1165968298 M * micah h01ger: is there a fixed fai build method available somewhere?
1165968597 M * h01ger no
1165968657 M * h01ger mugwump (mugwamp?) never had time to do it and i never cared,as i just use the debootstrap method and do the simply changes myself... he relies on his branch and couldnt really explain why...
1165968662 M * h01ger & hi micah :)
1165968684 M * h01ger (i also never cared as i had enough other stuff to do..)
1165968732 M * h01ger micah, but as i read -release these days, i'd say its too late for a fix now anyway. unless you make that RC which i dont really think it is...
1165968821 M * micah h01ger: hi! :) -- yeah, without any code at all, its unlikely the release team would be willing for someone to write it
1165968826 M * Bertl hmm, isn't that already integrated somewhere
1165968830 M * micah h01ger: i know I cant write it because I do not know FAI at all
1165968869 M * Bertl yep, vserver - build --help says:
1165968870 M * micah unfortunately the final freeze happened the day after I uploaded .212, so that isn't going to make it in to etch, but the version prior to that has most of the .212 functionality
1165968870 M * Bertl fai         ... -- [ -f <fai_vserver> ] [-n <nfsroot>] [-d <fai_dir> ] [ -a ]
1165968886 M * micah Bertl: yeah, I think the issue is that its broken
1165968888 M * micah let me see...
1165968908 M * h01ger micah, it checks for specical branch of fai...
1165968913 M * h01ger (last time i looked)
1165968918 M * micah Bertl: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=396593
1165968945 M * micah "...vserver-build.fai included in util-vserver only works with a development branch of fai, but not with the fai from trunk or the one in debian."
1165968975 M * micah what is the difference between the 'development branch' and 'trunk'?
1165969018 M * Bertl is there something which works with 'stable' debian out of the box?
1165969026 M * micah and is this change a trivial one? it might be, but I have no knowledge of FAI
1165969043 A * micah ignores Bertl's attempts at debian-baiting ;)
1165969054 Q * comfrey Quit: Lost terminal
1165969070 M * Bertl well, I think, debian folks are used to upgrading/backporting/patching by now ...
1165969072 M * h01ger micah, that development branch is a now outdated branch from mugwump. mugwump (mugwamp?) is not very much active in fai...
1165969101 M * h01ger Bertl, that branch is really a dead end i'd say
1165969104 M * Bertl well, it seems that sam was very interested a few months ago
1165969124 M * Bertl so maybe just contact him and see _what_ he uses atm
1165969137 M * micah h01ger: is it a simple change to something to have the build method not check for a specific version?
1165969137 M * h01ger i failed to understand why he made that check that way... (he touches a file "mybranch" or something and then checks for the existence of it.)
1165969147 M * h01ger micah, maybe
1165969160 M * h01ger i asked him, but he couldnt explain either
1165969178 M * h01ger and then i said, ok, i have my little script anyway, i dont need it...
1165969195 M * h01ger but contacting him is the way to go
1165969236 M * h01ger i think thomas lange (fai author and maintainer) has contacted him today, after i failed this bug. and he should be on the fai devel list, where i cc:ed it too. but mail him as well :)
1165969259 M * h01ger etch wont be released mañana :)
1165969316 A * h01ger sighs at downloads the latest util-vserver to take a look :) i was about to sleep... ;)
1165969322 M * h01ger s/at/and/
1165969350 M * daniel_hozac micah: actually, 0.30.211-6 will be really weird on 2.6.19.
1165969367 M * micah daniel_hozac: thats ok, 2.6.19 is not in etch
1165969380 M * daniel_hozac right, i'm just saying.
1165969381 M * micah unless that has changed
1165969403 M * h01ger micah, but there is a kernel update planned (as in 2.6.23 or something) for the medium lifetime of etch...
1165969405 M * micah daniel_hozac: but 0.30.212 wouldn't have that weirdness?
1165969417 M * micah h01ger: yeah etch+1 right?
1165969423 M * h01ger micah, no. etch+0.5
1165969430 M * micah h01ger: oh right
1165969434 M * daniel_hozac micah: no, with 0.30.212, there shouldn't be any quirks with 2.6.19 .
1165969436 M * h01ger :)
1165969451 M * micah h01ger: well, I need a good excuse to ask aba to let in .212 -- which he might becaues he uses vservers
1165969451 M * h01ger so _that_ would be a reason to go to 0.30.212 :)
1165969457 M * h01ger hehe
1165969460 M * h01ger yes
1165969468 M * micah well they might not let it until +.05
1165969545 M * micah daniel_hozac: the changes between 0.30.211-6 and 0.30.212-1 are pretty small actually
1165969556 M * daniel_hozac but oh so important.
1165969580 Q * lomonaco Quit: If you can't laugh at yourself, make fun of other people.
1165969598 M * micah yeah, but that fact is a useful one for making the case for allowing it in
1165969629 A * h01ger looked at vserver-build.fai and dislikes it: it needs the fai-nfsroot to be build. i dont use that anymore, i (almost) only use fai softupdates,which dont need nfsroots. so i would not use this anyway... and it greps for "VIRTUAL" in /usr/sbin/fai which only succeeds in that branch... (and as said, i didnt get why,looking at the branch now...)
1165969636 M * hardwire oooh creepy
1165969637 M * hardwire vwait
1165969644 M * hardwire I don't like that.. --timeout 30
1165969645 M * hardwire hehe
1165969656 M * hardwire zope isn't a timeout 30 kinda daemon
1165969659 M * hardwire squid isn't either ;)
1165969679 M * daniel_hozac h01ger: he had quite a few changes
1165969697 M * daniel_hozac removing things that just won't work inside a guest.
1165969706 M * daniel_hozac hardwire: that's why it's configurable.
1165969798 M * hardwire I just never knew about it
1165969841 M * h01ger daniel_hozac, yes (but thats rather cosmetical). but he also updated his branch the last time almost two month ago and *a lot* has happened since them in fai. micah, so i'd say, drop it.
1165969875 M * daniel_hozac h01ger: IIRC there were quite a few things that would make it fail.
1165969893 M * h01ger actually, mugwump has three development branches. and the string VIRTUAL (the one he greps for) is only in the one last updated 7th of july (!!)
1165969921 M * h01ger daniel_hozac, no. my setup-script doesnt do any such modifications and my vservers run fine
1165969934 M * daniel_hozac h01ger: because you run it on the host, i presume.
1165970000 M * h01ger yes. but i think he does the same, dunno.
1165970012 M * daniel_hozac no, it runs mostly within a guest.
1165970026 M * h01ger   [ -d /proc/self -a ! -e /proc/cmdline ] && export VIRTUAL=1      is what sets VIRTUAL in his /usr/sbin/fai
1165970117 M * h01ger daniel_hozac, actually no. i only run vserver-build on the host, then i prepare a script and execute it in the newly created guest
1165970211 M * daniel_hozac mugwump: ping?
1165970213 M * hardwire daniel_hozac: wild.. guess how long it took me to find out how vwait works.. 
1165970216 M * hardwire still.. looking.. 
1165970240 M * daniel_hozac the name is rather descript, IMHO.
1165970241 M * hardwire I suppose its sync-timeout for vshelper
1165970260 M * hardwire daniel_hozac: like what calls it.. etc.. 
1165970271 M * daniel_hozac the scripts, as with all the other utils.
1165970300 M * hardwire daniel_hozac: not trying to make anybody defensive.. but its just not intuitive if you aren't developing it.
1165970329 M * hardwire I still don't know what vwait actually does
1165970334 M * daniel_hozac so you know what wait does?
1165970345 M * hardwire ah.
1165970354 M * daniel_hozac vwait does the same, but for an entire context.
1165970357 M * hardwire I didn't make that connection by just adding a v to it.
1165970376 M * hardwire either way.. should I be editing /etc/vservers/vserver-name/apps/vshelper and changing sync-timeout?
1165970384 M * daniel_hozac yes.
1165970390 M * h01ger micah, mugwumps code is based on fai 2.10.1, we have fai 3.1.2 now. there where quite some changes since same (for example also a new fai install method, dirinstall, which would be useful for vservers as well). so whoever wants to use that vserver-build method whould need to downgrade&patch fai... so i would rather prefer/propose, mugwump (or we) fix vserver-build.fai and then those people need to upgrade util-vserver :)
1165970411 M * hardwire daniel_hozac: did your documentation guru get hit by a bus?
1165970442 M * daniel_hozac feel free to take his place.
1165970461 M * hardwire I am not prepared to document something I am not currently developing
1165970469 M * hardwire but did somebody just dissapear?
1165970473 M * daniel_hozac no.
1165970482 M * daniel_hozac there has never been anyone doing documentation, AFAIK.
1165970496 M * Bertl nobody likes to write documentation here :)
1165970499 M * daniel_hozac a lot of people have been pitching in here and there, some more than others.
1165970510 M * daniel_hozac indeed.
1165970522 M * daniel_hozac personally i don't mind keeping it up to date, once it exists.
1165970522 M * Bertl some of them have complained, but didn't improve anything ...
1165970526 M * daniel_hozac (like the flower page)
1165970550 M * daniel_hozac but i'm too deep in the code to make sense to any outsider, i think.
1165970594 M * Bertl and I did my best with the original Paper :)
1165970621 M * h01ger .oO( add a fancy "job offer" on the homepage ?! :)
1165970628 M * daniel_hozac hehe.
1165970642 M * hardwire daniel_hozac: I suffer personally with an issue to never document anything until I feel it is complete
1165970648 M * h01ger i didnt really know that was the case, though *now* it's obvious to me...
1165970654 M * hardwire that doesn't work in the open source world
1165970739 M * daniel_hozac i wonder if anyone has used util-vserver 0.30.212 yet... i figured people would be screaming at the top of their lungs when i broke their schedule files.
1165970827 M * Bertl works here for my tests ... but I haven't tested it much yet
1165970862 M * daniel_hozac yeah, should work for everyone but those who have dummy as their 6th line in schedule :)
1165970885 M * Bertl never added that, as it just looked stupid to me :)
1165970913 M * daniel_hozac it is. it's passed as the priority bias.
1165970934 M * daniel_hozac and since vsched used to just do x = atoi(optarg), that worked fine.
1165971042 M * h01ger g'night
1165971142 M * Bertl good night!
1165971324 M * kugg daniel_hozac: yepp it worked out fine and i learned a lot. :) Thanks for asking. Unfortunatly nobody finished the entire game, maybee its the wrong audince to little time and to hard levels. :/
1165971397 M * daniel_hozac hehe, that's too bad.
1165971423 M * kugg yepp, well i guess 10 levels could be to much too. 
1165971460 M * kugg I would hoverver like to trash the idea of level based wargame and make it more of a get root challenge
1165971472 M * kugg like king of the hill
1165971678 M * daniel_hozac hehe.
1165971695 M * daniel_hozac "whoever has root by <time> wins!"
1165971710 M * kugg well yea that would be cool i think
1165971751 M * Arcticfox what was your user name again *clicky* *ty* *click*
1165971825 J * derjohn2 ~aj@dslb-084-058-210-230.pools.arcor-ip.net
1165972041 M * hardwire are options for the util-vserver configuration files space delimited or = delimited?
1165972041 M * daniel_hozac which file?
1165972041 M * hardwire apps/vshelper
1165972041 M * daniel_hozac uh, that's a directory.
1165972051 M * hardwire /etc/vservers/vserver-name/apps/vshelper  
1165972056 M * daniel_hozac right.
1165972065 M * hardwire haha
1165972080 M * hardwire http://www.nongnu.org/util-vserver/doc/conf/configuration.html
1165972092 M * hardwire http://oldwiki.linux-vserver.org/VServerConfiguration
1165972103 M * hardwire it doesn't really appear to be a directory
1165972108 M * hardwire rather a file with options in it
1165972121 M * hardwire totally confusing
1165972122 J * comfrey ~comfrey@201.243.174.242
1165972123 M * hardwire but that answers it
1165972146 M * daniel_hozac the same layout is used consistently throughout the document.
1165972161 M * hardwire true
1165972174 M * daniel_hozac the fact that /etc/vservers/<vserver-name>/apps looked exactly the same way should be quite a strong hint.
1165972175 M * hardwire your right
1165972180 M * hardwire somebody needs to rewrite the documentation
1165972205 M * daniel_hozac why's that?
1165972225 M * Arcticfox the best doc is in the source
1165972227 M * Arcticfox *runs*
1165972229 M * hardwire your right
1165972248 M * hardwire daniel_hozac: I have nothing in my apps dir by default
1165972255 M * daniel_hozac and you shouldn't.
1165972268 Q * derjohn3 Ping timeout: 480 seconds
1165972303 M * hardwire I searched for timeout.. and it lead me to what appeared to be a file name
1165972315 M * hardwire had I rewound and read the entire thing I would be more in the know
1165972325 M * daniel_hozac sync-timeout is a filename.
1165972349 M * hardwire vshelper appeared to be a file having sync-timeout as a variable listed in it.
1165972372 M * hardwire you are right that interfaces is all one key per variable syntax.
1165972380 M * hardwire same with the name/etc.. 
1165972385 M * daniel_hozac feel free to write a stylesheet that list the file type right after the name.
1165972414 M * hardwire daniel_hozac: whats going on.. are you really this defensive about it.. can't you laugh about it with me?
1165972438 M * daniel_hozac you seem to just be complaining with absolutely no interest in fixing it
1165972468 M * daniel_hozac patches are accepted, you know.
1165972490 M * hardwire I have a lot of interest in fixing things with this product
1165972509 M * daniel_hozac sure don't show it ;)
1165972519 M * hardwire I am left with a feeling that I should have been smart and read the docs whenever I ask a question that you answer
1165972523 M * hardwire except.. 
1165972526 M * hardwire yeh.
1165972550 M * hardwire I have to run to the chiro
1165972550 M * hardwire bbl
1165973211 M * Arcticfox daniel_hozac: i got the previous issue resolved it was a typo from my side
1165973231 M * daniel_hozac the yum problem?
1165973245 M * Arcticfox no the eth
1165973333 M * Arcticfox i got another question though ..
1165973347 M * Arcticfox if i use eth0.32 everything is fine .. if i use vlan32 ...
1165973350 M * Arcticfox i get 
1165973382 M * Arcticfox creation of VLAN_PLUS_VID_NO_PAD devices is not supported
1165973403 M * Arcticfox create them before starting the vserver and use the 'nodev' flag then
1165973464 M * daniel_hozac ah, right.
1165973470 M * Arcticfox it works 100% both ways but i get the error on the later
1165973471 M * daniel_hozac i've forgotten to restructure that logic.
1165973591 M * Arcticfox i tried the same setup i run vserver in vmware .. brought the system to a standstill
1165973601 M * Arcticfox running 24 virtual hosts 
1165973616 M * daniel_hozac hehe.
1165973617 M * Arcticfox with some services.. like apache/mysql + X on some
1165973629 M * Arcticfox then i ran some simple benchmark tests
1165973631 M * Arcticfox *boom*
1165973665 M * daniel_hozac to be fair though, that's quite an apples and oranges comparison.
1165973684 M * Arcticfox i know but i mean for a linux on linux comparison
1165973935 J * Aiken_ ~james@tooax6-040.dialup.optusnet.com.au
1165973981 M * daniel_hozac would it be possible for you to test a fix to the vlan* problem?
1165974009 M * Arcticfox sure.. im actually running stock debian but i can just recompile their pkg
1165974010 M * Bertl morning Aiken_!
1165974042 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/delta-vlan-fix01.diff
1165974080 M * Arcticfox btw we are trying to add vserver support into the dist im involved with (arklinux) not really emphasis on server atm but desktop .. but i saw some usage for something like vserver there too
1165974125 M * Aiken_ hi Bertl
1165974127 M * daniel_hozac like moreubuntu? or something else?
1165974139 M * Aiken_ maybe getting somewhere with that driver issue
1165974149 M * Arcticfox no its build up from 100% scratch 
1165974170 M * daniel_hozac i meant more in the vserver use-cases on the desktop area.
1165974198 M * daniel_hozac (for the uninitiated, it's about multiple seats on one computer, with different distributions)
1165974199 M * Arcticfox yes and no .. i was thinking more for integrated things like webkiosks etc
1165974247 M * daniel_hozac how do you figure?
1165974257 Q * comfrey Ping timeout: 480 seconds
1165974264 Q * Aiken Ping timeout: 480 seconds
1165974271 M * Arcticfox daniel_hozac: it worked 100%
1165974274 M * Arcticfox your patch
1165974296 M * daniel_hozac ok, thanks.
1165974341 M * Arcticfox i can write a paper on it when i get there
1165974344 M * Arcticfox :)
1165974472 M * doener Bertl, daniel_hozac: Now that my Wii-arm is almost cured, I'm going back to the patch tracking thingy... One question: Would you prefer the patches to be placed on the box running the web-frontend, or just remote urls stored there? (CLI client would allow to store a prefix like "http://vserver.13thfloor.at/Experimental/" that can be automagically added in that case)
1165974621 M * daniel_hozac hmm, i can't decide which i would prefer. they both have their own sets of benefits.
1165974755 M * daniel_hozac so whichever you decide is fine by me.
1165974934 M * daniel_hozac i'm going to get some sleep now though, good night all!
1165975091 M * kugg god night!
1165975288 M * doener *lol* had hit Ctrl+s accidently and wondered we it took you so long to answer ;) Until I tried to type something *g*
1165975299 M * Arcticfox hehe
1165975303 M * Arcticfox been there
1165975323 M * doener happens about twice a week to me, should be used to it by now
1165975574 M * Arcticfox time to zz here
1165975598 M * doener good night
1165976030 Q * daniel_hozac Server closed connection
1165976041 J * daniel_hozac ~daniel@c-2c1472d5.010-230-73746f22.cust.bredbandsbolaget.se
1165977585 J * bronson ~bronson@adsl-75-36-147-248.dsl.pltn13.sbcglobal.net
1165978864 M * Bertl doener: hmm, I think a combination would be optimal
1165978906 M * Bertl i.e. in my case, most of the patches are local, so I'd like to feed them directly into the tracker
1165978921 M * Bertl but sometimes, we would like to add a 'remote' patch too
1165978938 M * Bertl (by just pointing the client there)
1165978955 M * Bertl but I'm perfectly fine if _all_ patches are stored as a copy
1165978980 M * Bertl in the tracker, with nice patch/diff coloring and annotation *hint*
1165979043 M * Bertl i.e. it would be really cool, if the patch would actually contain all comments and/or data/history at the beginning, and of course, automatic on-the-fly compression/decompression would be nice too ..
1165979107 M * Bertl most of that is already done on 13thfloor.at so integrating it there would be an option (but no requirement)
1165979183 M * doener as I'm going for Python and we're already using trac anyway, I planned to look into integrating its patch highlighting (if possible)
1165979309 M * Bertl okay, that would be cool
1165979860 J * RichyF ~RIchy@82-32-120-219.cable.ubr04.hawk.blueyonder.co.uk
1165979867 M * RichyF hi guys
1165979878 M * Bertl hey
1165980091 Q * dmax Ping timeout: 480 seconds
1165980171 Q * s0undt3ch Ping timeout: 480 seconds
1165980301 M * ntrs_ daniel_hozac, ping?
1165980629 J * dmax ~semaj@bl4-56-217.dsl.telepac.pt
1165980683 Q * Piet_ Ping timeout: 480 seconds
1165980741 J * s0undt3ch ~s0undt3ch@bl4-56-217.dsl.telepac.pt
1165987160 M * mugwump 13:10 < h01ger> mugwump (mugwamp?) never had time to do it and i never cared,as i just use the debootstrap method and do the simply changes myself... he relies on his  branch and couldnt really explain why...
1165987175 M * mugwump as I recall you once said "yes, that should be the default" after I explained it :)
1165987203 M * mugwump anyway, the reason to do it the way in the vserver branch is to make the vserver builds and the normal fai builds reproducable
1165987212 M * mugwump both methods have their advantages and disadvantages
1165987242 M * mugwump I've merged my changes both with 2.10.5 and trunk, but won't be able to push for a bit
1165987249 M * mugwump (on the road in Tasmania atm)
1165987405 M * Bertl hey mugwump! how u doing?
1165987599 Q * glut Server closed connection
1165987625 J * glut glut@no.suid.pl
1165988545 J * Zaki ~Zaki@88.213.58.243
1165989030 Q * Zaki Ping timeout: 480 seconds
1165989823 J * Zaki ~Zaki@88.213.58.243
1165989929 Q * Zaki Remote host closed the connection
1165989987 J * Zaki ~Zaki@88.213.58.243
1165989999 J * marcfi1 ~mef@c-68-39-177-97.hsd1.nj.comcast.net
1165990086 M * marcfi1 what's the delta between rc8 and rc9 for 2.0.2.2
1165990101 Q * Zaki Read error: Connection reset by peer
1165990136 M * Bertl hmm, sec
1165990151 J * Zaki ~Zaki@88.213.58.243
1165990208 M * Bertl marcfi1: removal of the __enter/leave entries
1165990506 Q * Zaki Remote host closed the connection
1165990739 J * Zaki ~Zaki@88.213.58.243
1165991402 Q * Zaki Quit: Leaving
1165991431 J * Zaki ~Zaki@88.213.58.243
1165991743 Q * s[e]bastian 
1165992633 J * dreamind ~dreamind@C2107.campino.wh.tu-darmstadt.de
1165992755 M * dreamind Hi
1165992759 M * Bertl hey
1165992781 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1165992888 Q * cdrx 
1165993611 M * dreamind Hi Bertl :)
1165993882 M * Zaki morning Bertl DreamerC 
1165993891 M * Zaki dreamind
1165993902 M * dreamind hi Zaki 
1165993919 A * dreamind needs some coffee ;)
1165993947 A * Zaki pours some coffee for dreamind
1165993956 M * dreamind thx :D
1165994832 J * FireEgl ~FireEgl@adsl-61-147-76.bhm.bellsouth.net
1165995075 Q * Zaki Ping timeout: 480 seconds
1165995200 J * meandtheshell ~markus@85-124-37-79.dynamic.xdsl-line.inode.at
1165995546 Q * thunder1 Remote host closed the connection
1165996498 Q * shedi Quit: Leaving
1165996829 J * Zaki ~Zaki@88.213.57.72
1165997494 J * dna ~naucki@52-224-dsl.kielnet.net
1165999366 J * kir ~kir@swsoft-mipt-nat.sw.ru
1165999468 J * dxTHVBBtk ~hollow@styx.xnull.de
1165999611 Q * Hollow Read error: Connection reset by peer
1165999693 T * Bertl http://linux-vserver.org/ | latest stable 2.0.2.1, 2.0.2.2-rc9, 2.2.0-rc5, devel 2.1.1.6, 2.3.0.4, 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 ;)
1165999773 M * Bertl okay, off to bed now .. have a good one everyone!
1165999778 N * Bertl Bertl_zZ
1166000330 Q * kir Remote host closed the connection
1166000873 M * harry btw. 2.6.19.1 is out
1166002295 M * h01ger mugwump, first of all, enjoy tasmania :) i still think that your method is the way to go once its ready, but i also do think it should not rely on the fai-nfsroot being build, because its unneeded. and for now its important what to do in regards of etch...
1166002324 M * h01ger (mugwump and sorry for mixing up your name a bit, i was tired...)
1166002334 M * h01ger +moin moin to all, too
1166003489 N * dxTHVBBtk Hollow
1166004034 M * waldi okay, I have to reject -rc9 for the update
1166004041 M * waldi changes in core structs
1166004084 M * h01ger uh
1166004405 M * waldi i did not properly check it yet
1166005056 J * TheSeer ~theseer@border.office.salesemotion.net
1166005101 J * djrise ~djrise2b@109-125.252-81.static-ip.oleane.fr
1166005110 M * djrise hi everyone
1166005404 Q * TheSeer Remote host closed the connection
1166005647 J * TheSeer ~theseer@border.office.salesemotion.net
1166005809 M * djrise i have a question about high avaibility on vserver
1166005817 M * djrise does anyone use that ?
1166006362 M * renihs ?
1166006447 Q * harry Ping timeout: 480 seconds
1166006794 M * anonc djrise - yeah - vservers on drbd managed by heartbeat
1166006867 M * cehteh mhm ... running the 2.6.19.1 patch over 2.6.19-vs2.2 ... the bets are open, will it apply?
1166006918 M * djrise anonc: cool
1166006938 M * djrise so a question
1166006979 M * djrise I have a vserver 
1166006981 M * djrise with postfix
1166007024 M * djrise the mails is in /var/mail mounted with nfs
1166007177 M * djrise does i need drbd to do a failover with another vserver on another machine or only ha ?
1166007810 Q * TheSeer Ping timeout: 480 seconds
1166007954 J * witchdoc ~witchdoc@3ece08a5.datahighways.de
1166007962 M * witchdoc hello folks
1166008139 M * witchdoc i use debian/etch with 2.6.18.5-vs2.0.2.2-rc9 and i mark my vserver with default, but the want start on boot :-/
1166008142 M * witchdoc cat /etc/vservers/dj01/apps/init/mark
1166008143 M * witchdoc default
1166008164 M * cehteh ... patch applied! besides the usualy makefile.rej
1166008193 M * cehteh mhm
1166008200 M * cehteh witchdoc: manual start works?
1166008213 M * cehteh vserver dj01 start
1166008213 J * DavidS ~david@85.125.165.34
1166008354 Q * akaihola Quit: Leaving.
1166008362 J * harry ~harry@d54C2508C.access.telenet.be
1166008473 M * renihs >>> Emerging (1 of 551)
1166008477 M * renihs hehe small update
1166008517 M * renihs ups
1166008519 M * renihs wrong channel
1166008629 M * Arcticfox hehe
1166008648 M * Arcticfox witchdoc: works 4 me
1166008846 M * witchdoc cehteh: yes
1166008860 M * cehteh mhm 
1166008861 M * witchdoc hmm strange ... after i compiled the kernel static
1166008869 M * witchdoc it works, but not for all
1166008881 M * witchdoc and i cp the mark to the other, it goes ...
1166008893 M * witchdoc grml -- "look ma, a noob...."
1166008901 M * cehteh maybe some typo in interfaces? :)
1166008909 M * cehteh fstab .. 
1166008946 M * witchdoc i will check this. its good to know thats not an generic problem with etch 
1166009059 M * daniel_hozac waldi: uh, that's to fix an off by one error that causes oopses...
1166009099 M * waldi daniel_hozac: where?
1166009110 M * daniel_hozac where what?
1166009126 M * daniel_hozac cehteh: you know where are patches for 2.6.19.1, right? :)
1166009160 M * cehteh daniel_hozac: i just build yesterday a 2.6.19-vs and wanted to install it this morning
1166009196 M * cehteh and well .. the .1 patch applied to my source tree so no problem
1166009197 M * daniel_hozac cehteh: yeah, Bertl_zZ uploaded patches against 2.6.19.1 with some proc fixes earlier this morning ;)
1166009215 M * cehteh uhm
1166009216 Q * Aiken_ Ping timeout: 480 seconds
1166009222 M * cehteh are the proc fixes important?
1166009235 M * cehteh still compiling ...
1166009252 M * waldi daniel_hozac: i'm not able to reconstruct the diff
1166009283 M * daniel_hozac reconstruct what diff?
1166009333 M * waldi between -rc8 and -rc9
1166009476 M * waldi the result is the following: http://stats.buildserver.net/fetch.php?&pkg=linux-2.6&ver=2.6.18-9%7Esnapshot.7994&arch=amd64&stamp=1165970481&file=log&as=raw
1166009534 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-2.0.2.2-rc8-rc9.diff
1166009606 M * daniel_hozac ntrs_: pong
1166009783 M * daniel_hozac waldi: well, i guess it's up to you if you want kernels that crash when you stop guests or not :)
1166011088 J * TheSeer ~theseer@border.office.salesemotion.net
1166011260 Q * ensc Killed (NickServ (GHOST command used by ensc_))
1166011270 J * ensc ~irc-ensc@p54B4E640.dip.t-dialin.net
1166011314 Q * dmax Quit: changing servers
1166011350 J * dmax ~semaj@bl4-56-217.dsl.telepac.pt
1166011878 M * brcc w
1166012527 M * h01ger daniel_hozac, does vs2.0.2.2-rc2 also have the problems with chrashing when stopping guests?
1166012577 M * daniel_hozac it shouldn't.
1166012625 M * h01ger thanks for the info/educated guess :)
1166012639 M * daniel_hozac it was introduced in rc7.
1166012670 M * h01ger s/educated guess// :)
1166012875 J * bronson_ ~bronson@adsl-75-36-147-248.dsl.pltn13.sbcglobal.net
1166012879 Q * bronson Quit: Ex-Chat
1166012894 Q * TheSeer Remote host closed the connection
1166013073 J * lilalinux ~plasma@80.69.41.2
1166014163 M * waldi daniel_hozac: hmm, is _vx_usage_stat used in core structs?
1166014225 Q * cehteh Quit: Terminated with extreme prejudice - dircproxy 1.0.5
1166014367 M * daniel_hozac waldi: probably, but that isn't changed.
1166014450 M * daniel_hozac waldi: _vx_cacct is changed though, and that's part of the vx_info, which task_struct has a pointer to.
1166014479 Q * djrise Remote host closed the connection
1166014510 Q * dreamind Quit: dreamind
1166014648 P * marcfi1 
1166015232 M * waldi daniel_hozac: okay, that is bad
1166015331 J * djrise ~djrise2b@109-125.252-81.static-ip.oleane.fr
1166015350 M * mnemoc hi, has anyone tried cowloop for vservers?
1166015428 M * daniel_hozac mnemoc: loopback with COW?
1166015441 M * daniel_hozac mnemoc: anything wrong with the COW in 2.1+?
1166015812 Q * duckx Quit: Client exiting
1166015831 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1166015833 J * duckx ~Duck@tox.dyndns.org
1166015894 M * mnemoc daniel_hozac: nothing, but it's at 2.1+
1166015929 M * daniel_hozac well, given that 2.0 won't be ported to 2.6.19+, i don't really see that as a problem.
1166015939 M * mnemoc outch
1166016756 M * mnemoc daniel_hozac: on 2.1+'s COW, how are the updates on the "master copy" (read-only one) handled?
1166016772 M * daniel_hozac there is no master copy.
1166016821 M * mnemoc so no way to update all the installations?
1166016844 M * daniel_hozac no.
1166016859 M * mnemoc :\
1166016872 M * mnemoc any suggestion?
1166016923 M * daniel_hozac hashification still works.
1166016951 M * mnemoc so update all of them and rehash?
1166016955 M * daniel_hozac sure.
1166016980 M * daniel_hozac you could temporarily remove iunlink+immutable and update the file, but it wouldn't be completely safe.
1166017011 M * daniel_hozac i.e. a guest could open it in write mode and then write some malicious stuff to it.
1166017150 M * mnemoc that's why i like the approach of using a master copy as read-only "layer" ... but unionfs doesn'T scale
1166017166 M * mnemoc i got lot of kernel panics when i tested that :\
1166017192 M * mnemoc it's only trivial to see what really belongs to the guest
1166017198 M * mnemoc and backup
1166017261 M * DavidS mnemoc: hard-linking backup programs like backuppc should work fine with hasified vservers ...
1166017262 J * TheSeer ~theseer@border.office.salesemotion.net
1166017392 M * harry /usr/local/sbin/vserver: line 669: popd: /tmp/vserver-init.wRUXRm: No such file or directory
1166017395 M * harry wtf?
1166017413 M * daniel_hozac version?
1166017512 M * harry 210 it seems
1166017518 M * harry upgrading as we speak :)
1166017675 M * harry make: *** [.limbak.stamp] Error 255
1166017677 M * harry wtf?
1166017717 Q * brc_ Ping timeout: 480 seconds
1166017837 M * harry daniel_hozac: any idea where that comes from?
1166017845 M * daniel_hozac start-vservers?
1166017856 M * daniel_hozac sounds like vserver limbak start/stop failed.
1166017863 M * harry yups
1166017868 M * harry it's the same for all vservers
1166017893 M * harry i run the vserver-default start
1166017913 M * harry (as i want my vservers to start at boot :))
1166017917 M * harry which they don't now :S
1166017951 M * daniel_hozac does vserver limbak start work?
1166017959 M * harry i'll check in a minute
1166017962 M * harry but i think so
1166017982 M * harry no it doesn't
1166017990 M * harry it did with 210 version of the tools
1166018005 M * harry vsched: non-numeric value specified for '--priority_bias'
1166018021 M * daniel_hozac there you go then.
1166018034 M * harry ?
1166018052 A * harry did nothing!
1166018062 M * harry it worked with previous tools
1166018065 M * harry doesn't work anymore
1166018065 M * daniel_hozac either convert to the split out scheduler configuration, or put a valid priority bias in your schedule file.
1166018068 A * harry wants working server :)
1166018081 M * daniel_hozac because the older tools didn't check their input, at all.
1166018083 M * harry is there a manual?
1166018094 M * daniel_hozac the great flower page, you mean?
1166018095 M * harry how do i ckeck stuff
1166018101 M * harry aha, it's all there...
1166018113 M * harry is there a link on the website?
1166018145 M * daniel_hozac to the flower page?
1166018152 M * harry yes
1166018158 M * daniel_hozac there should be, somewhere.
1166018177 A * harry looking at documentation section on linux-vserver site
1166018187 M * daniel_hozac probably easier to just go to http://www.nongnu.org/util-vserver/doc/conf/configuration.html
1166018203 M * daniel_hozac Bertl_zZ: i'm of the opinion that 2.1.1.6 is broken.
1166018216 M * harry would be nice for people to be able to know where to find docs
1166018221 M * harry starting from the website, wouldn't it
1166018234 M * daniel_hozac so, add a link.
1166018297 M * harry brrrr... i have to split up my scheduler info!
1166018317 M * daniel_hozac _or_ (for now) use a valid priority bias.
1166018336 M * daniel_hozac 0.30.213+ will remove support for the schedule file though, so probably better to upgrade now.
1166018344 M * harry hmm...
1166018362 M * harry so i need a sched dir now
1166018369 M * harry containing those files
1166018386 M * harry what's the advanced idle time ?
1166018390 M * daniel_hozac just the once you need.
1166018394 M * harry advance idle time
1166018399 M * daniel_hozac s/once/ones/ sigh.
1166018402 M * harry ;)
1166018408 M * witchdoc bye 
1166018411 Q * witchdoc Quit: ...
1166018426 M * daniel_hozac idle time should be explained in the scheduler documentation somewhere..
1166018429 M * harry which schedule options are "required"
1166018471 M * daniel_hozac technically, none.
1166018488 M * harry hmm... /me checks
1166018511 M * daniel_hozac but the ones from the schedule file are: fill rate, interval, tokens, tokens_min, tokens_max, and priority bias
1166018576 M * blizz is it possible that a context can determine its own rss/vss memory usage?
1166018586 M * harry bilbo:/usr/local/etc/vservers/limbak# echo 1 > sched/fill-rate
1166018586 M * harry bilbo:/usr/local/etc/vservers/limbak# echo 1 > sched/interval
1166018586 M * harry bilbo:/usr/local/etc/vservers/limbak# echo 10000 > sched/tokens
1166018586 M * harry bilbo:/usr/local/etc/vservers/limbak# echo 20000 > sched/tokens-max
1166018586 M * harry bilbo:/usr/local/etc/vservers/limbak# echo 100 > sched/tokens-min
1166018588 M * harry that makes sense?
1166018594 M * daniel_hozac no.
1166018597 M * harry ?
1166018607 M * daniel_hozac 1/1 means 100%.
1166018611 M * harry true
1166018615 M * harry it's a multiprocessor system
1166018616 M * harry ;)
1166018616 J * ntrs__ ~ntrs@68-188-55-120.dhcp.stls.mo.charter.com
1166018620 M * blizz daniel_hozac, answer to who? :P
1166018620 Q * ntrs_ Read error: Connection reset by peer
1166018621 M * harry i want it to get 1 cpu
1166018651 M * daniel_hozac 100% means all of all the processors.
1166018660 M * harry ?
1166018669 M * harry it used to be different, right?
1166018671 M * daniel_hozac http://wiki.linux-vserver.org/CPU_Scheduler
1166018675 M * daniel_hozac no.
1166018727 M * harry grmbl... someone told me that 100% was 100% of 1 cpu!
1166018729 M * daniel_hozac blizz: virt_mem and rss limits should let it see that.
1166018774 M * daniel_hozac (where that == the amount of used RSS)
1166018867 M * blizz are those proc files?
1166018891 M * daniel_hozac no, virt_mem is a context flag, and the rss limit is an rlimit.
1166018950 M * harry YESH
1166018953 M * harry it works
1166018955 A * harry kisses daniel_hozac 
1166019007 M * matti Oh my.
1166019016 M * matti harry, daniel_hozac: Get a room!
1166019016 M * matti ;p
1166019044 M * blizz and make it +p :-p
1166019059 M * matti ;-p
1166019709 M * djrise Can i talk with someone who are using ha clustering ?
1166020800 M * harry djrise: with vserver?
1166020810 M * djrise yes
1166020814 M * harry then: not me :)
1166020822 M * harry i just sync servers to DR site every night :)
1166020823 M * djrise ;-)
1166020829 M * harry if things go down, i start them remote
1166020839 M * harry (bad experiences with HA stuff ;))
1166020849 M * djrise yes but if your are in week end 
1166020851 M * djrise !
1166020868 M * harry then things just have to work
1166020893 M * harry if they don't they better pay me a lot, and then i'll fix it with gprs -> cellphone -> bluetooth -> laptop
1166020911 M * harry ;)
1166020917 Q * tokkee Server closed connection
1166020918 J * tokkee tokkee@casella.verplant.org
1166021348 M * harry Starting vservers of type 'default'.../usr/local/sbin/vserver: line 669: popd: /tmp/vserver-init.CE70aG: No such file or directory
1166021367 M * harry daniel_hozac: what can i do about this error in specific ? (it seems i get this error on all systems for some reason)
1166021390 M * daniel_hozac still 0.30.210?
1166021391 M * harry manual starting works fine
1166021395 M * harry that one is 211
1166021421 M * daniel_hozac you might want to try 0.30.212, where that should be fixed.
1166021429 M * harry i know it's fixed there
1166021437 M * harry but how comes it doesn't work in 211 ???
1166021443 M * harry it used to work (afaik)
1166021464 M * harry just upgrading all my systems is not what i allways want to do :S
1166021527 M * daniel_hozac it happens when the cwd goes away, for whatever reason.
1166021550 M * daniel_hozac usually it's due to namespace cleanup and starting guests from a mount that will be cleaned.
1166021561 M * harry if i change to 212 now... that means upgrading ALL my schedule things ?
1166021576 M * daniel_hozac if you have more broken files, yes.
1166021585 M * harry i have no broken files imho
1166021591 M * daniel_hozac obviously you do.
1166021595 M * daniel_hozac or you wouldn't get the error.
1166021600 M * harry can you tell me HOW they are broken?
1166021603 M * harry WHAT's broken?
1166021606 M * daniel_hozac vsched did.
1166021609 M * harry how do i know what's broken
1166021614 M * daniel_hozac --priority-bias wasn't an integer.
1166021623 M * harry i have a file: schedule
1166021633 M * daniel_hozac which means that your sixth line in the schedule file isn't an integer.
1166021640 M * harry true, it's "dummy"
1166021647 M * daniel_hozac there you go then.
1166021743 M * harry but that doesn't make my 211 tools work
1166021754 M * daniel_hozac ?
1166021829 M * harry i don't get it... a few days ago 211 tools seemed to work fine
1166021833 M * harry now they don't :S
1166021838 M * harry nothing changed afaik
1166021843 M * daniel_hozac so you changed something.
1166021851 M * daniel_hozac you're using vanilla 0.30.211?
1166021874 M * harry yeah
1166021877 M * harry well...no :)
1166021881 M * harry ns-cleanup :)
1166022233 Q * sc0tt Server closed connection
1166022236 J * sc0tt ~scott@209.51.169.84
1166022349 M * daniel_hozac there you go.
1166024584 M * derjohn daniel_hozac, devel 2.1.1.6 ?
1166024592 M * daniel_hozac what about it?
1166024596 M * derjohn you folks get a really fast release rate ....
1166024602 M * derjohn *got
1166024612 M * derjohn is there a 2.1.1.7 in sight ?
1166024615 M * daniel_hozac yes.
1166024620 M * daniel_hozac IMHO 2.1.1.6 is broken.
1166024629 M * daniel_hozac i don't know if it'll be fixed in place, or a 2.1.1.7 released.
1166024661 M * Medivh daniel_hozac: my problems seem to have disappeared after the kernel upgrade, all running fine since 1,5 days now
1166024672 M * daniel_hozac Medivh: awesome! thanks.
1166024697 M * derjohn then I wait a little more until I build a 2.6.19/2.1.1.x ... currently my archives provide 2.6.18.5/ 2.1.1.3, and I stick to them ...
1166024720 M * derjohn until 2.1.1.>=6 gets a little more stable ... 
1166024746 M * derjohn does the v6 patch work with  2.1.1.>=6 ?
1166024761 M * Medivh daniel_hozac: are the packages officially released yet? looking for the kernel-devel package... ;)
1166024762 M * daniel_hozac bonbons hasn't released one for 2.6.19, AFAIK.
1166024775 M * daniel_hozac Medivh: yeah, they're released.
1166024783 M * derjohn and the old one doesnt fit ?
1166024789 M * daniel_hozac no.
1166024798 M * daniel_hozac though i haven't looked at the rejects myself yet.
1166024813 M * derjohn well, then it doesnt fit in derjohn's release line at all ..
1166024860 Q * dmax Quit: changing servers
1166024883 J * Piet hiddenserv@tor.noreply.org
1166024884 J * dmax ~semaj@81.193.56.217
1166025123 Q * dmax 
1166025141 J * brc_ ~bruce@201.19.180.155
1166025228 J * s[e]bastian ~info@p54A96068.dip.t-dialin.net
1166025532 J * dmax ~semaj@81.193.56.217
1166025640 J * shedi ~siggi@inferno.lhi.is
1166025712 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net
1166026117 J * kir ~kir@swsoft-mipt-nat.sw.ru
1166026328 Q * DavidS Quit: Leaving.
1166026350 J * marcfiu ~mef@aegis.CS.Princeton.EDU
1166026901 Q * kevinp Quit: Leaving
1166027203 M * ntrs__ daniel_hozac, why do you think 2.1.1.6 is broken?
1166027218 M * ntrs__ daniel_hozac, do you also think 2.2.0-rc5 is broken?
1166027220 M * daniel_hozac because i was unable to boot with a similar patch ;)
1166027247 M * daniel_hozac no, 2.2.0-rc5 doesn't have the device mapping.
1166027291 M * daniel_hozac neither does 2.3. i don't think it was supposed to be included yet.
1166027970 J * stefani ~stefani@tsipoor.banerian.org
1166028477 Q * lilalinux Remote host closed the connection
1166028639 J * lilalinux ~plasma@80.69.41.2
1166029017 M * marcfiu hi daniel_hozac: looking for your fc5 kernel patches again... they are in http://cvs.hozac.com/viewvc/rpms/kernel/fedora-5/ ?
1166029123 M * daniel_hozac yep.
1166029394 M * marcfiu linux-2.6-vserver-2.0.patch
1166029401 M * daniel_hozac that's the one.
1166029485 M * marcfiu head of tree?
1166029492 M * daniel_hozac hmm?
1166029500 M * daniel_hozac of Fedora's, you mean?
1166029538 M * marcfiu do i click on the download link for the HEAD or the one that has the kernel-2_6_18-1_2255_fc5_vs2_0_2_2_0_rc9_1 tag?
1166029571 M * marcfiu same thing...
1166029571 M * daniel_hozac doesn't really matter.
1166029573 M * marcfiu never mind.
1166029579 M * marcfiu thanks
1166029797 Q * djrise Quit: Quitte
1166029993 J * bonbons ~bonbons@83.222.39.117
1166030551 M * marcfiu daniel_hozac: any surprises with FC5 2255 over 2239 or so?
1166030564 M * marcfiu I am about to upgrade and a heads up would be great.
1166030593 Q * alienn Quit: Leaving
1166030962 J * gerrit ~gerrit@bi01p1.co.us.ibm.com
1166031048 Q * michal` Ping timeout: 480 seconds
1166031108 M * daniel_hozac marcfiu: not that i know of.
1166031122 M * daniel_hozac marcfiu: let me know if you notice any.
1166031453 Q * tso Server closed connection
1166031465 J * michal` ~michal@www.rsbac.org
1166031469 J * tso ~tso@241-059.dyn-fa.pool.ew.hu
1166031643 M * waldi daniel_hozac: when does this crash occur?
1166031755 M * daniel_hozac waldi: when you're stopping a guest and the traffic has seen traffic classified as OTHER.
1166031773 M * daniel_hozac both h01ger and Snow-Man have gotten it with the Debian kernel, so it's definitely a problem ;)
1166031788 M * daniel_hozac s/traffic/guest/1
1166031793 M * waldi hmm
1166031817 M * daniel_hozac it's a simple off by one problem.
1166031832 M * daniel_hozac the array didn't grow when the number of types did.
1166031838 M * waldi yep
1166031993 M * daniel_hozac so, what're you going to do? upgrade to -rc9, or revert to the -rc6 cacct?
1166032095 M * daniel_hozac s/or/but/
1166032185 M * waldi each solution which don't changes task_struct should be okay
1166032213 M * daniel_hozac well, it's not changing the task_struct directly.
1166032221 M * waldi but indirectly
1166032223 M * daniel_hozac it's just a pointer.
1166032297 M * daniel_hozac IMHO the only sane option would be -rc9, but revert to -rc6 cacct.
1166032494 M * marcfiu daniel_hozac: what crash are you seeing?
1166032525 M * daniel_hozac well, i'm not seeing it.
1166032571 M * daniel_hozac i haven't run any of the affected kernels.
1166032640 M * daniel_hozac http://paste.linux-vserver.org/756 http://paste.linux-vserver.org/761
1166032726 Q * click Remote host closed the connection
1166033071 M * Snow-Man It's definitely a problem which needs to be fixed.
1166033095 M * Snow-Man Sorry if it changes the ABI but, well, that's life.
1166033183 J * click click@ti511110a080-2000.bb.online.no
1166034041 Q * s[e]bastian Ping timeout: 480 seconds
1166034046 J * s[e]bastian ~info@p54A96068.dip.t-dialin.net
1166034886 J * johnny ~johnny@ip70-160-117-211.hr.hr.cox.net
1166035717 M * Borg- daniel_hozac: what kernels are afected?
1166035767 M * daniel_hozac 2.0.2.2-rc7 and 2.0.2.2-rc8.
1166035810 M * Borg- erm wtf..
1166035817 M * Borg- www.linux-vserver.org ask me for password.
1166035838 M * Borg- it points at admin.sourcecode.de
1166035854 M * daniel_hozac then you must've typoed it.
1166035873 M * Borg- ahh right ;) im blind..
1166036527 Q * derjohn2 Quit: Verlassend
1166037379 N * Bertl_zZ Bertl
1166037383 M * Bertl morning folks!
1166037386 M * daniel_hozac morning Bertl!
1166037433 M * Bertl yes 2.1.1.6 is probably a little broken
1166037463 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-dmap-feat04.5.diff should fix most issues with it.
1166037464 M * Bertl but 2.2.0-rc* should be fine, no?
1166037474 M * daniel_hozac yeah, 2.2 and 2.3 should be fine.
1166037489 M * daniel_hozac but i guess the device mapping stuff should just be reverted, right?
1166037489 M * Bertl I wanted to disable the mapping, but I forgot
1166037521 J * lobo ~lobo@82.198.125.154
1166037536 M * daniel_hozac (04.5 puts the bits in the struct, should make it easier to grow on demand)
1166037725 M * Bertl okay, will integrate something similar, btw, what is the ACL stuff for?
1166037733 M * Bertl (not from 04.5)
1166037735 M * daniel_hozac hmm?
1166037766 M * daniel_hozac what ACL stuff?
1166037853 M * Bertl in 04, you have a comment 'typical ACL'
1166037864 M * Bertl *typically
1166037864 M * johnny when will 2.2 be marked as stable? is that available in the bugtracker?
1166037882 M * Bertl 2.2.x _is_ stable
1166037915 M * Bertl (per definitionem)
1166037931 M * daniel_hozac right, it's for the ACL entries.
1166037933 M * hardwire notice the even numbers
1166037949 M * daniel_hozac i.e. not remappings, just "this device can be OPEN|CREATEd".
1166038012 M * daniel_hozac johnny: what bug tracker is that?
1166038025 M * Bertl ah, okay, the comment confused me, but it was gone with later patches so I didn't ask :)
1166038035 M * johnny vserver bug tracker maybe?
1166038044 M * johnny not that i've ever been to it 
1166038060 M * daniel_hozac johnny: there is none, so that might be why.
1166038062 M * Bertl johnny: do you have a bug to report for 2.2.x?
1166038066 M * johnny aha 
1166038075 M * daniel_hozac Bertl: well, you restructured it so it wasn't necessary ;)
1166038079 M * johnny nope
1166038081 M * Bertl if so, please go ahead, so that we can fix it ...
1166038094 M * johnny i just figured you guys might perhaps be using a bug tracker to track releases :)
1166038106 M * johnny by making bugs dep on the 2.2 meta release bug
1166038108 M * Bertl yeah, we kind of plan to do that
1166038115 M * johnny for example
1166038116 M * johnny ok
1166038127 M * johnny well then.. i won't feel bad about asking you .. when is 2.2 going to be marked stable
1166038128 M * daniel_hozac usually we just fix the bugs though ;)
1166038145 M * Bertl hmm, dejavu again?
1166038150 M * johnny well yeah.. and you set up bugs for each requirement of the release
1166038159 M * johnny and then when they're all fixed.. you release it :)
1166038173 M * daniel_hozac except there aren't really any requirements.
1166038176 M * daniel_hozac we make those up as we go :)
1166038188 M * johnny well when will 2.2 be releeased then?
1166038204 M * johnny just want t know what's stopping it .. so i know whether to use it :)
1166038214 M * Bertl johnny: 2.2.x is stable and once we got enough feedback for 2.2.0-rc*, we'll do a first release
1166038215 M * daniel_hozac well, people testing it, i guess.
1166038227 M * johnny hmm i just wish i had an extra box 
1166038229 M * johnny to do one on
1166038234 P * lobo Abandonando
1166038236 M * johnny i only have 1 server that can be a vserver.. for now
1166038244 M * Bertl johnny: put it on your laptop :)
1166038251 M * johnny hmm..
1166038281 M * Bertl doesn't hurt what you are doing there, unless it is a bug, in which case, you should inform us and we'll fix it ASAP
1166038300 M * johnny but.. what needs to be tested about it?
1166038314 M * johnny just go through and install a vserver?
1166038317 M * Bertl typical usage pattern, inside and outside a guest
1166038331 M * Bertl put some things into a guest, build it from scratch
1166038341 M * Bertl check performance inside and outside for example
1166038343 M * johnny ok.. one more question then :)
1166038353 M * johnny what is the likely hood of it ever breaking my host server?
1166038362 M * daniel_hozac breaking?
1166038371 M * johnny is there ever a chance of that?
1166038374 M * Bertl breaking as in screwing up your filesystem or what?
1166038405 Q * dmax Read error: Connection reset by peer
1166038406 M * johnny anything that just affecting vservers :)
1166038410 M * johnny than*
1166038419 M * Bertl well, that risk is always there
1166038430 M * johnny sure.. but has it ever happened? :)
1166038430 M * Bertl it is also there when you install a mainline kernel
1166038433 M * johnny yes
1166038444 M * johnny what is the chance above the mainline kernel :)
1166038447 M * johnny issues
1166038448 M * Bertl yes, recently Aiken killed his system with AOE
1166038452 Q * s0undt3ch Read error: Connection reset by peer
1166038455 M * johnny AOE?
1166038458 M * Bertl (mainline issue)
1166038464 M * Bertl ATA over Ethernet
1166038472 M * Bertl got broken for Alpha as it seems
1166038472 M * johnny i mean past the mainline
1166038479 M * johnny mainline issues i know about
1166038490 M * Bertl we had no corruption yet, a few kernel oopses
1166038504 M * johnny hmm.. well maybe i can test it on there then
1166038514 M * daniel_hozac and usually problems get fixed within a few hours.
1166038514 M * johnny i doubt it will really kill my file system :)
1166038527 M * Bertl i.e. a stack trace and you have to reboot, that's the biggest issue we encountered yet, I'd say
1166038529 M * johnny right now.. only my personal stuff 
1166038537 M * johnny is in a vservfer
1166038548 M * johnny my dad's business stuff is still in the host
1166038590 M * johnny so breaking my stuff is no problem
1166038598 M * Bertl but the laptop suggestion wasn't that far fetched, because it is really a good idea to test Linux-VServer for every day purposes
1166038612 M * Bertl after all, similar virtualization is soon going into mainline
1166038630 M * johnny yeah.. but i guess i don't see the use of it 
1166038634 M * johnny for just me 
1166038638 M * johnny other than testing
1166038646 M * daniel_hozac what's 2.6.20 bringing in that area, btw? just pid spaces?
1166038656 M * hardwire yeh
1166038660 M * hardwire I wonder about that too
1166038692 A * johnny thinks about the equivalent of the root kits we've seen for windows
1166038696 M * hardwire werid.. uml vservers
1166038700 M * johnny the stealth ones
1166038706 M * hardwire weird.. even..
1166038831 Q * s[e]bastian Remote host closed the connection
1166038900 M * daniel_hozac Bertl: btw, what's missing for a 2.0.2.2 release? cleanup?
1166038947 M * Bertl yeah, shouldn't be too much left
1166038969 M * daniel_hozac arch/arm/kernel/irq.c can be removed completely.
1166038975 M * Bertl but I'm not sure if we shouldn't make that a 2.0.3 release actually
1166039005 M * daniel_hozac yeah, i guess that would be better.
1166039094 M * Bertl okay, I'm going to fix 2.1.x now, feel free to do some cleaning up on 2.0.x
1166039209 M * daniel_hozac i'm on it.
1166039710 M * Bertl btw, we got that one yesterday http://paste.linux-vserver.org/764
1166039727 M * Bertl it was the reason for the proc-fix04 cleanup
1166039747 M * Bertl though I'm not 100% convinced that they are related
1166039809 M * daniel_hozac -fix03, i guess?
1166039841 M * Bertl right
1166039857 M * Bertl not enough caffeine yet :)
1166039870 M * daniel_hozac what did addr2line give?
1166039997 M * Bertl sec
1166040014 M * Bertl fs/proc/array.c:146
1166040045 M * Bertl fs/proc/array.c:555
1166040056 M * Bertl include/linux/proc_fs.h:266
1166040067 M * Bertl eip and top 2 backtraces
1166040214 M * daniel_hozac a NULL task?
1166040270 M * Bertl I'm not sure what really happened
1166040297 M * Bertl at first glance, it looks like a mainline issue
1166040321 J * Hurga nobody@p508A8F78.dip0.t-ipconnect.de
1166040327 M * Hurga Hiya.
1166040339 M * Bertl welcome Hurga!
1166040346 M * daniel_hozac i agree.
1166040688 J * DreamerC_ ~dreamerc@61-224-134-2.dynamic.hinet.net
1166040937 A * hardwire finally reads the paper
1166040940 M * hardwire time to get started
1166040947 M * hardwire chapter 1.. word 1..
1166041097 Q * DreamerC Ping timeout: 480 seconds
1166041995 J * Aiken ~james@tooax8-091.dialup.optusnet.com.au
1166042099 M * Bertl morning Aiken!
1166042128 M * Bertl daniel_hozac: regarding dmap-feat04.5, what's the idea behind the cache_free in __hash_mapping()?
1166042152 M * daniel_hozac if we already have a default, we have to free it first.
1166042159 M * daniel_hozac or the memory is lost.
1166042178 M * Bertl yes, but that goes for all the other entries too, no?
1166042202 M * daniel_hozac well, do we support removal already?
1166042204 J * yarihm ~yarihm@84-75-123-221.dclient.hispeed.ch
1166042218 M * daniel_hozac and/or replacing.
1166042226 M * Bertl no, but we neither support defaults and replacing :)
1166042248 M * Bertl anyway, that explains it to me ... np
1166042297 M * daniel_hozac i suppose that should be split out to an __add_mapping or so.
1166042309 M * daniel_hozac (the default assignment)
1166042314 M * Bertl did you need the modifyable defaults in your tests?
1166042338 M * daniel_hozac nah, i don't need it at all with the default deny it is now.
1166042352 M * Aiken Bertl I might be operational again soon, just recieved an email a patch for my problem should be on the way soon :)
1166042360 M * daniel_hozac but i guess for emulating the old behaviour we'd need a DATTR_OPEN default.
1166042385 M * Bertl daniel_hozac: okay, maybe I add the defaults RSN
1166042386 M * daniel_hozac and being able to modify it just makes the most sense to me.
1166042403 M * daniel_hozac anything wrong with the default in 04.5?
1166042428 M * Bertl I'm still not convinced with the structure and the way the defaults are handled
1166042453 M * daniel_hozac how else would they be handled?
1166042486 M * Bertl I had something like this in mind:
1166042503 M * Bertl - hash search, if not found, check default
1166042527 M * Bertl - if default is missing change pointers and restart
1166042543 M * daniel_hozac hmm? change what pointers?
1166042583 M * Bertl in the original plans, it would have been the default entries and the hash 
1166042598 M * daniel_hozac ah, change to the main table, ok.
1166042625 M * Bertl also, I want to get rid of those spinlocks asap
1166042685 M * Bertl and the current vs_device_map doesn't make me happy either
1166042726 M * daniel_hozac ?
1166042740 M * Bertl - well, it has a fixed hash size, _and_ the bits?
1166042759 M * Bertl - it has only one default, while we need one for char and one for blk
1166042761 M * daniel_hozac right, the hash size is supposed to become dynamic :)
1166042777 M * Bertl it is bloated, and will definitely bloat vx_info :)
1166042800 M * daniel_hozac well, the vx_info should just have a pointer, no?
1166042816 M * daniel_hozac and in those cases we'd just start with the main table right away.
1166042824 M * Bertl yes
1166042837 M * Bertl (which also wasn't my original plan :)
1166042911 M * Bertl I guess I have to think a little more about it
1166042919 M * Bertl will fix up the obvious issues for now
1166042981 M * marcfiu Bertl: hello
1166043019 M * marcfiu Bertl: does vs2.0.2.2-rc9 include the vs_base.h related patch that I've submitted to you a few weeks ago?
1166043030 M * marcfiu Bertl: Looking at it now, but it doesn't appear to be.
1166043036 M * daniel_hozac it reverts the entire __(enter|leave)_vx_admin.
1166043086 M * marcfiu I see that.
1166043115 M * marcfiu however, it does not include the irq fix to vx_check, which is required for scheduling to work properly.
1166043143 M * Bertl hmm, have you verified that?
1166043234 M * daniel_hozac the reversion of __(enter|leave)_vx_admin should have the same effect.
1166043307 M * marcfiu daniel_hozac: you mean it takes out all of those __enter and __leave calls.
1166043317 M * daniel_hozac yes.
1166043331 M * marcfiu yes... but that was just one part of the fix.
1166043353 M * marcfiu the other part of the fix was to change __vx_check() to if one was in_interrupt()?
1166043377 M * Bertl yes, we decided to put that in 2.2.x and not 2.0.x
1166043417 M * daniel_hozac we're not sure if checking for in_interrupt in vx_check really is such a good idea.
1166043433 M * daniel_hozac i.e. if it actually gets us anything.
1166043444 M * marcfiu Bertl and daniel_hozac: so checking for in_interrupt() in __vx_check() isn't necessary?
1166043455 M * daniel_hozac maybe not.
1166043467 M * Bertl we don't know yet, it wasn't an issue in 2.0.x before
1166043474 M * marcfiu ugh...
1166043491 M * marcfiu so just taking away the __enter and __leave thing was enough to get the scheduler working?
1166043507 M * marcfiu Its been roughly two weeks and my memory doesn't serve me well.
1166043507 M * Bertl well, we decided not to change too much in 2.0.x as it is basically dead with 2.6.19
1166043523 M * marcfiu understood...
1166043525 M * daniel_hozac Bertl: hmm.
1166043540 M * daniel_hozac Bertl: 2.0.2.2-rc9 doesn't have the vx_capable in vfs_kern.
1166043543 M * marcfiu I just need to create one final stable version for PlanetLab before moving forward with 2.2.x (or whatever you suggest).
1166043564 M * daniel_hozac +_mount
1166043586 M * marcfiu any way, what I was about to say was that I thought without the in_interrupt() check that the scheduler still wasn't functioning correctly.
1166043595 M * marcfiu but that's an easy thing to test. :)
1166043604 M * daniel_hozac the scheduler should work fine either way.
1166043610 M * daniel_hozac it's other things that might break.
1166043617 M * Bertl for 2.0.x and 2.2.x now :)
1166043643 M * marcfiu Bertl: what do you mean with 2.0.x and 2.2.x now?
1166043681 M * Bertl 2.0.2.2-rc9 and 2.2.0-rc5 
1166043712 M * marcfiu since 2.0.x is dead with 2.6.19, when do you think there will be a "stable" release of 2.0.x for <= 2.6.19?
1166043751 M * daniel_hozac like, 2.0, 2.0.1, 2.0.2, 2.0.2.1, you mean?
1166043752 M * Bertl we are working on it (and it's < 2.6.19 actually)
1166043752 M * daniel_hozac :)
1166043781 M * marcfiu in other words, what is needed to go from 2.0.2.2-rc9 to 2.0.2.3 or 2.0-FINAL or whatever.
1166043789 M * marcfiu <2.6.19 is fine by me.
1166043815 M * Bertl probably 2.0.3, mainly testing
1166043817 M * marcfiu as mentioned, as soon as I can declare a stable 2.6.18 w/ 2.0.x, then I move on to 2.6.19 w/ 2.2.x
1166043831 M * Bertl and a little cleanup
1166044270 M * Wonka damn. got a linux 2.6.17.14-vs2.0.2.1 here with a dmesg full of "BUG: soft lockup detected on CPU#0!"
1166044378 Q * Wonka Quit: everything broken. bbl.
1166044933 Q * bonbons Quit: Leaving
1166045035 Q * lilalinux Quit: Leaving
1166045187 J * lilalinux ~plasma@80.69.41.2
1166045315 M * harry wonka...
1166045316 M * harry damn
1166045386 M * harry if that's the same bug i get with the JIT in java, it would prove it';s not grsec related, but definately in the mm of the kernel
1166045421 M * Hurga ouch. Which kernel?
1166045579 M * harry 2.6.17.13
1166045585 M * harry i thought it was a grsec problem
1166045599 M * harry don't have a "mess-me-up"-system yet
1166045609 M * harry so can't really do crashtesting on a production server
1166045632 M * harry was able to analyse the crashdump
1166045637 M * harry came somewhere in mm
1166046076 Q * lilalinux Quit: Leaving
1166046202 Q * Piet Remote host closed the connection
1166046481 M * Bertl and a little cleanup
1166046487 M * Bertl *sorry*
1166046703 M * harry ?
1166046717 M * Bertl wrong window ...
1166047310 M * daniel_hozac Bertl: http://people.linux-vserver.org/~dhozac/p/k/delta-2.0.2.2-clean01.diff
1166047327 M * daniel_hozac mostly cleanups, but i fixed the two things i found while reading through the patch too.
1166047333 J * lilalinux ~plasma@80.69.41.2
1166047350 M * Bertl daniel_hozac: okay, do you think you can live with that for now? http://vserver.13thfloor.at/Experimental/delta-dmap-feat05.diff
1166047374 Q * DreamerC_ Quit: leaving
1166047390 J * DreamerC ~dreamerc@61-224-134-2.dynamic.hinet.net
1166047461 M * Bertl okay, clean01 looks fine to me, will apply it now
1166047480 M * daniel_hozac that will still require CAP_MKNOD to create devices, right?
1166047488 M * Bertl yes
1166047499 M * Bertl but it should be fine to give that now
1166047508 M * daniel_hozac right.
1166047591 M * daniel_hozac shouldn't vs_device_perm be defined to -EPERM when CONFIG_VSERVER_DEVICE isn't set?
1166047629 M * Bertl good point, we should check the return value, but return EPERM
1166047636 M * Bertl (in general)
1166047722 M * daniel_hozac hmm, yeah, right now i think it'd be inverted.
1166047770 M * Bertl the problem is in the check, not the return value
1166047790 M * daniel_hozac ah, i see what you mean now.
1166047791 M * daniel_hozac yeah.
1166047800 M * Bertl i.e. we should define that vs_*_perm doesn't return an error, but says okay when it is permitted
1166047813 M * daniel_hozac right.
1166047824 M * Bertl so the error = in vfs_mknod is wrong
1166047831 M * daniel_hozac indeed.
1166047850 M * Bertl         if (S_ISCHR(mode) && !vs_chrdev_perm(dev, DATTR_CREATE))
1166047850 M * Bertl                 return -EPERM;
1166047855 M * Bertl looks better?
1166047878 M * daniel_hozac yep.
1166047961 M * daniel_hozac so which branches will get it? 2.1 and 2.3?
1166047965 Q * cdrx Ping timeout: 480 seconds
1166047986 M * Bertl yep, I'm keeping it in 2.1 for now, 2.3 will get the new networking stuff
1166047997 M * daniel_hozac ok.
1166048003 M * Bertl once 2.2.0 is out, 2.1 is dead, and we move it to 2.3
1166048009 M * daniel_hozac right.
1166048260 M * Bertl okay, updated the dmap-feat05 in place
1166048631 M * Bertl so anything you are missing for a 2.1.1.7
1166048655 M * daniel_hozac nah, it looks good to me.
1166048756 T * Bertl http://linux-vserver.org/ | latest stable 2.0.2.1, 2.0.3-rc1, 2.2.0-rc5, devel 2.1.1.7, 2.3.0.4, 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 ;)
1166048861 M * Bertl doener, daniel_hozac: btw, the uClibc irc channel is using this: http://cia.navi.cx/doc
1166049042 M * daniel_hozac sounds cool.
1166049078 M * Bertl yeah, and it seems to work, as far as I can tell from the channel
1166049104 M * Bertl there are two bots on oftc too
1166049200 M * Bertl okay, off for now ... back later ...
1166049204 N * Bertl Bertl_oO
1166050034 J * Wonka produziert@chaos.in-kiel.de
1166050299 M * Wonka did i say "this sucks" already?
1166050309 M * daniel_hozac yep.
1166050321 M * daniel_hozac is it reproducible with more recent kernels?
1166050328 M * Wonka another ext3 related error, this was
1166050338 M * Wonka not vserver specific, i think
1166050351 M * Wonka now running 2.6.19.1-vs2.2.0-rc5
1166050389 M * Wonka last kernel was 2.6.17.14-vs$something
1166051138 Q * gerrit Ping timeout: 480 seconds
1166051203 Q * meandtheshell Quit: Leaving.
1166052021 J * hardwire` ~hardwire@rdbck-4891.wasilla.mtaonline.net
1166052071 Q * hardwire Ping timeout: 480 seconds
1166052926 Q * lilalinux Remote host closed the connection
1166052992 J * lilalinux ~plasma@80.69.41.2
1166053327 M * bronson_ What's the best way to create a new vserver?  I used newvserver in the past but that doesn't appear to be a part of Debian's vserver-tools package.
1166053354 M * daniel_hozac vserver ... build.
1166053363 M * bronson_ heh, that's easy.
1166053365 M * bronson_ Thanks.
1166053383 M * daniel_hozac newvserver is part of vserver-debiantools though.
1166053427 M * bronson_ Ah, that's where it came from.
1166053428 M * Hurga Where can I find the patch for ipv6 for guests?
1166053448 M * daniel_hozac http://people.linux-vserver.org/~bonbons/ipv6/
1166053462 M * daniel_hozac note that there is none for 2.6.19 yet.
1166053540 M * Hurga ok, thanks.
1166053815 Q * dna Quit: Verlassend