1176249614 M * mjt dunno. I'd bind-mounted it 1176249653 M * mnemoc how would devpts react if the pts is open on a different namespace? 1176249705 M * mjt looks like devpts does not care 1176249723 M * mjt i was wrong about devpts 1176249735 M * mnemoc uhm? 1176249757 M * mjt the only purpose of devpts is to have dynamic entries - as many as there are pty pairs open 1176249775 M * mjt the rest is done by other parts of the kernel 1176249784 M * mjt which works with device numbers only 1176249802 M * mnemoc do you know anything like losetup to "pick a free pts"? 1176249834 M * mjt openpty() 1176249882 M * mnemoc so /sbin/ptssetup has to be coded :p 1176249899 M * mjt no 1176249984 M * mjt you call openpt() (it's posix_openpt() really) in the redirecting application, and call a shell script to start vserver from it. which in turn executes tty to get the name of already-open-pty-slave to bind to /dev/console inside guest 1176250032 A * mnemoc removing the dust from advanced programming the unix env 2nd. ed... 1176250067 M * mnemoc sounds like... easy :) 1176250071 M * mjt the 'next free' /dev/pts/* is created automatically when you call openpt 1176250086 M * mjt and there's no way to create it by other means 1176250111 M * mjt (actually opening /dev/ptmx which is done by openpt) 1176250197 M * mjt hmm. bind-mount doesn't work 1176250213 M * mjt but copying the device node does 1176250225 M * mnemoc :D 1176250314 M * blizz whatcha talking about? breaking context? 1176250331 M * mjt just random buzz... ;) 1176250361 M * blizz mkay :) 1176250458 M * mnemoc mjt: and screen /patch/to/that/pts/used/as/console works? minicom ? 1176250462 M * mnemoc path* 1176250540 Q * Piet_ Quit: Piet_ 1176251403 M * mjt dpm 1176251426 M * mjt don't have either installed. and it's just waaaaay too late here to try - 04:30 now ;) 1176251438 A * mjt goes to bed 1176251457 M * blizz pff. get some coffee 1176251459 M * blizz jk ;-) 1176251607 M * mnemoc 4:30? russia? 1176251613 M * mnemoc gn8 mjt 1176251651 M * mnemoc /whois is my friend :) 1176252027 J * adrien-modulis ~adrien@3-82-252-216-static.enter-net.com 1176252060 J * PowerKe_ ~tom@d54C13E4B.access.telenet.be 1176252193 J * mjt_ ~mjt@nat.corpit.ru 1176252323 Q * mjt synthon.oftc.net charm.oftc.net 1176252323 Q * PowerKe synthon.oftc.net charm.oftc.net 1176252323 Q * Greek0 synthon.oftc.net charm.oftc.net 1176252323 Q * yang synthon.oftc.net charm.oftc.net 1176252323 Q * kir synthon.oftc.net charm.oftc.net 1176252323 Q * glut synthon.oftc.net charm.oftc.net 1176252479 J * glut glut@no.suid.pl 1176252510 J * Greek0 ~greek0@85.255.145.201 1176252617 J * kir ~kir@swsoft-mipt-nat.sw.ru 1176252692 J * yang ~yang@cpe-213-157-253-172.dynamic.amis.net 1176253145 P * adrien-modulis 1176254700 Q * FireEgl Quit: ... 1176255170 Q * kir Server closed connection 1176255182 J * kir ~kir@swsoft-mipt-nat.sw.ru 1176255886 Q * dna Quit: Verlassend 1176260619 J * slack101 ~root@cpe-71-65-49-94.insight.res.rr.com 1176260630 M * slack101 im struggling 1176260637 M * slack101 reallly struggling 1176260754 M * slack101 i setup a slackwaree but setting it up is soo difficult 1176260796 M * slack101 download packages manually can find right version ..........then gotta compile and finding those packages in a newer state get kinda iffy 1176261165 J * FireEgl Proteus@2001:5c0:84dc:0:211:9ff:feca:b042 1176263453 M * slack101 if i want to give everyone the same distro can i do that ???? 1176263472 M * slack101 without wasting 15 users x 15 same distros installed ? 1176263595 Q * infowolfe Quit: Leaving 1176263845 M * Koffa clone it before adding the 15 users? 1176263866 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net 1176264090 M * slack101 Koffa: sint there a way they can share it and if a user changes a file it would jus make him that new file 1176264782 Q * softi42 Ping timeout: 480 seconds 1176265371 J * softi42 ~softi@p549D56AD.dip.t-dialin.net 1176266668 M * daniel_hozac mjt_: there's nothing theoretical about breaking out of a chroot. the method documented in chroot(2) works fine if you don't set the barrier.... 1176266709 M * daniel_hozac slack101: use unification and COW. 1176267596 J * quiksilv ~bob@203.176.96.250 1176267969 M * slack101 daniel_hozac: yea ......i will look into that later 1176268014 M * slack101 im trying to find a stage 4 gentoo so i dont waste time building one hmmmm 1176268044 M * daniel_hozac http://people.linux-vserver.org/~hollow/stages/? 1176268197 M * slack101 daniel_hozac: didnt it have something wrong with it ? 1176268202 M * slack101 i think i tried it before 1176269074 J * micah_ ~micah@micah.riseup.net 1176269140 Q * micah Ping timeout: 480 seconds 1176269182 M * quiksilv anyone know why when i stop my guest i get: 1176269183 M * quiksilv lockfile: open(): No such file or directory 1176269184 M * quiksilv Shutting down system logger: [ OK ] 1176269184 M * quiksilv lockfile: open(): No such file or directory 1176269189 M * quiksilv ? 1176269399 M * daniel_hozac quiksilv: do you have a /var/lock directory? 1176269418 M * daniel_hozac (or whereever you've configured the utils to do that sort of thing) 1176269435 M * daniel_hozac (hmm, or is that a mktemp -d directory?) 1176269455 M * daniel_hozac either way, use vserver --debug stop to see which one is causing the problems. 1176269468 M * quiksilv yea ive got a /var/lock on the guest 1176269474 M * daniel_hozac on the host. 1176269506 M * quiksilv yea i have a /var/lock and a /usr/local/util-vserver/var/lock for the vservers 1176269541 M * quiksilv ill try debug and see what comes out 1176269910 M * slack101 daniel_hozac: those stage4's provided i think are crippled 1176269957 J * sharkjaw ~gab@158.36.45.236 1176270402 N * DoberMann_ DoberMann 1176271454 M * quiksilv hmm ok when i do 'vserver stop' it gives "lockfile: open(): No such file or directory" twice and also creates two lock files in /tmp/ 1176274136 N * DoberMann DoberMann[PullA] 1176276821 J * chand ~chand@212.99.51.254 1176277720 J * meandtheshel1 ~markus@85-124-174-65.dynamic.xdsl-line.inode.at 1176278326 Q * chand Quit: chand 1176278348 J * chand ~chand@212.99.51.254 1176278737 M * daniel_hozac slack101: uh, what makes you think that? 1176280101 M * Hollow the stage4s only have the wrong profile set (in fact they have a development profile set which does not yet exist in the live tree) 1176280260 M * matti Hi Hollow :) 1176280289 M * Hollow hi matti 1176280354 M * Hollow daniel_hozac: did you read the backlog about my vserver-build suggestion? 1176280361 M * Hollow source vs. exec 1176280417 M * daniel_hozac yeah, but i'm not sure we want to let the initpost/initpre scripts to affect the build that much... 1176280507 M * Hollow why wouldn't we want that? you can lookup all config in /etc/vservers anyway, but some things (like debootstrap mirror, dist) are not saved in the configuration 1176280557 M * daniel_hozac right. 1176280576 M * daniel_hozac but return instead of exit in the scripts for failing is weird... 1176280591 M * daniel_hozac i.e. i would never remember to do that. 1176280614 M * Hollow set -e? 1176280642 M * Hollow you could just surround the pre/post scripts with set -e/set +e 1176280643 M * daniel_hozac hmm, yes, that would change as well. 1176280665 M * daniel_hozac vserver-build is already setting -e 1176280669 M * Hollow ah ok 1176280684 M * Hollow should work out-of-the-box then, no? :) 1176280686 M * daniel_hozac i'm not sure we want to fail the build due to a failing script. 1176280714 M * Hollow well, we may fail, but we probably don't want to remove the created build automagically 1176280718 M * daniel_hozac do you have some time to walk through the patch now? 1176280734 M * Hollow yeah 1176280759 J * dna ~naucki@154-250-dsl.kielnet.net 1176280816 M * daniel_hozac well, first, i don't think it should be so verbose. 1176280848 M * Hollow sounds like --verbose :) 1176280865 M * Hollow otoh, the gentoo scripts are verose too 1176280893 M * daniel_hozac yeah, but that's just the Gentoo-way ;) 1176280901 M * Hollow heh 1176280905 M * daniel_hozac +echo $(< "$cfgdir"/uts/nodename) | $_CHROOT_SH truncate /etc/hostname 1176280908 M * daniel_hozac looks really weird, IHMO. 1176280921 M * Hollow indeed 1176280924 M * daniel_hozac why not $_CAT "$cfgdir"... ? 1176280944 M * Hollow even better: $_CHROOT_SH truncate /etc/hostname < "$cfgdir"/uts/nodename 1176280968 M * daniel_hozac then you should probably get rid of the other UUOCs in there ;) 1176280974 M * daniel_hozac e.g. +$_CAT << EOF | $_CHROOT_SH truncate /etc/apt/sources.list 1176281013 M * Hollow guess that needs to be changed in the gentoo scripts sas well, i more or less did c&p 1176281029 M * daniel_hozac shouldn't the fstab have proc, devpts, etc. too? 1176281061 M * daniel_hozac e.g. something like echo "..." | $_CAT - "$cfgdir"/fstab | 1176281082 M * Hollow actually you shouldn't need fstab at all, but at least for gentoo the rc scripts fail if there is no entry in fstab, that's why we add a dummy... maybe debian works with an empty fstab if the runlevel scripts are removed 1176281111 M * daniel_hozac well, if it's not even required, then just remove it :) 1176281136 M * Hollow will have to test that first.. but it appeared in the newvserver script as well, so i though it might needed 1176281145 M * Hollow +be 1176281154 M * daniel_hozac ah. 1176281187 M * daniel_hozac why does splash_back have to be commented? 1176281214 M * Hollow pure cosmetic.. it wants to read /proc/cmdline 1176281230 M * daniel_hozac ah, okay. 1176281453 Q * Aiken Quit: Leaving 1176281534 M * Hollow daniel_hozac: will go through both gentoo and etch scripts later and reupload the patch 1176281548 M * daniel_hozac okay, thanks! 1176281820 J * bzed ~bzed@dslb-084-059-119-088.pools.arcor-ip.net 1176281896 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at 1176282481 J * Aiken ~james@ppp194-30.lns1.bne1.internode.on.net 1176282648 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1176282658 J * ensc ~irc-ensc@p54B4E22E.dip.t-dialin.net 1176282819 J * lilalinux ~plasma@80.69.41.3 1176285156 J * Piet hiddenserv@tor.noreply.org 1176285885 Q * shedi Quit: Leaving 1176287688 N * nebuchad` nebuchadnezzar 1176288234 Q * Aiken Quit: Leaving 1176288975 J * SoftIce ~phil@vc-196-207-45-253.3g.vodacom.co.za 1176288977 M * SoftIce bonjour 1176289037 M * SoftIce http://www.pastebin.ca/434317 1176289065 M * SoftIce Please could someody assist, I need /dev/tty9 to write logs, etc to. and it this amportal requires it for it to start correctly. 1176289082 M * daniel_hozac hmm? 1176289092 M * SoftIce how can I write to /dev/tty9 inside a vserver, and load the whole of dev into the vserver 1176289098 M * daniel_hozac you realize /dev/tty9 is a tty, right? 1176289106 M * SoftIce correct 1176289120 M * daniel_hozac well, just copy it into the guest if you really need it... 1176289134 M * SoftIce it likes to write logs on tty9 1176289134 M * daniel_hozac seems like ln -s null /dev/tty9 should do the trick too. 1176289136 M * SoftIce for extra loggin 1176289151 M * daniel_hozac those "logs" are lost pretty much immediately. 1176289161 M * daniel_hozac and only available if you're physically at the machine.. 1176289164 M * SoftIce daniel_hozac: sorry the ln -s null within the vserver 1176289179 M * daniel_hozac yes, exactly. 1176289193 M * SoftIce ye, I don't really need the logs myself, nor is their a monitor even connected to the box 1176289203 M * daniel_hozac so you don't really want tty9. 1176289209 M * daniel_hozac just use the symlink. 1176289227 M * SoftIce stty: standard input: Inappropriate ioctl for device 1176289232 M * SoftIce heh, its a bit smarter than that 1176289244 M * SoftIce what about an external symlink, can that work? 1176289257 M * Hollow what about using pseudo-terminals? 1176289267 M * daniel_hozac doesn't make much sense for logging, IMHO. 1176289281 M * daniel_hozac that should be going to a file anyway. 1176289285 M * Hollow sure, but at least it would be a terminal device ;) 1176289294 M * Hollow ln -s $(tty) /dev/tty9 ;) 1176289306 M * daniel_hozac only for the duration of that session... 1176289309 M * Hollow sure 1176289316 M * Hollow probably best to start it in a screen session or so 1176289324 M * SoftIce hmm 1176289340 M * daniel_hozac or: fix the startup script not to require tty9 :) 1176289350 M * Hollow indeed ;) 1176289383 M * SoftIce chown asterisk /dev/tty9 1176289392 M * SoftIce found where its writen in the script 1176289395 M * SoftIce let me see if I can remove it 1176289399 N * WorkRoey Roey 1176289426 M * Hollow SoftIce: did you read http://linux-vserver.org/Problematic_Programs#Asterisk ? 1176289445 M * SoftIce nope, but all ready fixed all the asterisk problems 1176289456 M * SoftIce this is stupid freepbx 1176289542 M * SoftIce hmm, I see that $FOPWEBROOT is not defined and is just being called, I presume that could be the problem 1176289742 M * Hollow off to university now, back later .. 1176290271 M * SoftIce toodaloo 1176290407 J * kir_ ~kir@swsoft-mipt-nat.sw.ru 1176290440 M * SoftIce hmmm, daniel_hozac: how would I build trixbox as a vserver ? 1176290451 M * daniel_hozac hmm? 1176290462 M * SoftIce trixbox is a customized centos 4.4 1176290471 Q * kir_ 1176290471 M * SoftIce with the modifications of asterisk , etc 1176290477 J * kir_ ~kir@swsoft-mipt-nat.sw.ru 1176290485 Q * kir_ 1176290491 Q * kir Quit: Leaving 1176290504 J * kir ~kir@swsoft-mipt-nat.sw.ru 1176290514 M * daniel_hozac so -m yum should be able to do it, no? 1176290530 M * daniel_hozac you just have to setup a configuration like centos4 for it. 1176290535 N * bzed bzed|afk 1176290562 M * SoftIce so I could just d/l the source for trixbox 1176290567 M * SoftIce and then -m yum 1176290568 M * SoftIce ? 1176290587 M * SoftIce and then point to the image 1176290641 M * daniel_hozac hmm? 1176290653 M * daniel_hozac you'd point it at the repository. 1176290693 M * SoftIce ahh, and it could be any trixbox repo ? 1176290748 M * daniel_hozac well, the os repo at least. 1176290753 M * daniel_hozac you'd probably want updates as well. 1176290793 Q * DavidS Quit: Leaving. 1176290964 M * SoftIce hmm, you don't happen to have a link, where I can get an example of installing a centos/fedora system on a debian box? 1176290992 M * daniel_hozac apt-get install yum; vserver ... build -m yum ... -- -d centos4 1176291045 M * SoftIce oh 1176291047 M * SoftIce :) 1176291135 M * SoftIce root@fs:/usr/lib/util-vserver/distributions# ls -al | grep centos 1176291135 M * SoftIce drwxr-xr-x 5 root root 4096 2007-04-06 19:53 centos42 1176291140 M * SoftIce hmf, is this outdated? 1176291147 M * daniel_hozac yes, you have very outdated utils. 1176291172 M * SoftIce thats what I hate about ubuntu 1176291180 M * SoftIce nothing is up to date 1176291316 M * SoftIce daniel_hozac: is there a documentation on how to force to update to the latest utils on this version of debian or should I say ubuntu 1176291328 M * SoftIce err how to force up to 1176291358 M * SoftIce or would I have to just install from sourcE? 1176291663 M * daniel_hozac probably easiest to install from source. 1176291970 M * SoftIce http://www.pastebin.ca/434352 1176291990 M * SoftIce strange seeing that I have the python-urlgrabber package installed 1176292073 M * daniel_hozac if you run python -c 'import urlgrabber', does it work? 1176292209 M * SoftIce ImportError: No module named urlgrabber 1176292216 M * daniel_hozac so, there you go. 1176292245 M * SoftIce stupid package 1176293167 M * SoftIce ok, mostly everything working :) 1176293651 Q * nebuchadnezzar Remote host closed the connection 1176293719 J * nebuchadnezzar ~nebu@zion.asgardr.info 1176293894 J * Psy-Q ~bar@despair.snm-hgkz.ch 1176293897 M * Psy-Q hello 1176293934 M * daniel_hozac hi 1176293939 M * SoftIce bonjour 1176293985 M * Psy-Q i've set rss limits of 10000 pages both via vlimit and via rlimits/rss and rlimits/rss.hard, but the vserver still seems to have all of the host machine's memory when i check using free? 1176294002 M * Psy-Q i've restarted the vserver of course 1176294065 M * daniel_hozac did you also set virt_mem in flags? 1176294122 M * Psy-Q ah, no i didn't. i went by this page of docs: http://linux-vserver.org/index.php?title=Memory_Limits 1176294155 M * daniel_hozac well, the limit is in effect even though free shows all of the messages. 1176294168 M * daniel_hozac s/messages/memory/ weird typo... 1176294204 M * Psy-Q oh! good. i was trying to use that small memtest script to find out how much could actually be allocated 1176294215 M * Psy-Q but i didn't get any conclusive results from that, i think i would have to read up on that one 1176294310 J * ema ~ema@rtfm.galliera.it 1176294419 Q * SoftIce 1176294534 M * Psy-Q daniel_hozac: thank you, when setting that flag, everything is reported correctly 1176294601 M * daniel_hozac there you go. 1176294623 M * Psy-Q i'm evaluating linux-vserver against openvz for future use at our university, we bought a horribly fat machine (8 cores, 16 gb RAM) to consolidate our old server farm on :) 1176294663 M * Psy-Q i hope i can compare notes with our openvz guy soon, though i'm leaning towards linux-vserver already simply because the kernel's included with debian 1176294687 M * Psy-Q and i'm very impressed with linux-vserver in general 1176294753 M * daniel_hozac thanks! 1176294923 M * harry WE RULE! ;) 1176295180 M * Psy-Q oh yeah! ;) 1176296360 Q * sharkjaw Quit: Leaving 1176296847 M * Psy-Q grmbl. vserver-copy as shipped with debian is a bit buggy 1176296933 M * daniel_hozac you shouldn't use it. 1176296944 M * daniel_hozac vserver ... build -m rsync or vserver ... build -m clone works much better. 1176297316 M * Psy-Q hmm, i was tempted by doing vserver ! 1176297333 M * Psy-Q i'm innocent! but i'll read the vserver manpage more carefully 1176297371 M * Psy-Q oh, it's not in the manpage 1176297373 M * Psy-Q that explains things 1176297392 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at 1176297793 Q * Piet Remote host closed the connection 1176297850 M * Psy-Q daniel_hozac: do you know where these options are documented? 1176298040 J * Piet hiddenserv@tor.noreply.org 1176298158 M * Psy-Q it seems the docs are a big area that needs to improve? 1176298205 Q * DavidS Quit: Leaving. 1176298353 J * awk ~awk@dsl-242-81-151.telkomadsl.co.za 1176298377 M * awk hi, hmm, does anyone have a source list for debian / ubuntu for the latest vserver-util, etc 1176298390 M * awk http://dpkg.u32.net 1176298397 M * awk that is the only 1 I know about, but very outdated 1176298769 M * Psy-Q isn't the stuff in etch quite up to date? 1176298774 M * Psy-Q or do you need one for sarge? 1176298840 M * awk I actually need 1 for ubuntu 1176298848 M * awk hmm, i'm looking on the ftp now 1176298860 M * awk what is the difference between 1176298860 M * awk util-vserver/ 09-Dec-2006 17:23 - 1176298860 M * awk vserver-utils/ 04-Aug-2006 11:13 - 1176300126 Q * FireEgl Quit: ... 1176300136 Q * awk Quit: . 1176300717 Q * bored2sleep Ping timeout: 480 seconds 1176301665 J * bored2sleep ~bored2sle@66.111.53.150 1176302212 Q * mire Ping timeout: 480 seconds 1176302212 Q * Hunger Ping timeout: 480 seconds 1176302247 J * maryy_ ~ss@88.233.190.228 1176302251 Q * maryy_ Excess Flood 1176302435 J * mire ~mire@133-169-222-85.adsl.verat.net 1176302671 J * Hunger Hunger.hu@Hunger.hu 1176303210 Q * Psy-Q Quit: ... 1176303543 Q * Val Ping timeout: 480 seconds 1176304114 J * ktwilight ~ktwilight@247.124-66-87.adsl-dyn.isp.belgacom.be 1176304114 Q * ktwilight_ Read error: Connection reset by peer 1176305412 J * stefani ~stefani@flute.radonc.washington.edu 1176305502 Q * ||Cobra|| Remote host closed the connection 1176307664 Q * phreak`` Ping timeout: 480 seconds 1176308307 Q * Hollow Ping timeout: 480 seconds 1176308537 M * Guy- um, what's up with http://www.linux-vserver.org/? 1176308551 M * Guy- I can connect, but it doesn't reply 1176308694 J * shedi ~siggi@ftth-237-144.hive.is 1176308723 M * sid3windr then I guess it's broken ;) 1176308748 M * AndrewLee looks like it cannot connect to database 1176308907 M * derjohn Bertl_vV, ping ... can you admin it ? 1176309186 M * sid3windr bertl is on vacation! ;) 1176309191 M * sid3windr I guess hollow's the admin, no? 1176309230 M * Guy- yes, but he's not around 1176309280 J * Hollow ~hollow@85.10.237.60 1176309397 J * bonbons ~bonbons@83.222.38.145 1176309427 M * derjohn abracadabra ... Hollow , there he is 1176309436 J * bronson ~bronson@66.237.74.66.ptr.us.xo.net 1176309576 N * bzed|afk bzed 1176310521 J * Piet_ hiddenserv@tor.noreply.org 1176310541 Q * ema Quit: leaving 1176310593 Q * Piet Remote host closed the connection 1176310790 Q * Hollow Ping timeout: 480 seconds 1176310798 Q * dna Quit: Verlassend 1176311977 J * Piet hiddenserv@tor.noreply.org 1176312025 J * Hollow ~hollow@styx.xnull.de 1176312060 M * Hollow daniel_hozac: ok, now i know what's up :) 1176312163 M * Hollow http://paste.linux-vserver.org/1407 :) 1176312344 Q * Piet_ Ping timeout: 480 seconds 1176312385 M * Koffa do I need vnet or smth like that to keep 'malicious guests' from stealing eg. network-access from other guests? or is setting up the .../guest/interfaces/0/{dev,ip} enough? 1176312891 N * DoberMann[PullA] DoberMann 1176313078 Q * chand Quit: chand 1176313448 M * daniel_hozac Koffa: guests are limited to their IP addresses. 1176313469 M * daniel_hozac Hollow: ah, OOM? 1176313475 M * Hollow yep. 1176313496 M * Hollow currently looking what some more ram costs :) 1176313531 M * daniel_hozac even out of swap?! 1176313535 M * Hollow ah, have to contact the support for prices .. 1176313551 M * daniel_hozac d'oh! 1176313559 M * Hollow indeed 1176313577 M * Hollow well, 2G swap .. maybe to less 1176313606 M * daniel_hozac nah... just seems strange that it's all used. what all are you running on that box? 1176313607 M * Guy- Hollow: how about using swapd to add swap dynamically when necessary? 1176313612 Q * bronson Quit: Ex-Chat 1176313624 M * daniel_hozac i.e. what the hell consumes 6 GiB RAM? 1176313631 M * Guy- Hollow: and yes, it's strange that it should use more than 6G... 1176313654 J * bronson ~bronson@66.237.74.66.ptr.us.xo.net 1176313665 N * mjt_ mjt 1176313665 M * Hollow Guy-: well, 2G ram + 2G swap = 4G :) 1176313684 M * Hollow daniel_hozac: do you know spamassassin? :) 1176313691 M * Hollow it's such a monster 1176313696 M * daniel_hozac it's just 2 GiB RAM? 1176313698 M * Hollow i just disabled spam filtering now .. OMG :) 1176313709 M * Hollow yeah 1176313720 M * daniel_hozac hmm, i must've misread that report then. 1176313721 M * Hollow way to much vservers on it 1176313734 M * Hollow or to few ram .. 1176313737 M * Hollow depends on the PoV :) 1176313740 M * daniel_hozac too many RAM-hogs, at least ;) 1176313757 M * Hollow guess i'll have a look at dspam soon, spamassassin annoys the hell out of me 1176313773 M * Guy- Hollow: I also use spamassassin but it doesn't eat all my RAM, what are you doing with it? :) 1176313796 J * boci^ ~boci@pool-0792.adsl.interware.hu 1176313811 M * Hollow don't know .. spamassassin never really worked for me 1176313817 M * Hollow i tried so many different setups 1176313819 P * boci^ 1176313821 M * Hollow it just doesn't work 1176313823 M * Hollow perl crap. 1176313832 M * daniel_hozac haha 1176313856 M * Guy- Hollow: I use it with amavisd-new and am more or less satisfied 1176313861 M * Hollow i spent the whole yesterday trying to figure out qmail-toaster at work, it's a perl scripts. 1176313876 M * Guy- Hollow: yes, I also have qmail :) 1176313888 M * Hollow eventually i said to my boss: we won't use it. for sure. :P 1176313898 M * Guy- Hollow: qmail+amavisd-new+spamassassin works pretty well 1176313899 M * Hollow dependency hell. 1176313905 M * Hollow qmail with amavis? hmm 1176313917 M * Hollow never tried that actually 1176313924 M * Guy- amavisd-new has a qmqp frontend 1176313931 M * Hollow i used maildrop, procmail, qmail-scanner, .. 1176313932 M * Guy- so qmail can use it like a qmqp-server 1176313934 M * Hollow oh nice 1176313957 M * Hollow ok, will give it a try :) 1176313971 M * Guy- you can also do trivial load balancing by putting several amavisd-new servers in control/qmqpservers 1176313992 M * Guy- (qmail needs a patch to use them in round-robin fashion though) 1176314027 M * Guy- Hollow: can I press you to a qmqp-patched amavisd-new.deb? 1176314048 M * Guy- Hollow: the patch is shipped with the source deb, so it's not much work 1176314051 M * Hollow oh no.. please don't talk about qmail on debian :P 1176314063 M * Hollow license-uber-suckage 1176314065 M * Guy- I have a pretty good qmail.deb too :) 1176314098 M * Hollow well, unfortunatelly i might need it at work, so lucky for you ;) 1176314395 M * AndrewLee How to build rpm based guest on debian host? I tried build -m yum, but got 'mount: mount point /etc/rpm does not exist'. 1176314405 M * daniel_hozac mkdir /etc/rpm 1176314412 M * daniel_hozac Debian's rpm package is broken. 1176314457 J * phreak`` ~phreak``@deimos.barfoo.org 1176314484 M * AndrewLee daniel_hozac: getting more error messages after 'mkdir /etc/rpm'. 1176314571 M * daniel_hozac such as? 1176314638 M * AndrewLee daniel_hozac: let me find a place to paste. hold on. 1176314664 M * daniel_hozac paste.linux-vserver.org? 1176314692 M * AndrewLee http://rafb.net/p/UVys0161.html 1176314751 M * AndrewLee daniel_hozac: http://paste.linux-vserver.org/1408 as well. 1176314800 M * daniel_hozac export LANG=C? 1176314888 M * AndrewLee looks like it works with LC_ALL=C now. :) 1176315185 M * AndrewLee nope, it doesn't work, let me post the message. 1176315394 M * AndrewLee http://paste.linux-vserver.org/1409 1176315436 M * AndrewLee looks like the 'yum' has problem. 1176315508 M * daniel_hozac looks fine to me. 1176315539 J * jkl jkl@c-67-173-253-237.hsd1.co.comcast.net 1176315947 M * AndrewLee daniel_hozac: wow, it starts. Don't know why it shows the message. 1176315947 M * AndrewLee daniel_hozac: thanks a lot. :) 1176316238 M * daniel_hozac np. 1176318428 Q * phreak`` Ping timeout: 480 seconds 1176318940 Q * Hollow Ping timeout: 480 seconds 1176319215 J * SoftIce ~phil@vc-196-207-45-253.3g.vodacom.co.za 1176319228 M * SoftIce why is http://linux-vserver.org so slow? 1176319233 M * daniel_hozac because it's down. 1176319242 M * SoftIce oh, makes sense :) 1176319252 M * SoftIce any mirror? 1176319265 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1176319307 M * SoftIce cant even access the google cache 1176319385 M * Koffa weird 1176319403 N * PowerKe_ Powerke 1176319466 M * Koffa ah, base href 1176319477 M * Koffa google doesn't squish that 1176319480 Q * DreamerC Remote host closed the connection 1176319482 Q * AndrewLee Remote host closed the connection 1176319486 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw 1176319493 J * DreamerC ~dreamerc@125-225-105-70.dynamic.hinet.net 1176319979 Q * brcc_ Ping timeout: 480 seconds 1176320108 J * Hollow ~hollow@85.10.237.60 1176320235 N * micah_ micah 1176320252 J * phedny_ ~mark@ip56538143.direct-adsl.nl 1176320282 M * AndrewLee I use vyum installed yum in centos4 guest, but somehow it doesn't work inside the guest. 1176320291 M * AndrewLee I got rpmdb: Program version 4.2 doesn't match environment version 1176320312 J * phreak`` ~phreak``@88.198.108.67 1176320652 Q * phedny Ping timeout: 480 seconds 1176320674 Q * Hollow Ping timeout: 480 seconds 1176320861 Q * phreak`` Ping timeout: 480 seconds 1176320907 M * daniel_hozac rm -f /var/lib/rpm/__db*; rpm --rebuilddb 1176321064 Q * Blissex Remote host closed the connection 1176321191 Q * Piet Quit: Piet 1176321259 J * Hollow ~hollow@85.10.237.60 1176321288 Q * dghill Ping timeout: 480 seconds 1176321383 J * phreak`` ~phreak``@deimos.barfoo.org 1176321557 M * AndrewLee daniel_hozac: It works!! Thanks!! :) 1176321890 Q * Hollow Ping timeout: 480 seconds 1176321898 M * AndrewLee Somehow 'minilogd' doesn't terminate when I stop the guest, does this a problem? 1176321933 Q * fs Remote host closed the connection 1176321984 J * fs fs@213.178.77.98 1176321989 Q * phreak`` Ping timeout: 480 seconds 1176322671 J * ntrs ntrs@68-188-55-120.dhcp.stls.mo.charter.com 1176322928 M * daniel_hozac AndrewLee: well, should be killed when the guest is stopped anyway. 1176323476 J * phreak`` ~phreak``@88.198.108.67 1176323803 J * Hollow ~hollow@styx.xnull.de 1176324287 Q * Hollow Ping timeout: 480 seconds 1176324387 Q * phreak`` Ping timeout: 480 seconds 1176324455 J * dna ~naucki@126-225-dsl.kielnet.net 1176324944 Q * meandtheshel1 Quit: Leaving. 1176325487 Q * bonbons Quit: Leaving 1176325516 N * DoberMann DoberMann[ZZZzzz] 1176325981 J * Hollow ~hollow@styx.xnull.de 1176326007 M * Hollow aaaaah 1176326017 M * doener beeeeeee 1176326031 M * daniel_hozac still OOM all the time? 1176326064 M * Hollow yeah.. i get the feeling someone is hogging memory, since spamassassin is disabled now :) 1176326083 M * Hollow so, i now have a always-on-top terminal witch watch vserver-stat open :) 1176326098 M * daniel_hozac hehe 1176326111 M * Hollow well, the linux-vserver guest is definitely eating the most mem :P 1176326122 M * doener which is probably the first to get OOM killed by Murphy ;) 1176326124 M * daniel_hozac what's eating the most htere? mysql? 1176326147 M * Hollow no, i have migrated all mysql instances to one guest .. 1176326162 M * daniel_hozac hmm, you changed the host key on helios? 1176326171 M * Hollow apache is eating mem 1176326177 M * Hollow no? 1176326202 M * daniel_hozac hmm, okay, nevermind, was my local wildcard DNS. 1176326214 M * daniel_hozac helios.dev.croup.de is returning SERVFAIL here though. 1176326274 M * Hollow yeah.. have to start some services manually now, since i pushed the reset button :) 1176326281 M * daniel_hozac ah, hehe. 1176326285 M * Hollow this is some weird bug in baselayout-vserver 1176326294 M * Hollow it should actually clean the state dir on boot 1176326300 J * yarihm ~yarihm@84-75-97-207.dclient.hispeed.ch 1176326324 M * doener Hollow: apache eating memory? Sounds familiar... 1176326364 M * Hollow well, the linux-vserver guest only runs cron, syslog, ftp and apache.. the former 3 don''t really eat much mem, and after boot the guest uses 250M :) 1176326380 J * Aiken ~james@ppp194-30.lns1.bne1.internode.on.net 1176326384 M * doener 250M RSS? woah! 1176326390 M * Hollow indeed 1176326399 M * daniel_hozac that just sounds wrong. 1176326430 M * daniel_hozac my webserver guest doesn't even use that much now. 1176326442 M * daniel_hozac sure, it's not at all as loaded as linux-vserver.org, but still... 1176326479 M * Hollow each apache process uses roughly 20M 1176326492 M * daniel_hozac surely most of that is shared though? 1176326497 M * Hollow and there are currently 13 apache procs 1176326512 M * doener 26M RSS, 6M SHR... that's lots of memory 1176326533 M * daniel_hozac indeed... 1176326557 M * daniel_hozac no wonder OOM likes that box. ;) 1176326562 M * Hollow well, we have php5, python, dav and svn :) 1176326574 M * daniel_hozac python is for trac, or? 1176326576 M * doener I'd point at php... 1176326577 M * Hollow yep 1176326583 M * Hollow yeah, php is the sucker 1176326602 M * Hollow well, i'll see what 2G more ram or so costs 1176326610 M * daniel_hozac what kind of RAM is it? 1176326614 M * doener most often, I've seen such behaviour due to too huge mysql result sets 1176326621 M * Hollow DDR i guess 1176326633 M * daniel_hozac dmidecode? :) 1176326635 M * Hollow mediawiki surely uses huge result sets :) 1176326653 M * Hollow we're approaching 300M now :) 1176326661 M * daniel_hozac stores the entire page in the database, does it? 1176326668 M * Hollow yup 1176326678 M * doener Hollow: could I get mysql access there to watch the process list a bit? 1176326698 J * brcc_ bruce@i.am.someasshole.com 1176326699 M * Hollow sure, but the mysql guest only uses 33M 1176326723 M * Hollow whoops, 330M now 1176326733 M * daniel_hozac the mysql guest?! 1176326737 M * Hollow no, helios :) 1176326743 M * daniel_hozac ah, heh. 1176326748 M * doener yeah, but large result sets take some time to transfer from mysql to the client, so they show up in the process list. That helps to get some idea where to look 1176326770 M * daniel_hozac mediawiki must suck... 1176326791 Q * dna Quit: Verlassend 1176326797 M * doener looking at the memory increase rate, it should be quite easy to figure out the offender, iff it is due to large mysql result sets 1176326852 M * daniel_hozac Hollow: could you reconfigure httpd to kill the children rather quickly? 1176326882 M * daniel_hozac i mean, to limit the amount of time that memory is hogged. 1176326883 M * Hollow yeah, will take a look at this, and probably also at some php caching/accelerating module 1176326898 M * Hollow althoutgh this should evnen hog more mem 1176326899 M * Hollow :P 1176326907 M * daniel_hozac i doubt the code itself is the problem. 1176326926 M * Hollow doener: query, btw 1176326927 M * Hollow ah 1176326929 M * Hollow nm :) 1176327249 J * john567 ~john@196.205.216.118 1176327306 M * john567 Hi. I'm a newbie to linux-vserver and was evaluating using it vs. UML and OpenVZ and had a few questions 1176327328 M * daniel_hozac i wouldn't really put UML in the same category... 1176327348 M * john567 I had read on the FAQ that linux-vserver currently doesn't support virtual interfaces but that an initiative known as ngnet (in 2004 and 2005) was supposed to change that 1176327372 M * john567 daniel: You're right. OpenVZ and linux-vserver as more like FreeBSD jails and Solaris containers 1176327384 M * daniel_hozac define "virtual interfaces" and what you need them for. 1176327427 M * john567 by virtual interfaces, I basically want the same thing that OpenVZ provides me with (each vps has a local loop back, it's own address that it can run iptables on, etc.) 1176327446 M * daniel_hozac a virtual network stack, then. 1176327456 M * john567 If it had some of UML's networking, that would be nice (things like completely separate from the host, hub and switch emulation, etc.) 1176327461 M * john567 yes 1176327463 M * daniel_hozac basically, ngnet is on hold pending the mainline virtualization in that area. 1176327478 M * john567 wow. So it's on hold for the last 1 1/2 years? 1176327514 M * daniel_hozac well, the demand isn't that high, and AFAIK there's been little or no funding for that particular feature... 1176327526 M * john567 I did some research and found it mentioned in 2004 and 2005 and then, all of a sudden, no one mentioned it. 1176327549 M * john567 daniel: Ooops. Sorry if I sounded demanding 1176327587 M * daniel_hozac the IP address isolation is sufficient in at least 90% of the cases. 1176327633 M * john567 you're right, but in scenarios such as creating labs for teaching, fooling around with iptables, etc. it's needed. 1176327661 M * daniel_hozac for labs, virtual machines are much better. 1176327678 M * daniel_hozac otherwise you can't change the kernel. 1176327681 M * john567 you mean like vmware, etc. 1176327690 M * daniel_hozac sure. 1176327714 M * john567 yes. but if the user's want to do everything else without fiddling with the kernel, I think that linux-vserver would scale much better 1176327731 M * daniel_hozac definitely. 1176327761 M * john567 I've used UML in the past but issues such as lack of support, etc. kind of put me off so I'm looking for an alternative 1176327773 M * daniel_hozac hmm? lack of support? 1176327783 M * daniel_hozac it's in mainline now, no? 1176327846 M * john567 lack of support as in on the mailing list. 1176327864 M * john567 also, their wiki / site is so out of date... 1176327993 M * daniel_hozac well, in a teaching environment, i'd go with virtual machines. students ought to compile their own kernel at least once ;) 1176328007 M * john567 so any idea when work on ngnet will be complete? 1176328013 M * john567 daniel: you're probably right 1176328020 M * daniel_hozac well, as i said, we're waiting on mainline for that. 1176328028 M * john567 but I also use it for my own test environment on a laptop with limited resources 1176328074 M * daniel_hozac there are already two patchsets for the level 2 virtualization. 1176328089 M * daniel_hozac (at least) 1176328099 M * john567 really? Could you kindly point me in the right direction? 1176328122 M * john567 Also, have you used either of them / recommend a certain one? 1176328135 M * daniel_hozac no, i haven't. 1176328161 M * daniel_hozac i don't know what the results of merging it with Linux-VServer were either. 1176328170 M * daniel_hozac http://lxc.sourceforge.net/network.php is somewhat outdated already, i think. 1176328250 M * daniel_hozac git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/linux-2.6-netns.git is the latest. 1176328275 M * john567 excellent. 1176328276 M * john567 thanks a lot 1176328316 M * john567 I had a quick look at lxc.sourceforget.net and it says that the network namespace patch applies to 2.6.20, which means that it should be up-to-date. 1176328320 M * daniel_hozac i don't know how complete it is yet, but from what i've gathered it's usable. 1176328336 M * john567 Thanks. 1176328344 M * john567 A final question if you would be so kind. 1176328349 M * daniel_hozac well, ebiederm's tree is against 2.6.21-rc6. 1176328364 M * john567 Does linux-vserver enable me to go into ACPI sleep? 1176328365 M * daniel_hozac (when he sent the email about it) 1176328373 M * daniel_hozac hmm? 1176328406 M * daniel_hozac we don't change anything related to that, so if it works without it, it should work fine with it... 1176328414 M * john567 you're probably wonder why but it's mostly because if I set up a test environment on my laptop, I like to keep it up and running. 1176328422 M * john567 excellent. I'll check it out. 1176328455 M * john567 I almost never reboot my machine (except for kernel updates) - my uptime on my laptop is usually 2 months or so 1176328461 M * john567 so having ACPI sleep would be really nice 1176328479 M * daniel_hozac as for lxc, the fact that it's against 2.6.20 doesn't mean it's updated. i believe it's based on an older revision of the patchset. 1176328498 M * john567 okay. so you recommend I stick with the other patch? 1176328507 J * DoYouKnow ~l33t@adsl-68-75-176-153.dsl.emhril.ameritech.net 1176328508 J * phreak`` ~phreak``@deimos.barfoo.org 1176328515 M * daniel_hozac no, on the contrary 1176328525 M * daniel_hozac i'd suggest you go with ebiederm's tree. 1176328559 M * john567 okay. I'll try it out tomorrow and will be back if I run into any issues :) 1176328582 M * john567 By the way, do you have any idea how to subscribe to the mailing list? 1176328602 M * daniel_hozac 2007-04-10T15:10:44 < Hollow> mailto:vserver-request@list.linux-vserver.org?subject=subscribe 1176328622 M * john567 thanks 1176328647 M * Hollow daniel_hozac: don't tell me you still have that in your scrollback buffer ;) 1176328662 M * daniel_hozac Hollow: nah, had to switch terminal and grep the irclog ;) 1176328667 M * Hollow heh 1176328721 M * john567 Anyway. Thanks a lot for your help daniel. gtg 1176328729 M * daniel_hozac you're welcome! 1176328751 Q * john567 Quit: using sirc version 2.211+KSIRC/1.3.12 1176329066 Q * infowolfe Read error: Operation timed out 1176329656 M * doener Hollow: hm, nothing obvious, probably not a problem with large result sets... which unfortunately makes the search harder :( 1176329694 M * Hollow i'll keep an eye to my watch terminal, but thanks anyway :) 1176329726 M * Hollow constantly using 300M now for some time 1176329749 M * daniel_hozac well, that's "good" :) 1176329826 M * Hollow yeah, kind of ;) 1176330124 J * chand ~chand@m167.net81-64-156.noos.fr 1176330124 Q * derjohn Read error: Connection reset by peer 1176330153 J * derjohn ~derjohn@80.69.41.3 1176330486 Q * chand Quit: chand 1176330952 J * chand ~chand@m167.net81-64-156.noos.fr 1176332510 Q * yarihm Quit: Leaving 1176332978 J * Piet hiddenserv@tor.noreply.org 1176333183 J * FireEgl Proteus@2001:5c0:84dc:0:211:9ff:feca:b042 1176334115 Q * slack101 Read error: Connection reset by peer 1176335733 J * DoberMann_ ~james@AToulouse-156-1-83-99.w86-196.abo.wanadoo.fr 1176335745 P * stefani I'm Parting (the water) 1176335835 Q * DoberMann[ZZZzzz] Ping timeout: 480 seconds