1165709591 Q * Piet_ Ping timeout: 480 seconds
1165710575 M * Aiken Bertl would you say AOE is driver/other or IO/other with the bug report?
1165710846 M * Bertl hmm, probably IO, but don't blame me :)
1165711018 M * Aiken I don't particularly like silent failures
1165711343 M * Bertl daniel_hozac: did you read the shutdown issue above?
1165711358 M * Bertl daniel_hozac: I mean the host reboot ...
1165711365 M * daniel_hozac ah, yeah.
1165711380 M * daniel_hozac yes, i agree, something needs to be done for that.
1165711509 M * daniel_hozac i guess a /usr/lib*/util-vserver/start-vservers --all --stop; for i in /proc/virtual/*; do vattribute --set --xid $i --flag ~persistent; vkill -s 15 --xid $i; sleep 2; vkill -s 9 --xid $i; done or so should do it.
1165711535 M * Bertl maybe that could be done from the vserver runlevel script on RL 6 or 0?
1165711557 M * daniel_hozac i think it should be a script separate from vservers-default.
1165711573 M * daniel_hozac but it should definitely be in an initscript.
1165711584 M * Bertl hmm, another one?
1165711601 M * daniel_hozac yes, i think so.
1165711616 M * daniel_hozac otherwise it'd be invoked on /etc/init.d/vservers-default restart, e.g.
1165711645 M * daniel_hozac and that script should just be managing the guests with the default mark.
1165711651 M * Bertl why? btw, the vserver-default is exactly one character too long for a nice chkconfig :)
1165711677 M * Bertl the runlevel scripts get both runlevel passed, no?
1165711689 M * Bertl a transition to 0 or 6 is final
1165711713 M * daniel_hozac well, i guess we could use that.
1165711740 M * daniel_hozac it would still be executed for all groups though.
1165711761 M * Bertl hmm?
1165711764 M * daniel_hozac (i.e. it might mess up the order of stopping the guests, but i guess depends should handle that)
1165711801 M * daniel_hozac if you have two groups of guests, one with mark default, and another with default-backup or so.
1165711823 M * Bertl you mean vserver-backup :)
1165711830 M * daniel_hozac right.
1165711878 M * daniel_hozac although i suppose it really shouldn't be a problem.
1165711882 M * Bertl well, another script is fine for me ...
1165711900 M * Bertl just having a disk check on every reboot is not :)
1165711908 M * daniel_hozac hehe.
1165711941 M * Bertl btw, renaming the scripts to vserver-default and vserver-legacy would make chkconfig look nice *hint*
1165711989 M * daniel_hozac but it should be plural, no? :)
1165712034 M * Bertl changing 'default' to def or std or mark would be an option too :)
1165712063 M * Bertl btw, vserver-mark would be more natural anyway IMHO
1165712071 M * Bertl (or vservers-mark)
1165712077 M * daniel_hozac well, renaming the initscript wouldn't be very nice.
1165712090 M * daniel_hozac people would have to re-chkconfig vservers-default on.
1165712100 M * daniel_hozac (if it was on)
1165712108 M * Bertl IIRC, you have to do that anyways
1165712119 M * daniel_hozac not after each upgrade.
1165712126 M * Bertl at least it is always off here
1165712138 M * Bertl and I have to turn it on after each rpm install
1165712150 M * daniel_hozac then your chkconfig is broken ;)
1165712198 M * Bertl the legacy OTOH, is always on
1165712222 M * daniel_hozac that is a bug.
1165712254 M * Bertl a known one?
1165712258 M * daniel_hozac it is now :)
1165712315 M * daniel_hozac fixed in trunk
1165712332 M * daniel_hozac btw, re vshelper, http://people.linux-vserver.org/~dhozac/p/uv/experimental/vshelper-warning.diff
1165712360 M * Bertl ntrs: I don't get a zombie here
1165712381 M * daniel_hozac does the patch look ok to you?
1165712424 M * Bertl looks good
1165712486 M * Bertl ntrs: please check with the following guest install:
1165712512 M * Bertl vserver sarge64 build -m debootstrap --context 164 --hostname sarge64.test.org --interface eth0:10.0.0.2/24 -- -d sarge -m ftp://ftp.jp.debian.org/debian-amd64/debian -- --arch amd64
1165713045 J * derjohn3 ~aj@dslb-084-058-199-027.pools.arcor-ip.net
1165713061 Q * derjohn2 Ping timeout: 480 seconds
1165713259 M * Bertl daniel_hozac: what was the reason why we brought the __enter/leave to 2.0.x in the first place?
1165714029 M * daniel_hozac hmm, wasn't it supposed to fix the MySQL issues?
1165714087 M * daniel_hozac hmm, seems to have been added in 2.0.2.2?
1165714109 M * daniel_hozac i.e. with the 2.6.18 port.
1165714168 M * Bertl yes, but I think it didn't fix the mysql issues, right?
1165714202 M * daniel_hozac well, it fixed them for derjohn3, IIRC.
1165714220 M * Bertl hmm, and how did we fix them for ntrs?
1165714231 M * daniel_hozac did we? i was under the impression there were still problems?
1165714245 M * daniel_hozac (a different set of problems, but problems nonetheless)
1165714247 M * derjohn3 well, I cannot remember how it was fixed 
1165714279 M * Bertl derjohn3: could you do a test run with a 2.6.18.5-vs2.0.2 kernel for me?
1165714294 M * Bertl (I mean, with a special version I upload)
1165714299 M * derjohn3 But it didnt occur and ntrs didnt complain for a long time , too
1165714301 M * ntrs Bertl, this is exactly why we (you)  need a detailed changelog
1165714309 M * derjohn3 Bertl,  2.0.2 ?
1165714315 M * ntrs 2.1.1.3 seems to fix the mysql issues.
1165714324 M * daniel_hozac all of them?
1165714325 M * Bertl ntrs: yeah, but derjohn3 doesn't want to do it *G*
1165714348 M * ntrs I can't confirm this 100% yet, I'll say they are definitely fixed when I don't see any for a month or so.
1165714355 M * daniel_hozac Bertl: quiet now, before someone suggests git ;)
1165714364 M * derjohn3 derjohn1-3 are confused by an illness called "versionitis"
1165714391 M * derjohn3 ntrs, 2.1.1.3 isnt a month old
1165714441 M * ntrs right, which is why I said I will consider it 100% fixed when the problem does not happen for a month.
1165714466 M * derjohn3 in my case it occured all the some 2 weeks
1165714493 M * derjohn3 I upgraded the machine to a new hardware and kernel about ... 2 weeks ago. no issues so far.
1165714500 M * Bertl so that won't help with testing either ...
1165714534 M * derjohn3 hm, sadly, yes, but I am very happy it that production dosnt to down anymore
1165714551 M * Bertl daniel_hozac: and I don't think we can assume that the enter fixed anything for derjohn3 if it only happens every 2 weeks, right?
1165714560 M * daniel_hozac i guess not.
1165714576 M * daniel_hozac it does seem to be rather inadvertent too.
1165714589 M * Bertl I think, we simply revert the enter for 2.0.x and that's it
1165714598 M * daniel_hozac i agree, sounds sane.
1165714638 M * Bertl where was the socket overflow patch?
1165714642 M * derjohn3 Bertl, what about that 2.0.2 ? where can I fetch the patch or kernel src ?
1165714658 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-cacct-fix01.diff
1165714664 M * Bertl derjohn3: I don't think it is relevant anylonger
1165714697 M * Bertl daniel_hozac: tx
1165714738 M * derjohn3 Bertl, ok ... but still I promised to bring up a a testing box for you
1165714767 M * derjohn3 Bertl, I'll have a look for it soon
1165714780 M * Bertl np, I have one atm, from princeton
1165714837 M * Bertl daniel_hozac: the unhash fix is probably relevant for 2.0.x too, no?
1165714878 M * daniel_hozac hmm, doesn't look that way.
1165714883 M * daniel_hozac or which unhash fix are you referring to?
1165714908 M * Bertl http://vserver.13thfloor.at/Experimental/delta-unhash-fix01.diff
1165714959 M * daniel_hozac well... __unhash_vx_info isn't called from anywhere but unhash_vx_info.
1165714983 M * daniel_hozac but i guess at least parts of that are relevant.
1165715018 M * daniel_hozac or is it already applied?
1165715025 M * Bertl that could be
1165715047 M * daniel_hozac it looks like it.
1165715052 M * Bertl if so, rc7-rc8 must contain that
1165715073 M * daniel_hozac yeah.
1165715074 M * daniel_hozac it does.
1165715080 M * Bertl okay, excellent
1165715122 M * Bertl so after reverting the enter/leave stuff, we should be back to 'stable' there (including the cacct fix)
1165715135 M * daniel_hozac i think so.
1165715198 M * daniel_hozac IIRC there was one other issue pending for 2.0...
1165715204 M * daniel_hozac i remember there being three.
1165715253 M * Bertl lock/private maybe?
1165715347 M * daniel_hozac that looks like it's ok (not present).
1165715392 M * daniel_hozac or, hmm... it's present for the network context, but not for the regular context.
1165715405 M * daniel_hozac (private, that is. lock isn't in the patch at all)
1165715721 J * FireEgl proteus@adsl-61-147-76.bhm.bellsouth.net
1165715771 M * Bertl yeah, maybe we should remove the private from nx there
1165716113 Q * Aiken Quit: Leaving
1165716388 J * Aiken ~james@tooax8-230.dialup.optusnet.com.au
1165717183 M * daniel_hozac Bertl: okay, i think i have the vserver-stat bug now.
1165717194 M * daniel_hozac (i.e. i can reproduce it)
1165717237 M * daniel_hozac (seems to have occured after reproducing the /proc issue...)
1165717287 M * Bertl okay, that is without the fix, yes?
1165717306 M * Bertl I mean without the xid -> n change
1165717315 M * daniel_hozac right.
1165717358 M * Bertl okay, I'm pretty sure it is the result of such a 'mistagged' dentry/inode
1165717383 M * Bertl if you can recreate it with that patch applied, we have another bug to hunt down
1165717397 M * daniel_hozac i'll try to find out which part of the process is breaking.
1165717418 M * Bertl all entries in such a xid are empty
1165717432 M * Bertl try cat /proc/virtual/<xid>/status
1165717486 Q * shedi Quit: Leaving
1165717489 M * daniel_hozac well, my kernel is modified to return -ENOENT.
1165717502 M * daniel_hozac but yes, that's what happens.
1165717650 M * daniel_hozac hmm, seems the run file has been removed for some weird reason.
1165717659 M * daniel_hozac OH!
1165717706 M * daniel_hozac there's a function somewhere in the scripts that cleans up run files from non-existing contexts... i believe that is using /proc.
1165717759 M * Bertl so that's probably the reason why it's gone
1165717770 M * daniel_hozac indeed, that explains it.
1165717883 M * daniel_hozac yeah, scripts/functions:getVserverStatus
1165720455 M * micah Bertl: I haven't had time to look at it yet
1165720465 M * micah but I did get the new util-vserver release uploaded today
1165720497 M * micah sorry for the delayed response
1165720529 M * Bertl micah: okay, would be nice if we could make it before 213 :)
1165720638 M * micah Bertl: I hope I have some time over the holidays, this pre-holiday stuff has been insane
1165724161 T * Bertl http://linux-vserver.org/ | latest stable 2.0.2.1, 2.0.2.2-rc9, 2.2.0-rc4, devel 2.1.1.5, 2.3.0.3, stable+grsec 2.0.2.1, devel+grsec 2.1.1 | util-vserver-0.30.212 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the Wiki, and we'll forget about the minute ;)
1165724202 M * Bertl off to bed now ... have a good one everyone ... cya!
1165724207 N * Bertl Bertl_zZ
1165725029 J * Aiken_ ~james@tooax6-014.dialup.optusnet.com.au
1165725358 Q * Aiken Ping timeout: 480 seconds
1165729459 M * mnemoc .oO( 2.3.x???! )o
1165736308 J * _dmax ~semaj@81.193.58.90
1165736406 Q * dmax Ping timeout: 480 seconds
1165736412 N * _dmax dmax
1165736426 Q * s0undt3ch Ping timeout: 480 seconds
1165741600 J * dna ~naucki@29-237-dsl.kielnet.net
1165743354 J * [flanque] ~nope@60-240-64-48-nsw-pppoe.tpgi.com.au
1165743404 Q * bronson Ping timeout: 480 seconds
1165745495 J * pmenier ~pmenier@ACaen-152-1-55-248.w86-192.abo.wanadoo.fr
1165746909 M * nebuchadnezzar Hello, there is no Changelog for 2.2.0-rc* ?
1165748521 M * derjohn3 nebuchadnezzar, no, not yet - even I am confused, and I was supposed to write the changelogs ..
1165749131 Q * derjohn3 Remote host closed the connection
1165749300 Q * [flanque] 
1165749464 M * Loki|muh any ideas why a reboot -f restarts the vserver to runlevel 3? a vserver <name> restart works as expected (runlevel 2)
1165751065 J * sebastian ~info@pD957DD0A.dip.t-dialin.net
1165752059 Q * ensc Killed (NickServ (GHOST command used by ensc_))
1165752068 J * ensc ~irc-ensc@p54B4EF9C.dip.t-dialin.net
1165752667 M * daniel_hozac nebuchadnezzar: there aren't even changelogs for the most recent 2.1 releases, so...
1165752805 M * daniel_hozac Loki|muh: that sounds strange, reboot -f should be doing pretty much the same thing as vserver <name> restart.
1165752823 M * daniel_hozac Loki|muh: i assume you set the runlevel in apps/init/runlevel.start, right?
1165752923 M * Loki|muh daniel_hozac: no I haven't set it manually... there was never a need to
1165752942 M * daniel_hozac so you're using a plain initstyle?
1165752966 M * daniel_hozac because 3 is the default runlevel for sysv.
1165752968 M * Loki|muh I guess I'm using fakeinit
1165753028 M * Loki|muh i have only set network information and a name, thats it.
1165753055 M * daniel_hozac then your default runlevel would be 3.
1165753087 M * Loki|muh ups, you're right
1165753099 M * Loki|muh hmmm, then my problem is somewhere else
1165753119 M * daniel_hozac what problem is that?
1165753161 M * Loki|muh when I type reboot -f, then only my dns server is started, no mysql server, no sshd
1165753183 M * daniel_hozac versions?
1165753196 M * Loki|muh old tools, 0.30.210
1165753214 M * Loki|muh vs2.1.1-rc20
1165753249 M * daniel_hozac with legacy support, i guess.
1165753261 M * Loki|muh yes
1165753484 M * daniel_hozac are you sure the guest is restarted?
1165753511 M * derjohn daniel_hozac, besides the kernel rebase: is there a difference between 2.1.1.3 and 2.1.1.5 ?
1165753546 M * daniel_hozac well, the kernel rebase meant we reworked the namespace commands
1165753555 M * daniel_hozac and added new versions.
1165753626 M * derjohn daniel_hozac, does that have any impact on 2.1.1.3/ 2.6.18.5 ? I mean, is there a "bug" in this combination?
1165753638 M * derjohn (well, I mean .. a known bug)
1165753648 M * daniel_hozac i don't think so.
1165753680 M * derjohn i assume the namespace stuff came with 2.6.19 , so you had to adapt to that with 2 "tries" .4 and .5 ?
1165753703 M * daniel_hozac well, we had 2.6.19-vs2.1.x-t[1-7] first.
1165753708 M * Loki|muh daniel_hozac: good point. the uptime shown by vserver-stat is reseted. is there a possibility to do any debugging from the host?
1165753734 M * daniel_hozac .5 fixes an issue with /proc/virt{ual,net}.
1165753743 M * daniel_hozac Loki|muh: well, that should be enough, i guess.
1165753760 M * derjohn this issue will be there with 2.1.1.3 on 2.6.18.5 i assume ?
1165753780 M * daniel_hozac Loki|muh: there's a file you can touch to get logging... /etc/vservers/.defaults/apps/vshelper/logfile
1165753798 M * daniel_hozac Loki|muh: that should get all the output from vshelper and the commands it executes.
1165753813 M * daniel_hozac derjohn: i doubt it. i'm fairly certain it came with the port.
1165753832 M * derjohn ok.
1165753857 M * daniel_hozac yeah, nid isn't used in 2.1.1.3.
1165753861 M * derjohn thats fine ... I strive toward having a kernel /vserver combination that simply works but is devel branch.
1165753873 M * derjohn so this one could be a good one
1165753877 M * daniel_hozac i guess you'll have to go 2.6.19 then :)
1165753904 M * daniel_hozac as 2.1.1.3 won't be getting any updates.
1165753908 M * derjohn daniel_hozac, i didnt see 2.9.19 i debian, but I could go from vanilla and make a oldocnfig for .config
1165753929 M * daniel_hozac i guess it's gonna be a while before any distribution updates to it.
1165753937 M * derjohn no updates ? thats fine , if there are simply no issues ;)
1165753950 M * daniel_hozac how likely do you think that is? ;)
1165753968 M * derjohn daniel_hozac, well, I would tend to wait a little before using this "perfect" kernel in production
1165753988 M * derjohn daniel_hozac, P("how likely do you think that is?") = 5%
1165755998 J * Piet hiddenserv@tor.noreply.org
1165756337 J * _mcp ~hightower@wolk-project.de
1165756382 Q * m4z Read error: Connection reset by peer
1165756382 Q * mcp Read error: Connection reset by peer
1165756392 N * _mcp mcp
1165756400 J * m4z m4z@bastard-operator.from-hell.net
1165756533 J * s0undt3ch ~s0undt3ch@81.193.58.90
1165756796 M * ensc just got an oops while trying out new utils http://paste.linux-vserver.org/756 (vs2.0.2.2-rc8)
1165756835 M * daniel_hozac i think that has the same cause as your previous oops.
1165756840 M * daniel_hozac it should be fixed in 2.0.2.2-rc9.
1165757513 Q * Aiken_ Quit: Leaving
1165758902 Q * blues Quit: leaving
1165760653 Q * sebastian 
1165761634 J * shedi ~siggi@inferno.lhi.is
1165762176 N * Bertl_zZ Bertl
1165762182 M * Bertl morning folks!
1165762299 M * eyck morning Bertl
1165762329 M * Bertl hey eyck! how are you?
1165762353 M * eyck hmm, is this an american question, or an european question?
1165762391 M * Bertl hehe .. well, best answer both variants :)
1165762418 M * eyck oh man, couldn't be worse. AND. great, thank you, how about you?
1165762467 M * Bertl why so, tell me AND fine to hear, tx! :)
1165762515 M * eyck american version sucks when it comes to socializing over beer ;) 
1165762590 M * Bertl 'nice day' - 'hmhmm' - 'and the sunshine' - 'yeah' ...
1165762621 M * Bertl so tell me, what bothers you ...
1165762673 M * eyck I'm about to loose my net connection, and a little chunk of montly income, and few servers, 
1165762723 M * Bertl ouch
1165762909 J * Simba_ ~Simba@mm-145-150-57-86.adsl.mgts.by
1165762946 Q * Simba_ 
1165763117 Q * Simbalin Ping timeout: 480 seconds
1165763299 J * bragon ~weechat@sd866.sivit.org
1165763307 M * bragon hello here
1165763309 M * daniel_hozac hi
1165763313 M * Bertl hey bragon!
1165763339 M * trippeh_ eyck: This is why I have 4 net connections :)
1165763360 Q * bragon 
1165763379 M * Bertl trippeh_: well, then go ahead, share one with eyck :)
1165763382 J * bragon ~bragon@sd866.sivit.org
1165763384 M * trippeh_ Primary work-sponsored ADSL2+, secondary cheap cable, then work-sponsored 3G/mobile and open neighbour wlans ;)
1165763384 M * bragon plop
1165763393 M * bragon i have a question
1165763418 M * bragon Is IPv6 Support for vserver done?
1165763537 M * daniel_hozac "done"?
1165763548 M * bragon hum
1165763554 M * daniel_hozac with bonbons' patch, you could have IPv6 support today, if that's what you mean.
1165763563 M * bragon cool
1165763570 M * bragon yes it's what i mean
1165763577 M * bragon because i found that
1165763584 M * bragon it's old but :
1165763598 M * bragon http://www.paul.sladen.org/vserver/archives/200211/0063.html
1165763634 M * daniel_hozac well, that's ancient...
1165763658 M * bragon yes
1165763666 M * bragon but first google answer
1165763668 M * bragon :)
1165763693 M * bragon ok
1165763717 M * bragon daniel_hozac do you have any tutorial to make ipv6 work under vserver ?
1165763779 M * daniel_hozac well, apply the patch to your kernel (2.6.18.5-vs2.1.1.3 seems like a good choice, as it hasn't been ported to 2.6.19 yet), make sure you have recent utils (0.30.212+), and you should be good to go.
1165763802 M * bragon ok
1165763819 M * bragon so just apply the bonbons' patch + vserver patch to my kernel and it roulz
1165763837 M * daniel_hozac yep.
1165763896 M * bragon thank you for these answer
1165763899 M * bragon :)
1165763909 M * bragon say hello to Bertl for me :))
1165763913 M * bragon see you latter
1165763934 M * daniel_hozac feel free to stick around...
1165763955 M * daniel_hozac note that you'll probably want my rediff of bonbons' patch, http://people.linux-vserver.org/~dhozac/p/k/delta-2.1.1.2.2-ipv6.diff
1165764030 M * bragon ok
1165764153 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk
1165764270 Q * bragon Remote host closed the connection
1165766264 J * sebastian ~info@p54A95016.dip.t-dialin.net
1165767173 Q * ensc Ping timeout: 480 seconds
1165768193 Q * maharaja Read error: Connection reset by peer
1165768262 J * Simba_ ~Simba@mm-145-150-57-86.adsl.mgts.by
1165768275 M * Bertl wb Simba_!
1165768283 M * Simba_ ?
1165768301 M * Bertl well, you connected about 90min ago :)
1165768313 M * Simba_ ÇÇ
1165768899 J * ensc ~irc-ensc@p54B4EF9C.dip.t-dialin.net
1165768904 Q * Simba_ Read error: Operation timed out
1165769194 M * Bertl daniel_hozac: hmm, I wonder, aren't you missing indoe.h in the headers?
1165769209 M * daniel_hozac hmm?
1165769242 M * Bertl well, it is not part of the installed headers
1165769250 M * Bertl (kernel headers)
1165769309 M * daniel_hozac why would it be missing?
1165769326 M * Bertl it defines the flags, no?
1165769330 M * daniel_hozac it's not something that was used before i moved to headers_install.
1165769341 M * daniel_hozac well, the utils keep their own set of flags.
1165769349 M * daniel_hozac in lib/vserver.h
1165769365 M * Bertl hmm, I guess that should be moved then ...
1165769383 M * Bertl anyway, I'll add inode.h to the exported headers
1165769395 M * daniel_hozac yeah, it should be included.
1165770481 J * Simbalin ~Simba@deb30.mgts.by
1165771279 J * bronson ~bronson@adsl-75-36-147-248.dsl.pltn13.sbcglobal.net
1165771349 J * dreamind apwdsl@p548AC27B.dip0.t-ipconnect.de
1165771390 P * dreamind 
1165772238 J * Chris1976 ~Chris1976@212.183.134.67
1165772254 M * Bertl welcome Chris1976! wb Simbalin! :)
1165772450 P * Chris1976 
1165775325 J * Val ~val@v41.org
1165775337 M * Val Hi
1165775347 M * Bertl hey Val!
1165775358 M * Val Hi Bertl :)
1165775376 M * Val i see there is a new 2.6.18 vs 2.0.2.2-rc9 patch available
1165775383 M * Bertl yup
1165775419 M * Val i tried to apply it on 2.6.18.5 (-7) from Debian sources but i got lot of rejects ...
1165775503 M * daniel_hozac i guess waldi will have updated packages RSN.
1165775524 M * Val daniel_hozac : util-vserver ?
1165775534 M * daniel_hozac hmm?
1165775551 M * Val hmm sorry, i fall in there :)
1165775563 M * daniel_hozac i'm referring to linux-image-vserver
1165775568 M * Val ok
1165775575 M * daniel_hozac micah uploaded 0.30.212 yesterday.
1165775581 M * Val oh well
1165775583 M * Val good
1165775878 M * Val got some reversed (or previously applyed) patch detected : arch/alpha/Kconfig, arch/i386/kernel/microcode.c, arch/ia64/sn/kernel/bte.c, block/scsi_ioctl.c, drivers/char/agp/generic.c, drivers/char/agp/intel-agp.c, drivers/media/Kconfig, drivers/net/tg3.c, drivers/pcmcia/ds.c, drivers/scsi/scsi_lib.c, fs/fuse/dir.c, init/version.c, net/bluetooth/hci_sock.c, net/bridge/br_ioctl.c, net/dccp/ipv6.c, net/ieee80211/softmac/ieee80211softmac_io.c, net/ipv4/ipvs/ip_vs_core.c...
1165775908 M * daniel_hozac i am fairly certain we don't patch those files.
1165775915 M * Bertl me too!!
1165775918 M * Val ok
1165775920 M * Val well
1165775937 M * Val may be i did a mistake
1165775942 M * Val ... let's retry :)
1165775949 M * waldi daniel_hozac: too lat for the upload today
1165775978 M * Bertl btw, I just realized that the 2.0.2.2-rc9 extraversion is wrong
1165775996 M * Val well... a disk space error when applying patch
1165775999 M * Val sorry
1165776071 Q * bj Read error: Operation timed out
1165776078 Q * m4z Read error: Operation timed out
1165776129 Q * michal` Read error: Operation timed out
1165776148 Q * meebey Read error: Operation timed out
1165776222 M * Bertl hum, the rc9 seems to got broken somehow
1165776224 J * bj ~bj@insanefactory.com
1165776224 Q * bogus Remote host closed the connection
1165776243 Q * phreak`` Remote host closed the connection
1165776246 J * phreak`` ~phreak``@styx.xnull.de
1165776300 M * daniel_hozac Bertl: hmm? how do you mean?
1165776316 M * Bertl I guess I already know what happened, will update it shortly
1165776331 M * Bertl the rc9 is based on 2.6.18.3 instead of 2.6.18.5
1165776337 M * Val hu
1165776346 M * Bertl Val: give me a minute, you get a new patch
1165776347 M * Val that's why i got all these rejects
1165776351 Q * brcc Ping timeout: 480 seconds
1165776360 M * Val Bertl : ok, fine :)
1165776399 Q * pmenier Quit: KVIrc 3.2.0 'Realia'
1165776498 J * michal` ~michal@www.rsbac.org
1165776536 J * m4z m4z@bastard-operator.from-hell.net
1165776693 M * daniel_hozac Bertl: well, there's not a single offset or anything, so it's not noticable...
1165776715 J * bogus ~bogusano@fengor.net
1165776740 M * Bertl daniel_hozac: it revert 2.6.18.5 to 2.6.18.3+vserver :)
1165776745 J * meebey meebey@booster.qnetp.net
1165776753 M * Bertl Val: okay, updated patch at the same location
1165776760 M * Val ok, thx
1165776794 M * Bertl daniel_hozac: so I assume, the debian is actually a 2.6.18.3 something
1165776802 M * daniel_hozac Bertl: ah! i see.
1165776806 M * Bertl daniel_hozac: and thus, reverting 2.6.18.5 stuff doesn't work
1165776872 M * daniel_hozac fixed version up now?
1165776876 M * Bertl yep
1165776891 M * Val rc9 patch apply well
1165777273 M * hardwire heh
1165777279 M * hardwire CoralCDN uses vservers
1165777283 M * hardwire I like that
1165777393 M * waldi Bertl: 2.6.18.5 since one hour
1165777426 M * Bertl what where?
1165777450 M * daniel_hozac Bertl: fs/namei.c:do_lookup, are these checks circumvented if the filesystem has a revalidation function? or am i just misinterpreting again?
1165777491 M * Bertl 2.6.19?
1165777501 M * daniel_hozac (or is it just my tree again...)
1165777514 M * daniel_hozac 2.6.18.
1165777606 M * Bertl which checks?
1165777626 M * Bertl ah, I see what you mean
1165777701 M * Bertl yeah, when d_revalidate is there, the PROC check is not done here
1165777723 M * Bertl (same goes for the normal vx_check())
1165777739 M * daniel_hozac okay.
1165777751 M * daniel_hozac where is it done instead?
1165777775 M * daniel_hozac or isn't it done at all?
1165777775 M * Bertl good question ... probably not at all
1165777816 M * daniel_hozac should we move the checks into the done: branch?
1165777822 M * Bertl won't work
1165777850 M * Bertl (because of the inode checks)
1165777876 M * daniel_hozac what?
1165777888 M * Bertl our checks are inode based
1165777900 M * Bertl done can be called without any relevant inode
1165777940 M * daniel_hozac right... can't we just conditionalize that like it is now?
1165777958 M * daniel_hozac except we rename the current done: to really_done: or something.
1165777965 M * daniel_hozac and have done: do the checks.
1165777987 M * daniel_hozac (if there is an inode)
1165778003 M * Bertl probably, the question is, what cases are currently affected by the check (block) and what checks are missed?
1165778017 M * Bertl *blocked
1165778110 M * daniel_hozac NFS seems to be the only filesystem using d_revalidate which is xid tagged.
1165778160 M * daniel_hozac and for proc, it only seems to be set for the fd and process directories.
1165778173 M * daniel_hozac (which should be protected through other means already)
1165778285 M * Bertl we could pobably move the done up, and let the inode check branch to a done2
1165778339 Q * sebastian Ping timeout: 480 seconds
1165778354 M * Bertl I'll draft up a patch
1165778731 J * sebastian ~info@p54A969ED.dip.t-dialin.net
1165779104 M * Bertl daniel_hozac: something like this? http://vserver.13thfloor.at/Experimental/delta-lookup-fix01.diff
1165779164 M * Bertl but definitely needs some testing
1165779228 M * daniel_hozac yeah.
1165779382 Q * h01ger Quit: h01ger
1165779416 J * h01ger ~holger@socket.layer-acht.org
1165779447 J * Medivh- ~ck@dslb-084-058-254-245.pools.arcor-ip.net
1165779450 M * Medivh- hi
1165779462 M * Bertl hey Medivh-!
1165779526 M * Medivh- i just installed a new vserver machine (running 2.6.18/vs2.0.2.1, daniel_hozac's rpm)... getting some strange error messages here...
1165779535 M * Medivh- _loads_ of "vxW: xid=1 did hit the barrier."
1165779603 M * Medivh- plus stuff like this: http://paste.linux-vserver.org/757
1165779610 M * Medivh- wondering if anyone got an idea? ;)
1165779661 M * Bertl probably daniel_hozac will updated the rpms soon ...
1165779684 M * daniel_hozac that's what i'm working on right now ;)
1165779727 M * Medivh- any problems known with that version then? ;)
1165779789 M * Medivh- i also have the problem that some stuff "crashed" in vservers, i.e. the processes keep running fine, but are unresponsive to anything from the network... usually bind or mysql which doesn't respond any more, maybe that got something to do with the error messages, ionno...
1165779892 M * daniel_hozac well, yeah. the hard CPU scheduler doesn't work, e.g.
1165780431 Q * Johnnie Ping timeout: 480 seconds
1165780657 M * Medivh- could that be causing the messages i'm getting, too?
1165780734 M * Bertl maybe
1165780871 J * DreamerC_ ~dreamerc@61-217-226-202.dynamic.hinet.net
1165780951 M * Bertl Medivh-: what's smaug?
1165781083 M * Medivh- hm, no idea... just checked, customer is running it in her vserver
1165781169 Q * mire Quit: Leaving
1165781281 Q * DreamerC Ping timeout: 480 seconds
1165783352 Q * Borg- Quit: leaving
1165783552 J * brcc bruce@i.am.someasshole.com
1165783959 J * bonbons ~bonbons@83.222.39.117
1165784122 M * daniel_hozac bonbons: have you looked at porting the IPv6 patch to 2.6.19?
1165784138 M * bonbons daniel_hozac: not yet
1165784151 M * daniel_hozac ok.
1165784182 M * bonbons but I may get some time tomorrow, let's see
1165784315 M * Bertl bonbons: if you do, please start with 2.3.x
1165784373 M * bonbons Bertl: ok, will take note of it
1165784512 M * daniel_hozac Bertl: when you're adding headers to Kbuild, add device.h and device_cmd.h as well.
1165784526 M * Bertl already did that
1165784540 J * Borg- borg@aprogas.student.utwente.nl
1165784567 M * daniel_hozac okay.
1165785841 J * Aiken ~james@tooax6-160.dialup.optusnet.com.au
1165786096 Q * sebastian Remote host closed the connection
1165786117 M * Bertl morning Aiken!
1165786313 M * Medivh- hmm, ipv6 patch? there is one?
1165786328 M * Aiken hi Bertl
1165786373 M * daniel_hozac yes.
1165786418 M * daniel_hozac it's been around for something like 6 months now.
1165786424 M * daniel_hozac 5 at the very least.
1165786449 J * raw squid@wwwbox.uni-mb.si
1165786472 M * raw hi, is there a german support channel?
1165786479 M * daniel_hozac no.
1165786596 M * Bertl raw: but I think you will manage the english :)
1165786636 M * raw okay, i'll try in english :) i want to rent a vserver from a ISP. i asked them and they told me that they use linux-vserver and it's not possible (from their side) to get a other guest system then suse 10. i prefer a other linux distribution and want to replace the suse with a my favirote by myself.
1165786636 Q * bonbons Quit: Leaving
1165786658 M * Bertl raw: probably not a real problem
1165786673 M * Bertl raw: do they provide some 'recover all' when you screw up?
1165786720 M * raw no, but i have done this @ home using vmware and a ssh with chroot only.
1165786755 M * raw oh i have read wong :) they provide a free reinstall if i damage my system
1165786766 M * Bertl okay, good
1165786784 M * Bertl in that case, build/upload your distro to a subdir
1165786786 M * raw so i can allways retry
1165786804 M * Bertl then get a busybox or so up
1165786841 M * raw uses vserver the distrubutions kernel? or is it something like a 'shared' kernel?
1165786857 M * daniel_hozac it's shared.
1165786875 M * raw so i must'nt worry about bootloader and kernel, okay :)
1165786927 M * Bertl once you've uploaded your distro and the busybox
1165786935 M * raw and just replace the / filesystem and it should work. must my distribution fill some requirements to work? (i want to use archlinux)
1165786937 M * Bertl simply start the busybox (static compiled)
1165786947 M * raw yes Bertl
1165786957 M * Bertl then move everything into a subdir, and the distro dir into root
1165786965 J * Piet_ hiddenserv@tor.noreply.org
1165786967 M * Bertl after that, restart the guest (remote)
1165786990 M * Bertl make sure that init/runlevel scripts are suse conform
1165787014 M * raw what does suse conform mean?
1165787018 M * Bertl raw: do you already have the guest?
1165787027 M * raw nope
1165787044 M * raw i'll get it tomorrow
1165787047 M * Bertl okay, you can probably figure most thinks when you have access to the guest
1165787053 M * Bertl *things
1165787082 M * Bertl but I wonder, who is only supporting SuSE?
1165787106 M * raw ?h only debian, sorry i was wrong
1165787110 M * raw http://weccos.de
1165787119 M * Bertl ah, okay, debian sounds more like it
1165787151 M * Bertl so what you need to figure once you have the guest is, if it uses sysv or plain init style
1165787160 M * raw what matter do the initscripts for vserver?
1165787175 M * Bertl you can start a guest with or without init
1165787194 M * Bertl in the init-less case, only the runlevel scripts are called
1165787203 M * Medivh- just got a "vxW: pid_task(7879,0): task ffff8100526f8080[#103,7880] did lookup ffff810037fc6040[#0,7879]" again, pid 7879 is a mysqld running in a vserver, now it hangs... could that really be related to the kernel version?
1165787219 M * Bertl Medivh-: yes
1165787224 M * raw ah, i understand
1165787263 M * raw so if the ISP uses init-less i must take care that the init scripts are guest-system-debian-like
1165787293 M * raw (its debian 3.1)
1165787296 M * daniel_hozac i.e. has an /etc/{init,rc}.d/rc script that handles it.
1165787301 M * Medivh- Bertl: i see... any info under what circumstances these probs occur?
1165787313 M * Bertl raw: yes, but then it is probably sufficient to have a rc which starts init
1165787315 M * daniel_hozac Medivh-: which Fedora? 6 i guess?
1165787316 Q * Piet Ping timeout: 480 seconds
1165787327 M * Medivh- daniel_hozac: yep, FC6 on x86_64
1165787405 M * daniel_hozac ok, that might take some time... i'm building FC5 right now.
1165787445 M * Bertl daniel_hozac: do you build inside a guest?
1165787449 M * daniel_hozac yeah.
1165787473 M * daniel_hozac (that's the reason i want the device ACL/mapping stuff ;))
1165787511 M * raw how i can find out which init style is used?
1165787518 Q * Val Quit: zZz bye
1165787521 M * raw (i have a debian 3.1 installed local)
1165787632 M * raw debian uses /etc/rc[0-6].d
1165787642 M * daniel_hozac yes.
1165787761 M * raw archlinux uses just some scripts called: /etc/rc.sysinit, rc.singe rc.multi rc.shutdown and so on
1165787761 M * daniel_hozac Bertl: i think i have a vdevmap ready now... :)
1165787784 M * daniel_hozac (no config support yet though)
1165787821 M * Bertl sounds good, I'm almost done with the kernel side :)
1165787838 M * raw Bertl?
1165787863 M * Bertl raw: well, as I said, you need to provide _a_ script, nothing special
1165787882 M * Bertl raw: maybe you need to map the passed args (runlevel values)
1165787892 M * Bertl raw: but that's it ...
1165787906 M * Bertl (iif it uses sysv init style)
1165787923 M * raw okay...
1165787956 M * raw i'll try my luck, thank you for helping me :)
1165787970 M * Bertl you're welcome! feel free to hang around ...
1165788040 Q * FireEgl Quit: ...
1165788051 J * FireEgl proteus@adsl-61-147-76.bhm.bellsouth.net
1165788055 Q * FireEgl Read error: Connection reset by peer
1165788074 J * FireEgl ~FireEgl@adsl-61-147-76.bhm.bellsouth.net
1165788077 Q * Piet_ Quit: Piet_
1165788093 Q * FireEgl Read error: Connection reset by peer
1165788433 J * marcfiu ~mef@h46058c7c.area7.spcsdns.net
1165788440 M * marcfiu hello
1165788441 M * Bertl hey marcfiu!
1165788450 M * marcfiu back from tokyo... hard trip.
1165788463 M * Bertl but probably interesting ...
1165788467 M * marcfiu how are things with the new network stack & ipv6.
1165788483 M * marcfiu definitely interesting... my first time to asia.
1165788497 M * marcfiu electric city in tokyo is something to experience.
1165788518 M * Bertl yeah? let's hear!
1165788544 M * marcfiu just tons of stores with electronic gadgets...
1165788585 M * marcfiu and all of that mixed together with weird horror / soft-porn animation stuff.
1165788610 M * daniel_hozac lol
1165788632 M * marcfiu then in the back alleys one can pick up ripped "everything".
1165788655 M * marcfiu dvds, software, fake apple shuffle (with Apple logo and copyright)...
1165788667 M * marcfiu a bit overwhelming.
1165788719 M * Bertl yeah, sounds like ...
1165788725 M * marcfiu Didn't have much time to see much more.... arrived Thursday evening and just returned and am waiting for the train.
1165788762 M * marcfiu Was in electric city @ 10:00 in the morning and the hustle and bustle started immediately when most stored opened around 10:30am.
1165788914 M * marcfiu Bertl: while in Tokyo I talked to the same folks who are interested in IPv6.  
1165788935 M * marcfiu They are also interested in full LSM support within a vserver.
1165788943 M * marcfiu Where are things with IPv6?
1165788956 M * Bertl hmm, LSM inside a guest?
1165788975 M * Bertl you probably mean context aware LSM, right?
1165788988 M * Bertl ipv6 is work in progress ... as expected
1165789463 M * marcfiu ok
1165789487 M * marcfiu standing room only on the %$^&^^% train.... time to go.
1165789492 M * marcfiu bye
1165789498 M * Bertl okay, cya
1165789504 P * marcfiu 
1165789863 Q * Blissex Remote host closed the connection
1165790456 J * Johnnie ~jdlewis@jdlewis.org
1165790545 Q * Aiken Quit: Leaving
1165790858 J * Aiken ~james@tooax6-160.dialup.optusnet.com.au
1165791177 J * mire ~mire@104-166-222-85.adsl.verat.net
1165792159 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at
1165792309 M * raw if vserver is used in init-less mode, it just starts the runlevel-3 initscripts right? under /etc/rc3.d/* ?
1165792325 M * Bertl it starts the 'main' rc script
1165792338 M * Bertl (with init like arguments)
1165792351 M * raw where is the main rc script located?
1165792368 M * Bertl that is different between distros, and adjustable
1165792396 M * raw how i can find out?
1165792402 M * Bertl but IIRC, recent tools scan for the script, daniel_hozac has the details
1165792420 M * daniel_hozac /etc/{rc,init}.d/rc
1165792428 Q * Johnnie Ping timeout: 480 seconds
1165792482 M * raw oh good
1165792582 M * raw debian(sysv-style) has a /etc/init.d/rc script. archlinux (bsd-style) not. so i just need to create a /etc/init.d/rc scritp that starts all my runlevel-3-scripts locatet in /etc/rc.d/ right?
1165792643 M * daniel_hozac the script has to understand the argument.
1165792650 M * daniel_hozac the first argument will be the runlevel to enter.
1165792668 M * daniel_hozac on 1-5 you should start your services.
1165792682 M * daniel_hozac on 0 and 6, it's asking you to shutdown.
1165792766 M * raw but there is no 2nd argument
1165792771 J * trigger_si JavaUser@88-110-225-223.dynamic.dsl.as9105.com
1165792781 M * trigger_si this is nuts
1165792783 Q * Medivh- Ping timeout: 480 seconds
1165792787 M * trigger_si how did i get on here
1165792795 M * trigger_si what are we chatting about
1165792804 M * Bertl trigger_si: see topic :)
1165792845 Q * dna Quit: Verlassend
1165792929 M * raw i think copiyng the /etc/init.d/rc script from debian to /etc/rc.d/rc on archlinux and making a symlink from /etc/init.d/ to /etc/rc.d should work, right?
1165792948 M * Bertl maybe, try at home
1165792951 M * daniel_hozac why would it?
1165792970 M * raw the script looks very simple
1165792978 M * daniel_hozac yes, but it requires a certain layout.
1165793022 M * daniel_hozac i.e. /etc/rc<runlevel>.d/S<NN><service> for services to start, and /etc/rc<runlevel>.d/K<NN><service> for services to stop.
1165793042 M * raw yes, and i will try to make this layout using symlinks
1165793073 M * trigger_si where am i
1165793098 M * trigger_si is this place for information techs
1165793117 M * waldi no, for geeks
1165793139 M * trigger_si oo rihgt
1165793141 M * trigger_si right
1165793146 M * trigger_si cool
1165793151 M * trigger_si u know how to hack
1165793165 M * raw i try it @ home :) thanks for help again :)
1165793451 Q * trigger_si Quit: Powered by IBIRC from [http://www.munjmedia.com/]
1165793992 J * mrrm ~urkel@tor-irc.dnsbl.oftc.net
1165794386 Q * mrrm_ Ping timeout: 480 seconds
1165795055 J * zxcv geoff@S01060040f4a589b9.vs.shawcable.net
1165795064 M * Bertl welcome zxcv!
1165795067 M * zxcv hey