/etc/apt/sources.list"): No such file or directory... I have a gentoo host, what needs to be there?
1219794018 M * daniel_hozac that's inside the guest, not on the host.
1219794040 M * hparker oh
1219794070 M * hparker uhm
1219794095 M * daniel_hozac so somehow you ended up with a guest that has no /etc/apt... an indication that something went very wrong.
1219794103 M * hparker http://paste.linux-vserver.org/12397
1219794127 M * hparker I don't doubt that, just used gentoo guests till now
1219794163 M * daniel_hozac that doesn't show any debootstrap outupt.
1219794178 M * daniel_hozac what debootstrap version are you downloading?
1219794182 M * hparker uhm
1219794198 M * hparker that one :P I've not seen debian in like 10 years
1219794209 Q * ViRUS Quit: Leaving
1219794236 M * hparker But, I changed something to do with that because of the error I had before, let me fix that
1219794237 M * daniel_hozac it's in the output, just masked on the pastebin.
1219794270 M * daniel_hozac http://ftp.debian.org/debian/pool/main/d/debootstrap/debootstrap_1.0.10_all.deb should work.
1219794299 M * hparker In /etc/vservers/.defaults/apps/debootstrap/uri ?
1219794310 M * daniel_hozac rooijan: well, you have some IPs as /16, some as /24, and a ton of routes not specifying a source IP...
1219794315 M * daniel_hozac hparker: right.
1219794319 M * hparker k
1219794353 M * rooijan daniel: I understand.  This is a production setup.  Worked for a year until update and reboot on Ubuntu 7.04 host this weekend.
1219794374 M * daniel_hozac most likely by accident :)
1219794408 M * daniel_hozac you'll want to correct those IPs.
1219794465 M * rooijan my problem is slightly bizarre.  my backup server identifies my host as one of the guests when talking back to it.  When I shut the guest down I can make this work.  however the guests are all in production so can't really shut down to get a clean backup here.
1219794477 M * hparker daniel_hozac: http://rafb.net/p/JNpuZN25.html
1219794532 M * daniel_hozac hparker: right, use -- --arch amd64
1219794571 M * hparker Does that go after -- -d ?
1219794578 M * daniel_hozac at the very end, yes.
1219794587 M * hparker k
1219794613 M * daniel_hozac rooijan: not unexpected. your guests are using /24s, which is more than /16...
1219794676 M * rooijan ok let me try fixing it.  although I have several hosts in production and always laid out guests class C and hosts class B.  I will experiment quick.
1219794686 M * hparker Just returns me to the command prompt and nothing bult in /vservers
1219794748 M * hparker hparker@saturn ~ $ sudo /usr/sbin/vserver freeside.pcsrvc.com build -m debootstrap --context 111 --hostname freeside.pcsrvc.com --interface 10.0.0.111/24 -- -d lenny -m http://ftp.us.debian.org/debian &> debian -- -arch amd64
1219794764 M * hparker with or without the -- before amd64
1219794769 M * rooijan daniel_hozac you are brilliant
1219794782 M * hparker that he is ;)
1219794792 M * daniel_hozac hparker: &> debian has to be after it ;)
1219794809 M * hparker After what? :P
1219794814 M * daniel_hozac -- --arch amd64
1219794820 M * daniel_hozac otherwise the shell will eat that up.
1219794836 M * rooijan weird that this broke now.  must be something is more sensitive after my update.  I will have to fix a lot of production instances :(
1219794880 M * daniel_hozac rooijan: it's fine to use that setup, as long as the guest's /24s aren't in the host's /16.
1219794918 M * hparker bash: syntax error near unexpected token `newline'
1219794966 M * daniel_hozac hparker: when running sudo /usr/sbin/vserver freeside.pcsrvc.com build -m debootstrap --context 111 --hostname freeside.pcsrvc.com --interface 10.0.0.111/24 -- -d lenny -m http://ftp.us.debian.org/debian -- --arch amd64 &> debian ?
1219794999 A * hparker runs
1219795029 M * rooijan hmm my guest /24's are always in my flat /16.  I will fix and later investigate if this could be NAC policy.  Thanks I can get my host backed up now by just fixing the on guest to /16
1219795034 M * rooijan You are very helpful
1219795086 M * daniel_hozac i try...
1219795118 M * hparker daniel_hozac: that command runs and drops me at that prompt.. but, nothing in /vservers or /etc/vservers for the guest
1219795157 M * rooijan Well people like me fighting the good fight and up against PHB's allday to justify linux-vserver vs vmware fixing quickly is really a big plus.  Their Gartner crap alwasy say we will be stuck with no support. Gmmffff