1157847017 Q * serving Ping timeout: 480 seconds 1157850153 J * Viper0482 ~Viper0482@p54976828.dip.t-dialin.net 1157850171 M * daniel_hozac derjohn: oh crap, i just noticed a problem with that patch (2.6.17.13 rebase). 1157850280 Q * Viper0482 1157850513 M * daniel_hozac derjohn: i updated the patch in place, http://people.linux-vserver.org/~dhozac/p/k/delta-dm-fix01.diff is the delta. major thinko when i "fixed" that conflict... 1157850634 Q * gerrit Ping timeout: 480 seconds 1157851291 Q * Skram Quit: leaving 1157852609 Q * mcp Read error: Connection reset by peer 1157852623 J * mcp ~hightower@wolk-project.de 1157853470 J * FCOJ ~mordur@dsl-201-4.hive.is 1157853737 J * serving ~serving@86.108.118.82 1157854107 Q * FCOJ Remote host closed the connection 1157854491 N * Bertl_oO Bertl 1157854501 M * Bertl evening folks! was a long day ... 1157854550 M * Bertl ah, I missed two kernel releases? .. nice 1157854663 M * matti Heh. 1157854850 M * matti Bertl: .13 looks very minor, but .12 looks quite impressive. 1157854966 M * matti Eh. 1157854972 M * matti Anyone want coffee? 1157855004 M * Bertl no thanks, not for me right now :) 1157855023 M * node_ way too late for coffee, i could use some beer though 1157855052 M * matti node_: You're in USA, right? 1157855072 M * node_ yes, eastern time zone 1157855082 M * matti Uhm, too bad. 1157855092 M * matti I'm looking for somebody from UK. 1157855108 M * node_ ah 1157855143 M * matti node_: :-) 1157855200 M * node_ what part of the UK are you in? 1157855228 M * matti I am not in UK yet. 1157855242 M * matti I'll move from Poland to UK at the end of the month. 1157855260 M * node_ ahh 1157855272 M * node_ i was just at polish wedding in london, how ironic 1157855280 M * matti New work, new place to live... 1157855298 M * matti Maybe I'll find my place there... 1157855375 M * matti Ottery St. Mary 1157855379 M * matti Little town. 1157855452 M * matti Or should I say "village" - to be more precisely. 1157855486 M * matti http://maps.google.co.uk/maps?f=q&hl=en&q=ottery+st+mary&ie=UTF8 1157855516 M * matti Near Exeter. 1157855549 M * matti Quite peaceful and beautiful. 1157855558 M * matti :) 1157855698 M * matti node_: Who was from polish? Bride? 1157855730 M * matti Hmm... 1157855732 M * node_ good deal 1157855742 M * matti ? 1157855746 M * matti Good deal? 1157855760 M * matti ;-) 1157855778 M * matti Eh. 1157855857 M * node_ the bride was the british one 1157855867 M * node_ but the ceremony was in polish 1157855877 M * matti Hm... 1157855907 M * matti I hope, you have a good time there. 1157855909 M * matti :) 1157855942 M * node_ oh, this all happened last month 1157855953 M * matti I see. 1157857745 J * jesse_ ~wenchien@59-105-176-11.adsl.static.seed.net.tw 1157857770 N * jesse_ wenchien 1157858406 M * Bertl aloha wenchien! 1157858424 M * Bertl okay folks! I'm off to bed for now ... have a good one everyone! 1157858440 N * Bertl Bertl_zZ 1157860772 Q * brc_ Ping timeout: 480 seconds 1157863270 N * nokoya _nokoya 1157863277 N * _nokoya nokoya 1157863504 Q * id23 Ping timeout: 480 seconds 1157864032 J * id23 ~id@p50811905.dip0.t-ipconnect.de 1157867902 Q * Piet Quit: :tiuQ 1157868166 J * brc_ ~bruce@201.19.192.104 1157868496 J * meandtheshell ~markus@85-124-36-238.dynamic.xdsl-line.inode.at 1157870142 Q * node_ Ping timeout: 480 seconds 1157871891 M * essobi_ or ext3 1157871898 M * essobi_ Ot 1157872131 Q * serving Read error: Connection reset by peer 1157872316 J * davidsmind ~david@149.130.davidsmind.com 1157873022 Q * davidsmind Remote host closed the connection 1157873652 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157875673 Q * Viper0482 Remote host closed the connection 1157876149 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157876437 J * bonbons ~bonbons@83.222.36.236 1157876824 Q * Viper0482 Remote host closed the connection 1157877651 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157878113 Q * Viper0482 Remote host closed the connection 1157878464 Q * Hollow Quit: Konversation terminated! 1157880334 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157881445 P * Viper0482 und weg 1157881729 J * Hollow ~hollow@2001:a60:f026::1 1157881865 M * Hollow cehteh: i was able to migrate all data to my lvm system, only 20M of / still reside on the broken disk, and i made backups.. xfsdump rocks :) 1157883981 Q * transacid Quit: reboot 1157884001 J * michael ENETDOWN@fw-ext.konaktiva.tu-darmstadt.de 1157884002 M * michael hi 1157885374 J * _mcp ~hightower@wolk-project.de 1157885399 J * mcp332 ~hightower@wolk-project.de 1157885406 Q * _mcp Read error: No route to host 1157885416 Q * mcp Read error: No route to host 1157885426 N * mcp332 mcp 1157888152 J * spq_ ~spq@dslb-084-063-017-160.pools.arcor-ip.net 1157888261 M * spq_ hi there, i have some questions about the linux-vserver... i have setup a kernel and the util-vserver(all newest stable releases). now my question: what to do now? i cant find any information how to create vservers from images. 1157888300 M * spq_ is there any beginners howto which is not distribution specific? (im using debian but with the tools self compiled 1157888449 M * Loki|muh vserver etch build -m debootstrap -- -d etch 1157888454 M * Loki|muh would be a start ;) 1157888472 M * Loki|muh this way you create a debootstrapped etch vserver-image 1157888596 M * spq_ hm, but how to install a image e.g. one found here: http://mirrors.sandino.net/vserver/images/ 1157888757 M * spq_ ah and doesnt thevserver build command copy my installation?! 1157888761 M * spq_ of the host? 1157888766 M * spq_ that is not what iwant 1157888799 M * Loki|muh vserver build has more than one way to build a image 1157888811 M * Loki|muh i'm just searching for the old howto 1157888850 M * daniel_hozac with really recent util-vserver packages (i think Gentoo is the only one who has the patch yet), you can use vserver ... build -m template ... -- -t 1157888854 M * spq_ where can i find infos about vserver build? what all the parameters do etc 1157888862 M * daniel_hozac vserver - build --help 1157888900 M * spq_ daniel_hozac, no doesnt work, it beins to create a vserver named - 1157888906 M * daniel_hozac no it doesn't. 1157888931 M * daniel_hozac it outputs the help message. 1157888940 M * Loki|muh spq_: http://oldwiki.linux-vserver.org/alpha+util-vserver <-- section "vserver creation" 1157888965 M * spq_ bicheris:~# vserver - build --help [... wait 2seconds - ctrl+c ... ] bicheris:~# ls -ld /vservers/-/ 1157888967 M * spq_ drwxr-xr-x 3 root root 17 2006-09-10 13:47 /vservers/-/ 1157888972 M * spq_ it does :) 1157889009 M * daniel_hozac uh, what util-vserver version is that exactly? 1157889013 M * daniel_hozac 0.30? 1157889029 M * spq_ how to get theversion? 1157889046 M * daniel_hozac vserver --version 1157889050 M * spq_ ive d/led version 0.30.210 if i mention correctly 1157889071 M * daniel_hozac then it wouldn't be creating a guest named -. 1157889085 M * spq_ no --version prints the help screen 1157889091 M * Hollow /usr/lib/util-vserver/vserver-build --help should work too 1157889093 M * spq_ can it be that its broken? 1157889110 M * Loki|muh spq_: maybe you have old versions installed via apt? 1157889127 M * spq_ i ve installed both 1157889143 M * spq_ but im using my self compiled via /usr/local/sbin/vserver --version 1157889150 M * spq_ the debian one works 1157889171 M * spq_ the self made doesnt, it doesnt react like described 1157889171 M * daniel_hozac so you're using 0.30 then. 1157889186 M * spq_ vserver 0.30.204 1157889190 M * spq_ from debian (apt'ed) 1157889207 M * daniel_hozac i meant your self-made version. 1157889223 M * daniel_hozac and you'll want to use the version from backports.org. 1157889227 M * spq_ but the binary in /usr/local/sbin/ should be a 0.30 1157889254 M * spq_ ah right its util-vserver-0.30 1157889400 M * spq_ hm, from wheredid i get this old one? 1157889427 M * spq_ http://www.13thfloor.at/~ensc/util-vserver/files/stable/ 1157889431 M * spq_ shit :/ 1157889443 M * spq_ THOUGHT IT WOULD BE THE NEWEST 1157889448 M * spq_ sry, caps 1157889467 M * spq_ ok thx, i ll install the new one 1157889667 J * lilalinux ~plasma@dslb-084-058-209-107.pools.arcor-ip.net 1157889897 Q * romke Quit: leaving 1157890484 Q * lilalinux Remote host closed the connection 1157891175 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157891806 M * spq_ ok the newtoolswork nice 1157891994 M * Loki|muh :) 1157892249 Q * michael Quit: michael 1157892262 M * spq_ http://www.ampaste.net/3673 1157892264 M * spq_ hm 1157892267 M * spq_ i get this error 1157892550 Q * Viper0482 Quit: one day, i'll find this peer guy and then i'll reset his connection!! 1157892591 M * daniel_hozac what kernel are you using? 1157892634 M * spq_ uhm 1157892648 M * spq_ i hate lilo :) it sitched back to myold one thx :) 1157892839 J * s0undt3ch_ ~s0undt3ch@bl7-247-194.dsl.telepac.pt 1157892999 Q * s0undt3ch Read error: Operation timed out 1157893000 N * s0undt3ch_ s0undt3ch 1157893233 M * derjohn daniel_hozac, I didnt get a reject with your 1st patch. so it was something "internal" you resolved wrongly? 1157893265 M * daniel_hozac yeah, one of my fixes was completely broken. 1157893322 Q * Loki|muh Ping timeout: 480 seconds 1157893382 M * derjohn he, the one we talked about (dm). Do you have an completely updates patch avail? 1157893411 J * michael ENETDOWN@fw-ext.konaktiva.tu-darmstadt.de 1157893412 M * derjohn daniel_hozac, I can work with that one to, but as I provide my dpatch to public it would be nice to have working one. 1157893415 M * michael hi 1157893432 J * Loki|muh loki@satanix.de 1157893434 M * derjohn daniel_hozac, if not: dont care, I just publish binary and pre-patched-src :) 1157893438 M * daniel_hozac derjohn: i updated it in place. 1157893442 M * derjohn michael, hiho! 1157893446 M * derjohn daniel_hozac, k thx ! 1157893447 M * daniel_hozac so if you refresh it, it should be the fixed one. 1157894465 J * Piet hiddenserv@tor.noreply.org 1157894519 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157894519 Q * Viper0482 1157896077 M * matti :-) 1157896652 J * vrwttnmtu ~eryktyktu@82-69-161-137.dsl.in-addr.zen.co.uk 1157896707 M * vrwttnmtu Hello all - just a quick question - is there support for post-start, and post-stop scripts? The docs refer to the old single config file style? 1157896733 M * daniel_hozac of course, see the flower page. 1157896733 M * vrwttnmtu http://oldwiki.linux-vserver.org/linux-vserver_administrators_gide 1157896740 M * vrwttnmtu OK 1157896742 M * vrwttnmtu :) 1157896773 M * daniel_hozac hmm, i've never seen that page before. 1157896807 M * daniel_hozac ah, 2.4 stuff. 1157896809 M * vrwttnmtu /etc/vservers/vserver-name/scripts 1157896816 M * vrwttnmtu Just for anyone else searching the logs 1157897146 M * waldi hmm, what is the reason that SECURE_MOUNT and SECURE_REMOUNT to be seperate? it is possible to emulate the later with using bind mounts 1157897207 M * daniel_hozac secure_mount is the "larger" of the two, IMHO. 1157897347 Q * vrwttnmtu Quit: Leaving 1157897589 M * spq_ /proc/uptime can not be accessed. 1157897622 M * spq_ what to do? its an error i get when i start a vserver 1157897644 M * spq_ Usually, this is caused by procfs-security. 1157897680 M * spq_ i should read the faq, which is not available. 1157897806 M * daniel_hozac why isn't it available? 1157897935 M * spq_ http://www.linux-vserver.org/index.php?page=Linux-Vserver+FAQ 1157897959 M * spq_ cause it doesnt exist in the new wiki and the content of that page in the old wiki is useless 1157898215 M * daniel_hozac hmm. 1157898222 M * daniel_hozac oldwiki is very messed up. 1157898332 M * daniel_hozac http://oldwiki.linux-vserver.org/Proc-Security has the info you want though. 1157898416 M * spq_ vcontext: vc_set_cflags(): Operation not permitted 1157898428 M * spq_ fixed the last error with the vprocunhide 1157898440 M * spq_ but now this errorpops up 1157898453 M * spq_ vcontext: vc_set_cflags(): Operation not permitted 1157898494 M * daniel_hozac paste the output of http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh to paste.linux-vserver.org. 1157898665 M * spq_ http://www.ampaste.net/3674 1157898684 M * spq_ looks good for me 1157898730 M * daniel_hozac so how did you configure grsec? 1157898738 M * daniel_hozac i assume you followed the instructions, right? 1157898748 M * spq_ which instructions? 1157898754 M * daniel_hozac harry's instructions. 1157898764 M * daniel_hozac http://people.linux-vserver.org/~harry/_README_ 1157898767 M * spq_ give me a link andi ll recompile the kernel 1157898789 M * spq_ why isnt this linked anywhere in the wiki? :/ 1157898902 M * spq_ i ve had enabled CONFIG_GRKERNSEC_HIGH 1157899001 M * daniel_hozac http://www.paul.sladen.org/vserver/archives/200512/0260.html 1157900439 M * Loki|muh spq_: as you surely read, the wiki was changed these days and the new is not fully filled... but as it is world writeable you're welcome to edit it :) 1157900457 M * spq_ hm yeai think i will 1157900475 M * spq_ first have to test my now running vserver :) 1157900573 J * transacid ~transacid@transacid.de 1157903021 Q * Piet Quit: :tiuQ 1157903798 J * dna_ ~naucki@70-209-dsl.kielnet.net 1157903820 J * Alissa atnuzdmxqo@151.81.1.237 1157903833 M * Alissa hi all :D 1157904059 M * daniel_hozac hello 1157904202 M * Alissa daniel_hozac ! ;* 1157904357 J * Piet hiddenserv@tor.noreply.org 1157904688 Q * Piet Quit: :tiuQ 1157904912 J * Piet hiddenserv@tor.noreply.org 1157905010 M * Alissa daniel_hozac vserver support quota and quotatools? :) 1157905057 M * daniel_hozac yes, but you need to have one device per guest. 1157905070 M * Alissa uhmmmm 1157905099 M * Alissa i can create a device and mount this on loop, or setting fstab to quota? 1157905105 M * daniel_hozac http://oldwiki.linux-vserver.org/Standard+non-shared+quota 1157905161 M * Alissa thank's ;* 1157905810 M * harry spq_: check the config then... the ones that are NOT SET... should be disabled, or vserver doesn't work ... 1157905839 M * harry wou could also completely disable chroot protections... just to be sure 1157905852 M * harry and if it still doesn't work, don't use grsec ;) 1157905880 N * Bertl_zZ Bertl 1157905881 M * harry and see if you can reproduce the error, ==> isolate the problem 1157905883 M * spq_ harry, ive already read your page 1157905888 M * Bertl morning folks! 1157905892 M * doener morning Bertl 1157905895 M * harry and? 1157905898 M * spq_ and modified my config 1157905902 M * spq_ now it works 1157905906 M * spq_ thx! 1157905909 M * harry kinkyness!! ;) 1157905912 M * doener Bertl: did you see the CAP_IPC_OWNER discussion? 1157905916 M * matti Bertl: :))))) 1157905926 M * matti Hi Alissa :)) 1157905932 M * harry nice to know my patch is used by so many people 1157905933 A * matti hugs harry :) 1157905936 M * harry now i know it's a good thing :) 1157905943 M * matti Yes yes :) 1157905953 M * harry i knew it for a long time, but it's nice to know that i'm not the only one ;) 1157905955 A * matti hands harry a cookie ;) 1157905956 M * Bertl doener: yep, think we should try :) 1157905958 M * Alissa matti ^^ 1157905979 M * harry all tnx to spender, pax "team" ;) and bertl and all the other vserver guys! 1157905986 M * doener Bertl: "try" as in "get ensc to enable it in .211"? 1157906011 M * matti harry: And coffee makers all over the world! 1157906011 M * matti ;p 1157906025 A * matti is addicted to coffee a bit ;p 1157906026 M * Alissa Bertl ! ;* 1157906028 M * Bertl well, more as in motivate daniel_hozac to take over the project (and/or contact enrico, I tried for a week now but failed) 1157906028 M * doener matti: call 0xc0ffee 1157906035 M * matti doener: Opps... 1157906036 M * Bertl hey Alissa! 1157906042 M * Bertl morning matti! 1157906042 M * matti doener: Last time I forgot to ret ;p 1157906046 M * matti Bertl: :)))))))) 1157906063 M * doener matti: well, ret happens somewhere else, so we're not supposed to see it ;) 1157906074 M * matti doener: Oh my. 1157906088 M * matti doener: ;-) 1157906111 M * doener daniel_hozac: go go go :) 1157906124 M * daniel_hozac lol 1157906131 M * matti Hey. 1157906133 M * matti Guys. 1157906140 J * dreamind ~dreamind@C2107.campino.wh.tu-darmstadt.de 1157906141 M * matti Anyone have LinkedIn profile? 1157906142 M * dreamind Hi 1157906154 M * daniel_hozac wow, looks like enrico is on now. 1157906158 M * dreamind anyone working on a patch for 2.6.17.13? 1157906173 M * daniel_hozac which branch? 1157906180 M * dreamind stable? 1157906192 M * doener dreamind: IIRC daniel_hozac has a patch for that 1157906204 M * dreamind ah ok 1157906207 M * doener and I'm (still) working on my port to .18-rc6 1157906216 M * daniel_hozac stable is just a few offsets. 1157906223 M * daniel_hozac you'll only have to fix the Makefile manually. 1157906225 A * Bertl hasn't even started yet :/ 1157906230 M * doener daniel_hozac: wasn't there some dm problem? 1157906230 M * dreamind ah ok I'll try myself 1157906235 M * daniel_hozac doener: only on devel. 1157906244 M * doener i c 1157906260 A * doener has read "stable" as in "-stable" ;) 1157906278 M * daniel_hozac hehe. 1157906279 M * dreamind daniel_hozac: I have a rejectfile for fs/binfmt_elf.c 1157906292 M * daniel_hozac dreamind: hmm, really? 1157906295 M * dreamind if I try to apply the 2.6.17.11 patch to 2.6.17.13 1157906300 M * dreamind yup but I use grsec + vserver 1157906313 M * daniel_hozac dreamind: well, that's why then ;) 1157906319 M * dreamind hm :( 1157906325 M * doener Bertl: looks like mostly trivial stuff, except for ebiederm's proc cleanup... took me some time to grasp the changes 1157906349 M * doener dreamind: you'll have to bug harry about that ;) 1157906362 M * dreamind hm, 'k 1157906419 M * daniel_hozac Hollow: we have a go from ensc to move to svn.linux-vserver.org. 1157906442 A * Hollow goes bananas 1157906447 M * Hollow yay! 1157906459 M * Hollow :P 1157906462 M * matti ;ppp 1157906530 J * ensc ~irc-ensc@p54B4F939.dip.t-dialin.net 1157906539 M * ensc hi 1157906562 M * Hollow hi enrico 1157906584 M * ensc daniel_hozac asked me to add util-vserver to linux-vserver.org's SVN 1157906588 M * ensc it is ok with me 1157906607 M * ensc but I would like to import a svn dump generated by cvs2svn 1157906611 M * Hollow yep, he just told me :) for cvs2svn i need the cvsroot 1157906629 M * Hollow at least i didn't seemto get it with a working copy 1157906634 M * ensc rsync cvs.savannah.nongnu.org::sources/util-vserver/ 1157906660 M * Bertl heya ensc! 1157906668 M * ensc Bertl: hi 1157906684 M * Hollow ah, great 1157906688 M * Bertl I'm glad that you found the time to pay a visit here 1157906702 M * ensc Bertl: sorry, I was away at the beginning of this month 1157906716 M * ensc your messages get lost in irssi restarts 1157906716 M * Bertl ah, well, that probably explains it 1157906752 M * Bertl ensc: so do you have a moment to talk about the future of util-vserver? 1157906756 M * ensc ok 1157906791 M * ensc it is ok, when other people get write access to the repository 1157906811 M * ensc e.g. daniel or you 1157906821 M * Bertl first, what we want to know is, how you see it, i.e. do you want to keep maintaining it? are you swamped with work and have no time for that? or did you get sick of util-vserver? 1157906848 M * ensc I just do not have time at the moment to implement new features or go through the patches 1157906879 M * ensc I still have interest in util-vserver 1157906884 M * Bertl okay, and you would be fine, if daniel_hozac would help you there, without changing the tool name? 1157906896 M * ensc yes, it is ok 1157906909 M * Bertl (or would you prefer a branch/spinnoff solution) 1157906951 M * ensc no; I do not see sense in a branch (at least with the currently existing patches) 1157906961 M * ensc things might change when people want to contribute a perl daemon ;) 1157906968 M * doener ugh 1157906973 A * Hollow grins 1157906982 M * Bertl ensc: thing is, like with jack back then, we definitely do not want to cut you out of the loop in any way 1157907013 M * Bertl ensc: on the contrary, we are pretty happy for further contributions and/or design ideas 1157907044 M * Hollow would be no problem to make you an account on the new svn server 1157907047 M * ensc my main reasons to rewrite vserver-tools were, to make them secure 1157907075 M * ensc I do not want that new code will be added blindly to util-vserver which might open leaks 1157907113 M * Bertl do you have a good or bad feeling with daniel_hozac's work as far as tou know it? 1157907131 M * Bertl (no need to say that this is what we want too :) 1157907144 M * doener ensc: if you don't mind to give a short explanation (and remember how it works), why does the bind mount in namespace-enabled vservers actually protect against chroot breakouts? I spent hours and hours to understand that and failed... :( 1157907150 M * ensc Bertl: I discussed some of the patches with him, and they are of good quality 1157907207 M * ensc doener: basically, a chdir("..") will be relative to the last seen '/' 1157907235 M * ensc doener: with a 'mount --bind / /vservers/...', you change the last seen '/' 1157907260 M * ensc unfortunately, it is a weak explaination and covered by personal experience only 1157907301 M * ensc Hollow: what do I need to get a SVN account? 1157907306 M * daniel_hozac ensc: btw, about the vshelper sync problems, http://people.linux-vserver.org/~dhozac/p/uv/experimental/util-vserver-0.30.210-vshelper.patch solves it along the planned lines, right? 1157907407 M * Hollow ensc: i just need your ssh publiy key 1157907964 M * Bertl okay, here are some things I would like to see in the near future ... 1157907973 M * Bertl (not necessarily features :) 1157908021 M * Bertl - an efford/plan/concept to wiki-ize the Flower Page without losing the 'create it from source' option 1157908055 M * Bertl (I hope mediawiki is powerful enough to accomplish that) 1157908092 M * Bertl - we need to prepare for the 2.2 branch and vci/version checks 1157908121 M * Bertl - newly assigned flags and ccaps should get a name too 1157908134 M * Bertl - the scheduler support is very important (IMHO) 1157908136 M * Hollow Bertl: well, we rather need an xslt template doing wiki markup i guess 1157908136 M * waldi Bertl: this file is autogenerated? 1157908161 M * Bertl waldi: as far as I understood, yes, please ask ensc for details 1157908193 M * ensc yes; it is an XML file 1157908292 M * ensc an xslt stylesheet generating wiki syntax should be possible 1157908307 M * mnemoc Bertl: http://wikipediafs.sourceforge.net/ 1157908329 M * ensc but I do not see much sense in it; or: how to I get changes from wiki back into the XML source? 1157908380 M * Bertl that's where mediawiki could help, what about having specially marked 'sections' with static context? Hollow, is that possible? 1157908416 M * Bertl or maybe we break down the config into smaller chunks, and allow comments below/above? 1157908477 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157908495 M * ensc or, give up the XML stuff completely and maintain it in wiki only? 1157908501 Q * Viper0482 1157908510 M * Bertl ensc: would be an option too I guess? 1157908526 M * Hollow Bertl: well, as i said, i think it's best to make an xslt generating wiki markup 1157908532 M * Bertl ensc: is the xml used for anything besides the docu? 1157908536 M * Hollow wiki only would be ok for me too 1157908546 M * ensc Bertl: no 1157908568 M * mnemoc http://wikipediafs.sourceforge.net/ 1157908569 M * harry i'll do the grsec + vserver for 2.6.17.13 tomorrow, dreamind 1157908573 M * ensc I wanted to generate e.g. DocBook documentation out of it but it was never completed 1157908578 M * Bertl ensc: so there would be no loss in moving that to the wiki (except for the in-package docu) 1157908585 M * dreamind harry: I'll already have it patched myself :) 1157908588 M * harry there is also a small pax fix that i'll include aswell 1157908590 M * dreamind harry: seems to work so far 1157908591 M * doener if we keep the xml file and have the wiki site locked, maybe some svn hook could automagically update it 1157908596 M * Loki|muh gna 1157908603 M * Loki|muh ups, sorry :( 1157908604 M * dreamind harry: ah ok, will use that tomorrow evening... 1157908606 M * harry dreamind: i don't expect much problems with thtat.. ;) 1157908621 M * ensc which ways exist to export wiki into ... mmh... html, docbook, perhaps man-pages? 1157908621 M * harry so your "fixes" will probably be as good as mine 1157908689 M * Bertl html is probably easy, just requires a printable css 1157908716 M * Bertl (same goes for pdf and ps, no idea about docbook) 1157908735 M * ensc Bertl: I thought the same thing about other Wikis (moinmoin) and found out that the 'save as' feature in firefox was the best solution :( 1157908747 M * Hollow btw, http://ensc.de/id_rsa-linux-vserver.pub 1157908749 M * Hollow argh 1157908752 M * Hollow i'm sorry 1157908764 M * Bertl the 'old' wiki (tavi) did provide latex and postscript output natively 1157908773 M * Hollow sometimes my copy&paste doesn't work :/ 1157908779 M * Hollow http://svn.linux-vserver.org/viewvc.py/util-vserver/ 1157908797 M * ensc I hope, my 'rm ...' was fast enough.... ;) 1157908823 M * Hollow :) 1157908824 M * doener well, in the worst case you have access to some unknown boxes now ;) 1157908830 M * Hollow hehe 1157908832 M * Bertl http://wiki.enneenne.com/index.php/MediaWiki_PDF_export 1157908842 M * ensc ;) 1157908853 M * Bertl ensc: hiding public keys eh? :) 1157908862 M * michael bye 1157908863 M * ensc Bertl: it is only 2048 bits ;) 1157908864 Q * michael Quit: michael 1157909294 M * Hollow ensc: for savannah, anonymous checkout can be done with http://svn.linux-vserver.org/svn/util-vserver, the link currently in svannah is only the webbrowser 1157909498 M * ensc Hollow: ok; updated. But I can not change the "CVS" wording 1157909543 M * Hollow yeah, guess that's ok, no? 1157909560 M * Hollow both urls are webbrowsable... 1157909579 M * Hollow the native one being quite ugly.. 1157909580 M * Hollow ;) 1157910710 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1157910755 Q * Blissex Read error: Connection reset by peer 1157910779 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1157911706 M * ensc btw, when I am here: I upgraded some (hashified) machines with FC[24] guests to FC5 recently. I ran a 'find /vservers/.hash -links 1 | xargs rm -f' command. With 'df -i /vservers' I saw that lot of inodes were removed, but 'df /vservers' before and after the cleanup did not show a big difference. After rebooting with enforcing an fsck, I got the full space (perhaps 500-1000MB) back. 1157911774 M * Bertl aha, interesting ... 1157911805 M * mugwump open file handles? 1157911810 M * daniel_hozac was it really the fsck that solved it? 1157911830 M * ensc mugwump: I did a 'vserver ... restart' for almost every vserver 1157911861 M * Bertl did the fsck report anything related? 1157911864 M * ensc daniel_hozac: I am not sure whetehr a plain reboot would have done this too 1157911887 M * ensc no, fsck ran without output 1157911893 M * ensc (resp. problems) 1157911924 M * Bertl very interesting ... which filesystem? 1157911938 M * ensc ext3 1157911944 Q * ntrs Read error: Connection reset by peer 1157911953 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1157911957 M * Bertl but you did wait the 30 seconds or so for the ext sync 1157911998 M * ensc yes, it were more than the 30 seconds... and I did some 'sync' 1157912000 M * daniel_hozac ensc: live upgrades? 1157912008 M * ensc daniel_hozac: yes 1157912065 M * daniel_hozac what does "almost every vserver" mean? were there at least one previously FC2 and one FC4 still running? 1157912088 M * ensc there were some FC5 vservers already which were kept as is 1157912113 M * daniel_hozac IMHO the open file handles is what we chalked it up to when micah had similar problems with Debian guests. (they weren't upgraded, just hashified while running) 1157912122 M * daniel_hozac s/IMHO/IIRC/ 1157912148 M * daniel_hozac so all the guests running FC2 and FC4 previously were restarted? 1157912254 M * ensc on the machines, there was an older kernel (2.6.17.4 vs2.0.2-rc25) running during the upgrade 1157912261 M * ensc yes, all old vservers were restarted 1157912384 J * Piet_ hiddenserv@tor.noreply.org 1157912394 M * ensc I do not have a reproducer; just wondered why I am running out of space although all old files were removed 1157912461 M * ntrs how is the personality set in the config for 32 bit guests on a 64 bit host? 1157912723 M * ntrs anyone? 1157912727 M * Bertl PER_LINUX_32BIT or PER_LINUX32 IMHO 1157912731 Q * Piet Ping timeout: 480 seconds 1157912740 M * ntrs in what file? 1157912749 N * Piet_ Piet 1157912758 M * Bertl in the personality file in the config (see flower page) 1157912836 M * Bertl I have no 64bit x86 machine atm, so you have to try .. 1157912855 M * daniel_hozac linux32 works here. 1157912964 M * Hollow shouldn't it work without? 32bit chroot works on amd64 too here 1157912987 M * ntrs I am seeing some weird syscall 311 missing messages from the kernel. 1157912998 M * daniel_hozac what kernel is that? 1157913048 M * ntrs 2.0.2-rc20 with 2.6.16.16 1157913067 M * ntrs some applications expec for syscall 311 to be present 1157913071 M * ntrs expect 1157913075 M * Bertl ntrs: those are simply not implemented in the 32bit compat for x86_64 1157913093 M * Bertl maybe the next kernel release will have them ... 1157913103 M * daniel_hozac 2.6.17 has it. 1157913116 M * daniel_hozac glibc started using it too quickly, but it falls back gracefully. 1157913146 M * ntrs I see, so no real need for the personality setting? 1157913157 M * daniel_hozac it won't help with that problem. 1157913211 M * ntrs ok, so is there a reason to use the personality setting then? 1157913259 M * daniel_hozac not that i know of, but i haven't tried without it. 1157913347 M * doener I haven't tried with ;) 1157913352 J * _are_ are@dslb-084-057-169-021.pools.arcor-ip.net 1157913371 M * Bertl welcome _are_! 1157913398 M * _are_ Hi 1157913451 M * mnemoc Bertl: now that the conversetaion finished, take a look at http://wikipediafs.sourceforge.net/ ,-) 1157913516 M * Bertl well, I already had a look there, but I'm not sure that this is the way we want to go :) 1157913542 M * mnemoc :) 1157913551 M * _are_ some FUSE implementation to whatebver in discussion? 1157913575 M * mnemoc personally i hate to write on web interfaces.... i prefer my beloved vi 1157913601 M * mnemoc but it's also an alternative to keep generated content 1157913607 M * mnemoc like the weed page 1157913970 M * Bertl ensc: regarding the hadlink removal, I think that it must be some still referenced files, might be from the journal or so, as we do not change the filesystem behaviour in this regard (could even be some quota or ext3 oddity) but if that can be recreated somehow, we can investigate it 1157913976 M * Bertl *hardlink 1157914052 M * Alissa Bertl have I activated quota on the vservers, but is it normal that webmin doesn't succeed in understanding that quota is activated and gives me that any filesystems is applicable? 1157914083 M * Bertl depends on many things, like vroot, filesystem, fstab/mtab entry 1157914111 M * Bertl but it could be that the way webmin accesses quota requires the vroot hack we did recently 1157914128 M * Alissa I have used this: http://oldwiki.linux-vserver.org/Standard+non-shared+quota 1157914152 M * Alissa (I'm use a loop mount poit (filesystem of vserver is a file mounted in loop) 1157914177 M * Bertl okay, what filesystem? 1157914184 M * Alissa ext3 1157914209 M * Bertl inside the guest, what does quotaon -vaug and quotaoff -vaug say? 1157914213 M * _are_ hmm, when i hear filesystem: I had configured some box to mount the vservers root directory on vserver-start, but some vserver-version later and some tools version later it didn't work anymore. Is this supported at all or ... 1157914214 M * _are_ just had been an accidental feature back then? 1157914251 M * Bertl _are_: it is supported natively with recent tools. you just have to add an entry to the configuration fstab 1157914287 M * _are_ had been like that, then some update happened with i think util-vserver 204 -> 210 and it failed 1157914316 M * Bertl hmm, works here with 0.30.210 ... how does it fail for you? 1157914351 M * _are_ it claimed no vserver root existed 1157914390 M * Bertl daniel_hozac, ensc: any ideas? 1157914405 M * _are_ when i mounted the root manually and then started it worked again. unfortunately I had been in a hurry back then so had no time for a closer look. 1157914495 M * _are_ i should have access to a box that has the very same setup sometime next week, but with old tools and old patch, perhaps I find the exact error message again then 1157914524 M * Bertl i think that would be interesting .. 1157914779 M * matti _are_: :) 1157915328 A * _are_ just installed wikipediafs, looks nice, only missing a search interface (or I have not found it, yet) 1157915838 M * matti wikipediafs? 1157915841 M * matti OMG. 1157915844 M * matti :) 1157916212 M * daniel_hozac Hollow: so how do i get an authenticated checkout? 1157916237 J * are|lunch are@dslb-084-057-188-099.pools.arcor-ip.net 1157916241 M * doener Hollow: any news on the imap thing? how would I access that? 1157916261 M * daniel_hozac doener: i thought the conclusion was "use .forward" :) 1157916278 M * doener must have missed that 1157916339 M * daniel_hozac (or a mail client that lets you specify the command to use) 1157916364 M * daniel_hozac but i think he uninstalled cyrus. 1157916458 M * Bertl I wouldn't suggest to go for cyrus imap, it's a PITA from handling and configuration 1157916499 M * Bertl IMHO .forward is the best choice, but if we figure that some imap/pop is required, I'd suggest to use that simple imapd 1157916656 Q * _are_ Ping timeout: 480 seconds 1157916775 M * Alissa . VServer Context Quota Author: Herbert Pötzl 1157916781 M * Alissa Bertl is you? :D 1157917227 M * gdm Bertl: just as a follow up from my email to the list, i have updated kernels on the boxes and will be puttin gthem in a place where they should be accessible (well, one of them) if it crashes. but this is definitely seeming to be a problem of servers becoming unresponseive when (maybe) one of the vservers has a high load of connections. 1157917246 M * gdm hello, by the way :-) 1157917287 M * gdm i may try to migrate some of the wiki stuff over, particularly the memory pages i had worked on before, as well - but i have very little time at the moment 1157917313 M * Bertl Alissa: yup 1157917331 M * Hollow daniel_hozac: svn+ssh://dhozac@svn.linux-vserver.org/var/svn/util-vserver/trunk 1157917336 M * Hollow doener: not really... 1157917344 M * Hollow doener: probably best to use .forward 1157917361 M * Bertl gdm: could you give me a quick overview of the issues here? 1157917366 M * Hollow doener: i can of course set a password for you and you can use imap normal with password 1157917483 M * gdm Bertl: sorry about the lack of context ;-) http://archives.linux-vserver.org/200609/0002.html - 'vserver hosts "just stop responding" - cause??' 1157917521 M * gdm basically host system becomes completely unresponsive if one vserver is under high number of connections from the web 1157917534 M * gdm even via serial console, is not accessible 1157917543 M * Bertl what about the P4 does it run an SMP kernel? 1157917549 M * gdm there do not seem to be any errors anywhere in logs etc that we can find 1157917571 M * Bertl what kind of kernel 'debug' options are enabled on those kernels? 1157917618 M * gdm i think it may have been... i have now used 2.6.17.11 and turned smp off. can't remember about the de3bug options 1157917640 M * gdm but i am going to make a new kernel for the dual xeon, so i can make sure that the debug options are on if that helps 1157917680 M * gdm how can i find out what debug optoins are enabled? is it in the config in /boot ? 1157917740 M * Bertl well, it is in the config of your kernel, no idea _how_ you build the kernels and _what_ config you use 1157917775 M * gdm ok, /me is no kernel whizz, still very new.. i'm just looking now 1157917792 N * are|lunch _are_ 1157917810 M * Bertl do you build the kernels yourself? 1157917869 M * gdm yes, i followed http://oldwiki.linux-vserver.org/KernelHowto 1157917899 M * gdm ok, just looking at the .config file now 1157917916 M * Bertl okay, then you have a .config file in the kernel build tree 1157917930 M * gdm # CONFIG_DEBUG_KERNEL is not set 1157917945 M * Bertl you probably want to adjust a few things there, so maybe just upload it somewhere and I give you some hints 1157917963 M * gdm what, the whole config file? 1157917970 M * gdm i can do that... one sec 1157917978 M * gdm (this is the p4 by the way) 1157918016 Q * Adrinael Ping timeout: 480 seconds 1157918024 M * Bertl paste.linux-vserver.org can be used for that (for example) 1157918065 M * Alissa or phpfi.com :P 1157918114 M * gdm http://paste.linux-vserver.org/357 1157918115 Q * Blissex Read error: Connection reset by peer 1157918366 M * Bertl how much memroy has that machine? 1157918387 M * Bertl and is the cpu HT or EMT64 capable? 1157918413 M * Bertl /proc/cpuinfo should provide some clues here 1157918461 M * gdm is that one of the 'flags' ? 1157918486 M * gdm RAM: 1Gb 1157918526 M * gdm one of the flags: ht 1157918564 M * Bertl yes, ht means hyper threading capable, lm means emt64 or so 1157918594 M * gdm ok, ht - but not lm 1157918637 M * Hollow daniel_hozac: what about the vshelper patch? do you think i cann add it our ebuild? 1157918652 M * Bertl gdm: good, the question now is, does your chipset support HT? 1157918673 M * Bertl gdm: i.e. did you see two cpus with an smp/smt kernel? 1157918740 M * gdm oh, i don't recall. i guess i can reboot the machine (it's next to me) into the old kernel 1157918748 M * gdm should i do that? 1157918855 M * Bertl won't hurt to know that 1157918874 M * Bertl check in the bios (in this case) what APIC/MP options are set 1157918880 M * gdm ok, give mea minute or too 1157918900 M * Bertl np 1157919098 M * spq_ is there a way to easier read that "great flower page" - the one with the touch of thc :) 1157919139 M * gdm spq_: change the css. use a proper web browser if you can't ;-) 1157919180 M * gdm Bertl: where would i find the APIC/MP options? 1157919218 M * gdm i can't see any mention of them anywhere... not standard CMOS or advanced BIOS. power management setup? frequency/voltage control? 1157919244 M * waldi hmm, is there a tool to cleanup hash dirs from vhashify? 1157919408 M * waldi hmm, find -links 1 should do it 1157919416 M * Bertl gdm: look for interrupt mode and/or MPS settings 1157919837 M * Bertl gdm: might as well be that your board/bios dosn't support that ... which IMHO would disqualify it for a server setup, but YMMV 1157919891 M * gdm yeah, it is a 1/2us and seems to have stuff for gaming. but nevermind, it's what i got so i have to use it 1157919932 M * Bertl okay, what chipset does it have? 1157920132 M * gdm 81845GV 1157920152 M * gdm i found this, which i'm pretty sure is correct one: http://www.pcquest.com/content/hardware/2004/104050809.asp 1157920352 M * gdm yes, that is the correct one 1157920365 M * gdm http://www.giga-byte.co.uk/Motherboard/Products/Products_GA-8I845GV-C.htm 1157920468 M * Bertl okay should support HT and APIC 1157920486 M * gdm ok, so do i still need to boot into old kernel? 1157920490 M * Bertl make sure to disable any power saving functions 1157920510 J * Viper0482 ~Viper0482@p54975825.dip.t-dialin.net 1157920511 M * Bertl and select PNP OS in the bios 1157920529 P * Viper0482 1157920618 M * daniel_hozac Hollow: i wouldn't add it yet, halt is still not fixed. 1157920641 M * Hollow ok, i'll wait :) 1157920753 M * dreamind nite folks :) 1157920827 Q * dreamind Quit: dreamind 1157920970 M * gdm Bertl: sorry for delays, am working with other ppl (remote) to try and get your answers and to do as you say 1157921247 M * Bertl np 1157921359 M * Bertl in any case, check for a bios update too 1157921564 Q * bonbons Quit: Leaving 1157921587 M * gdm yesk, there is one 1157922080 M * gdm Bertl: ok, ,think we have it like that... but will do bios update at later stage 1157922122 M * gdm reboot should be to 'old' or 'new' kernel? 1157922159 Q * Alissa 1157922165 M * Bertl gdm: better do the bios update right now, because we get the correct ESDT then 1157922323 M * gdm ah, ok.. will have a look at that then. 1157923165 Q * dna_ Quit: Verlassend 1157923710 M * gdm Bertl: just realised i can't update the bios. i don't have a floppy disc here with me where i am, so will have to wait fora nother time 1157924042 Q * meandtheshell Quit: bye bye ... 1157924590 M * spq_ whats the best way to give a vserver a defined amount of diskspace? 1157924608 J * node_ ~dave@c-69-143-148-254.hsd1.md.comcast.net 1157924643 M * daniel_hozac disk limits. 1157924649 M * daniel_hozac http://oldwiki.linux-vserver.org/Disk+Limits 1157924753 J * Aiken ~james@tooax7-083.dialup.optusnet.com.au 1157924759 M * gdm damn, you beat me to it ;-) 1157924804 M * gdm spq_: http://wiki.linux-vserver.org/List_of_old_Documentation_pages should be a useful link to bookmark - index to all 'old' documentation pages 1157924805 M * spq_ but thx to u both :D 1157924817 M * spq_ ah nice double thx;) 1157924873 M * Bertl gdm: okay, so let's boot the older SMP kernel and see what it shows 1157924879 M * gdm ok, will do 1157925247 M * gdm Bertl: ok, now i still only see one cpu (you mean when i do cat /proc/cpuinfo right?) 1157925274 M * gdm and SMP is definitely set in this one 1157925781 M * gdm Bertl: was there anything else i should check here? 1157925930 M * Bertl please upload the dmesg output of the bootup and the contents of /proc/interrupts and /proc/cpu if possible 1157925950 M * gdm ok, one sec 1157926079 M * gdm http://paste.linux-vserver.org/358 1157926430 M * Bertl okay, looks like it has (and uses) an apic, but the HT support is not there (in the bios, probably) 1157926449 M * Bertl in your .config, add the EMBEDDED option 1157926471 M * Bertl (easiest way is to simply remove that line and call 'make oldconfig') 1157926473 M * gdm wait, that is the old one, yes? you mean add EMBEDDED in new one? 1157926496 M * Bertl yes, in the new kernel config, you uploaded previously 1157926503 M * gdm this was new one: http://paste.linux-vserver.org/357 1157926505 M * gdm ahh, ok 1157926551 M * Bertl disable the PM there 1157926563 M * Bertl and enable the DEBUG option 1157926589 M * Bertl no need for SMP/SMT as it seems (for now) so no spinlock debugging and such 1157926616 M * Bertl raise the memory split to 2/2 (that's what the embedded is for) 1157926678 M * matti :] 1157926702 M * matti This can be also disabled: 1157926704 M * matti # 1157926704 M * matti CONFIG_UID16=y 1157926704 M * matti # 1157926705 M * matti CONFIG_VM86=y 1157926716 M * Bertl gdm you might need a few boots/tries to get the apic working with non-smp 1157926720 Q * eyck Remote host closed the connection 1157926737 M * gdm Bertl: no probs, that is why i brought the server home ;) 1157926742 M * matti Also this: 1157926744 M * matti # 1157926744 M * matti CONFIG_LEGACY_PTYS=y 1157926744 M * matti # 1157926744 M * matti CONFIG_LEGACY_PTY_COUNT=256 1157926779 M * gdm Bertl: you agree with matti i take it 1157926785 M * matti # 1157926785 M * matti CONFIG_BLK_DEV_IDETAPE=m 1157926785 M * matti # 1157926786 M * matti CONFIG_BLK_DEV_IDEFLOPPY=m 1157926790 M * gdm matti: sorry, not that i don't believe you ;-) 1157926794 M * matti This is very obsolete. 1157926853 M * matti You have SCSI enabled, but you didin't selected any drivers. 1157926867 M * matti I assume, that you're using some USB stick. 1157926869 M * matti Then. 1157926875 M * gdm er, no 1157926886 M * matti CONFIG_MD_LINEAR=y - waste of time. 1157926888 M * gdm er, don't know, rather ;-) 1157926893 M * matti # 1157926907 M * matti CONFIG_I2O=y 1157926914 M * matti You really have any I20 devices? 1157926920 M * matti This is very rare. 1157926930 M * matti Well. 1157926936 M * matti Sorry for interrupting :) 1157926973 M * matti CONFIG_8139TOO_PIO=y 1157926986 M * matti Believe me, you better want MMIO than PIO :) 1157926994 M * matti # 1157926997 M * matti CONFIG_SHAPER=m - very bad idea. 1157927005 M * matti Old and obsolete in favor of QoS in kernel. 1157927017 M * matti # 1157927024 M * matti CONFIG_MOUSE_SERIAL=m 1157927026 M * matti Hmm... 1157927039 M * matti Uh. 1157927051 M * matti Lot of strange options you have selected, indeed :) 1157927060 M * Bertl well, if you haven't used legacy ptys till now, no need to do so 1157927070 M * gdm most i just went with defaults that were there from prev. person who compiled kernel 1157927083 M * gdm i am not kernel whizz at all - am learning pretty fast tho, it seems! 1157927100 M * Bertl cutting down the config to stuff you actually need/know is a good approach 1157927102 M * matti CONFIG_HPET_RTC_IRQ=y - this may brake things. 1157927128 M * matti Since /dev/rtc will not produce any IRQ requests on HPET capable hardware. 1157927155 M * matti Which most of the time more problematic than useful. 1157927166 M * matti Also. 1157927183 M * matti All i2c devices are probably not really required. 1157927202 M * matti SPI aslo is very rare. 1157927237 M * matti Dallas 1-wire is 100% not necessery ;-p 1157927247 M * matti CONFIG_IBM_ASM=m 1157927265 M * matti And lot other things. 1157927277 M * matti Bertl: Correct me if I am wrong somewhere. 1157927390 M * Bertl basically what I suggest to do is to check each option, if you don't know it, read the help '?' if you still don't get it, stick with the default 1157927419 M * matti Which is not always good idea. 1157927429 M * matti If you're not sure, what option is really for. 1157927430 M * gdm that's what i did beffore 1157927431 M * matti Google it. 1157927433 M * Bertl for hardware, it's simple, everything you have no connector for, you do not need to enable 1157927433 M * gdm stick with default 1157927433 M * ray6 dallas 1-wire is neede for javaring-login-tokens, what's 100% not neccessary about that? :))) 1157927461 M * matti ray6: Really? 1157927466 M * matti ray6: I didn't know. 1157927488 M * matti ray6: Basically, he probalby didn't need this either. 1157927488 M * ray6 matti: no I'm not really sure. They use the Dallas 1-wire BUS but I expect you usually address them completely in userspace :) 1157927506 M * ray6 matti: but for sure - nobody uses javarings ;) 1157927510 M * matti ray6: This is some pice of hardware? 1157927529 M * ray6 matti: sure. You can't connect software using 1 wire :)) 1157927549 M * matti ray6: So, we can assue, that this hardware sucks ;p 1157927557 M * matti ray6: Who is using 1-wire right now? 1157927563 M * matti ;D 1157927575 M * matti Ancient history. 1157927575 M * matti ;p 1157927577 M * matti I guess. 1157927578 M * matti :> 1157927588 M * ray6 matti: it's a java CPU inside a ring! When it came out it was one of the coolest pieces of hardware a geek could wear :) 1157927589 M * matti ray6: But, thanks for pointing this ouy :) 1157927598 M * matti ray6: Hahahahaha. 1157927609 M * matti ray6++ :) 1157927653 M * ray6 http://www.useit.com/papers/javaring.html :) (I just wanted to mention that Dallas 1-Wire actually is a bus for which devices exist. Possibly because I imported a few back then myself from the US :) 1157927692 M * matti ;-p 1157927712 M * matti You're crazy dude :) 1157927715 M * matti 8-) 1157927901 M * ray6 (and completely off-topic now :) Never used the javaring in a vserver :) 1157927903 M * gdm Bertl: it didn't give me the option to disable PM 1157927920 M * gdm Bertl: but gives me option to turn off PM_LEGACY 1157927933 M * matti gdm: You can keep PM, but disable only the legacy option. 1157927943 M * matti gdm: If this is quite new hardware. 1157927953 M * matti gdm: Choose ACPI - it is needed mose of the time. 1157927955 M * gdm 2004 1157927964 M * matti gdm: And on newest hardware probably is not broken. 1157927966 M * Bertl gdm: okay, make that 1157927967 M * matti Even more. 1157927975 M * gdm 22:08 < Bertl> disable the PM there 1157927976 M * matti You'll neet pmtmr. 1157927982 M * matti s/neet/need/ 1157927989 M * gdm ok, i say no 1157927992 M * matti Since HPET is rare. 1157927994 M * Bertl no, PM should be turned off in the bios 1157928000 M * matti And TSC can be tricky. 1157928005 M * matti Indeed. 1157928022 M * matti I am not an kernel expert. 1157928033 M * gdm matti: nor am i!!!! :) 1157928036 M * matti But I spent lot of time reading docs and mailing lists ;] 1157928041 M * Bertl could well be that your system falls into power saving when there is not much work to do 1157928055 M * spq_ how can i create a vserver without installing a distro? (i want to use an image) 1157928057 M * Bertl and does not recover from a certain PS state 1157928069 M * Bertl spq_: what kind of image? 1157928078 M * matti gdm: Bertl is right on PM and BIOS> 1157928085 M * spq_ something from here: http://mirrors.sandino.net/vserver/images/ 1157928097 M * matti gdm: Disable _every_ possible suspensed and such kind of options 1157928224 M * gdm ? 1157928235 M * gdm oops, wreong windwo - sorry 1157928239 M * matti ;p 1157928253 M * gdm i was trying to get the kernel help ;-) 1157928264 M * matti Hehehe. 1157928322 M * gdm PIO or MMIO, matti? 1157928327 M * gdm u said MMIO, correct? 1157928335 M * matti Yes. 1157928344 M * matti Use MMIO instead of PIO. 1157928364 M * matti You can also enable MSI-X 1157928385 M * gdm traffic shaper? 1157928391 M * Bertl spq_: well, you simply build a 'skeleton' guest with your desired config 1157928398 M * spq_ ah ok 1157928406 M * gdm hmm, am saying no 1157928410 M * Bertl spq_: then you extract the tar into the new guest 1157928427 M * Bertl spq_: make sure that the dev and /tmp is cleaned up though 1157928438 M * Bertl (i.e. contains what the skeleton does) 1157928442 M * spq_ ok 1157928462 M * spq_ so i extract everything except tmp and dev 1157928516 M * matti gdm: You're free to ask ;] 1157928651 M * gdm matti: what about VT (virtual terminal) then? 1157928659 M * gdm matti: i'm connected via serial at present 1157928687 M * matti ptys and tty are quite handy, so leave it alone ;) 1157928698 M * matti You can drop legacy BSD one. 1157928800 M * gdm i can turn off HPET right 1157928874 M * gdm HPET_RTC_IRQ i mean. deffault is no 1157929000 M * gdm legacy 16-bit UID syscall wrappers? default is Y 1157929207 M * Bertl leave it as is 1157929219 M * Bertl what you want to enable is the nmi watchdog 1157929222 M * gdm yea, i did (lucky!) 1157929235 M * Bertl and get a serial console/line for the machine 1157929247 M * gdm i have on connected at the moment 1157929253 M * Bertl you might need to test that the nmi watchdog actually works 1157929255 M * gdm which i am using - doing this through. 1157929258 M * Bertl good 1157929267 M * Bertl also check that the magic sysreq works 1157929274 M * gdm nmi watchdog i don't know about, hope i left it inthere 1157929281 M * gdm i am now doing make modules 1157929367 M * Bertl there is a small howto in the kernel Documnetation dir, how to enable and verify the nmi watchdog 1157929388 M * gdm ok, will look at that in a minute 1157929452 M * gdm i have trouble with the sysrq as i am connected from a mac laptop and the keyboard is not quite "normal" 1157929499 M * Bertl well, figure how to send a 'break' 1157929502 M * matti Bertl: In 2.6 there's softdog and hangcheck. 1157929534 M * Bertl yes, and we had a bunch of false positive from the softdog recently :) 1157929537 M * gdm yeah, i shall 1157929547 M * matti Bertl: Hehehe. 1157929611 M * matti Bertl: What you think. Is "irqpoll" handy? 1157929695 M * Bertl boot option, yes? 1157929702 M * matti Yes. 1157929716 M * Bertl well, it can fix certain misconfigurations 1157929890 M * Bertl but it can also be the source of trouble 1157930062 M * matti I see. 1157930593 M * spq_ why does thedefault skeletton create a 1mb tmpfs /tmp ? 1157930621 M * spq_ 16mb 1157930648 M * Bertl because we consider that appropriate for small files 1157930664 M * spq_ why a tmpfs at all? 1157930674 M * daniel_hozac why not a tmpfs? 1157930681 M * Bertl increases performance by reducing the I/O 1157930682 M * spq_ why not normal disk 1157930697 M * spq_ hm ok 1157930703 M * spq_ but it isnt necessary 1157930706 M * spq_ right? 1157930708 M * Bertl tmpfs is stored in memory and backed by swap 1157930710 M * daniel_hozac of course not. 1157931497 Q * spq_ Ping timeout: 480 seconds 1157931541 M * gdm sorry, maybe a silly question, but do i need to do "make install" before "make modules_install" or afterwards? 1157931569 M * Bertl well, depends on what you want to do (and what install does for your distro) 1157931585 M * Bertl basically modules_install copies the modules in place 1157931593 M * ray6 I would expect it doesn't make a difference? I usually do modules_install first. You should do both before you reboot :)) 1157931599 M * gdm ok, i just got told it doesn't matter 1157931608 M * Bertl you can then simply copy the bzImage from the arch dir and be done 1157931615 M * ray6 possibly make install want's to build an initrd on some systems and needs the modules for that? 1157931683 M * gdm ok, the moment of truth... 1157931851 M * gdm gdm@kompost:~$ uname -r 1157931852 M * gdm 2.6.17.11-vs2.0.2-irc_help 1157931855 M * gdm :-) 1157931866 M * gdm thanks Bertl, thanks matti 1157931883 M * gdm so now i'm gonna have to set some limits for the vservers, i think 1157931893 M * gdm and then will get it networkd again so it can be 'live' 1157931901 M * gdm and we shall see if it crashes ;-) 1157931903 M * Bertl have you tested the nmi and sysrq yet? 1157931921 M * gdm oops, forgot about them. 1157931935 M * gdm i need to set the nmi config in grub, will go do that now 1157931943 M * gdm and need to do some research about the sysrz