1206490574 Q * julius_ Remote host closed the connection 1206491025 Q * Piet Quit: Piet 1206491050 Q * mrfree Ping timeout: 480 seconds 1206494093 Q * transacid Ping timeout: 480 seconds 1206495515 J * transacid ~transacid@transacid.de 1206495810 M * nox lol Bertl this was the 2^13 time someone asked you this question, this is worth a price 1206495825 M * nox but it is really hard 2 find 1206495852 M * Bertl maybe we should make a wiki page for that 1206495884 M * nox or force hium 2 use 2.2 1206495923 M * nox but i will try to write it down in the wiki this days 1206496013 M * nox btw ipv6 works like a charme 1206496031 M * nox i am still completly happy with vserver <3 1206496063 M * Bertl glad to hear 1206496108 M * nox still lots of probs to be cleared with 6.24? 1206496456 M * Bertl yep, didn't have much time yet 1206496695 M * nox ic 1206496717 M * nox well you spent a couple of manyears already to the project 1206496726 M * nox thx 4 that 1206496772 M * nox 4 my servers there is no actual need for .24 1206496898 M * Bertl well, I'm pretty sure we'll get there .. just takes time :) 1206496976 M * nox take your time 1206497076 M * Bertl will do, thanks for the acknowledgment ... 1206497549 M * nox :) 1206497555 M * nox sleep well 1206498067 M * Bertl yep, good idea ... off to bed now ... have a good one too! 1206498071 N * Bertl Bertl_zZ 1206500984 M * Supaplex heh I asked #debian-kernel that a few hours ago. the changelog has it on packages.debian.org 1206501005 M * Supaplex sadly 2.3 isn't in sid yet. hopefully it makes it before the freeze. 1206501583 Q * mire Ping timeout: 480 seconds 1206503299 Q * nenolod Quit: nenolod: I want to make ellsworth the second silicon valley | its the most absurd impossible goal I can think of 1206503458 J * nenolod ~nenolod@ip70-189-74-69.ok.ok.cox.net 1206503490 J * besonen_mobile_ ~besonen_m@71-220-235-129.eugn.qwest.net 1206503531 Q * Aiken Remote host closed the connection 1206503542 J * Aiken ~james@ppp121-45-199-100.lns1.bne1.internode.on.net 1206503868 Q * besonen_mobile Ping timeout: 480 seconds 1206504262 J * etteyafed ~etteyafed@68-113-163-245.dhcp.plt.ny.charter.com 1206504308 M * etteyafed I setup my /etc/vserver/servername/sched/ directory and the limits don't seem to be taking effect. 1206504320 M * etteyafed Is there something I need to do to activate them? 1206504907 M * etteyafed Does anyone have a working scheduler setup to manage vserver CPU resources? 1206504949 M * etteyafed If so could you pastebin me a directory listing and cat a couple of the files? 1206505327 Q * balbir_ Remote host closed the connection 1206507370 J * balbir ~balbir@122.167.204.163 1206509072 M * etteyafed I swear I have read through every doc that references the sched dir and there is no reason that this shouldn't work, cat /proc/virtual/4040/sched even outputs the proper data 1206513464 J * cryptronic ~oli@p54A3B6BF.dip0.t-ipconnect.de 1206514524 J * JonB ~NoSuchUse@77.75.164.169 1206514535 J * Slydder ~chuck@194.59.17.53 1206514541 J * sharkjaw ~gab@64.28.12.166 1206514980 J * dna ~dna@223-197-dsl.kielnet.net 1206515175 M * Slydder morning all 1206515298 M * daniel_hozac etteyafed: did you set SCHED_HARD in flags? do you have CONFIG_VSERVER_HARDCPU=y in your kernel? 1206515376 Q * cryptronic Quit: Leaving. 1206515525 M * etteyafed well I don't know about the kernel you built it 1206515545 M * etteyafed but I tried setting te flag and I think I did it wrong 1206515555 M * daniel_hozac that definitely works, i use the CPU limiting. 1206515558 M * etteyafed I out sched_hard in the flags file 1206515570 M * etteyafed put* 1206515587 M * etteyafed is it supposed to be SCHED_HARD=1? 1206515594 M * etteyafed or SCHED_HARD? 1206515601 M * daniel_hozac no, just sched_hard on a line of its own. 1206515627 M * etteyafed I did that, the results should be visible in top though right? 1206515653 M * daniel_hozac yes. 1206515687 M * daniel_hozac you don't have idle-time enabled, do you? 1206515739 M * etteyafed I do, but it is setup properly. if I cat /proc/virtual/####/sched everything looks fine 1206515766 M * daniel_hozac if you have idle-time enabled and you don't have other processes using the CPU, the guest will be able to consume 100% CPU. 1206515773 M * daniel_hozac it being idle and all. 1206515806 M * etteyafed I thought it would abide by the fill-rate2 and interval2 settings in that case 1206515808 M * etteyafed oops 1206515819 M * daniel_hozac that's only if you have multiple idle-time guests. 1206515820 M * etteyafed so I should remove idle-time 1206515877 M * etteyafed ok, so when I get the other guests up I can put idle time back and it should use the R2 and T2 files 1206515904 M * daniel_hozac if the other guests also want the CPU, yes. 1206515918 M * etteyafed I wish there were more people helping you guys on this to write docs. I don't have the time myself, or I would. 1206515989 M * etteyafed About how many devs are there that commit regularly? 1206516014 M * daniel_hozac 3 or 4, depending on your definition of regularly. 1206516054 M * etteyafed Well, that's not very many for a project like this really. 1206516081 M * etteyafed And I am sure most of you have other jobs. 1206516100 M * daniel_hozac yep. 1206516285 M * etteyafed Well we all are very grateful for your efforts. Are you looking for devs, or do you (and the others) prefer not to have the added headaches of a larger community? 1206516325 Q * Aiken Remote host closed the connection 1206516361 M * daniel_hozac there are lots of things to do, so more manhours wouldn't be a bad thing. 1206516393 M * daniel_hozac but we mostly need documentation. 1206516427 M * etteyafed I can see that. Its mostly there, just missing details and examples. 1206516568 M * etteyafed I wish the kernel could change the way it handles process permissions on the fly. IE make processes invisible to users other than root and their owners without having to patch the kernel and make a custom build. I like to keep up with security updates without having to patch and recompile all the time. 1206516991 Q * balbir Ping timeout: 480 seconds 1206516997 J * DavidS ~david@85.125.165.34 1206517065 M * etteyafed DO I need to delete the fill-rate2 and interval2 files also to get rid of the idle-time problem I was having? 1206517174 M * etteyafed the answer is yes I do 1206517189 M * etteyafed great. it works now 1206517270 M * daniel_hozac no, just removing idle-time is sufficient. 1206517292 M * daniel_hozac fill-rate2 and interval2 won't be used in that case. 1206517298 M * etteyafed It wouldn't work until I removed them 1206517319 M * etteyafed the process was still getting 100% of the cpu 1206517329 M * etteyafed well the cpu it was running on 1206517443 Q * dna Quit: Verlassend 1206517643 J * balbir ~balbir@122.167.193.12 1206517817 J * tobifix ~tobifix@IVV7KNALLER.UNI-MUENSTER.DE 1206517825 M * tobifix good morning ;) 1206519034 Q * nkukard synthon.oftc.net tachyon.oftc.net 1206519794 J * nkukard ~nkukard@196.212.73.74 1206520165 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it 1206520192 Q * mrfree Remote host closed the connection 1206520341 M * gebura hi 1206520345 J * friendly12345 ~friendly@ppp121-44-224-29.lns2.mel4.internode.on.net 1206520907 N * DoberMann[ZZZzzz] DoberMann 1206521738 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it 1206522003 J * jsambrook ~jsambrook@aelfric.plus.com 1206522499 J * Julius ~julius@p57B27570.dip.t-dialin.net 1206522633 J * rgl ~rgl@lx2-84-90-11-254.netvisao.pt 1206522636 M * rgl hello 1206522651 M * rgl JonB, are you there? 1206522667 M * JonB yes 1206522690 M * rgl JonB, great :D 1206522710 M * rgl JonB, you know what? my raid array is fubar again :( 1206522741 M * rgl JonB, the server magically rebooted, and when it come back again, no raid :/ 1206522748 M * JonB thats not good 1206522758 M * JonB can you see the partitions? 1206522766 M * rgl and now when I do a mdadm --assemble /dev/md3 /dev/sda7 /dev/sdb7 1206522782 M * rgl it just says: mdadm: cannot open device /dev/sdb7: Device or resource busy !!! 1206522796 M * rgl lsof | grep sdb7 shows nothing :/ 1206522801 M * daniel_hozac run mount. 1206522809 M * rgl and mdadm --examine /dev/sdb7 is okay. 1206522812 M * daniel_hozac or rather, cat /proc/mounts 1206522830 M * JonB rgl: cat /proc/mdstats and see if some raid is already started 1206522899 M * rgl it at http://pastie.caboo.se/170776 1206522939 M * rgl before the reboot there was a single message at /var/log/messages that says Mar 26 02:11:10 host mdadm: DeviceDisappeared event detected on md device /dev/md/3 1206523010 M * JonB rgl: fdisk -l /dev/sda and /dev/sdb 1206523015 J * Aiken ~james@ppp118-208-122-28.lns4.bne4.internode.on.net 1206523015 M * JonB rgl: also do a dmesg 1206523022 M * JonB and use pastie 1206523124 M * rgl http://pastie.caboo.se/170776 1206523131 M * rgl dmesg does not say anything. 1206523150 M * JonB strange 1206523153 M * rgl (there is no new stuff after this) 1206523184 M * JonB cat /dev/sdb7 >> /dev/null 1206523197 M * rgl btw, no vm are running. 1206523221 M * rgl seems to run okay. 1206523228 M * JonB rgl: abort it 1206523232 M * JonB rgl: pvscan 1206523239 M * JonB and vgdisplay -v 1206523273 M * JonB i'm thinking that the Volume Group might have started to use sdb7 and not the raid device 1206523309 M * rgl I've though that too. but vgdisplay not pvscan show sda7 or sdb7 1206523316 M * rgl err 1206523325 M * rgl I've though that too. but vgdisplay and pvscan do not show sda7 or sdb7 1206523339 M * JonB have you tried rebooting again? 1206523345 M * rgl no 1206523378 M * rgl should I? 1206523385 M * JonB i dunno 1206523389 M * JonB i would try it 1206523395 M * JonB "windows trick" 1206523400 M * JonB but it might work 1206523437 M * JonB mdadm: cannot open device /dev/sdb7: Device or resource busy 1206523458 M * JonB i understand this as the system thinks that sdb7 is busy 1206523478 M * JonB but it does not seem to be mounted, or in a VG 1206523500 M * rgl true. so odd. 1206523605 M * rgl going to try reboot. 1206523734 M * rgl omg!!! 1206523740 M * rgl the array is now up after reboot 1206523817 M * JonB heh 1206523825 M * JonB "windows trick" 1206523899 M * rgl but now vserver is getting wacko on me again :( 1206523909 M * rgl vserver ocelot start 1206523916 M * rgl Unknown tag; use '-l' to get list of valid tags 1206523930 M * rgl /proc/uptime can not be accessed. Usually, this is caused by etc :( 1206523952 M * JonB rgl: did you run /etc/init.d/util-vserver start ? 1206524012 M * rgl JonB, I think so; it was on the boot scripts 1206524022 M * rgl should I run it again? 1206524030 M * JonB i dunno 1206524087 Q * ruskie Read error: Connection reset by peer 1206524365 M * rgl I started it, but still no good :( 1206524372 J * ruskie ruskie@ruskie.user.oftc.net 1206524375 M * rgl daniel_hozac, can you help? 1206525154 M * tobifix i have a verser host running debian, now i wanted to create a gentoo guest as explained here: http://www.gentoo.org/proj/en/vps/vserver-howto.xml#doc_chap3 1206525168 M * tobifix if i try to start the guest, it throws a lot of errors 1206525169 M * tobifix :( 1206525220 M * tobifix here is an output: http://nopaste.de/p/aHxevWsFhb 1206525240 M * tobifix i don't know how to fix this :( 1206525251 M * Hollow tobifix: which stage did you use? 1206525255 M * tobifix 3 1206525285 M * tobifix stage3 and 64bit 1206525290 M * Hollow there are a lot of stage3's ;) 1206525293 M * tobifix as my host also has 64bit 1206525296 M * Hollow whee di you download it? 1206525297 M * tobifix Hollow, wait 1206525298 M * tobifix ;) 1206525317 M * tobifix i used this: stage3-amd64-2007.0.tar.bz2 1206525337 M * tobifix as its the normal 64bit stage3, ;) 1206525343 M * Hollow won't work ... use a stage4 from the url in the howto (http://people.linux-vserver.org/~hollow/stages/ ) 1206525365 M * tobifix aah, ok, thanks a lot dude 1206525365 M * tobifix ; 1206525367 M * tobifix ;) 1206525395 M * rgl daniel_hozac, Hollow: how odd, after I reboot with "default" on apps/init/mark config setting it works. when I run env - /usr/sbin/vserver ocelot start it works too.. 1206525471 M * Hollow rgl: hm? 1206525576 M * hparker Hollow: Did you roll up a new stage yet? 1206525583 M * rgl Hollow, I'm having trouble running the vserver utility :( 1206525604 M * Hollow hparker: no, i'm waiting for a fixed openrc ebuild to appear in portage ... 1206525621 M * rgl Hollow, it does not work when I run it with strait from command line, but works with env - /usr/sbin/vserver ! 1206525655 M * hparker Hollow: heh... I think that may of happened... Maybe :P 1206525660 M * rgl well, kinda of, on one guest it fails with vnamespace: vc_enter_namespace(): Invalid argument 1206525781 M * Hollow rgl: which version of util-vserver? 1206525826 M * hparker Hollow: btw, latest ~ of util-vserver and kernel works peachy ;) 1206525864 M * rgl Hollow, 0.30.214 1206525886 M * rgl Hollow, but as been hapenning to me before that version too. 1206525915 M * rgl now I have one vserver that I cannot enter... it says: vnamespace: vc_enter_namespace(): Invalid argument :( 1206525919 M * hparker rgl: .215 works fine 1206525920 M * Hollow *shrug* 1206526193 Q * DavidS Quit: Leaving. 1206526727 N * Bertl_zZ Bertl_oO 1206526750 J * ktwilight ~ktwilight@87.66.205.10 1206527028 Q * ktwilight_ Ping timeout: 480 seconds 1206527248 M * rgl JonB, oh thx a bunch :-)) 1206527265 M * rgl well, everyone :D 1206527288 M * rgl Bertl_oO, hey :) 1206527653 M * rgl from http://pastie.caboo.se/170776 it seems my util-vserver is linked against glibc? 1206528352 M * snooze is it possible to do some sort of traffic shaping of vservers? 1206528515 M * JonB snooze: i see no reason why you cant use normal traffic shaping 1206528633 M * snooze i really should learn about traffic shaping/qos in linux.. overall 1206528636 M * snooze guess that would help a bit :P 1206528838 Q * whisk Ping timeout: 480 seconds 1206529186 J * mire ~mire@37-174-222-85.adsl.verat.net 1206530165 Q * mrfree Remote host closed the connection 1206530535 J * dna ~dna@216-230-dsl.kielnet.net 1206530791 Q * friendly12345 Quit: Leaving. 1206531177 Q * ruskie Read error: Connection reset by peer 1206533685 J * cryptronic ~oli@p54A3B6BF.dip0.t-ipconnect.de 1206533698 Q * mire Ping timeout: 480 seconds 1206533782 J * ruskie ruskie@ruskie.user.oftc.net 1206534052 Q * Aiken Remote host closed the connection 1206534184 Q * balbir Read error: Operation timed out 1206534253 M * Bertl_oO rgl: why do you think so? 1206534323 M * rgl Bertl_oO, even though "Use dietlibc: yes" there is a "vserver(2) syscall#: 236/glibc 1206534323 M * rgl " which I don't known what it is. 1206534341 M * Bertl_oO it is the way the syscall is invoked 1206534356 M * rgl ah, so al'ls good? 1206534365 M * Bertl_oO looks good to me :) 1206534382 M * rgl :D 1206534384 M * Bertl_oO (except for the /var/lib/vservers :) 1206534406 M * rgl what should that read? 1206534425 M * Bertl_oO on every distro except debian it is /vservers 1206534438 M * rgl I'm on ubuntu *G* 1206534446 M * sid3windr /var/lib/vservers is fine! ;) 1206534480 M * rgl what about the vserver enter error: "vnamespace: vc_enter_namespace(): Invalid argument"? 1206534512 M * rgl I have 3 guests running, but I cannot enter that one 1206534537 M * Bertl_oO without Linux-VServer debugging enabled, hard to tell 1206534560 M * daniel_hozac sid3windr: actually, that means that if you lack the chroot barrier, the guest will have access to your host's /var. 1206534574 M * sid3windr as compared to your host's / ? 1206534605 M * daniel_hozac no, that's semi-protected by the bind mount. 1206534622 M * sid3windr which bind mount? 1206534634 M * daniel_hozac the rbind from to / 1206534644 M * Bertl_oO so we finally have a security reason for putting it in / ? 1206534645 M * sid3windr ahh 1206534650 A * Bertl_oO rejoices 1206534653 M * sid3windr interesting :) 1206534663 M * rgl Bertl_oO, you mean util-vserver compiled with debug info? or the kernel itself? 1206534682 M * tobifix rgl, i think he means the kernel 1206534686 M * daniel_hozac well, IIRC, the host's / is still accessible. it just needs some trickery. 1206534690 M * Bertl_oO rgl: the kernel 1206534726 M * rgl Bertl_oO, how can I do that? is there Kconfig knob? 1206534745 M * Bertl_oO rgl: the question is, will this remain after a reboot? 1206534837 M * rgl Bertl_oO, I'll try it this night. 1206534906 M * Bertl_oO do you get any messages in dmesg? 1206534959 M * rgl nope 1206535007 J * balbir ~balbir@122.167.209.231 1206535039 M * Bertl_oO rgl: what kernel version? 1206535061 M * rgl Bertl_oO, 2.6.22.18-vs2.2.0.6 1206535093 M * rgl (yeah, I have to upgrade the kernel and vserver to one version up) 1206535199 M * Bertl_oO do you have a /proc/virtual//nsproxy ? 1206535335 M * rgl Bertl_oO, yes 1206535531 M * Bertl_oO could you upload that for me for your guests? 1206535623 M * rgl Bertl_oO, upload what? 1206535630 M * rgl Bertl_oO, the contents os nsproxy? 1206535631 M * Bertl_oO the contents of that file :) 1206535702 M * rgl Bertl_oO, http://pastie.caboo.se/170829 1206535703 Q * balbir Ping timeout: 480 seconds 1206535753 M * Bertl_oO is that the one you cannot enter? 1206535757 M * rgl yesd 1206535767 M * Bertl_oO could you upload one you can enter too? 1206535851 Q * ruskie Read error: Connection reset by peer 1206535861 M * rgl Bertl_oO, http://pastie.caboo.se/170829 1206535883 M * Bertl_oO as expected, you guest doesn't use a namespace 1206535891 M * Bertl_oO what util-vserver version is that? 1206535892 M * rgl Bertl_oO, humm, no Namespace column! 1206535929 M * rgl 0.30.214 1206535955 M * Bertl_oO daniel_hozac: any fixes in this regard? i.e. trying to enter a namespace of a guest which has none? 1206535964 M * Bertl_oO (in 0.30.215 I mean) 1206535988 M * Bertl_oO rgl: did you add nonamespace at some point (for this guest) and later remove it? 1206536006 M * rgl Bertl_oO, no ideia. 1206536007 M * daniel_hozac i don't see how you'd get a guest without namespaces. 1206536030 M * rgl Bertl_oO, I don't remeber messing with that. 1206536037 M * Bertl_oO daniel_hozac: isn't that an option? 1206536077 M * rgl Bertl_oO, is that some file we can create inside the guest configuration? 1206536093 M * Bertl_oO I thought so, but daniel_hozac should know for sure 1206536220 M * Bertl_oO nonamespace (it's still listed on the flower page) 1206536242 M * Bertl_oO try the following for your guest: 1206536257 M * Bertl_oO touch /etc/vservers//nonamespace 1206536270 M * Bertl_oO then see if you can enter it 1206536335 M * rgl vserver piranha enter 1206536335 M * rgl No command given; use '--help' for more information. 1206536351 M * rgl no error message, but still, it does not work. 1206536362 M * Bertl_oO that sounds like a bug in util-vserver ... 1206536401 M * Bertl_oO but to check my theory, could you create a new guest with the nonamespace flag set (after creation, before startup) 1206536419 M * bXi whats the latest kernel that stable runs on? 1206536430 M * Bertl_oO 2.6.22.x 1206536434 M * bXi hmm 1206536448 M * bXi that doesnt include ipv6 right? 1206536459 M * Bertl_oO stable no, but devel does 1206536471 M * bXi i'm running 2.3.0.17 i think 1206536482 M * bXi oh a bit newer 1206536487 M * bXi 2.3.0.24 1206536515 M * bXi had a reproducable kernel panic now 1206536532 M * bXi havent seen the panic itself yet since i dont have a screen connected 1206536540 M * bXi but it has something to do with my sata stuff 1206536547 J * ruskie ruskie@ruskie.user.oftc.net 1206536575 J * mire ~mire@37-174-222-85.adsl.verat.net 1206536586 M * Bertl_oO upgrading to latest devel wouldn't hurt, but if it isn't Linux-VServer related, it won't help much :) 1206536590 M * rgl Bertl_oO, I can't do that on that particular box (no storage left :/). but I can at home. 1206536620 M * bXi i'll check the output tonight 1206536629 M * bXi and let you see it 1206536634 M * rgl Bertl_oO, I just need to create the guest, touch nonamespace and vserver start it correct? 1206536639 M * bXi you can probably see if its vserver related or not 1206536923 Q * JonB Quit: This computer has gone to sleep 1206536956 M * daniel_hozac Bertl_oO: sure, but since rgl didn't know bout it, that seems unlikely to be the cause ;) 1206537041 M * Bertl_oO well, the question is, what happens if you start a 'nonamespace' guest and then remove the nonamespace flag? 1206537071 M * rgl oh running a guest without a namespace seems bad :/ 1206537071 M * daniel_hozac then you won't be able to enter it, as expected. 1206537074 M * Bertl_oO and more important, why is util-vserver (with nonamespace) complaining? 1206537120 M * daniel_hozac it seems likely that the guest's filesystem is actually mounted from /etc/vservers//fstab, thus not entering the namespace means it's not available. 1206537149 M * rgl daniel_hozac, yes it is. 1206537202 M * rgl daniel_hozac, like all my other guests. they all have the fstab file. 1206537222 M * rgl is that a no no? 1206537250 Q * Hunger charon.oftc.net neutron.oftc.net 1206537250 M * Bertl_oO sounds more like an userspace issue to me than a kernel one, but I might be wrong ... fact seems to be that your guest doesn't have a namespace right now .... 1206537272 M * Bertl_oO okay, bbl, have to grab a few things ... 1206537325 M * rgl daniel_hozac, what can I do to fix the "not available" you are talking about? 1206537444 J * Hunger Hunger.hu@213.163.11.138 1206537461 Q * Hunger Killed (cation.oftc.net (Nick collision (new))) 1206537461 J * Hunger Hunger.hu@Hunger.hu 1206537462 Q * Hunger Max SendQ exceeded 1206537478 J * Hunger Hunger.hu@Hunger.hu 1206537721 J * DavidS ~david@85.124.122.215 1206537880 M * DavidS daniel_hozac: hi, http://oldwiki.linux-vserver.org/some_hints_from_john mentions a bug about not being able to map tun devices into vservers. is this still so? 1206538040 M * daniel_hozac rgl: nothing. if the context has no namespace assigned, you can't enter it. 1206538080 M * daniel_hozac DavidS: hmm? the only limitation AFAIK is that you cannot set the peer address using the utils. 1206538092 M * rgl daniel_hozac, humm, so how to fix this issue I'm having? 1206538098 M * daniel_hozac rgl: restart the guest. 1206538131 M * rgl daniel_hozac, how? 1206538166 M * rgl (vserver piranha stop does not work either) 1206538166 M * daniel_hozac vkill, ssh, etc. whichever works for you. 1206538436 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it 1206538532 M * DavidS daniel_hozac: seems like i missed the point of "If you add "ip addr add peer dev tun0" there is no way to map the tun0 interface into a guest, even not with a 'nodev' option. (bug confirned to be reproducible by daniel_hoczac)" 1206538549 Q * ruskie Read error: Connection reset by peer 1206538574 M * daniel_hozac DavidS: that problem has been fixed since about 2 hours after it was found :P 1206538651 M * rgl daniel_hozac, does not fix the problem. the next time I start the guest, its created without a namespace. 1206538842 M * DavidS *sigh* 1206538847 M * DavidS that's how rumors start 1206539252 M * daniel_hozac rgl: really? reproducibly? 1206539312 M * rgl daniel_hozac, yes really. maybe if I reaboot this whole mess is gone :/ 1206539329 M * daniel_hozac unlikely. 1206539363 M * daniel_hozac could you tar up /etc/vservers for me? 1206539372 M * glen_ wtf is ln -0 ? 1206539374 M * glen_ ln -0s 1206539386 M * glen_ http://oldwiki.linux-vserver.org/some_hints_from_john ln -0s" uses a Vserver extention to create a unified link. 1206539392 M * rgl what bugs me is that until yesterday this worked fine; but the server rebooted and not this mess is hapenning :( 1206539397 M * glen_ first i tought it's typo, but it's written twice on that page 1206539411 M * rgl daniel_hozac, I can. just a sec. 1206539415 M * glen_ ln: invalid option -- 0 1206539424 M * daniel_hozac glen_: you'd have to ask derjohn 1206539782 Q * gebura Quit: Quitte 1206539958 J * virtuoso_ ~s0t0na@ppp91-122-138-181.pppoe.avangarddsl.ru 1206540186 Q * sharkjaw Quit: Leaving 1206540368 Q * virtuoso Ping timeout: 480 seconds 1206540397 J * balbir ~balbir@122.167.199.79 1206540488 J * fatgoose ~samuel@76-10-149-199.dsl.teksavvy.com 1206540554 J * gebura ~gebura@77.192.186.197 1206541118 J * ftx ~ftx@space-ext.gatworks.de 1206541523 J * JonB ~NoSuchUse@77.75.164.169 1206541842 Q * mire Ping timeout: 480 seconds 1206541913 J * jokkmokk ~hans@mail.netcare.at 1206542049 Q * fatgoose Quit: fatgoose 1206542095 M * jokkmokk hi, if i try to mount a samba share from within a vserver i get "mount error 12 = Cannot allocate memory", but in the host it works fine. The vserver version is 2.2.0.1. Interestingly it works perfectly with the vserver package form debian etch/stable (vs2.0.2.2-rc9). Any idea? 1206542119 M * JonB jokkmokk: did you give your guest the capability to mount stuff 1206542128 M * Bertl_oO jokkmokk: you are using cifs, which tries to spawn a kernel thread 1206542141 M * Bertl_oO jokkmokk: if you manage to use smbfs, that is not required 1206542154 M * jokkmokk yes, the capabilities are set 1206542162 M * jokkmokk yes I'm using cifs 1206542174 M * Bertl_oO jokkmokk: in any case, you can do the mount in the guest config's fstab or from the host (for the guest= 1206542181 M * Bertl_oO s/=/) 1206542242 M * jokkmokk I see. But I don't understand why this works in debian etch/stable. I'd like to use cifs to use the noatime flag 1206542268 M * Bertl_oO it works in etch, because that one uses smbfs :) 1206542283 M * jokkmokk "mount -t cifs" uses smbfs? 1206542312 M * Bertl_oO maybe? do I know debian? .. check with strace -fF (inside the guest on the mount) 1206542344 M * Bertl_oO last time somebody tried, it used smbfs (which works, because it doesn't spawn a kernel thread) 1206542378 M * jokkmokk does smbfs support the noatime flag? 1206542397 M * jokkmokk because if it doesn't it's not possible, since I'm mounting with noatime flag and this works perfectly 1206542400 M * Bertl_oO I would guess yes, that is usually above the fs layer 1206542430 M * Bertl_oO but let's double check with strace, shall we? 1206542461 N * Bertl_oO Bertl 1206542969 M * jokkmokk alright, how do I see if it's using cifs or smbfs in the strace output? 1206542998 Q * Slydder Quit: Leaving. 1206543019 M * Bertl there should be a mount line, which has the arguments passed and returns ENOMEM 1206543036 M * Bertl (or in the case where it works, a file handle) 1206543068 M * Bertl nah, 0 on success 1206543265 M * jokkmokk yeah, it definitely uses cifs 1206543289 M * Bertl okay, then I would like to see the two straces, the one which is working and the other, which isnt 1206543333 M * Bertl please use paste.linux-vserver.org or similar 1206543347 M * jokkmokk kk, mom 1206543586 M * jokkmokk this is the strace output for the mount that fails: http://paste.linux-vserver.org/11887 and this is the successful one: http://paste.linux-vserver.org/11886 1206543729 J * fatgoose ~samuel@76-10-149-199.dsl.teksavvy.com 1206543844 M * Bertl that's indeed strange ... 1206543900 M * Bertl are both mounts 'remote' or is one share on the same host? 1206543941 M * Bertl do you get a message in 'dmesg'? 1206543941 M * jokkmokk both are remote 1206543951 M * jokkmokk no, nothing in dmesg 1206543984 M * jokkmokk well, actually i get lots of stuff from dmesg in the vserver that can mount. the other is empty 1206544112 M * jokkmokk hmm...I'm getting the same error no matter if I use smbfs or cifs 1206544191 M * Bertl well, we have to figure what the difference between those two mounts are ... 1206544220 M * Bertl i.e. is it a property of the guest system (mount) or of the server providing the filesystem 1206544261 M * Bertl what kernel version are you running, and are warnings disabled? 1206544270 Q * DavidS Quit: Leaving. 1206544536 J * dowdle ~dowdle@scott.coe.montana.edu 1206544581 M * jokkmokk i'm afraid this is a custom patched kernel and maybe that's the problem. I'll try to isolate this and come back if I have more questions, thanks for your great support in the meantime. 1206544602 Q * tobifix Quit: Leaving 1206544607 M * Bertl you're welcome! keep us updated ... 1206544624 Q * jokkmokk Quit: Verlassend 1206545367 N * ensc Guest2682 1206545367 Q * Guest2682 Remote host closed the connection 1206545377 J * ensc ~irc-ensc@77.235.182.26 1206546041 J * ruskie ruskie@ruskie.user.oftc.net 1206546353 M * rgl daniel_hozac, found any problem with my /etc/vserver config? 1206546816 Q * ruskie Read error: Connection reset by peer 1206546845 J * julius_ ~julius@p57B277E3.dip.t-dialin.net 1206546985 J * ruskie ruskie@ruskie.user.oftc.net 1206547283 Q * Julius Ping timeout: 480 seconds 1206547493 Q * ftx Ping timeout: 480 seconds 1206547813 N * virtuoso_ virtuoso 1206548247 Q * mrfree Ping timeout: 480 seconds 1206548411 Q * doener Quit: leaving 1206548633 Q * JonB Quit: This computer has gone to sleep 1206548867 J * tobifix ~tobifix@muedsl-82-207-217-090.citykom.de 1206548876 M * tobifix re 1206548878 M * tobifix ;) 1206548912 M * Bertl wb 1206550186 J * Slydder ~chuck@dslb-088-075-070-146.pools.arcor-ip.net 1206550493 M * derjohn daniel_hozac, glen_ : i did not write that, someone must have changed that 1206550660 J * Infinito ~argos@200-101-127-188.gnace701.dsl.brasiltelecom.net.br 1206551104 M * tobifix Hollow, thanks alot for your gentoo stages, they work like a charm ;) 1206551198 M * Hollow :) 1206551350 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1206551431 M * daniel_hozac rgl: could you do a vserver --debug piranha start? 1206551459 Q * Slydder Quit: Leaving. 1206552073 Q * gebura Quit: Quitte 1206552077 M * rgl daniel_hozac, sure. I'm doing it now. 1206552119 M * glen_ err, what was the case with portmap using in vserver? 1206552249 J * ftx ~ftx@dslb-084-060-242-093.pools.arcor-ip.net 1206552384 M * rgl daniel_hozac, I've placed it in private. 1206552466 M * rgl daniel_hozac, look for "WITHOUT" to get into the vserver start that does not have the "nonamespace" setting set. 1206552470 Q * Infinito Quit: Leaving 1206552499 M * glen_ and how to figure out what address is known as "localhost" 1206552516 M * glen_ as 'ip a' doesn't show them in the order itnerfaces/* are 1206552527 M * rgl daniel_hozac, err look for WITHOU (I forgot the trailing T ..) 1206552706 M * Bertl glen_: well, actually that depends on what addresses you use :) 1206552783 M * Bertl glen_: the one listed in /etc/hosts as localhost will be used as localhost (with a normal resolver) 1206552805 J * bronson ~bronson@adsl-68-122-117-135.dsl.pltn13.pacbell.net 1206552830 M * glen_ wasn't there something 127.0.0.1 address will be really itnerfaces/0 address? 1206552870 M * Bertl that is not really related to 'localhost' but yes, for kernels <= 2.2.x, the first ip will be used to map 127.0.0.1 1206552893 M * Bertl for devel kernels, the 'lback' address will be used 1206553054 M * glen_ i think i have vs2.1 1206553078 M * Bertl that is the old devel series, and you might consider upgrading that 1206553083 M * glen_ # vserver-info 2>/dev/null|grep VS-API VS-API: 0x00020101 1206553101 M * glen_ in that devel what's the status of looback? 1206553120 M * glen_ i can't upgrade right now anyway 1206553226 M * daniel_hozac rgl: vserver-info /etc/vservers/piranha CFGDIR; echo $? 1206553230 M * daniel_hozac what does that output? 1206553258 M * rgl /etc/vservers/piranha 1206553258 M * rgl 0 1206553302 M * Bertl glen_: no idea, that is the API, not the actual kernel version, but I'd assume it doesn't know lback 1206553349 M * rgl daniel_hozac, is this line normal: exec /usr/sbin/vnamespace --new -- /usr/sbin/vserver ----nonamespace --debug piranha start ? I mean, the ----nonamespace argument seems to have extra "--". 1206553350 M * daniel_hozac rgl: so why is that failing during vserver ... start? 1206553376 M * daniel_hozac is /etc/vservers/piranha a symlink or something? 1206553376 M * Bertl a missing space? 1206553381 M * glen_ kernel is 2.6.16.x 1206553393 M * daniel_hozac no, all 4 dashes are supposed to be there. 1206553409 M * Bertl really? funny option 1206553449 M * daniel_hozac it's an internal option. 1206553489 M * Bertl ah, so be it ... you're the expert here ... 1206553502 M * Bertl besides that ... nap attack! :) 1206553504 M * zbyniu glen_: you're asking in context to put .22 to ac and compatibility with .16? 1206553510 N * Bertl Bertl_zZ 1206553719 M * rgl daniel_hozac, I have no ideia why its failing :/ 1206553740 M * glen_ zbyniu, no. don't care for compat 1206553750 M * glen_ blah. some application is still trying to use external ip 1206553752 M * glen_ bind(5, {sa_family=AF_INET, sin_port=htons(956), sin_addr=inet_addr("0.0.0.0")}, 16) = -1 EACCES (Permission denied) 1206553774 M * glen_ application that registers service in portmap 1206553776 M * rgl daniel_hozac, humm I got these on dmesg: vserver-info[19536:#0] trap stack segment rip:4028ea rsp:7269702f73726576 error:0 1206553792 M * glen_ can i perhaps make for app,lication invisible eth0 inside vserver? 1206553803 M * waldi rgl: nice stack pointer corruption 1206553828 M * glen_ some chbind inside vserver perhaps? 1206553866 M * daniel_hozac rgl: wait, you're using Ubuntu? 1206553874 M * rgl daniel_hozac, yes. 1206553890 M * rgl daniel_hozac, I'm using dietlibc from linux-vserver.org though. 1206553893 M * daniel_hozac rgl: get the Debian util-vserver.deb. 1206553922 M * glen_ bah. can't run chbind inside vserver 1206553922 M * glen_ chbind --ip 192.168.2.21 ip a 1206553922 M * glen_ chbind: vc_set_ipv4root(): Operation not permitted 1206554067 M * rgl daniel_hozac, I'm using the package from here: http://archive.ubuntu.com/ubuntu/pool/universe/u/util-vserver/util-vserver_0.30.214-6build1.dsc 1206554092 M * daniel_hozac don't. 1206554115 M * daniel_hozac get 0.30.215 from http://ftp.debian.org/debian/pool/main/u/util-vserver/. 1206554163 M * rgl daniel_hozac, just download the .deb from there? http://ftp.debian.org/debian/pool/main/u/util-vserver/util-vserver_0.30.215-2_amd64.deb ? 1206554174 M * daniel_hozac yes. 1206554187 M * rgl lets see if it works. 1206554267 M * rgl util-vserver depends on libc6 (>= 2.7-1); however: 1206554267 M * rgl Version of libc6 on system is 2.6.1-1ubuntu10. 1206554275 M * rgl no go :( 1206554289 M * daniel_hozac try the 0.30.214 backport for etch. 1206554314 M * rgl you known where is it? 1206554327 M * rgl backports.org? 1206554330 M * daniel_hozac http://backports.org/debian/pool/main/u/util-vserver/ 1206554658 M * rgl daniel_hozac, works! 1206554693 M * rgl now there is a namespace in all guests :)) 1206554741 M * daniel_hozac just... don't ever use Ubuntu packages. 1206554759 M * etteyafed What is this namespace nonsense? 1206554761 M * daniel_hozac they've been broken for years. 1206554826 M * daniel_hozac etteyafed: same name, multiple times per kernel. 1206554834 M * daniel_hozac i.e. what it sounds like :) 1206554841 M * glen_ any ideas why inside vserver this fails, and what's the workaround: 1206554841 M * glen_ bind(4, {sa_family=AF_INET, sin_port=htons(876), sin_addr=inet_addr("0.0.0.0")}, 16) = -1 EACCES (Permission denied) 1206554842 J * JonB ~NoSuchUse@77.75.164.169 1206554851 M * rgl daniel_hozac, but those are not even ubuntu packages; ubuntu packages have "unbuntu" in their name. util-vserver is sync from debian. maybe I should not have used dietlibc from linux-vserver? or maybe gcc is broken on ubuntu? 1206554852 M * daniel_hozac you have no IP addresses assigned? 1206554865 M * daniel_hozac gcc is broken on Ubuntu. 1206554891 M * glen_ i have one ip from eth2 1206554898 M * glen_ 20:08:13 root[pts/20]@fotik /# ip a 1206554898 M * glen_ 4: eth2: mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:0a:5e:1a:e5:b3 brd ff:ff:ff:ff:ff:ff inet 192.168.2.21/24 brd 192.168.2.255 scope global secondary eth2 1206554964 M * daniel_hozac no CAP_NET_BIND_SERVICE? 1206554965 N * DoberMann DoberMann[PullA] 1206555037 M * glen_ 20:06:27 root[pts/12]@black laborint/interfaces# zcat /proc/config.gz |grep -i cap_net_bind_service 1206555040 M * glen_ 20:10:24 root[pts/12]@black laborint/interfaces# 1206555046 M * glen_ its vs2.1 on 2.6.16.60 kernel.. 1206555058 M * daniel_hozac uh... 1206555069 M * daniel_hozac try grep Cap /proc/self/status 1206555084 M * glen_ # grep Cap /proc/self/status 1206555084 M * glen_ CapInh: 0000000000000000 1206555084 M * glen_ CapPrm: 00000000344c14ff 1206555084 M * glen_ CapEff: 00000000344c14ff 1206555100 M * glen_ i added this: 1206555101 M * glen_ # cat bcapabilities 1206555101 M * glen_ CAP_NET_ADMIN 1206555131 M * daniel_hozac so you are running that as root in the guest, right? 1206555139 M * glen_ no... 1206555172 M * daniel_hozac you realize that doesn't work on a regular Linux box either, right? 1206555202 M * glen_ um 1206555208 M * glen_ it worked on regular box 1206555278 M * daniel_hozac you have boxes without capabilities? strange. 1206555292 M * daniel_hozac though, hmm, even that should require uid = 0. 1206555337 M * glen_ what i see that 0.0.0.0 doesn't work: 1206555337 M * glen_ bind(4, {sa_family=AF_INET, sin_port=htons(2373), sin_addr=inet_addr("192.168.2.21")}, 16) = 0 1206555340 M * glen_ bind(5, {sa_family=AF_NETLINK, pid=0, groups=00000000}, 12) = 0 1206555343 M * glen_ bind(5, {sa_family=AF_INET, sin_port=htons(714), sin_addr=inet_addr("0.0.0.0")}, 16) = -1 EACCES (Permission denied) 1206555350 M * daniel_hozac no, it's ports <1024. 1206555354 M * glen_ err. umm, 714 < 1024 1206555490 M * glen_ it's odd that i got it running in some other vserver, portmap and the application using portmap 1206555498 M * glen_ the same 2.6.16 kernel with vs2.1 1206555523 M * daniel_hozac as root, i presume. 1206555543 M * glen_ well that bind < 1024 is maybe side effect of something else failing 1206555579 M * rgl daniel_hozac, hymm util-vserver binaries do not like gcc ssp et al, thats why I got the generated binaries in ubuntu broken? 1206555621 M * daniel_hozac rgl: yep 1206555672 M * rgl daniel_hozac, in theory I could disable all that using CFLAGS and build? 1206555720 M * daniel_hozac i believe there are people who have successfully done it. 1206555765 M * daniel_hozac just remember to do the same for dietlibc. 1206555813 M * rgl its recommended to use linux-vserver.org dietlibc version? or a stock/ubuntu/debian one should do? 1206555830 M * daniel_hozac get the latest release. 1206555840 M * daniel_hozac i.e. upstream. 1206555843 M * rgl from upstream? 1206555844 M * rgl ah ok. 1206555918 Q * balbir Ping timeout: 480 seconds 1206555949 M * rgl util-vserver will use only stuff from dietlibc and its own code, correct? 1206555996 M * daniel_hozac mostly. 1206556456 J * balbir ~balbir@122.167.181.20 1206556464 M * rgl I see. 1206556486 M * rgl thx for all the help daniel_hozac ! :D 1206557028 Q * balbir Ping timeout: 480 seconds 1206557075 N * C14r_ C14r 1206557688 J * balbir ~balbir@122.167.176.239 1206558374 J * DavidS ~david@86.59.70.100 1206558500 N * DoberMann[PullA] DoberMann 1206558796 M * tam When I build a new vserver, I frequently get the following error... If I rerun the same command again, sometimes it builds the vserver, other times I get the error... I keep running the build until it's successful... any one have any idea what's going on? 1206558800 M * tam rpm-fake-resolver: vc_ctx_migrate(): No such process 1206558803 M * tam rpm-fake.so: failed to initialize communication with resolver 1206558842 M * daniel_hozac with util-vserver 0.30.215? 1206558857 M * tam util-vserver: 0.30.213; Jul 27 2007, 10:19:17 1206558870 M * tam on this box, i have 215 on another box 1206558882 M * tam I'll upgrade and try again 1206559626 M * tam ok, yes it happens with util-vserver: 0.30.215 1206559647 M * tam Kernel: 2.6.22.18-vs2.2.0.6 1206559716 M * daniel_hozac hmm, i have not been able to reproduce it. 1206559730 M * daniel_hozac could you run with RPM_FAKE_DEBUG=65535 vserver ...? 1206559789 M * tam Sure will 1206559794 M * tam of course, sometimes it works the first time 1206559802 M * tam i'll let it build it, delete it and try again 1206559988 N * Bertl_zZ Bertl 1206560020 J * ntrs ~ntrs@77.29.69.241 1206560326 Q * dowdle Remote host closed the connection 1206560380 M * Bertl YAY! it seems that the ro bind mounts are finally merged upstream! (the Dave Hansen version :) 1206560636 M * tam daniel- I can't get my vserver build to fail when adding the debug. 1206560643 M * tam I'll keep trying. 1206561977 M * daniel_hozac Bertl: sweet! 1206561982 M * daniel_hozac that took long enough... 1206562063 J * Aiken ~james@ppp118-208-122-28.lns4.bne4.internode.on.net 1206562211 M * zbyniu yep, but imo record belongs to posix file capabilities, almost 10 years... 1206562332 M * Bertl zbyniu: posix is not involved there :) 1206562722 M * zbyniu afair posix caps and linux caps are more diffrent than equal, but based on 1206562725 J * haxier ~haxier@111.85-84-169.dynamic.clientes.euskaltel.es 1206562754 M * Bertl yes, but read only bind mounts have nothing to do with them ... at least IMHO 1206562996 M * haxier Hi 1206563107 M * haxier I have some vservers using dummy0 interfaces. Host makes the routing and applies SNAT to output packets 1206563138 M * Bertl welcome haxier! okay, sounds good (well except for the unnecessary dummy0) 1206563192 M * haxier Unnecesary? Hmmm... vservers are in another subnet 1206563211 M * haxier I don't like too much ethernet aliases 1206563242 M * Bertl yeah, that's fine ... it helps folks to 'pretend' that there are 'virtual' interfaces ... doesn't hurt so it's fine 1206563259 M * haxier Yes, that's the point :) 1206563288 M * haxier Well, the host have some other interfaces: eth0, eth1, eth2, tun0, tap0... 1206563364 M * haxier But I don't want vservers "know" what interfaces are out there, so in the guests 'route -n' only shows dummy0 related entries 1206563412 M * Bertl could be eth0 as well (would make it even more 'natural' no?) 1206563434 M * Bertl but that's not your problem ... yes? 1206563442 M * haxier Yes 1206563443 M * haxier :) 1206563443 M * zbyniu Bertl: yes, i mean loooong time to put into mainstream 1206563507 M * haxier My main problem is guests access to tun/tap devices. I don't want to make a tun/tap device for each guest 1206563525 M * Bertl hmm, okay 1206563527 Q * Aiken Remote host closed the connection 1206563580 N * weasel Weasel 1206563598 M * haxier tun/tap gets dynamic IP so i must replace the IP in /etc/vservers//interface 1206563606 M * Bertl haxier: so ... what process is connecting to the tun? 1206563618 M * haxier OpenVPN, QEMU... 1206563628 M * Bertl running on the host? 1206563638 N * Weasel weasel 1206563646 M * haxier Yes 1206563661 M * Bertl okay, and you get (somehow) dynamic ips to that interface, yes? 1206563672 M * haxier yes 1206563688 M * Bertl good, so simply add those IPs to the guests you want to participate 1206563704 M * Bertl (you can do that dynamically when the guest is running) 1206563726 M * haxier with ifconfig? 1206563735 M * Bertl nah, with naddress :) 1206563753 M * Bertl (check out naddress --help) 1206563764 Q * JonB Quit: This computer has gone to sleep 1206563779 M * haxier I'm looking... i don't know that command 1206563815 M * zbyniu btw caps, you know that in test patch for .24 CAP_CONTEXT==CAP_SETFCAP? 1206563827 M * Bertl yep 1206563845 M * zbyniu just missing support for 64bit caps? 1206563871 M * Bertl yep 1206563872 J * larsivi ~larsivi@144.84-48-50.nextgentel.com 1206563890 M * zbyniu ok 1206563928 J * JonB ~NoSuchUse@77.75.164.169 1206564115 M * haxier bertl: looking the FAQ... well, naddress will solve my problem 1206564156 M * haxier bertl: a guest must have a tun/tap device to access the host tun/tap device? 1206564180 M * Bertl nah, there is nothing like the beforementioned 'virtual itnerfaces' in Linux-VServer 1206564206 M * Bertl Linux-VServer uses IP isolation, so you assign an address, and the interface shows up (magically) 1206564324 M * haxier bertl: so i define some interface with /etc/vservers/guest01/interfaces/nodev? 1206564325 M * zbyniu ...and even if you restart network ip address may disappear but still works (full magic) ;-] 1206564361 M * Bertl haxier: would be an option, if you know the IP in advance 1206564380 M * Bertl haxier: otherwise, you add it with beforementioned naddress command 1206564443 M * haxier bertl: oh, I understand now: there's no need to define additional interfaces, linux-vserver will do the trick 1206564610 M * haxier can't the guest have only one dummy address with only one dummy interface, and be the host the one to make the routing to ethX, tunX and so on? 1206564638 M * Bertl it is _always_ the host which does the routing and such 1206564661 M * Bertl but if you want the guest to 'utilize' an IP, you have to assign it 1206564679 M * Bertl if you do the same with S/DNAT, that's fine too 1206564943 M * haxier bertl: yes, i know how routing works... perhaps I must re-read the frozentux iptables manual before asking questions like this 1206564964 M * haxier well, one last question 1206564970 M * Bertl sure 1206565021 M * haxier if I run 'route -n' in the guest i see the full routing table of the host 1206565076 M * haxier can't this be filtered so only relevant devices and routes are shown? 1206565130 M * Bertl depending on the kernel, it should be already filtered 1206565215 M * haxier i see entries with '*' Ifaces and '0.0.0.0' gateways. Can this be hidden? 1206565257 M * Bertl sure, a lot of things _can_ be done ... but in this specific case, you need to make some modifications to the kernel :) 1206565374 J * Piet ~piet@tor.noreply.org 1206565431 M * haxier bertl: opps, kernel patching is not the best way for me 1206565433 M * haxier iptables -A FORWARD -i $LAN -j LOG --log-level debug 1206565438 Q * DoberMann Remote host closed the connection 1206565440 M * haxier Arghh! 1206565455 J * DoberMann ~james@cap31-6-88-180-72-76.fbx.proxad.net 1206565460 J * yarihm ~yarihm@84-75-103-252.dclient.hispeed.ch 1206565465 M * haxier bertl: the naddress command will solve the problem. Thanks for the help 1206565489 M * Bertl you're welcome! 1206565703 M * ntrs when I start a daemon using the vsysvwrapper like v_httpd for example, it always binds to eth0. How do I make it bind to eth1? Does anyone know how I can do that? 1206566217 J * ktwilight_ ~ktwilight@200.208-66-87.adsl-static.isp.belgacom.be 1206566298 M * daniel_hozac ntrs: that's legacy cruft. just use chbind --nid 42 --ip /etc/init.d/httpd restart 1206566300 Q * ktwilight Ping timeout: 482 seconds 1206567348 J * Aiken ~james@ppp59-167-114-199.lns3.bne4.internode.on.net 1206567356 M * Bertl wb Aiken! 1206567369 M * Aiken hi 1206567401 Q * Piet Quit: Piet 1206567943 Q * bonbons Quit: Leaving 1206568383 J * ktwilight ~ktwilight@200.208-66-87.adsl-static.isp.belgacom.be 1206568578 Q * ktwilight_ Ping timeout: 480 seconds 1206568909 J * ntrs_ ~ntrs@77.29.64.94 1206569333 Q * ntrs Ping timeout: 480 seconds 1206569374 M * haxier for my question about tun/tap interfaces, if interest, the solution is as simple as a 'iptables -t nat -A POSTROUTING -s -o tunX -j MASQUERADE' 1206569421 M * haxier Sorry for asking about this kind of things (iptables related, not vserver) 1206569426 M * Bertl yes, but you might consider using SNAT instead of MASQ (as it is more flexible and precise) 1206569470 M * haxier bertl: yes, but tunX gets a dynamic IP so MASQUERADE works well with this 1206569507 M * Bertl i.e. depending on the setup, MASQ might do unexpected things ... 1206569563 M * Bertl the dynamic address is not an issue, if handled properly, after all the tun ip is assigned at a specific point (script) which can be used to adjust the iptables rule as well 1206569590 M * Bertl but if it works for you ... that's fine for me too :) 1206569751 M * haxier bertl: yes, you're rigth (again) :-) Thanks for the help 1206569752 Q * julius_ Ping timeout: 480 seconds 1206569759 M * Bertl np, have fun! 1206569775 M * Bertl and feel free to hang around ... if you like that channel :) 1206569866 M * Bertl s/that/this/ :) 1206570530 M * haxier oh, thanks, I use vservers a lot, asked sometimes in this channel 2.0.3-rcX patches for 2.6.16.x kernels (now in .60) 1206571102 M * haxier time to sleep, bye 1206571102 Q * haxier Quit: I am going away 1206571128 N * DoberMann DoberMann[ZZZzzz] 1206571674 Q * tobifix Remote host closed the connection 1206571709 J * mire ~mire@37-174-222-85.adsl.verat.net 1206571887 Q * JonB Ping timeout: 480 seconds 1206571922 Q * DavidS Quit: Leaving. 1206571967 J * Infinito ~argos@200-101-127-188.gnace701.dsl.brasiltelecom.net.br 1206572161 Q * rgl Quit: Saindo 1206572201 Q * mire Quit: Leaving 1206572500 J * dna_ ~dna@216-230-dsl.kielnet.net 1206572738 Q * ftx Remote host closed the connection 1206572763 J * ftx ~ftx@dslb-084-060-242-093.pools.arcor-ip.net 1206572883 Q * dna Ping timeout: 480 seconds 1206573304 Q * cryptronic Quit: Leaving. 1206573538 J * esa ~esa@ip-87-238-2-45.static.adsl.cheapnet.it 1206573553 Q * esa` Ping timeout: 480 seconds 1206574407 Q * larsivi Remote host closed the connection 1206575135 Q * ftx Ping timeout: 480 seconds