1189296161 M * Bertl let me note here, that this will be the same on a non Linux-VServer machine as well :) 1189296307 M * yvonne the first part of what you wrote works, the second doesn´t: http://paste.linux-vserver.org/6530 1189296374 M * yvonne I dont´t care about if it is the same on a "normal" system at the moment - I simply want that it works ;-) 1189296424 M * Bertl what util-vserver do you use? I used 0.30.214 here 1189296462 M * yvonne 0.30.212-1 1189296553 J * Hubbert_ ~chatzilla@vpn015.rz.uni-mannheim.de 1189296672 M * Bertl so, I'd suggest to try with recent mainline tools first 1189296684 M * Bertl if that works fine, you can file a bug report to debian 1189296708 M * yvonne :-D 1189296718 M * yvonne what are recent mainline tools? 1189296726 Q * michal` Ping timeout: 480 seconds 1189296728 M * Bertl 0.30.214 1189296754 M * Bertl http://ftp.linux-vserver.org/pub/utils/util-vserver/util-vserver-0.30.214.tar.bz2 1189296762 M * yvonne can I change that without that my vserver crashes? 1189296782 M * Bertl yes, you should even be able to find a debian package for that 1189296827 M * Bertl http://groups.google.com/group/linux.debian.changes.devel/browse_thread/thread/543c2312d2c69a6f/80f928e2218fc762?lnk=raot 1189296851 Q * Hubbert Ping timeout: 480 seconds 1189296861 N * Hubbert_ Hubbert 1189296874 M * Bertl http://packages.debian.org/unstable/net/util-vserver 1189297073 J * michal` ~michal@www.rsbac.org 1189297219 M * yvonne so I have to do it like that: 1189297253 M * yvonne dpkg --purge util-vserver 1189297255 M * yvonne and than simply install that newer package? 1189297302 M * Bertl I think upgrading should work too 1189297316 A * Bertl is not using debian that much 1189297338 M * yvonne the whole system? 1189297348 M * Bertl no, just the package :) 1189297538 M * yvonne it does not become newer 1189297562 M * Bertl did you add the sid repository? 1189297577 M * yvonne what´s that? 1189297589 M * Bertl well, it looks like this version is in sid 1189297599 M * Bertl you are currently using etch repositories 1189297629 M * yvonne you mean the sources list? 1189297635 M * Bertl yup 1189297647 M * yvonne deb http://ftp.freenet.de/debian/ etch main non-free contrib 1189297649 M * yvonne deb http://security.debian.org/ etch/updates main contrib non-free 1189297669 M * eyck don't add sid repository, just wget http://ftp.au.debian.org/debian/pool/main/u/util-vserver/util-vserver_0.30.213-1_ia64.deb 1189297677 M * eyck and then dpkg -i the .deb 1189297688 M * Bertl actually the 0.30.214-2 1189297696 M * eyck yes. 1189297699 M * Bertl but yeah, probably that's the best choice 1189297706 M * eyck it's not in debian yet 1189297710 M * eyck I don't know why 1189297717 M * Bertl http://packages.debian.org/sid/util-vserver 1189297717 M * eyck (fror x64) 1189297723 M * Bertl it seems it is 1189297740 M * eyck oh, I'm looking at itanium, sorry 1189297748 M * Bertl http://ftp.de.debian.org/debian/pool/main/u/util-vserver/util-vserver_0.30.214-2_amd64.deb 1189298012 M * Bertl hmm, seems to be built against a newer libc 1189298164 M * yvonne http://paste.linux-vserver.org/6531 1189298195 M * Bertl as I said, so you probably have to rebuild the package 1189298198 M * eyck it isn't ulibc? 1189298215 M * Bertl dietlibc for most parts, but some parts ar glibc too 1189298233 M * eyck I've got 0.30.213-2 backported to etch available 1189298259 M * Bertl well, maybe that works .. maybe the problem isn't even related to the tools 1189298273 M * eyck http://eyck.forumakad.pl/~eyck/uv/util-vserver_0.30.213-2_i386.deb 1189298286 M * yvonne we cant´t start our virtual server now :-( 1189298292 M * Bertl eyck: if you like to help yvonne, you could try this sequence: http://paste.linux-vserver.org/6529 1189298316 M * Bertl yvonne: because the tools are now uninstalled 1189298352 M * eyck I have amd64 guest on hand, and dovecot install works ok 1189298431 M * Bertl eyck: your package is i386 (JFYI) 1189298452 M * Bertl http://www.backports.org/debian/pool/main/u/util-vserver/util-vserver_0.30.213-1~bpo.1_amd64.deb 1189298506 M * eyck oh, of course it is. hmm, it's too late for me, obviously. good night. 1189298516 M * Bertl have a good one! 1189298522 M * Hubbert gn 1189298684 M * Hubbert ok 1189298691 M * Hubbert we can start vserver again 1189298724 M * Bertl try with a new isntall (as shown on my paste) 1189298744 M * Bertl you can remove guests with: vserver etch64 delete 1189298790 M * Bertl (adjust the interface/ip accordingly) 1189298821 M * yvonne http://paste.linux-vserver.org/6532 1189298824 M * Bertl also, the following things could help if you host setup is kind-of broken 1189298837 M * Bertl mkdir -p /etc/vservers/.defaults/files 1189298843 M * Bertl cp /etc/resolv.conf /etc/vservers/.defaults/files/ 1189298857 M * Bertl and prepend the following to the build command: 1189298870 M * Bertl env -i PATH=$PATH 1189298878 M * Bertl i.e. env -i PATH=$PATH vserver ... 1189298880 M * yvonne :-D it took several weeks to reach the current state of installation and config, I really do not want to delete all that :-( 1189298896 M * Bertl well, the problem is _inside_ your guest 1189298907 M * Bertl it is a debian (not Linux-VServer) issue 1189298920 M * Bertl probably even a config issue on your side 1189298934 M * Bertl the idea is to try with a new guest to verify that 1189298946 M * yvonne hmm 1189298954 M * Bertl you do not have to use the same name for that, just assure that the guest's ip can reach the outside 1189298972 M * Bertl (either by using a public ip or by using S/DNAT) 1189298995 M * yvonne but I´m afraid we are too tired now 1189299024 M * Bertl np with me :) 1189299038 M * yvonne are you also in this channel tomorrow? 1189299058 M * yvonne I also tried to add you in icq ;-) 1189299076 M * Bertl I don't use icq anymore ... 1189299088 M * yvonne oooh :-( 1189299095 M * Bertl but I'm usually around, you can tell by my nick 1189299108 M * yvonne ok :-) 1189299117 M * Hubbert thank for your help 1189299122 M * Bertl you're welcome! 1189299126 M * Hubbert and good night 1189299143 M * yvonne you are a linux-god :-) 1189299148 T * Bertl http://linux-vserver.org/ | latest stable 2.2.0.3, 2.0.3-rc3, devel 2.3.0.19, stable+grsec 2.0.2.1, 2.2.0.3 | util-vserver-0.30.214 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the Wiki, and we'll forget about the minute ;) 1189299191 Q * Hubbert Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.6/2007072518] 1189299208 M * yvonne have a good night! cu! 1189299214 M * Bertl cya 1189299265 Q * yvonne Quit: ChatZilla 0.9.78.1 [Firefox 1.5.0.12/2007073111] 1189299905 Q * Johnnie Ping timeout: 480 seconds 1189300464 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.ct.comcast.net 1189302184 J * DoberMann_ ~james@AToulouse-156-1-94-71.w90-30.abo.wanadoo.fr 1189302293 Q * DoberMann[PullA] Ping timeout: 480 seconds 1189302883 J * friendly12345 ~friendly@ppp59-167-90-160.lns1.mel6.internode.on.net 1189304911 J * virtuoso ~s0t0na@pppoe-208.2.110.89-adsl.spbnit.ru 1189305322 Q * virtuoso_ Ping timeout: 480 seconds 1189308125 Q * Aiken Quit: Leaving 1189308835 J * Aiken ~james@ppp121-45-255-55.lns2.bne4.internode.on.net 1189309798 Q * AndrewLee Quit: poweroff 1189314876 Q * hparker Remote host closed the connection 1189315086 J * hparker ~hparker@linux.homershut.net 1189315566 J * yarihm ~yarihm@84-74-20-183.dclient.hispeed.ch 1189316332 M * Bertl okay, off for tonight ... have a good one everyone! 1189316337 N * Bertl Bertl_zZ 1189319913 J * dna ~dna@150-238-dsl.kielnet.net 1189324879 Q * hparker Quit: g'nite 1189325373 J * bonbons ~bonbons@2001:960:7ab:0:20b:5dff:fec7:6b33 1189327564 N * DoberMann_ DoberMann 1189328867 J * fatgoose ~samuel@206-248-135-39.dsl.teksavvy.com 1189329992 J * ktwilight ~ktwilight@95.221-66-87.adsl-static.isp.belgacom.be 1189330120 Q * fatgoose Quit: fatgoose 1189330124 J * Julius ~julius@p57B26A4C.dip.t-dialin.net 1189330177 M * Julius guten morgen 1189330392 Q * ktwilight_ Ping timeout: 480 seconds 1189331062 J * DavidS ~david@p54811BD3.dip0.t-ipconnect.de 1189331851 J * meandtheshell ~markus@85.127.103.128 1189331922 J * jmcaricand ~jmcarican@d77-216-139-133.cust.tele2.fr 1189332622 M * matti Hello :-) 1189332875 Q * meandtheshell Quit: Leaving. 1189332907 Q * dsoul Read error: No route to host 1189332911 J * dsoul darksoul@vice.ii.uj.edu.pl 1189334042 J * meandtheshell ~markus@85.127.111.224 1189334195 Q * duckx Remote host closed the connection 1189334221 J * duckx ~Duck@tox.dyndns.org 1189338698 Q * nou Read error: Connection reset by peer 1189338768 Q * Aiken Quit: Leaving 1189338792 J * nou Chaton@causse.larzac.fr.eu.org 1189340022 Q * Julius Remote host closed the connection 1189340429 J * Julius ~julius@p57B26A4C.dip.t-dialin.net 1189340795 Q * nou Ping timeout: 480 seconds 1189340867 J * nou Chaton@causse.larzac.fr.eu.org 1189341370 Q * nou Ping timeout: 480 seconds 1189341371 J * nou Chaton@causse.larzac.fr.eu.org 1189341829 Q * meandtheshell Quit: Leaving. 1189341900 Q * nou Ping timeout: 480 seconds 1189342146 J * nou Chaton@causse.larzac.fr.eu.org 1189342302 P * DoberMann 1189342746 Q * nou Ping timeout: 480 seconds 1189342747 J * nou Chaton@causse.larzac.fr.eu.org 1189343261 Q * nou Ping timeout: 480 seconds 1189343437 J * nou Chaton@causse.larzac.fr.eu.org 1189343995 Q * nou Ping timeout: 480 seconds 1189344835 Q * phreak`` Quit: leaving 1189344888 J * phreak`` ~phreak``@deimos.barfoo.org 1189345049 P * friendly12345 1189345512 J * nou Chaton@causse.larzac.fr.eu.org 1189347451 Q * Julius Ping timeout: 480 seconds 1189347660 N * Bertl_zZ Bertl 1189347665 M * Bertl morning folks! 1189347665 Q * FireEgl Read error: Connection reset by peer 1189347673 M * daniel_hozac morning Bertl! 1189347745 Q * Baby Remote host closed the connection 1189348664 J * FireEgl FireEgl@FireEgl.CJB.Net 1189349680 Q * jmcaricand Quit: KVIrc 3.2.4 Anomalies http://www.kvirc.net/ 1189349961 J * ema ~ema@fw.galliera.it 1189351343 M * Bertl wb ema! 1189352638 J * fatgoose ~samuel@206-248-135-39.dsl.teksavvy.com 1189352660 N * ensc Guest3103 1189352669 J * ensc ~irc-ensc@p54B4F646.dip.t-dialin.net 1189352767 Q * Guest3103 Ping timeout: 480 seconds 1189353233 M * igraltist hi 1189353265 M * igraltist i have some question again about the function in mm/mmap.c 1189353283 M * igraltist and this is what does this vx_vmpages_add(mm, len >> PAGE_SHIFT) ? 1189353300 M * daniel_hozac performs the necessary accounting. 1189353307 M * daniel_hozac i.e. what it sounds like. 1189353361 M * igraltist will this working if i replace all this 1189353362 M * igraltist if ((mm->pax_flags & MF_PAX_SEGMEXEC) && (vm_flags & VM_EXEC)) 1189353362 M * igraltist mm->total_vm += len >> PAGE_SHIFT; 1189353377 M * igraltist with vx_vmpages_add ? 1189353412 M * daniel_hozac no. 1189353428 M * daniel_hozac you should only replace the second line. 1189353443 M * igraltist ok this mm->total_vm += len >> PAGE_SHIFT; 1189353444 M * Bertl what's the purpose of adding to total_vm on some condition? 1189353455 M * daniel_hozac yeah, that seems rather... awkward. 1189353460 M * Bertl I mean, it doesn't really make sense to me 1189353474 M * Bertl either the memory was allocated or not 1189353501 M * Bertl you will end up getting negative values when you free up that space 1189353522 M * igraltist i dont know i more a user more not, but i like to have pax support :) 1189353523 M * Bertl (unless you have an identical check for exactly those pages) 1189353562 M * daniel_hozac but not grsec? 1189353568 M * Bertl igraltist: yeah, well, you might ask the pax devels and have them explain it to you? 1189353603 M * igraltist no i dont need grsec 1189353670 M * daniel_hozac isn't grsec disableable on compile time? 1189353721 M * phreak`` daniel_hozac: depends. you can enable grsec on compile time, but tell the system to not enable any sysctl controlable function on boot 1189353829 M * igraltist http://rafb.net/p/TdAkZb61.html 1189353837 M * igraltist this situation i have 1189353890 M * igraltist first the vx_vmpages_add replace it and the next line again it if pax enabled is the replacment again there 1189353909 M * Bertl doesn't make sense 1189353934 M * Bertl this would account the len twice, when (mm->pax_flags & MF_PAX_SEGMEXEC) && (vm_flags & VM_EXEC) 1189353934 M * igraltist yes i stuck on this point 1189354005 M * daniel_hozac IIRC that's a PaX feature. 1189354016 M * Bertl hmm? 1189354018 M * daniel_hozac i remember reading something about it halving the available address-space. 1189354050 M * daniel_hozac for some non-executable page feature. 1189354067 M * Bertl well, in this case, I would adjust the shift 1189354093 M * Bertl otherwise you end up doing quite complex accounting twice in a row (adding unecessary overhead) 1189354121 M * daniel_hozac indeed... 1189354237 M * Bertl something like: 1189354239 M * Bertl http://paste.linux-vserver.org/6534 1189354263 M * Bertl note again, that you need to handle that identical in the free 1189354285 M * daniel_hozac right 1189354357 M * igraltist ok thanks is added :) 1189354404 M * Bertl note that you probably have to define the dup earlier and you should reuse it at least over the function 1189354468 M * Bertl also note that this code was a wild guess based on what daniel_hozac remembers reading about pax :) 1189354483 M * igraltist means this return the dup? 1189354531 M * Bertl no, but unless you compile with c99 enabled, which is not the case in the kernel, you cannot simply declare a variable in mid-function 1189354653 Q * fatgoose Quit: fatgoose 1189354724 J * fatgoose ~samuel@206-248-135-39.dsl.teksavvy.com 1189354729 Q * fatgoose 1189355503 M * bXi Bertl: http://apparoat.nl/Screenshot.png 1189355655 M * daniel_hozac ? 1189355681 J * Baby ~miry@195.37.62.208 1189355713 M * bXi just a random screenshot of my desktop 1189355765 M * igraltist only one virtual desktop :) 1189355776 M * igraltist i need 10 pieces 1189355857 M * bXi i've hidden my other virtual desktops 1189355890 M * bXi if i need to see them all i press super+e 1189356683 Q * arachnist Quit: brb/bbl 1189357454 J * arachnist arachnist@088156185052.who.vectranet.pl 1189358269 J * hparker ~hparker@linux.homershut.net 1189358339 M * Bertl wb hparker! 1189358365 M * hparker Afternoon Bertl 1189358885 Q * ema Quit: leaving 1189361166 J * meandtheshell ~markus@85.127.111.224 1189361287 M * hparker Is it safe to symlink guests from other partitions under /vserver? 1189361354 M * bXi isnt there a symlink in /etc/vservers/$NAME/ ? 1189361358 M * daniel_hozac there is. 1189361505 M * hparker Here's the situation, I have a 500 gig raid1 for vservers, but I'd prefer not to make it one huge partition.. Would hate to run out of inodes at some point or anything else 1189361571 M * hparker so I thought I'd use it as 5 100 gig partitions 1189361573 M * daniel_hozac well, IIRC the number of inodes scales with the filesystem size, so you're just as likely to run out if you use lots of smaller filesystems... 1189361628 M * eyck plus, you can tune the amount of available inodes on given filesystem on the fly.. 1189361731 M * hparker oh? 1189361775 A * hparker didn't find that info 1189361812 M * daniel_hozac man mke2fs: 1189361812 M * daniel_hozac -i bytes-per-inode 1189361813 M * daniel_hozac Specify the bytes/inode ratio. mke2fs creates an inode for every bytes-per-inode bytes of space on the disk. 1189361866 M * hparker Right, that's what I was going to play with 1189361881 M * hparker And from mke2fs.conf: inode_ratio = 8192 1189362346 M * hparker Ok, I'll gamble and make it one partition 1189362867 M * eyck I wouldn't recommend that, but it seems to be 'common sense' in daniel's world, so go ahead, have fun. 1189362900 M * hparker eyck: How would you do it? 1189362921 M * daniel_hozac eyck uses one filesystem per guest because he can get more RAM and disk whenever it's needed :) 1189362967 M * hparker I considered that, it's on lvm2 1189362982 M * hparker But... sounds like it could get to be a pita 1189362983 M * eyck I do use one filesystem per guest, for reason that are sadly misunderstood 1189363016 M * hparker Also sounded like a real easy way to set space limits 1189363053 M * hparker eyck: Well, being new to the channel I haven't had a chance to misunderstand, I'd like that chance ;) 1189363057 M * eyck I also use more then one filesystem for some guests, but I try not to speak about it, that would regarded as heavy heresy 1189363059 M * hparker pm if you want 1189363169 M * eyck you loose daniel's optimisations, but you gain hard and simple independence between guests, plus things like hot-backup, easy snapshotting, clean and fast with lvm+xfs (xfsdump/xfsrestore) combination, 1189363205 M * eyck plus, sometimes you need guaranteed IO, and that's not easy with single filesystem 1189363243 M * eyck for example I sometimes dedicate few spindles to guests with high demands and low tolerance for delays 1189363302 M * eyck basically, everything you usually do when your guests are not almost-identical copies of each other 1189363454 M * eyck if they are, single fs for everything beats everything else, and as I heard, you have to go with ext2/3 then. 1189363470 M * hparker yeah.. this is all a pair of spindles, so I don't get that.. and the snapshotting in lvm looked neat... why did you use xfs, the dump/restore or other reasons? 1189363487 M * eyck dump/restore plus online resizing. 1189363493 M * hparker ahh 1189363509 M * daniel_hozac yeah, because ext2/3 don't have that at all :) 1189363512 M * eyck I can't afford to shutdown my machines every 6 months just to add a pair of spindles 1189363584 M * hparker Any experience with xfs and maildir storage? 1189363587 M * eyck daniel_hozac: oh, goodie, didn't know they finally fixed that 1189363609 M * eyck yes, and no. I don't host large maildirs, those that I host, I haven't had any problems. 1189363644 M * eyck plus, I've got some critical paths that go through many-million files in directory scenarios, so that should be a bit tougher then normal maildir usage 1189363646 M * hparker i've always used ext2/3 on my servers.. Played with xfs on my desktop once but didn't see any difference with ext3 in what i use it for 1189363665 M * eyck xfs makes little sense on desktop 1189363668 M * hparker ahh 1189363673 M * eyck it doesn't like flaky hardware :( 1189363681 M * hparker heh 1189363702 M * eyck I do use it on laptop, but wouldn't recommend it 1189363745 M * hparker it worked fine... I just know the couple of problems I've had with ext2/3 I've been able to at least fix well enough to recover data, so keep falling back to it 1189363826 M * eyck I had problems on desktop with xfs, little less-then-ideal hdds, few bad sectors.. and recovery required 1.5G of ram 1189363846 M * hparker ahh 1189363849 M * eyck (this was when normal desktops had 256M, and beefy ones had 512M) 1189363888 M * hparker heh 1189363892 M * hparker hrrmm 1189363918 M * eyck on servers I've though - perfect. with very heavy usage, that made ext2/3 go cry in a corner 1189366793 M * arachnist i expect from my filesystem to be stable and reliable 1189366834 M * arachnist i wouldn't like to see my data gone foobar because power went out in a bad moment 1189366874 M * daniel_hozac nobody mentioned reiserfs, did they? :) 1189366875 M * arachnist xfs used to append stuff with some random data on hard reboots last time i used it, and i never had any problems with ext3 1189366904 M * arachnist i used to use reiserfs for a while, but only for my portage tree 1189367049 M * hparker I'm playing with ext4 for the portage tree on a couple of systems 1189367086 M * hparker Can't say i've seen any difference, but it hasn't blown up ;) 1189367112 Q * the-dude Quit: Is that a kangeroo? 1189367681 J * the-dude ~martijn@senturparks.xs4all.nl 1189368730 Q * the-dude Remote host closed the connection 1189368865 Q * bonbons Quit: Leaving 1189368920 J * comfrey ~comfrey@host-64-179-9-253.buf.choiceone.net 1189369220 Q * yarihm Quit: Leaving 1189369509 J * the-dude ~martijn@senturparks.xs4all.nl 1189370037 Q * the-dude Remote host closed the connection 1189370100 Q * ktwilight Read error: Connection reset by peer 1189371422 M * Bertl nap attack ... back later ... 1189371428 N * Bertl Bertl_zZ 1189371755 J * Aiken ~james@ppp121-45-255-55.lns2.bne4.internode.on.net 1189371784 Q * dna Quit: Verlassend 1189372366 Q * DavidS Quit: Leaving. 1189373878 J * the-dude ~martijn@senturparks.xs4all.nl 1189374461 Q * comfrey Ping timeout: 480 seconds 1189374647 J * comfrey ~comfrey@host-64-179-9-253.buf.choiceone.net 1189374850 J * ktwilight ~ktwilight@95.221-66-87.adsl-static.isp.belgacom.be 1189376497 Q * meandtheshell Quit: Leaving. 1189377940 Q * comfrey Ping timeout: 480 seconds 1189379458 Q * quasisane Remote host closed the connection 1189382388 M * hparker Is it normal for du -h to report 0 used and 0 free after setting a max disk limit with dlimit?