1119139223 M * Bertl well, there are pros and cons for both ... 1119139258 M * Bertl ad 2) the important part is, that the helper is indirectly blocking the context 1119139271 M * Greek0 am I understanding 2) correctly? what are the pros? 1119139283 M * Bertl i.e. while the helper is executing (synchronously, which is the future default) 1119139301 M * Bertl the context can not disappear, because the sys_reboot() did not return 1119139330 M * Bertl except for that, which requires the helper to fork of the waiter, you understood it quite fine 1119139361 M * Greek0 ah, I was thinking wrong of 1) it turns out. 1119139366 M * Bertl the advantage of the second approach is that the context is created anew 1119139374 M * Greek0 I thought 1) was triggered through a sys_reboot too 1119139392 M * Bertl yes, it is ... 1119139425 M * Greek0 hmm. ah 1119139429 M * Bertl but 1 is not really doable with an init process for example 1119139440 M * Bertl well, not with the typical init at last ... 1119139440 M * Greek0 the difference is that in 2 you take the whole context down. 1119139455 M * Bertl yes, with all accounting, sockets, etc ... 1119139462 M * Greek0 what happens in 1) with init? 1119139491 M * Bertl as I said, 1) is probably not suitable for the 'usual' init 1119139504 M * Bertl but many (most?) guests do not use a separate init 1119139545 M * Greek0 would a guest-init survive in 1)? 1119139561 M * Bertl OTOH, if the init would be designed so that it 'restarts after a reboot syscall, it would be fine with an init too 1119139609 M * Bertl probably the default init throws an error if sys_reboot() returns ;) 1119139615 M * Greek0 hehe 1119139626 M * Greek0 ok, I think I understood it 1119139632 M * Greek0 I got 1) wrong a bit. 1119139656 J * Alissa ali@151.81.2.14 1119139667 M * Alissa hi all 1119139670 M * Greek0 hi 1119139690 M * Alissa I have a question =) 1119139719 M * Alissa vserver suport only more ip's v4? 1119139732 M * Alissa ifconfig don't work's 1119139733 M * Alissa =/ 1119139837 M * Bertl hi Alissa! 1119139847 M * Alissa hi Bertl =)* 1119139871 M * Bertl hmm, let me rephrase your questions (to make sure I udnerstood them) 1119139903 M * Bertl a) does linux-vserver support only ipv4? 1119139917 M * Alissa sorry 1119139920 M * Bertl b) does linux-vserver support assigning more than one ip? 1119139925 M * Alissa my english it's very bad =PP 1119139939 M * Bertl c) why does ifconfig not work/show anything inside a guest? 1119139962 M * Bertl d) something completely different? :) 1119139976 M * Alissa Bertl b and c questions =) 1119140012 M * Bertl okay, ad b) yes, up to 16(15) ipv4 addresses are supported for each guest 1119140042 M * Bertl (the limit is arbitrary, and can be raised, but it's a good tradeoff of overhead and typical usage) 1119140061 M * Alissa Bertl b) whith ifconfig or other command (or config)? 1119140078 M * Bertl ad c) because ifconfig is several years old, and can not keep up with recent (1 year) kernel development 1119140099 M * Bertl b) you 'assign' the ips to a guest, they are created on the host 1119140118 M * Alissa Bertl sorry I not understand.... 1119140128 M * Alissa 'assign' ?!? 1119140133 M * Bertl no problem, I try to rephrase it ... 1119140157 M * Bertl when you create a guest, you decide which ips you want it to use 1119140165 M * Alissa exactly 1119140171 M * Alissa I use a debian script 1119140172 M * Bertl for example, guest1 (10.0.0.1, 10.0.0.2 and 10.0.0.3) 1119140190 M * Alissa debian newserver modified of sarge edition.. 1119140202 M * Bertl forget the newvserver script 1119140214 M * Bertl it's outdated and does weird/wrong stuff 1119140238 M * Bertl the best thing you can do is to update to a recent kernel patch, and recent tools 1119140264 M * Bertl we are currently trying to resolve this with the debian folks 1119140271 M * Alissa I'm using kernel 2.4.30 and utils 1.2.* 1119140285 M * Alissa utils 1.2.10 1119140307 M * Bertl 1.2.10 is the kernel patch, the tools are probably 0.30 or 0.30.204, no? 1119140314 M * Alissa ups 1119140317 M * Alissa sorry =P 1119140320 M * Alissa yep ^^ 1119140321 M * Bertl np 1119140338 M * Bertl okay, the 'script' wrote some config file in /etc/vservers/* 1119140371 M * Bertl unfortunately the config is broken too (on debian), so you have to adjust it manually 1119140409 M * Bertl but in this config file (we now use a config tree, btw) you can configure those ips 1119140443 M * Alissa Bertl you have the sample config with more ip? 1119140444 M * Bertl when the guest is started, then depending on the config, those ips are created and become visible inside the guest 1119140466 M * Alissa or links? 1119140471 M * Bertl sec 1119140530 M * Bertl IPROOT="eth0:10.0.0.1/24 eth0:10.0.0.2/24" 1119140530 M * Bertl IPROOTDEV="" 1119140549 M * Bertl this for example defines two ips, which will be created on vserver startup 1119140590 M * Greek0 /24? 1119140606 M * Bertl hmm, with older tools it would be; 1119140611 M * Alissa Greek0 this is a lan, very Bertl? 1119140613 M * Bertl eth0:10.0.0.1/255.255.255.0 1119140616 M * Alissa =) 1119140628 M * Greek0 ah, ic 1119140634 M * Alissa I'm using vserver for a hosting web etc etc 1119140640 M * Alissa it's a GOD project 1119140660 M * Bertl so you like it ... good! 1119140674 M * Alissa no, no 1119140678 M * Alissa GOD 1119140679 M * Alissa =) 1119140686 M * Alissa uhmmm 1119140696 M * Alissa the translation it's difficuty.. 1119140709 M * Alissa an italian is: E' un porgetto da Dio 1119140717 M * Alissa *progetto 1119140720 M * Alissa ^^" 1119140753 A * Bertl .o( doesn't mean anything to me :) 1119140774 M * Alissa an image on marlok.dk site is a tested image for vserver? 1119140783 M * Greek0 plan from god, according to babelfish 1119140815 M * Bertl k, the marlow images are pretty old, but they should work 1119140832 M * Alissa cool 1119140857 M * Alissa Bertl you is a l33t or a g33k? =D 1119140886 M * Bertl hmm .. what about a l33t g33k? :) 1119140902 M * Bertl no, I'm just a developer ... 1119140943 M * Alissa Bertl l33t or g33k = Italian GURU or a Sistem helper for unix sistem ^^ 1119141089 M * Greek0 Bertl: motivated for one last question? 1119141117 M * Bertl sure ... 1119141151 M * Greek0 how do you want reboot != stop+start in case 2)? 1119141190 M * Greek0 or was that just "could be implemented, if we went for 1)"? 1119141194 M * Bertl usually, stop would do the following things: 1119141213 M * Bertl 1) execute various service stop scripts (inside the geust) 1119141237 M * Bertl (or at least talk to the init process?) 1119141249 M * Bertl 2) kill off all processes inside the guest 1119141259 M * Bertl 3) do some administrative stuff on the host 1119141282 Q * gaba_ Quit: leaving 1119141282 M * Bertl 1) and 2) is _almost_ done in a reboot-from guest case 1119141304 M * Bertl so no need to do it again (especially the scripts) 1119141333 M * Greek0 well, but you have to check if all processes are killed, and do that if there are some left.. 1119141334 M * Bertl wile the 3) part has to be done, so that the context can be started again ... 1119141358 M * Bertl the vkill is what I put into _almost_ ;) 1119141374 M * Greek0 hehe 1119141406 M * Greek0 so you mean reboot = we can save some work that's already been done by the guest. 1119141435 M * Greek0 I thought you ment reboot = we don't have to take the whole thing down completely.. which somehow clashed with my opinion what 2) should do.. 1119141469 M * Bertl the first one is original case 2, and the second one is my case 1) ;) confused? 1119141472 M * Greek0 makes sense now.. 1119141504 M * Greek0 I think I understand everything, but I don't get your number game 1119141516 M * Bertl okay, then just forget it ;) 1119141523 M * Greek0 1) and 2) are ascii art 1 and ascii art 2 for me. 1119141528 M * Greek0 the 2 images you painted above 1119141534 M * Bertl yep 1119141548 M * Bertl AA1 shows the not-fully-shutdown reboot case 1119141558 M * Greek0 yes 1119141567 M * Bertl while AA2 shows the reduced 'reboot' case 1119141574 M * Greek0 mhm 1119141584 M * Greek0 and what are the other 1,2 ? 1119141588 M * Bertl AA1 is applicable from inside _and_ outside 1119141613 M * Bertl while AA2 has different pathes for inside and outside reboot 1119141681 M * Bertl but they can be combined to: 1119141689 M * Bertl running running 1119141689 M * Bertl ________ ______________ 1119141689 M * Bertl \ / 1119141689 M * Bertl reboot \ hot restart / 1119141689 M * Bertl _________\ ____________ / 1119141691 M * Bertl \ / 1119141694 M * Bertl \ / 1119141697 M * Bertl \_ stop _/ 1119141705 M * Bertl 1119141717 M * Alissa night all 1119141720 M * Greek0 AA2 inside is clear, for AA2 outside, you kick off the shutdown.. is it guaranteed that sys_reboot is called by init then? 1119141721 M * Bertl which probably makes it easier to see the difference ... 1119141722 M * Alissa night Bertl 1119141723 M * Alissa =) 1119141725 M * Greek0 n8 1119141729 M * Alissa thank's for all 1119141734 M * Bertl night Alissa! 1119141738 M * Greek0 or how do you know that the shutdown scripts finished? 1119141803 M * Bertl from outside, either you are 'calling' the shutdown scripts yourself (no-init-case) or talking to init, in which case you have to wait for the helper 1119141804 N * Alissa Alissa`not`here 1119141824 M * Greek0 ie. if you are shutting down the guest from the outside, how do you control the shutdown process? that the shutdown scripts are executed, processes killed, and when you can take down the ctx? 1119141841 M * Greek0 ok 1119141865 M * Bertl in the init/helper case, the helper can then kill init 1119141878 M * Bertl which in turn will wake the waiter/sleeper ;) 1119141902 M * Greek0 hu? I thought the waiter waited for the helper to die? 1119141906 M * Greek0 not for init 1119141918 M * Bertl it waits for the context to be disposed ... 1119141926 M * Greek0 ah. bam 1119141928 M * Greek0 now it clicks 1119141942 M * Greek0 and that happens once sys_reboot returns. or not returns, or whatever. 1119141961 M * Bertl when the last process dies ... 1119141985 M * Bertl so let me list the cases once again: 1119142000 M * Bertl first the outside shutdown procedures: 1119142028 M * Bertl S-N (shutdown-noinit): 1119142045 M * Greek0 I just imagine myself in a comic with a huge "?" floating over my head ;) 1119142052 M * Bertl - tools call the various shutdown scripts 1119142086 M * Bertl hmm, no scratch that 1119142099 M * Greek0 hehe 1119142100 M * Bertl - tools start the sleeper/waier 1119142109 M * Bertl - tools call the various shutdown scripts 1119142126 M * Bertl - after some timeout (maybe 30-40 secs) a vkill is done 1119142147 M * Bertl - all administrative stuff is reset for that context 1119142163 M * Bertl meanwhile, there is a point where the waiter awakes 1119142190 M * Bertl in the shutdown case, it basically should make the context available again (i.e. remove the 7var/run/whatever) 1119142197 M * Bertl that's it ... 1119142207 M * Bertl S-I (shutdown with init) 1119142217 M * Bertl - tools start the sleeper/waiter 1119142231 M * Bertl - tools signal init to make a shutdown 1119142244 M * Bertl - guest init calls the various shutdown scripts 1119142255 M * Bertl - guest init invokes sys_reboot(halt) 1119142277 M * Bertl - vshelper is invoked, and does a vkill on init 1119142317 M * Bertl - the waiter/sleeper awakens, and declares the context available 1119142357 M * Bertl (the administrative shutdown on the host can happen either in the main thread after signaling init, or in the vshelper or in the waiter) 1119142378 M * Bertl now the shutdown from inside: 1119142398 M * Bertl - guest init calls shutdown scripts 1119142408 M * Bertl - guest init calls sys_reboot(halt) 1119142418 M * Bertl - vshelper is invoked, and does a vkill on init 1119142424 M * Bertl wrong! 1119142437 M * Bertl - vshelper is invoked, and spawns the waiter 1119142448 M * Bertl - vshelper kills the guest init 1119142456 M * Bertl - the waiter/sleeper awakens, and declares the context available 1119142472 M * Bertl this setup, if you want to unify all of them somewhat 1119142489 M * Bertl means that you have to start the waiter from the vshelper 1119142504 M * Bertl and you have to do the administrative shutdown either in the helper or waiter 1119142510 M * Alissa`not`here uhmm 1119142515 M * Alissa`not`here devo metterli in crontab... 1119142531 M * Bertl Greek0: everything clear so far? 1119142545 M * Greek0 pretty much, yes. well. 1119142555 M * Bertl okay, now the reboot cases ... 1119142558 M * Greek0 wait 1119142573 M * Greek0 administrative shutdown means tear the context down? 1119142610 M * Bertl it means whatever is required to undo what was done on context startup (from the host side) 1119142621 M * Greek0 you'll have to do that in the waiter, since the helper is keeping the context alive, or? 1119142641 M * Bertl everything except the actual context exit 1119142677 M * Bertl you can decide what you put where, actually 1119142680 M * Greek0 is the context exit done before or after that administrative stuff 1119142699 M * Bertl the vshelper will run as last thing before the context exits 1119142714 M * Bertl and the waiter will run right after the context exited 1119142740 M * Greek0 hmm. do I have to care about the mounts in the guest namespace? 1119142746 M * Bertl not really 1119142756 M * Greek0 ok, so it really doesn't matter all that much? 1119142773 M * Bertl well, there are pre/psot scripts and such stuff 1119142781 M * Bertl also you might want to save for example dlimits 1119142803 M * Bertl all those 'persistent' things have to be considered here 1119142825 M * Bertl with ngnet the interface creation/destruction will be part of that equation too 1119142827 J * eXplasm2 explasm@p549F4C06.dip.t-dialin.net 1119142833 M * Bertl welcome eXplasm2! 1119142870 M * Greek0 well, it's all scripted anyway, so if we make it modular it probably shouldn't be too much of a problem to move it around later on.. 1119142917 M * Bertl probably ... 1119142933 M * Greek0 ok, I'm pretty sure I understand everything about the shutdown and most of the reboot now.. I'll have a look into the sources tomorrow. 1119142945 M * Bertl okay, have fun! 1119142947 M * Greek0 I'm off to bed now, thanks for your patience. 1119142963 M * Bertl actually it was my pleasure! 1119143004 M * Greek0 cu 1119143009 M * Bertl cya 1119143286 Q * eXplasm Ping timeout: 480 seconds 1119143396 M * Doener wow... quite some discussion... 1119143409 A * Doener was surfing a little... 1119143590 M * Bertl where? 1119143658 M * Doener the web ;) mostly heise.de, somehow I enjoyed reading stuff from Jörg Schilling in the forum threads on the SchilliX news 1119143679 M * Doener discussion = here... 1119143697 M * Doener (just in case that we had a misunderstanding ;) 1119143735 M * Bertl ah, and I was thinking about actual 'surfing' ... 1119143996 M * Doener any plans for tonight? 1119144017 M * Bertl not yet ... 1119144025 M * Bertl suggestions? 1119144055 M * Doener no, searching for ideas myself ;) 1119144097 M * Bertl well, we could try to patch up a release candidate based on 2.6.12? 1119144122 M * Bertl or we could take another look at ngnet? 1119144150 M * Doener hm? 1119144232 M * Doener what did the first one mean? 1119144287 M * Bertl well, we have a patch for 2.6.12, but we need to break it down, check it, add the changes we did so far, etc ... 1119144335 M * Bertl not saying that this is terribly complicated ... 1119144362 M * Bertl it was more the late saturday night thing :) 1119144412 M * Doener ah, yeah, if you don't mind "teaching" me how to do that, let's do that 1119144430 M * Doener (teaching: how to break down the patch nicely) 1119144664 M * Bertl hmm, well, the first part is done by one of my magic scripts 1119144691 M * Bertl the second part is checking the result and sorting ambiguous stuff my hand ... 1119144792 M * Doener ok, then I guess I'll simply let you do that part ;) 1119144805 M * Bertl btw, did you look at the vxdebug stuff? 1119144853 M * Doener IIRC yes, but let me double check... 1119144998 M * Doener yes, I did... don't understand the kernel patch. 1119145045 M * Doener well, the shell script neither, as i'm not used to gawk 1119145102 M * Bertl did you try it? 1119145194 M * Doener readelf: Error: Invalid option '-x' 1119145209 M * Bertl hmm, interesting ... 1119145216 M * Doener funny, readelf --help lists -x 1119145232 M * Bertl did you compile the kernel with the info support? 1119145257 M * Bertl did you pass it as first argument or call the script inside the kernel build tree? 1119145324 M * Doener ah, damn... tried in the patched source tree: no vmlinux, then switched into a tree with vmlinux, but an unpatched... 1119145429 M * Bertl option -G is interesting ... 1119145460 M * Doener that's the documentation generation? 1119145468 M * Bertl yep 1119145483 M * Doener oh boy... now i forgot to enable debugging... 1119145644 M * Doener ok, now it works :) 1119145670 M * Bertl good :) 1119145711 M * Doener the number in front of some rows are the second and third argument to VC_CMD, right? 1119145731 M * Doener s/number/numbers/ 1119145744 M * Bertl no, they are the bits to set in /proc/sys/vserver/debuf_* 1119145750 M * Bertl *debug_* even 1119145792 M * Bertl but we could easily add entries for the syscall commands too 1119145798 M * Doener ok 1119145847 A * Doener somehow forgot to turn brain on today... 1119145994 M * Bertl so, how is OpenSolaris? ;) 1119146128 Q * shuri Remote host closed the connection 1119146151 M * Doener didn't try it out... AFAIK the stuff that sun made available don't get you a running system. SchilliX 0.01 just has multi-user console mode for x86 and i don't plan to spend a few hours/days/whatever just to get that stuff running 1119146172 M * Doener "that stuff" = more than just what SchilliX provides atm 1119151274 M * Doener i'm off to bed, good night Bertl! 1119151279 N * Doener Doener_zZz 1119151298 M * Bertl night Doener_zZz! 1119152508 J * Doener` ~doener@p54873EE8.dip.t-dialin.net 1119152940 Q * Doener_zZz Ping timeout: 480 seconds 1119153206 M * Bertl okay, off to bed now too .... cya later 1119153215 N * Bertl Bertl_zZ 1119153384 M * mugwump gosh, it really is right out of phase here in NZ 1119159193 J * Aiken ~james@tooax8-125.dialup.optusnet.com.au 1119162530 J * Aiken_ ~james@tooax6-117.dialup.optusnet.com.au 1119162845 Q * Aiken Ping timeout: 480 seconds 1119168295 J * Hollow|mobile ~bene@p54978A5E.dip0.t-ipconnect.de 1119168299 M * Hollow|mobile morning 1119168617 M * albeiro morning mobile Hollow :) 1119169099 M * DaPhreak morning Hollow|mobile, albeiro 1119169100 M * DaPhreak .. 1119172292 M * Greek0 mugwump: well, so was their "good night" here (same timezone as Bertl and presumably Doener). people who work get up at that time ;) 1119172847 Q * Aiken_ Quit: Leaving 1119175225 N * Alissa`not`here Alissa 1119175230 M * Alissa hi all 1119175231 M * Alissa =) 1119176198 J * Aiken ~james@tooax6-117.dialup.optusnet.com.au 1119177312 Q * Hollow|mobile Quit: Leaving 1119177581 N * Doener` Doener 1119177590 M * Doener morning! 1119177595 M * DaPhreak morning Doener 1119177597 M * DaPhreak ! 1119179777 J * _are_ ~are@dsl-084-056-146-139.arcor-ip.net 1119179867 M * Greek0 what happens if I run a klogd inside the guests. does it get data to log from the kernel? 1119179933 M * Alissa Greek0 hi =) 1119179957 M * Greek0 hi Alissa 1119180185 Q * are|afk Ping timeout: 480 seconds 1119180300 M * Doener Greek0: earlier versions simply denied access, but IIRC recent versions virtualize /dev/kmsg(? is that the right one?) 1119180347 M * Doener at the moment that just means that vservers can access that device, but don't get any messages 1119180490 M * Alissa vserver support IPv6? 1119180522 M * Doener no, not yet 1119180554 M * Alissa uhmmm =/ 1119180754 Q * rs Quit: rs 1119180907 M * Greek0 Doener: thanks 1119181968 M * Greek0 hmm. I have no device nodes in my devpts mount in the vserver. this results in an EPERM warning when entering the vserver. 1119181984 M * Greek0 ideas why this happens and what I should do against it? 1119182062 J * neofutur ~neofutur@neofutur.net 1119182083 M * Doener Greek0: when you ssh to the vserver you should get a pty 1119182127 M * Doener the tools don't allocate a pty inside the vserver, but just keep the pty from outside the vserver, thus you don't see the device node and have no rights to modify it, if you use "vserver xxx enter" 1119182181 M * Greek0 orest:/home/vservers/sid/etc# vserver sid enter 1119182181 M * Greek0 mesg: /dev/pts/1: Operation not permitted 1119182186 M * Greek0 orest:/# 1119182199 M * Greek0 ^-- this is my problem (a purely cosmetic one, so to say) 1119182224 M * Greek0 I'm in the guest then, but the error message isn't very nice. 1119182333 M * Greek0 and /dev/pts is empty 1119182451 M * Doener yep, because you are actually using a pty from the host 1119182484 M * Doener btw, did you switch to root using "su" or "su -"? 1119182507 M * Greek0 in the host? `su` 1119182509 M * Doener "su -" should have already changed permissions on the pty, and the error message should not show up 1119182534 M * Greek0 uhm. where is the difference in su and su - 1119182536 M * Greek0 ? 1119182555 M * Greek0 ah, login shell? 1119182597 M * Doener changes the environment, so that it is like it would be if the user you switch to had logged in directly 1119182613 M * Doener as a side-effect(?) it lowers permissions on the pty 1119182681 M * Greek0 ic.. 1119182720 M * Doener of course it's just a workaround, not a solution 1119182759 M * Greek0 ic 1119184460 Q * Aiken Quit: Leaving 1119184673 M * Greek0 I've added some stuff to . anyone cares to review? 1119184722 M * Doener i'm already at it ;) (wiki has a mailing list that notifies of changes) 1119184737 M * Greek0 hehe 1119184956 M * Doener hey, the flower page has some nice stylesheets ;) 1119184972 M * Doener the default one is ehrm... strange... but some of the others are nice 1119185031 M * DaPhreak good morning Doener ;) 1119185039 M * DaPhreak boring is the best :P 1119185045 M * Doener daniel_hozac: hm, once again? 1119185057 M * Doener oops... DaPhreak was meant... 1119185072 M * daniel_hozac heh. 1119185073 M * DaPhreak in german you would say "Ja guten morgen auch" Doener 1119185074 M * Greek0 Doener: do you know how to switch stylesheets in firefox? 1119185095 M * Doener View->Page Style 1119185097 M * DaPhreak Greek0: View -> Page Style 1119185165 M * Greek0 ah, nice 1119185183 M * Greek0 I just curl'ed it on my own webserver and viewed it there, without pictures and stylesheets ;) 1119185307 M * id p?iv?? #vserver 1119185540 J * yarihm ~yarihm@217-162-204-252.dclient.hispeed.ch 1119186743 M * Hollow heya 1119186815 M * Hollow Doener: does VCMD_vx_info return a vcmd_vx_info_v0 struct? 1119186854 M * Doener no, an int 1119186865 M * Hollow every VCMD returns an int? 1119186888 M * Doener yes, as the syscall returns an int... 1119186902 M * Hollow mhm 1119186913 M * Doener you provide a pointer to a vcmd_vx_info_v0 to VCMD_vx_info and the syscall fill sit 1119186922 M * Doener s/fill sit/fills it/ 1119186935 M * Hollow ah, ok 1119187219 M * daniel_hozac Hollow: any specific reason you're not using libvserver? 1119187234 M * Hollow daniel_hozac: just playing around 1119187395 M * SiD3WiNDR weedpage is nice style 1119187395 M * SiD3WiNDR :) 1119187504 M * Hollow yeah 1119187586 Q * monrad Quit: Leaving 1119188001 J * monrad ~monrad@213083190130.sonofon.dk 1119188229 M * Greek0 hmm. according to the flower page /etc/vservers/.default/apps/vshelper/logfile should point to a logfile. but that logfile has to exist too, so something is logged. 1119188290 Q * Beirdo Ping timeout: 480 seconds 1119188311 M * Greek0 I'd rather like it if /etc/ ...vshelper/logfile was used as logfile if it's a real file, or if it's a symlink, no matter if the target exists or not.. 1119188324 M * Greek0 are there any drawbacks with this? 1119188876 M * _are_ most policies I know don't allow logfiles under /etc, especially not if they are expected to grow 1119188911 M * _are_ and I know quite a few people sort of relying on /etc not growing much 1119189113 M * Doener hm, those semantics would take care of that... 1119189154 M * Doener no logfile: no log, logfile is symlink: have the target of the symlink be the actual log, logfile is real file: log directly to it 1119189183 M * Doener I wouldn't want logs below /etc, but I don't see any problems by just allowing it 1119189195 M * Doener anyways, got to go, back later! 1119189200 N * Doener Doener|gone 1119190931 M * Greek0 the point is that currently, if you just do ln -s /var/log/vshelper.log /etc/vservers/.default/apps/vshelper/logfile, it does not work. 1119190953 M * Greek0 you have to do touch /var/log/vshelper.log to get actual logging. 1119191005 M * Greek0 but it's a bit of a pain to do this with the way findObject is called internally.. 1119191482 M * _are_ here it doens't work because /etc/vservers/.default/apps/vshelper doesn't exist on my host, but I have not tried to do vshelper logging, either, so far 1119191565 M * _are_ so as at my end you have to do some manual work for logging anyway, so can't see the big difference. Which is no reason not to implement it in some sane way, but probably this is mostly distributors stuff 1119191591 M * _are_ they should add the appropriate logrotate (or whatever they use) entries at the same time, too 1119191706 M * Greek0 well. If the flower page says that file is where logging goes, I wouldn't expect that I have to do an additional touch to finally get logging 1119191723 M * Greek0 it's just about least surprise in the UI 1119193701 J * Alissa`Umount ali@151.81.3.202 1119193840 J * shuri shuri@64.235.209.226 1119194096 Q * Alissa Ping timeout: 480 seconds 1119194097 N * Alissa`Umount Alissa 1119195929 J * p ~p@p5091BF1F.dip.t-dialin.net 1119195991 M * p hello. is there an easy way to set quota for vserver? i just want to limit the disk space of a vserver 1119196011 M * Alissa p use a quota disk kernel 1119196022 M * Alissa set a viserver for a gorup 1119196036 M * Alissa *group 1119196135 M * p how can i set a group for a vserver? 1119196143 M * Alissa exactly 1119196162 M * Alissa creating a group "vservers" 1119196177 M * Alissa and set quota to 1GB or 2Gb etc 1119196192 M * Alissa and for any vserver if you create 1119196209 M * Alissa add an user to a "vservers" group 1119196211 M * Alissa =) 1119196222 M * Alissa and starting vserver for this user ^^ 1119196264 M * Greek0 http://www.13thfloor.at/vserver/d_addons/overview/ <-- quota stuff 1119197546 M * daniel_hozac do you want quotas inside a vserver or a way of specifying "this vserver may not grow over this limit"? 1119197669 M * micah I could've sworn that I made this kernel with vserver support, but when I do setattr --barrier /vservers I get "Function not implemented" does this mean that I dont have vserver support? 1119197725 M * Greek0 micah: what arch? 1119197742 M * micah Greek0: i386 1119197754 M * Greek0 there's a problem in -rc4 with arch/um 1119197777 M * micah yea, not using that version 1119197791 M * micah there is something in /proc I can look for, xid or something? 1119197833 M * daniel_hozac /proc/virtual 1119197889 M * micah thats it, thanks 1119199395 M * Greek0 anyone familar with the vshelper tool? I'm digging into it a bit, but in every case I test it always resolves to method=sync 1119199415 M * Greek0 but there's are async/reboot/.. cases in the code.. 1119200173 N * Bertl_zZ Bertl 1119200183 M * Bertl evening folks! 1119200184 M * FaUl hey bertl 1119200212 M * Bertl p: what you want is called 'Disk Limits' and it is an integral part of the 2.0-* release ... 1119200240 M * Bertl micah: which filesystem, what tools? 1119200389 Q * greedy Quit: using sirc version 2.211+KSIRC/1.3.12 1119200425 M * Greek0 what do I have to throw into the guests /dev, so they can reboot on their own (ie. shutdown -h now should work inside)? 1119200432 M * micah Bertl: turns out to be me booting the wrong kernel 1119200442 M * Bertl Greek0: nothing ... 1119200473 M * Bertl Greek0: shutdown, reboot, halt is usually a link to init 1119200492 M * Bertl 'reboot -f' will 'just' call the sys_reboot() 1119200512 M * Bertl reboot will contact init via initctl (which is created by init) 1119200539 M * Bertl which in turn will do whatever is scripted in inittab 1119200566 M * Greek0 do I need a special init method for this to work? 1119200605 M * Bertl hmm .. you need a real guest init for the non -f version 1119200619 M * Greek0 mm, that's probably why it won't work ;) 1119200646 M * Bertl otherwise 'reboot -f' is the way to go 1119200667 M * Greek0 yup, just tried that 1119200710 M * Greek0 I'm really puzzled here. 1119200742 M * Greek0 in vshelper there is action, that's what it gets via argv from the kernel. and then it reads the method via /var/run/vserver/. 1119200751 M * Greek0 and later it dispatches based on method 1119200778 M * Greek0 which should be reboot/sync/async, acording to the code. 1119200781 M * Bertl yes, which leads to strange effects ... 1119200786 M * Greek0 but it's just always sync here. 1119200790 M * Greek0 no matter what I do. 1119200837 M * Bertl it might not be the worst idea to rewrite the helper completely 1119200838 M * Greek0 but when I cat the /var/run/vserver/, it says nothing of sync.. 1119200875 M * Greek0 if you ask me enrico went quite some way to make sure the code isn't very readable. 1119200906 M * Greek0 I hate those "conditionals by ||" 1119200952 M * Bertl well, I have to admit that it's very hard to read, but OTOH, he did it very modular/extensible 1119200971 M * Greek0 all you can do with them can be done with "if" too by nature 1119200990 M * Greek0 the only difference is, in the if case the code is executed if the condition is true. 1119201015 M * Greek0 with || it's the other way round, which always takes surprisingly much thought, even for simpler expressions 1119201054 M * Bertl yep, a very common sequence is: 1119201075 M * Bertl && || 1119201105 M * Greek0 interestingly enrico always uses || ;) 1119201270 M * Hollow ha, it works :) 1119201281 M * Bertl congrats! 1119201297 M * Hollow i got vx_info working with the struct as third argument ;) 1119202026 N * Doener|gone Doener 1119202034 M * Doener evening folks! 1119202064 M * Bertl evening Doener! 1119202291 M * DaPhreak heya Doener, Bertl! 1119202314 M * Bertl evening DaPhreak! 1119202455 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net 1119202508 M * Bertl welcome tchan! 1119202537 M * Alissa Bertl :) :* 1119202650 M * Bertl hey Alissa! 1119203006 M * Greek0 since when is enrico unreachable actually? 1119203073 M * Bertl hmm, last contact I had was in early may 1119203154 M * Greek0 urgs 1119205555 Q * monrad Ping timeout: 480 seconds 1119205759 Q * shuri Quit: Leaving 1119207492 J * skceb skceb@bang-bang.feuer-frei.com 1119207593 M * skceb hi, im building an debian vserver, debootstrapping right now. but for ~15 minutes its stuck at "I: Configuring base-config...", no cpu usage, no eth0 traffic.. is that normal? 1119207658 M * Alissa skceb no 1119207671 M * Alissa deboot strap it's blocked or crashed 1119207727 M * Bertl skceb: check if you can reach the sources ... (nameservice, internet, etc) 1119207786 M * skceb internet is working, and i can resolve host names 1119207806 M * Bertl skceb: how do you build the guest? 1119207809 M * skceb i cancelled it now, i used this cmd line: vserver DebianSid build -m debootstrap --force -- -d sid -m ftp://ftp.at.debian.org/debian/ 1119207833 M * skceb refering to http://linux-vserver.org/Step-by-Step+Guide+2.6 1119207850 M * Bertl okay, looks good so far, did you enable a source in /etc/vservers/*/sources.list ? 1119207863 M * Bertl (or whatever that file is called :) 1119207898 M * skceb ii dont have /etc/vserver 1119207906 M * skceb u mean apt sources.list? 1119207918 M * Doener there's no such thing for debootstrap 1119207920 M * Bertl yes, but the one in /etc/vservers/.... 1119207929 M * Bertl hmm, no? 1119207937 M * Doener he specified the repository on the command-line 1119207950 M * skceb i created /usr/local/etc/vservers/.defaults/apps/debootstrap/uri and put "http://ftp.debian.org/debian/pool/main/d/debootstrap/debootstrap_0.3.1.2_all.deb" in it 1119207955 M * Bertl ah, k, please take over ... 1119208028 M * Doener if he wouldn't have done, there's a hardcoded default. debootstrap does not depend on the packagemanagement, but works on a lower level 1119208071 M * Doener hm, let me check if the debootstrap process is running in a virtualized environment... 1119208206 M * Doener hm, AFAICT it is not... 1119208407 M * Doener skceb: hm, could you try again with: vserver build -m debootstrap --context --hostname -- -d sid -m ftp://ftp.at.debian.org/debian/ 1119208445 M * Doener replace and by values of your choice... xid can be anything >1 and <49152 1119208447 M * Bertl jsut replace and with appropriate values 1119209684 M * skceb thx, trying it now 1119210046 M * Greek0 make sure you have the latest debootstrap from sid, (the .2 thing) 1119210061 M * Greek0 the .1 version had a bug that prevented the system from being setup correctly 1119210418 M * skceb where to specify? 1119210433 M * skceb now its stuck @ "I: Checking component main on ftp://ftp.at.debian.org/debian..." for about 5 minutes 1119210469 M * skceb the host system is sarge btw, but i assume that that doesnt matter?! 1119210548 M * Bertl could you try a debootstrap to some dir (without the util-vserver tools?) 1119210764 M * skceb how? 1119210804 M * Bertl debian folks probably know that better than I :) 1119210822 M * skceb i know the gentoo stage1 install, so i think to start i would need all sources and some kind of script 1119210885 M * skceb well thx so far 1119210975 M * Greek0 skceb: that took some time too here 1119210985 M * Greek0 though I had excessive network traffic during that time 1119211011 M * Greek0 I could imagine that's what joeyh ment when he said that debootstrap is still buggy in sid 1119211035 M * Greek0 just use the debootstrap method with vserver, and wait. 1119211038 M * aba Greek0: debootstrap in sid is massivly changed currentl. 1119211059 M * aba use the sarge version if you wantto be happy for now 1119211062 M * Greek0 aba: I noticed when debootstrap failed on me yesterday :) 1119211092 M * skceb i tried again, hang at Configuring base-config..., but i will wait an hour or so :P 1119211108 M * Greek0 it didn't take _that_ long here 1119211128 M * skceb and i think im using debootstrap from sarge, hostsystem is sarge 1119211130 M * aba skceb: hm, is your output connected wrong? Can you run it with set -x? 1119211160 M * skceb how do you mean that? 1119211179 M * skceb x-window? i dont have any graphical stuff installed 1119211225 M * Greek0 aba: you mean that progress notification is being printed, just in some log, and not to stdout? 1119211310 M * aba that is just a guess that debconf wants to write at some fd which is not connected. 1119211319 M * aba it's a wild guess, and it might be wrong 1119211453 M * Greek0 it works reasonably well, one just has to have patience. 1119211471 M * Bertl maybe his network conenction is just -very_ slow? 1119211479 M * skceb no 1119211484 M * skceb i got 5mbits dowbn 1119211492 M * skceb and watching iptraf 1119211494 J * monrad ~monrad@213083190130.sonofon.dk 1119211497 M * Bertl hmm ... 1119211503 M * Bertl skceb: which tools? 1119211505 M * skceb total: o.o kbits/sec 1119211522 M * skceb http://www.13thfloor.at/~ensc/util-vserver/files/alpha/util-vserver-0.30.207.tar.bz2 1119211535 M * Bertl should be fine then ... 1119211542 M * skceb :P 1119211556 M * Bertl well, you can add --debug 1119211577 M * skceb well it passes a lot of stuff, and always "hang" at configuring base-config 1119211581 M * Bertl and look where it ends/hangs .. then move the --debug there .. until you reach a debootstrap 1119211590 Q * knoppix_ Read error: Connection reset by peer 1119211595 M * skceb but there is no cpu usage.. 1119211602 M * skceb that what im wondering about 1119211609 M * aba skceb: does a plain debootstrap work? 1119211611 M * Bertl then execute this debootstrap with strace -fF -o debootstrap.trace ... 1119211637 M * skceb i have no idea how to do it @ aba 1119211639 M * Bertl and see where it hangs/stalls 1119211657 M * skceb ill add --debug next time 1119211661 M * skceb for now i wait .P 1119211702 M * aba skceb: as root "debootstrap sarge $dir $url", where dir is the directory it should be installed into, and url the next mirror 1119211738 M * aba or even better with either an strace before, or set -x before, 1119211812 M * skceb does the vserver stuff need debootstrap? cause i just realize its nor installed P 1119211841 M * Bertl hmm, yeah :) 1119211847 M * skceb cool 1119211991 M * skceb seems to work now *g 1119212019 M * Bertl congrats! ;) 1119212023 M * skceb vserver utils configure script doesnt check if debootstrap is installed 1119212036 M * Bertl please file a bug report on savannah ... 1119212056 M * skceb savannah? 1119212071 M * Bertl https://savannah.nongnu.org/bugs/?group=util-vserver 1119212104 M * skceb ok 1119212112 M * Bertl tx 1119212464 M * skceb done 1119212592 M * skceb hm, still a bunch of new errors, exciting ;) will habe a look at it 2morrow. thx for help and bye 1119212605 M * Bertl cya 1119212658 J * knoppix_ ~knoppix@dsl-213-023-135-094.arcor-ip.net 1119213112 Q * yarihm Quit: Leaving 1119213484 M * Greek0 OK, I've had a look at util-vserver today, though I still don't understand some things. for example where the sync method comes when shutting down, but I'm quite confident I can find this out tomorrow. 1119213508 M * Greek0 but somehow unifying stop and reboot will take quite some refactoring afaics 1119214845 J * Aiken ~james@tooax6-117.dialup.optusnet.com.au 1119214901 M * Greek0 Bertl: does the kernel mess with /var/run/vshelper/? (yes, I know how dumb that question sounds) 1119214964 M * Greek0 I'm just completely lost now where that file is changed. 1119214965 M * p If i want to limit disk space for a vserver by using user/group quota, how can i start that vserver under the uid/gid? 1119214967 M * Bertl hehe, no it doesn't 1119214990 M * Bertl p: you can't .. each guest can use all users/groups 1119215013 M * Bertl p: as I mentioned previously, what you want to use are Disk Limits 1119215032 M * Greek0 Ok, here's what I'm doing: 1119215048 M * Greek0 vserver sid start, file in /var/run/vshelper/ gets created. 1119215057 M * p Bertl: is that aviable in vs1.2.10? 1119215071 M * Bertl yes, as an addon patch ... 1119215083 M * Greek0 content /etc/vserver/sid\ndefault 1119215109 M * Greek0 vserver sid enter, inside: reboot -f 1119215110 M * Bertl okay 1119215158 M * Greek0 as soon as I do that, the /var/run/vshelper/ file changes to /etc/vserver/sid\nsync\n/tmp/vserver-stop.S1KCLT/pipe 1119215169 M * p Bertl: i think i've found something like this but this doesnt match my kernel version (2.4.30) 1119215195 M * Greek0 in the first line in /sbin/vshelper, before doing _anything_ else, i cat /var/run/vshelper/ >> logfile 1119215202 M * Greek0 and it's already the modified file 1119215243 M * Bertl p: http://www.13thfloor.at/vserver/s_addons/quota/ try the latest here, if it doesn't apply cleanly (ontop of the vserver patched kernel) elt me know 1119215258 M * Bertl Greek0: the thing is, there is not jsut a single reboot call 1119215263 M * p ok thx 1119215274 M * Greek0 Bertl: ? 1119215289 M * Bertl Greek0: the reboot syscall has several options, and IIRC init does call some of them on reboot 1119215303 M * Greek0 more sys_reboots() from the guest? or does the kernel dispatch it to more applications? 1119215317 M * Bertl I'd suggest to use a 'simple' logging helper for the first few tests 1119215334 M * Bertl i.e. something which does echo "$@" >>/tmp/helper.log 1119215342 M * Bertl exit 0; 1119215378 M * Alissa the vserver do not support tun0? 1119215397 M * Greek0 ok, knowing that the helper just gets called several times already unpuzzles quite a few things. 1119215398 M * Bertl well, they support it .. but they do not use it themselves 1119215648 M * Greek0 Bertl: thanks a lot, that cleared things up :) 1119215650 M * mugwump morning all 1119215662 M * Greek0 good night all 1119215665 M * Greek0 cu 1119215678 M * Bertl night Greek0! 1119215739 M * Bertl Alissa: yes, the guests 'allow' you to use tun0 the way you use eth0 or whatever ... but they do not actively use it themselves 1119215760 M * Alissa Bertl ok than's 1119215773 M * Alissa I modify the config of the virtual server? 1119215824 M * Bertl a tun device has to be created, this is _your_ part, once it is there, you can use it in the config like you do with eth0 now 1119217467 J * Alissa`Umount ali@151.81.3.199 1119217862 Q * Hollow Remote host closed the connection 1119217872 J * Hollow ~Hollow@home.xnull.de 1119217876 Q * Alissa Ping timeout: 480 seconds 1119217885 M * Hollow heya 1119217906 N * Alissa`Umount Alissa 1119217920 M * Bertl hey Hollow! 1119218083 Q * duckx Quit: Client exiting 1119218235 M * Hollow Bertl: why are some commands prefixed with vx_? 1119218361 M * Bertl like? 1119218398 M * Hollow vcmd_vx_set_vhi_name 1119218426 M * Hollow but it's not vcmd_vx_set_cflags e.g. 1119218434 M * Bertl sec 1119218476 M * Hollow vcmd_vx_info too 1119218489 M * Bertl ah, no, vx_info means vx_info ;) 1119218494 J * wolli ~helmut@chello212186065148.1.14.vie.surfer.at 1119218497 M * Hollow i.e.? 1119218529 M * Bertl bwelcome wolli! 1119218536 M * Bertl *welcome even ;) 1119218560 M * Bertl Hollow: the VCMD_vx_set_vhi_name is a mistake ... 1119218577 M * Bertl but the VCMD_vx_info is intentional 1119218587 M * Bertl as it returns the vx_info structure/infos 1119218588 M * Hollow so what does vx mean? 1119218602 M * Bertl vx and nx are for context and network(context) 1119218617 M * wolli hi all 1119218635 M * Bertl Hollow: Vserver-conteXt, Network-conteXt 1119218646 M * Hollow ok.. and vc is? 1119218660 M * Bertl Vserver-Command 1119218711 M * Hollow then what is the difference between vx_set_namespace and vc_set_namespace e.g.? 1119218775 M * Bertl vx_set_namespace() is a kernel internal helper 1119218791 M * Bertl while vc_set_namespace() is the syscall command 1119218812 M * Hollow so i should only use vc functions 1119218825 M * Bertl you should not be able to use either ... 1119218844 M * Hollow because of #ifdef KERNEL, right? 1119218851 M * Bertl yep 1119218855 M * Hollow good... :) 1119218861 M * Doener the single interface you have is the syscall 1119218877 M * Bertl vserver(VCMD_set_namespace, ...) 1119218885 M * Hollow yup 1119218939 M * Doener i'm off to bed... got to get up early tomorrow... 1119218946 M * Hollow n8 Doener 1119218948 M * Doener have a good one! 1119218953 M * Doener night! 1119218957 N * Doener Doener_zZz 1119218959 M * Bertl night Doener_zZz! 1119218977 M * Hollow i realized that programming with vserver isn't that difficult :P 1119219004 M * Hollow but i got shocked by the utils in the first place ;) 1119219028 M * Bertl I take this as a compliment for the API :) 1119219061 M * Hollow yeah, not too complicated, especially with your syscall wrapper ;) 1119219107 M * Hollow the wrapper handles the syscall on all archs, right? 1119219213 M * wolli I am missing a remove function in the utils 1119219235 M * Bertl wolli: we too, but I added a stub to the openfoundry version 1119219256 M * Bertl wolli: you might want to look into that and fill in the required commands ... 1119219299 M * Greek0 err 1119219309 M * Greek0 isn't that was vserver remove does? 1119219329 M * Greek0 err. destroy 1119219336 M * Bertl which version do you currently look at? 1119219383 M * Greek0 svn from uh.. the url you gave me yesterday 1119219394 M * Bertl that's the open foundry version *G* 1119219404 A * Greek0 tries to remember the name :) 1119219406 M * Bertl i.e. you are looking at my stub there ... 1119219417 M * Greek0 yep, just noticed your (c) in the header 1119219441 M * Bertl filling in a few commands would probably make that work as expected 1119219457 M * Greek0 :) 1119219460 M * Bertl but I was too lazy to search for the variables expanding to the pathes 1119219493 M * wolli wher can I find the open foundry version? 1119219541 M * Bertl http://www.paul.sladen.org/vserver/archives/200505/0028.html 1119219575 M * Bertl http://www.paul.sladen.org/vserver/archives/200505/0022.html 1119219578 M * Bertl (actually) 1119219578 M * wolli 0.30.207 has a lot of dead code, i.e. not working functions 1119219604 M * wolli thx, will look into it 1119219620 M * Bertl agreed there is a lot of stuff broken or missing actually 1119219666 M * wolli e.g. copy or move would be nice 1119219722 M * Bertl yes, IIRC I added a stub for that too :) 1119219739 M * Greek0 :) 1119219933 M * Greek0 Bertl: is there a vserver devel mailing list? 1119219950 M * Bertl just use the main ML, all developers are subscribed 1119219966 M * Bertl (and I think it will not hurt the others to read something devel related) 1119220009 M * Hollow off to bed.. cu tomorrow 1119220016 M * Bertl night, cya! 1119220267 M * Greek0 hmm. you know quick way to get a working diff from svn (trunk -> working tree)? 1119220280 M * Greek0 svn diff does produce something that doesn't really apply again with patch 1119220317 M * Bertl hmm .. it should ... 1119220398 M * Greek0 hmm. it does in scripts with -p1. 1119221522 Q * eXplasm2 Quit: Verlassend 1119221690 Q * mugwump Ping timeout: 480 seconds 1119222070 Q * neofutur Quit: leaving 1119222073 J * neofutur ~neofutur@neofutur.net 1119222085 Q * neofutur Quit: 1119222102 J * neofutur ~neofutur@neofutur.net 1119222110 M * Bertl neofutur: hmm? 1119222128 M * neofutur just changed my irssi config ;) 1119222151 M * neofutur I came here for irclogs which helped me with grsec/selinux problems 1119222166 M * neofutur but finally I'm talking more on #frsecurity than here 1119222175 M * Bertl np 1119222179 M * neofutur ;) 1119222190 M * neofutur your irclogs helped me much 1119222197 A * neofutur thanks god for irclogs ;) 1119222200 M * Bertl yeah, we like them too ... 1119222236 M * matti :) 1119222241 M * neofutur i'm not yet using vserver, only vhosts 1119222242 A * matti logs every channel too ;p 1119222245 M * neofutur but vserver will come 1119222269 M * Bertl neofutur: yeah they are the 'future' :) 1119222290 M * neofutur by now I'm just needing a kernel-secure i can trust 1119222300 M * neofutur i means secured without SELiux . . . 1119222330 M * neofutur so I came for #vserver irclogs and /list gave me #grsecurity . . . 1119222579 M * badiane hello 1119222591 M * Bertl hey badiane! 1119222597 M * badiane if I nhi 1119222600 M * badiane hi 1119222606 M * badiane how's everyone doing 1119222611 M * badiane question? 1119222627 M * Bertl fine! *answer* :) 1119222630 M * badiane if I need to set some routing and rules for each vserver do 1119222637 M * badiane I place thing in the post-start 1119222644 M * badiane or pre-start.d directory 1119222656 M * Bertl yep, you can do that 1119222675 M * badiane so the pre or the post 1119222720 M * Bertl usually the pre, as you want to be 'outside' the guest 1119222730 M * badiane tres cool 1119222835 M * badiane I've been experiencing some weirdness 1119222845 M * badiane I have a vserver connected to my wireless pci card 1119222874 M * badiane and giving out dhcp addresses to 3 laptops an a tempermental sl-5500 zaurus running gpe 1119222903 M * badiane the linux machines access the web easily (though I've experienced some delays and hanging) 1119222910 M * badiane but the windows machines just can't get out 1119222919 M * badiane when I used their crappy nslookup tool 1119222935 M * badiane it can't use the cable company provided dns numbers 1119222946 M * badiane could it be routing problems? 1119223172 M * badiane I have used tcpdump to watch the dns requests coming from the linux machine and the ones coming from 1119223214 M * badiane the windows machine and the linux ones 1119223252 M * badiane the windows one seem to send the same kind of requests but for some reason do not receive anything from the dns severs 1119223296 M * Bertl hmm, well, I don#t know how this is vserver related, but why not setup a forwarding nameserver on one of the linux machines? 1119223345 M * badiane yeah I have installed dnscache but haven't configured it yet 1119223621 Q * knoppix_ Quit: Verlassend 1119224902 M * badiane a question that is outside the vserver spectrum 1119224944 M * badiane I noticed while watching tcpdump that the word domain was appended to the mindspring.com name server 1119224953 M * badiane when the requests are going out 1119224971 M * badiane is that correct or is there something weird going on 1119225023 M * Bertl it depends on the resolver config 1119225140 M * badiane I have the search line as mindspring.com