1105315697 M * nayco ok, I got my first vserver running, but still no eth... I'll continue tommorow. Thanks all for your help ! 1105315812 M * nayco bye. 1105315813 Q * nayco Quit: Bonne nuit ! 1105316296 J * _Plug ~plug@217.112.88.40 1105316296 Q * Plug Read error: Connection reset by peer 1105320316 M * Bertl night folks! 1105320322 N * Bertl Bertl_zZ 1105320570 Q * lilo Quit: brb 1105320849 Q * chrish01 Quit: chrish01 1105321579 J * lilo ~lilo@lilo.usercloak.oftc.net 1105322862 J * chrish01 ~chrish01@69.90.131.10 1105326090 Q * SiD3WiNDR Ping timeout: 480 seconds 1105326209 J * SiD3WiNDR luser@bastard-operator.from-hell.be 1105331431 Q * Schak Ping timeout: 480 seconds 1105332480 Q * lilo Quit: bbiab 1105333271 J * lilo ~lilo@lilo.usercloak.oftc.net 1105337201 Q * chrish01 Read error: Connection reset by peer 1105338413 J * chrish01 ~chrish01@69.90.131.10 1105344754 N * Bertl_zZ Bertl 1105344758 M * Bertl morning folks! 1105344766 M * chrish01 hey Bertl 1105344776 M * Bertl still up? 1105344793 M * chrish01 yup yup...amost 12:30am here 1105344883 M * Bertl hmm, after that comes 1:00am ? 1105344980 M * chrish01 hehe ya 1105347696 Q * chrish01 Quit: Leaving 1105347706 J * mhepp ~mhepp@r72s22p13.home.nbox.cz 1105347724 M * Bertl welcome mhepp! 1105347747 M * mhepp hello! 1105347764 M * mhepp what is this ng8.7 in the topic? 1105347793 M * Bertl ng is short for ngnet which is short for next generation networking ... 1105347807 M * mhepp :) 1105347813 M * Bertl it is a 'new' approach to network virtualization 1105347831 M * mhepp new network virtualization? 1105347855 M * Bertl well, gives you virtual interfaces and routing and arp and such stuff ... 1105347866 M * mhepp oooh! 1105347881 M * mhepp very good thing! 1105347925 J * chrish01 ~chrish01@69.90.131.10 1105347926 M * Doener arp as well? why would you need that? 1105347952 J * tchan_ ~tchan@c-24-13-81-164.client.comcast.net 1105347958 M * Bertl because it would be kind of strange if 10.0.0.1 asks for 192.168.0.2 on the same lan? 1105348017 M * Bertl (and also the inter server connections require virtual arp) 1105348028 M * chrish01 it would be strange, but should be legal 1105348065 Q * tchan Ping timeout: 480 seconds 1105348066 M * chrish01 especially if say the context is routing for multiple networks 1105348080 M * nox imho only with a /0 netmask 1105348806 M * Bertl actually a smart inter server arp-ing seems to be the trickiest part ... 1105348874 M * chrish01 especially since both could be connected to seperate networks with same IP space 1105349339 J * rs rs@ice.aspic.com 1105349349 M * rs hello folks 1105349386 M * Bertl who are you? ;) 1105349400 M * Bertl happy new year rs! (or did we already say that?) 1105349422 J * _are_ ~are@mail.foehl.de 1105349422 Q * sannes Read error: Connection reset by peer 1105349448 M * _are_ Hi 1105349468 M * rs Bertl: I think so, but happy new year again! :) 1105349553 M * _are_ server crasehed between 2:28 and 3:17 today. but this time I have parts of a kernel oops, just need to get some access to the picture we took of it which is sort of complicated today. 1105349576 M * Bertl k, will look at it when available ... 1105349588 M * Bertl okay, folks, moving out, back later ... 1105349593 N * Bertl Bertl_oO 1105349629 J * Val ~val@gj403.loria.fr 1105349632 M * Val Hi :) 1105349637 M * _are_ Hi Val 1105350675 J * tchan ~tchan@c-24-13-81-164.client.comcast.net 1105350905 Q * Val Quit: grmpf 1105351011 Q * tchan_ Ping timeout: 480 seconds 1105351987 J * Val ~val@gj403.loria.fr 1105353277 N * chrish01 chrish|zzz 1105354909 M * _are_ I seem to run into trouble with qemu within a vserver, I start qemu with some image and just nothing happens. 1105354943 M * _are_ teh display should be forwarded to my laptop via X11 forwarding. 1105355006 M * _are_ now 3 possibilities: a) qemu won't work with X11-forwarding, but '-nodisplay' doesn't work, either or b) qemu won't work correctly within vserver or c) qemu won't work correctly with opterons 1105355110 J * click click@80.65.50.35 1105355117 Q * click Remote host closed the connection 1105355152 M * _are_ can anyone confirm qemu runs on a), b) and c) or any subset? 1105355290 M * _are_ for a): I used ssh to the vserver and ssh back to my laptop and started qemu on that laptop, so forwarded from laptop->vserver->laptop and this works, it seems, so it is b) or c) not working 1105355486 J * BWare ~bware@212.26.196.41 1105355716 J * click click@80.65.50.35 1105356481 J * sannes ~ace@home.skarby.no 1105368885 M * _are_ umpf, my server lost all rbind mounts all of a sudden, no rbind worked anymore, had to --bind mount all directories 1105370866 J * Pazzo ~thomas@host130-250.pool8172.interbusiness.it 1105371755 Q * no_maam_ Ping timeout: 480 seconds 1105372438 M * Pazzo I've had a nice Ooops with 2.6.9-vs1.9.3: "attempt to access beyond end of device" / "dm-0: rw=2, want=1610612744, limit=1550057472" / "Unable to handle kernel paging request at 1105372438 M * Pazzo virtual address f3ffffff". this host is a backup server booting from a sw raid1 with all data on a 800GB LVM2 ext3 "disk". could this ooops be vserver related??? I'm running similar configured servers without vserver patch without any problems... 1105372491 M * Zoiah "attempt to access beyond end of device" is nearly always a configuration problem. (Broken partition tables, using the wrong device, etc.) 1105372594 J * no_maam ~erik@datenzone.de 1105372833 Q * mhepp Remote host closed the connection 1105372840 M * Pazzo hmmm... 1105372868 M * Pazzo partition tables seem to be right, lvm is up and running... 1105372887 M * Pazzo (4 pv's, 1vg, 1lv) 1105372902 M * Pazzo strange thing... 1105373123 M * meebey Zoiah: but it should not oops, in any case 1105373123 Q * sannes Read error: Connection reset by peer 1105373151 M * _are_ true, it gives log messages and some tool might die, but usually doesn't oops here, either 1105373175 M * _are_ and as I do drbd setups I had many wrong devices in earlier times ;) 1105373226 M * Pazzo :) 1105373251 M * meebey use ksymoops 1105373269 M * Pazzo kernel: dm-0: rw=2, want=1610612744, limit=1550057472 1105373282 M * Pazzo dm is the device-mapper, what is 0? 1105373292 M * Pazzo devicemappers first device? 1105373296 M * _are_ device 0, I'd say 1105374133 M * _are_ if there is anyone who can give me a hint on a kernel oops with vserver, opteron and megaraid, I currently have pictures of the Oops, unfortunately i think the last column is missing: ... 1105374133 M * _are_ http://vgn.lihas.de/bugs/opteron-megaraid-ops-1.jpg http://vgn.lihas.de/bugs/opteron-megaraid-ops-2.jpg 1105374189 M * _are_ to my understanding it crashes in megaraid-mbox while some 'find' attempts to access the scsi disks 1105374258 M * _are_ hmm, far more than last column missing, it seems. anyone can make any sens of the snippet i have? 1105374289 J * Mega\work ~Megabart2@host111-101.pool80182.interbusiness.it 1105374305 M * Mega\work hello 1105374310 M * _are_ Hi Mega\work 1105374410 M * Mega\work I've a question, can i create new vserver copying an exiting vserver? 1105374418 M * _are_ yes 1105374430 M * _are_ (at least I do so and it seems to work) 1105374506 M * Mega\work I've created the first vserver with the command newvserver -v ecc 1105374525 M * _are_ i use alpha tools and new configuration, the script i use for copying is at http://lihas.de/vcopyserver.sh 1105374571 M * _are_ last tuple of the IP address prefixed with 10 is the context id here 1105374583 M * _are_ have to run, will be back in 6h 1105374590 Q * _are_ Quit: Disconnecting 1105374603 M * Mega\work but this command get the package from internet and request too many time to create a new vserver 1105374668 M * Mega\work request 0 require 1105374672 M * Mega\work request = require 1105374949 M * Pazzo Mega\work: vserver vs1 stop, cp -a /vservers/vs1 /vservers/vs2, cp -a /etc/vservers/vs01 /etc/vservers/vs02, have a look into /etc/vservers/vs02, modify the "run" and "vdir" links and modify name, uts/nodename and interfaces/whatever 1105374972 M * Pazzo (if you are using vs1.9.x and alpha-utils 1105374973 M * Pazzo ) 1105375029 M * Pazzo if you are running stable vs on 2.4 with old utils changes in etc will be different (easier :-) 1105375132 M * Mega\work Pazzo, I using the debian package for install utils 1105375153 M * Mega\work and the kernel is 2.4 with grsec 1105375296 M * Mega\work but I had thought to compying the directory of vserver 1105375434 M * Mega\work excuse me if I'm OT 1105375467 M * Mega\work Someone use bind in a dns server? 1105375895 M * Pazzo Mega\work: there should be no problem with copying the vserver directory - reread my post! bind8 should run without any problems, bind9 needs to be recompiled with "--disable-linux-caps" 1105375935 M * Mega\work Pazzo, my problem is another! 1105376060 M * Mega\work At the moment, when i link a client computer in a lan, for resolv the name of new client, i must edit the zone file and insert the ip of new client and the name in a lan 1105376134 M * Mega\work is it a method for automatically associate the hostname at the client ip? 1105376225 M * Pazzo ????? 1105376238 M * Pazzo sorry, I didn't understand your problem... 1105377610 Q * grecea Read error: Connection reset by peer 1105377730 J * grecea ~grecea@h-195-22-237-74.mdl.net 1105378364 M * Mega\work thank's Pazzo, I have resolved 1105378367 M * Mega\work :D 1105379762 J * ensc ~ircensc@ultra.csn.tu-chemnitz.de 1105380321 J * sannes ~ace@home.skarby.no 1105380505 M * sannes Pazzo : running reiserfs? and have made a partition without reboot inbetween, seem to remember an issue .. 1105380794 Q * rs Quit: leaving 1105381067 Q * Mega\work Quit: Leaving 1105381120 M * Pazzo sannes: running ext3 1105381261 M * Pazzo sannes: I did pvcreate, vgcreate, lvcreate, mkfs.ext3, mount, change fstab, reboot (to check if it is reboot-proof - that's not always sure with todays "stable" kernels :o) 1105381266 M * Pazzo everything worked fine 1105381344 M * Pazzo this backup file server has many inodes and hard links, something like 250/300gb of the 800gb partition is used (without hard links ther would be some tb :-) 1105381370 M * Pazzo after two weeks -> ooops, probably while running rsync 1105381443 M * Pazzo (I have to leave now, will be back in something like 90 minutes) 1105381449 N * Pazzo PazZzzzoooo 1105382273 M * we2by I have the following problem 1105382298 M * we2by I have 3 vservers. and I have a share on the host OS, it is /vserver/share 1105382313 M * we2by I want vserver1 has read access to it 1105382323 M * we2by and I want vserver2 read+write access 1105382329 M * we2by and vserver3 write access 1105382335 M * we2by is this posible? 1105382698 M * nox mount -o bind,ro /vserver/share /vserver/1/share 1105382705 M * nox mount -o bind,rw /vserver/share /vserver/2/share 1105382722 M * nox should work fine 1105382978 M * we2by cool 1105384230 M * Snow-Man Where are the 2.6.10 kernel patches again? 1105384393 J * _BWare_ ~bware@s4s.xs4all.nl 1105384463 M * Snow-Man http://vserver.13thfloor.at/Experimental/patch-2.6.10-vs1.9.3.17.diff ; is that decent? :) 1105385237 M * dominance frost: yep 1105385782 N * chrish|zzz chrish01 1105385900 Q * sannes Read error: Connection reset by peer 1105386033 N * Bertl_oO Bertl 1105386041 M * Bertl evening folks 1105386059 M * Bertl nox: hmm, dream on ... 1105386463 M * Bertl okay, night folks! 1105386468 N * Bertl Bertl_zZ 1105386493 M * chrish01 night Bertl_zZ 1105386975 Q * chrish01 Quit: Leaving 1105386996 N * _Plug Plug 1105388175 N * PazZzzzoooo Pazzo 1105388312 Q * _BWare_ Ping timeout: 480 seconds 1105388669 J * weasel ~weasel@weasel.noc.oftc.net 1105388675 J * monrad ~monrad@213083190130.sonofon.dk 1105389477 Q * BWare Ping timeout: 480 seconds 1105390257 J * _are_ ~are@dsl-084-056-145-111.arcor-ip.net 1105390278 J * sannes ~ace@home.skarby.no 1105390306 M * sannes nox : if you have bme yes, but on vanilla, no.. 1105390316 J * chrish01 ~chrish01@69.90.131.10 1105390623 M * _are_ hi 1105390665 M * chrish01 hi _are_ 1105390746 J * BWare ~bware@212.26.196.41 1105390754 M * sannes :) 1105390755 M * sannes hiello 1105390860 Q * serving Read error: Connection reset by peer 1105391555 Q * sannes Read error: Connection reset by peer 1105391668 J * sannes ~ace@home.skarby.no 1105393424 J * xTomx ~x@p5480DBBD.dip.t-dialin.net 1105393816 J * nayco ~nayco@lns-vlq-47-nan-82-252-247-66.adsl.proxad.net 1105393820 M * nayco 'llo ! 1105394055 M * Pazzo hi nayco 1105394065 M * nayco ;-) 1105394158 M * nayco I managed to run my first vserver today, with apache in it... 1105394221 M * we2by nice job nayco 1105394442 M * Pazzo nayco: happy with linux-vserver? 1105394486 M * nayco well, this is the start, I just displayed a single index.html... ;-) 1105394601 M * nayco But but it work. What I find sad is the lack of support for Mandrake (Ok, this may be not the best distro for server, but that's not the point) 1105394629 M * we2by Pazzo, is it true that each vserver will take minimum 128mb ram? 1105394648 M * nayco I mean, not in the kernel, but the util-vserver package seem to work better for RH or deb... 1105394655 M * Pazzo we2by: nope. who told you so?? 1105394672 J * yarihm ~yarihm@80-218-1-3.dclient.hispeed.ch 1105394673 M * nayco s/seem/seems/ 1105394712 M * we2by Pazzo, I can´t remember, it was said int his channel 1105394729 M * nayco anyway, I rooled back to the stable version of the package, and now, I _can_ build my vservers ;-) 1105394751 M * nayco s/rooled/rolled/ ;-) 1105394887 Q * grecea Ping timeout: 480 seconds 1105394916 M * nayco Is it ensc That maintains the Mdk packages ? 1105394924 M * Pazzo we2by: maybe someone was talking about uml (usermode linux) - every uml system eats up all the memory it is given. each vserver uses exactly as much memory as the processes running in it's context are using 1105395215 M * Pazzo nayco: I've no idea about the mdk packages. I was running stable vserver kernels at the beginning but switched to devel soon. I'm using nothing but debian and building kernel and userspace packages by myself as I've found no useful alternative... the "official" debian packages are really strange... 1105395399 M * nayco Well, the mdk devel packages do not work alone, I had to create files and dir by hand, and when started, the vserver has no IP (No alias on eth0 is created, in fact !). So I tried today the stable 0.30 package, and It worked... But I still got many errors when starting (I think about services acting as they where alone on the system, but which are prevented some actions by the kernel, 127.0.0.1=>localhost not working, etc...) 1105395556 M * nayco Sadly there is no system image for Mdk (I could try to make one, and upload it somewhere, but I think I don't know enough to do it alone). 1105395659 M * nayco And no script to download from the mdk mirrors (like there is for Rh, Fc, deb, suse...). I'd like to contribute to this... Is it Bertl who manages the utils-vserver packages ? 1105395756 J * nayco_ ~nayco@lns-vlq-47-nan-82-252-247-66.adsl.proxad.net 1105395767 Q * nayco_ Quit: 1105395809 M * nayco oops 1105395833 M * we2by cool 1105395938 M * Pazzo nayco: bertl is the big kernel hacker :) enrico is the userspace man - but I don't know if he is also "responsible" for packages... 1105396105 M * _are_ packages are done by whoever likes to do them, e.g. dominance did unofficiall debian package. If you know how to do Mandrake ones, rarely anyone wil object 1105396132 M * _are_ hoewever, i think Bertl announced yesterday he will do MDK 10 packages himself when he manages to install MDK 10 on his box 1105396183 M * dominance _are_: my pacakges are unofficial for now.. but they are targetted for Debian.. i.e. they will be uploaded to experimental when the time has come 1105396216 M * _are_ dominance: i know, nonetheless you just started doing them an dlater found out the debian package maintainer likes them :-) 1105396231 M * chrish01 dominance, great! any idea when they will allow for experimental integration? 1105396279 M * nayco ok 1105396329 M * dominance _are_: not really.. i did them FOR debian.. and the objective was to make them be liked by ola.. whatever it takes.. 1105396355 M * dominance chrish01: we're working on it.. that's about as good as i can fix the date so far =) 1105396381 M * _are_ :-> 1105396384 M * nayco Great, I'll try Bertl's packages as soon as possible to report bugs 1105396386 M * chrish01 dominance, great...thanks for taking the time to do that 1105396415 Q * yarihm Read error: Connection reset by peer 1105396467 M * nayco I got another question: Is it possible to create a vserver in an existing directory (Typically a mountpoint to a dedicated partition) without having vserver...build complain ? 1105396472 M * dominance chrish01: hehe, sure =) 1105397329 M * Pazzo dominance: are you talking about http://backend.verfaction.de/~kk/util-vserver/ ? 1105397429 M * _are_ nayco: i can't see a real problem there, no need to ever call vserver build. you just need to construct the config data (e.g. copy it from some other server and modify it) and you may want to change the init process 1105397444 M * dominance Pazzo: yep 1105397574 M * nayco Yes, You mean that the first vserver is the "hardest" (The only one ?) to build.... But what about factorising disk space by unifying If have to copy my template vs to a new one ? The hardlinks will be copied as full file to the new partition, no ? 1105397607 M * _are_ new partition = new file, hardlinks are only possible on same filesystem 1105397636 M * nayco Oh, yes... I must admit that I _never_ use them. (The hard links.) 1105397654 M * _are_ wel, for unifying vservers they are essential 1105397714 M * nayco So, I may consider using a large partition for all my vservers (I mean the system, the data can be eslewhere), instaed of one parttion for each vserver... But is it as secure ? 1105397749 M * nayco _are_: Yes, I read about that... This is a great idea. 1105397859 M * _are_ 'as secure' as in 'as hard to break in like into vservers on differente partitions: yes, to my understanding. 1105397910 M * _are_ 'as secure' as in: as easy to recover in case of a filesystem failure: no, less moving data on 1 partition is always the better condition for filesystem recovery 1105397967 J * serving ~serving@213.186.173.201 1105397996 M * _are_ nn 1105398001 N * _are_ are|afk 1105398715 J * chrish01_ ~chris@69.90.131.10 1105398727 Q * chrish01_ Quit: 1105398728 M * nayco ok, thx. I'll try with one partition for my vserver's sytems, and maybe make one partition each for their data. 1105399016 N * chrish01 chrish|lunch 1105400072 M * Pazzo dominance: I downloaded your package for doing a diff against my one. we did it more or less the same, but there are some differences: 1105400072 Q * we2by Read error: Connection reset by peer 1105400077 J * we2by ~we2by@dc5146d009.adsl.wanadoo.nl 1105400086 M * Pazzo why "conflicts: vpopmail-bin"? 1105400121 M * Pazzo and why does util-vserver depend on vlan? 1105400173 M * dominance vpopmail-bin comes from the current debian main deb 1105400181 M * Pazzo who did the cleanup in /usr/lib/util-vserver/functions (replace alle the rm's and cp's with variables)? 1105400184 M * dominance and depend on vlan is imported from the upstream spec 1105400201 M * dominance i.e. ensc does think vlan is a good idea for util-vserver 1105400213 M * dominance the current development snapshot does have only the suggests: vlan 1105400223 M * dominance no more depends.. maybe i should update *g* 1105400230 M * Pazzo :) 1105400244 M * dominance the source is cvs from some days ago 1105400250 M * dominance all in the ndim_rollup.dpatch =) 1105400261 M * Pazzo we did a little modification to /usr/lib/util-vserver/vserver-build: 1105400276 M * Pazzo 59: debootstrap ... -- -d [-m [-s