--- Log opened sob maj 22 00:00:02 2004 --- Day changed sob maj 22 2004 00:00:02< Bertl> JonB: hmm, let me check ... 00:00:15< pflanze> fine, worked 00:00:19< pflanze> even man ip worked :)) 00:00:28< pflanze> ip addr del 10.0.5.1/24 dev lo 00:00:38< ensc> ah ok... but I would not call this a manpage 00:01:15< pflanze> (it's so cool that there are two sets of tools where the traditional one does not show essential info..) 00:01:30< pflanze> ensc: on my debian sarge "man ip" looks perfectly well. 00:01:50< ensc> my manpage says 00:01:52< ensc> DESCRIPTION 00:01:52< ensc> This utility allows you to configure your network interfaces in various 00:01:52< ensc> ways. 00:01:52< ensc> OPTIONS 00:01:55< ensc> For the complete command reference please look at the following docu- 00:01:57< ensc> ment: 00:02:00< ensc> /usr/share/doc/iproute-2.4.7/ip-cref.ps 00:02:07< ensc> then a SEE ALSO and AUTHORS section 00:02:07< pflanze> nah, I have a fine manpage 00:02:41< ensc> after working with ip, I understand 'sbin/ip addr help' also 00:02:43< Bertl> JonB: hmm, now it's working, don't know why it didn't work when I checked it, sorry that I bothered you ... 00:03:53< pflanze> do you want to fetch it? -> netcat 195.226.6.71 1442 > ip.8.gz 00:04:01< JonB> Bertl: maybe because i restarted it? 00:04:17< JonB> Bertl: anyway, you need to be able to test 00:04:39< pflanze> ensc: (only valid for one more minute :) netcat 195.226.6.71 1442 > ip.8.gz 00:04:55< Bertl> JonB: btw, how is your new project? 00:06:10< JonB> Bertl: starting tomorrow *cough* 00:06:11< JonB> ;-P 00:06:18< JonB> Bertl: i was at work today 00:09:10< pflanze> Ok, both aliases are being set up by vserver start now, seems to work fine, I can see both aliases with their ip from within the vserver using "ifconfig", 00:09:15< pflanze> but still: TCP 10.0.5.1:25->195.226.6.71:52200 (ESTABLISHED) 00:11:16< pflanze> This is my interfaces/ contents: 00:11:18< pflanze> ./loc/dev: 00:11:18< pflanze> lo 00:11:18< pflanze> ./loc/ip: 00:11:18< pflanze> 10.0.5.1 00:11:18< pflanze> ./loc/mask: 00:11:20< pflanze> 255.255.255.0 00:11:23< pflanze> ./loc/name: 00:11:25< pflanze> 5loc 00:11:29< pflanze> ./pub/dev: 00:11:31< pflanze> lo 00:11:33< pflanze> ./pub/ip: 00:11:37< pflanze> 192.168.5.1 00:11:39< pflanze> ./pub/mask: 00:11:41< pflanze> 255.255.255.0 00:11:43< pflanze> ./pub/name: 00:11:45< pflanze> 5pub 00:12:18< pflanze> ifconfig says (eth0 and lo stripped): 00:12:20< pflanze> lo:5loc Link encap:Local Loopback 00:12:20< pflanze> inet addr:10.0.5.1 Mask:255.255.255.0 00:12:20< pflanze> UP LOOPBACK RUNNING MTU:16436 Metric:1 00:12:20< pflanze> lo:5pub Link encap:Local Loopback 00:12:21< pflanze> inet addr:192.168.5.1 Mask:255.255.255.0 00:12:23< pflanze> UP LOOPBACK RUNNING MTU:16436 Metric:1 00:12:47< pflanze> from within the vserver. 00:12:54< pflanze> /proc/self/status says: 00:12:55< pflanze> ipv4root: 0105000a/00ffffff 0105a8c0/00ffffff 00:12:55< pflanze> ipv4root_bcast: ffffffff 00:13:25< hiaslboy> Bertl: I posted about modules, should I just compile unix.ko into the kernel to get rid of troubles? 00:14:08>> JonB [~NoSuchUse@129.142.112.33.ip.tele2adsl.dk] has quit [Quit: Leaving] 00:16:05< Bertl> hiaslboy: atm yes, will be fixed in the next release 00:17:13< Bertl> hiaslboy: but if you want to help a little, you could compile with everything (network and process related) compiled as modules, and report all unresolved symbols ... 00:17:49< hiaslboy> does it make sense to test a vserver-kernel with uml ? or bochs? has anybody tried yet? 00:18:18< hiaslboy> Bertl: I will do that tomorrow or Sunday - not sure which day. 00:18:22< Bertl> there is uml support (should be) and I'm testing with QEMU (== bochs^3) 00:22:54< hiaslboy> bertl: what is the way to find out about unresolved symbols? depmod ? 00:23:09< Bertl> yep, should be sufficient ... 00:27:04< hiaslboy> Bertl:WARNING: /lib/modules/2.6.6-vs1.9.0/kernel/net/unix/unix.ko needs unknown symbol free_vx_info 00:27:04< hiaslboy> WARNING: /lib/modules/2.6.6-vs1.9.0/kernel/net/unix/unix.ko needs unknown symbol free_nx_info 00:27:04< hiaslboy> WARNING: /lib/modules/2.6.6-vs1.9.0/kernel/net/unix/unix.ko needs unknown symbol nxlist_lock 00:27:05< hiaslboy> WARNING: /lib/modules/2.6.6-vs1.9.0/kernel/net/unix/unix.ko needs unknown symbol vxlist_lock 00:27:30< hiaslboy> I think that I have everything compiled as modules ... 00:27:54>> hiaslboy is now known as hiaslboy_zZ 00:27:56< Bertl> ah okay ... so the unix module is the only one? strange ... I'll check ... thanks 00:41:52 * pflanze has just looked at http://www.iku-ag.de/unternehmen/vserver.jsp 00:42:16< pflanze> funny, they do not even mention the word "vserver". 00:42:53>> Doener` [~doener@p5082DFC6.dip.t-dialin.net] has joined #vserver 00:43:05< Bertl> well, they are the experts, they do not need to mention it ;) 00:44:02< pflanze> you just mean that ironically, right? 00:44:20< pflanze> or were they part developing it? 00:44:52< Bertl> yes, that is my sarcasm ... 00:46:18< pflanze> You don't know the guys on the pictures? 00:47:52< Bertl> no, never heard of them before, just stumbled over the page ... 00:49:57>> Doener [~doener@pD9E1253D.dip.t-dialin.net] has quit [Ping timeout: 480 seconds] 01:07:33>> shuri [~shushushu@cpu183.adsl.qc.bellglobal.com] has quit [Read error: Connection reset by peer] 01:26:31>> taxcollector [~taxcollec@192.16.167.161] has joined #vserver 01:44:28>> cypromis [chuck-the-@62.212.85.27] has joined #vserver 01:45:25< Bertl> hi cypromis! hey ryan! 01:48:09< taxcollector> Howdy! 01:50:44< taxcollector> Bertl: are those unixbench results I posted to the mailing list sufficient? 01:51:26< Bertl> hum obviously missed them ... investigating now 01:52:21< Bertl> ah, yes, looks good, I assume higher values are better ... 01:52:57< taxcollector> Yes 01:53:55< cypromis> hi 01:54:06< cypromis> I am back to using vserver after some years 01:54:15< cypromis> and tried to install freevps on whiteboxlinux 01:54:33< cypromis> everything is ncie besides that destroying a security context reboots the box 01:54:34< cypromis> lol 01:54:34< cypromis> :) 01:59:28>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has joined #vserver 01:59:40< Napalm> hiya everyone 01:59:43< Napalm> hiya bert 02:00:09< Napalm> hi ensc, _id and taxcollector 02:00:12< Napalm> hows things? 02:00:56< Bertl> good, and for you? 02:01:06< taxcollector> Hey Napalm: things are good 02:01:16< Bertl> cypromis: well, freevps always had some stability issues ... 02:01:42< Bertl> cypromis: have you tried the linux-vserver releases? 02:03:23< cypromis> yeah but with redhats funky kernel that's not easy 02:03:24< cypromis> :) 02:03:32< cypromis> on other linuxes I wouldn't use the freevps 02:03:33< cypromis> stuff 02:03:34< cypromis> :) 02:03:40< taxcollector> Which version of RH? 02:03:51< cypromis> I use whiteboxlinux 3.0 02:03:57< cypromis> which is basicaly the same as RHEL 3.0 02:03:58< infowolfe> Bertl, in 1.2.6 do we have a way to limit memory? 02:04:13< taxcollector> Perhaps you know this then... 02:04:30< infowolfe> say i've got a process that i *need* to give reserved memory to (so that the blasted java processes that somebody is running won't lag me out) 02:04:30< taxcollector> For RHEL3 you can't use the generic 2.4 kernel and have everything work out of the box. 02:04:39< cypromis> yep I know 02:04:40< taxcollector> e.g. nslookup will core dump 02:04:50< cypromis> freevps has patched RHEL kenerels 02:04:54< cypromis> they are even quite up to date 02:04:58< cypromis> when RHEL releases a new one 02:04:59< taxcollector> Getting generic 2.6 to work too more effort than I wanted to expend 02:05:02< cypromis> a freevps one is out next day 02:05:09< taxcollector> s/too/took 02:05:20< cypromis> yeah I scrapped the 2.6 idea 02:05:37< infowolfe> Bertl, ping! 02:05:39< cypromis> but since I want to standardise on something where less customers make bla bla bla 02:05:40< taxcollector> I cobbled together a patch for RHEL3 kernel for vs1.3.9.1 02:05:43< cypromis> rhel seemed like a good idea 02:05:49< cypromis> and till now it is very solid 02:05:57< cypromis> hiya 02:06:02< cypromis> can I try that one ? 02:06:08< taxcollector> Sure... 02:06:15< Bertl> infowolfe: pong! 02:06:38< cypromis> anyway it loks like freevps stuff works 02:06:39< taxcollector> It isn't well tested at the moment and cheated by skipping a couple of entries in /proc... 02:06:45< cypromis> only destroying a vserver is impossible 02:06:46< cypromis> lol 02:06:52< cypromis> hahaha 02:07:00< taxcollector> and if you have problems with it bug me instead of Bertl 02:07:06< cypromis> :) 02:07:11< cypromis> haven't bugged anybody yet 02:07:19< cypromis> it's 2am and I am rebooting the box for the 15th time 02:07:31< cypromis> switched off ht to see if that helps 02:07:32< cypromis> :) 02:07:34< infowolfe> Bertl, may i /msg? 02:07:39< taxcollector> http://archives.linux-vserver.org/200405/0030.html 02:07:44< Bertl> infowolfe: yep 02:08:42< cypromis> hmmm will require some work 02:08:47< cypromis> redhat is allready at -15 02:08:47< cypromis> :) 02:09:08< taxcollector> I know; this e-mail is from the start of the month 02:09:14< cypromis> ok 02:10:53>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection] 02:10:56< cypromis> will try to patch mi hugememory kernel with it 02:10:59< Napalm> hey taxcollector 02:11:09< taxcollector> cypromis: cool 02:11:13< taxcollector> Napalm: what up? 02:11:51< Napalm> Taxcollector: i've just decided to redo my vserver installation, and go for vs1.9.0 02:12:05< taxcollector> 1.9.1 you mean? ;) 02:12:48< Napalm> well im not sure whats better 02:12:50< Bertl> 1.9.0 is a good choice, it seems pretty stable ... 02:13:00< Napalm> im sticking with stablility then 02:13:03< Napalm> ;) 02:13:18< taxcollector> I stand corrected. :) 02:13:19< cypromis> which proc entries are bogus there ? 02:13:41< taxcollector> grep the patch for FIXME RHEL 02:14:21< Napalm> bert: does 1.9.0 use the new config setup? 02:16:45< Napalm> whats the split-2.6.6-vs1.9.0.tar file used for? 02:17:08< taxcollector> It breaks down the changes in the larger patch file into manageable pieces 02:17:13< taxcollector> Same changes, just separated out 02:17:15< Bertl> 1.9.x does work with older and newer tools 02:17:29< Bertl> newer features will require newer tools ... 02:18:11< Napalm> hmm sounds like a good idea, im thinking of using the new setup so that i can develope tools for the future so on 02:18:18< Bertl> taxcollector: ahh, good point, something I wanted to ask regarding those patches 02:18:38< Napalm> how do i get the current kernel config? 02:18:57< Bertl> Napalm: for a 2.6 kernel? 02:19:23< taxcollector> \/boot/config-XXX if it is a RH install 02:19:42< Napalm> well im running rh9 with 2.4.26-vs1.27 and ive downloaded 2.6.6 and vs1.9.0 patch 02:19:54< Napalm> ive patched the source files and now i need to run the config 02:20:05< taxcollector> You have to do special stuff to run 2.6 on RedHat 9 02:20:09< Bertl> taxcollector: broken out/down and split(up) could you advise on the nomenclature ... 02:20:25< Napalm> huh? 02:20:28< Bertl> Napalm: do not use a 2.4 kernel config for configuring 2.6 kernel 02:20:35< Napalm> ok 02:20:40< Napalm> so start fresh 02:20:41< Napalm> ? 02:20:48< Bertl> check the hardware configured for your system 02:21:07< Bertl> and start with some default 2.6 config (2.6 does this with make defaultconfig) 02:21:10< taxcollector> Bertl: Split is probably more accurate since we aren't talking about physical objects I suppose 02:21:34< infowolfe> Bertl, http://www.blogdigger.com/strace.txt 02:21:40< infowolfe> oh, and this luser is a moron... 02:21:49< infowolfe> the guy's running his site out of "/root/web" 02:22:09< taxcollector> "broken out" is quite right for this case 02:22:26< Napalm> lol 02:22:36< taxcollector> s/is/isn't 02:22:41< Napalm> idiot, i did'nt think anyone was that bad 02:23:24< Napalm> any ideas on the direction i should take this bert? or taxcolelctor? 02:23:29< Napalm> *taxcollector 02:23:53< Bertl> so using split-blabla for those tars sounds good to you? 02:23:53< taxcollector> Napalm: Depending on your expertise, you should decide if you want to hack rh9 to host 2.6 or just go with fedore core 02:24:05< Napalm> erm, 02:24:09< taxcollector> Bertl: yup 02:24:35< Napalm> i have no idea 02:24:50< taxcollector> Google the for the existing howtos 02:24:54< Napalm> :s 02:25:03< taxcollector> see if you want to do what they suggest 02:28:02< cypromis> hmmm quite some edit 02:28:02< cypromis> s 02:28:24< taxcollector> Yours or mine? 02:30:46< cypromis> mine 02:30:47< cypromis> fork.c 02:30:49< cypromis> unistd.h 02:30:51< cypromis> and some others 02:31:01< cypromis> will see if it works lol 02:31:17< taxcollector> You mean for -15? 02:32:50< cypromis> yep 02:32:58< cypromis> also some stuff you did for ppc64 does not go in 02:33:02< cypromis> but I just ignore it 02:33:56< taxcollector> OK; generally speaking my patch doesn't differ very much from the stock 1.3.9.1 patch... 02:34:19< taxcollector> Data structures were called by different names mostly, RH added their own system calls which messed up the diff, etc. 02:34:42< taxcollector> Main difference was that RH incorporated O(1) patch from 2.6... 02:35:24< Bertl> hmm, how did you 'adapt' the scheduler stuff to the O(1)? 02:35:26< taxcollector> That code I took from the 1.9.x tree which inherited its logic from the original vserver ck1 patch that used to be maintained 02:35:41< Bertl> okay, thanks for the info ;) 02:36:03< ensc> taxcollector: thx for the bugreport 02:36:09< taxcollector> You're welcome 02:36:49< taxcollector> ensc: If I boot off a non-vserver kernel, vprocunhide causes the boot process to hang. Any ideas? 02:37:24< cypromis> other solution is to use freevps kernel with the standard vserver utils 02:37:29< cypromis> I'll try both 02:37:29< cypromis> :) 02:37:41< taxcollector> For sure... 02:38:06< ensc> taxcollector: disable it, and execute 'bash -x /usr/lib/util-vserver/vprocunhide' and see where it hangs 02:38:45< taxcollector> It doesn't hang when I run it normally after boot though 02:39:14< taxcollector> i.e.; via a manual /etc/init.d/vprocunhide start 02:40:17< taxcollector> I must go walk over and reboot it manually; back in a few... 02:40:37< ensc> mmh... add a 'set -x' at top of vprocunhide 02:40:51< taxcollector> OK 02:43:48< cypromis> hmm config files should stay the same 02:43:50< cypromis> so 02:43:52< cypromis> compiling now 02:44:52< cypromis> util-vserver does not need linuxconf anymnore ? 02:53:53< taxcollector> ensc: disregard what I said about vprocunhide hanging the boot process; either it only happened with the unpatched version or it was a figment of my imagination 02:54:54< Bertl> in the latter case, you have quite a vivid imagination ;) 02:55:24< taxcollector> :) 02:58:14< cypromis> do I need the util-vserver-legacy package ? 02:58:28< taxcollector> I don't use it 03:02:28< netrose> Bertl: ensc: pflanze: back to the debian sarge installation 03:02:55< cypromis> ok time to reboot 03:02:56< cypromis> brb 03:03:07< Bertl> netrose: yes sir! ;) 03:03:26< netrose> There is definitely some problem with the utils and the vserver build for debian sarge. 03:03:34< netrose> It breaks like this: 03:03:35< netrose> Errors were encountered while processing: 03:03:35< netrose> apt-utils 03:03:35< netrose> W: Failure while configuring base packages. This will be attempted 5 times. 03:03:35< netrose> dpkg: dependency problems prevent configuration of apt-utils: 03:03:35< netrose> apt-utils depends on libdb4.2; however: 03:03:36< netrose> Package libdb4.2 is not installed. 03:03:36< netrose> dpkg: error processing apt-utils (--configure): 03:03:38< netrose> dependency problems - leaving unconfigured 03:04:06< netrose> Then tries that 5 times and finally ends and the server cannot be restarted 03:04:40< netrose> if I copy the Fake start-stop-daemon to the real one it starts fine but there are all kinds of problems later on. 03:05:05< netrose> It seems that not everything is installed via the build script. It seems as if some things are missing. 03:05:28< pflanze> netrose: have you read the link I pasted about the "Howto install Debian remotely"? 03:05:32< pflanze> might help you 03:06:04< netrose> No i did not. I didn't think it would help as I am not installing it on a separate server but within a vserver. 03:06:23< pflanze> Generally: I have never really met a vserver specific problem using debian. 03:06:37< pflanze> It's just debian basics. 03:06:54< pflanze> Why it broke when it built the first image, I don't know. 03:07:14< pflanze> Maybe you could do as I did and just copy a preexisting installation? 03:07:22< pflanze> Do you use Debian yourself? 03:07:43< taxcollector> Later dudes. 03:07:45>> taxcollector [~taxcollec@192.16.167.161] has quit [Remote host closed the connection] 03:08:20< pflanze> (Maybe you would be better off joining #debian or the like for getting help.) 03:08:21< netrose> No, I do not use Debian. Do you have an existing good Sarge vserver template that I could download. 03:08:42< pflanze> I did use a woody installation and then did a dist-upgrade. 03:08:56< netrose> Will that work? 03:09:02< netrose> I guess I could do that too/ 03:09:11< ensc> netrose: mmmh, debootstrap seems to have been changed... you will have to remove dev and rename dev.X to dev 03:09:27< pflanze> For someone knowing a little bit about debian, a dist-upgrade works easily. 03:09:38< netrose> ensc: I noticed that and I did that. 03:10:00< netrose> But after that there are all kinds of other dependency problems. 03:10:20< netrose> It looks as if it didn't install completely (all the packages) 03:10:21< pflanze> netrose: if you have dependency problems, the first step is to type "apt-get -f install" 03:10:40< pflanze> and usually say yes to the suggested changes. 03:11:02< netrose> Did that it won't even ask anything. It just fails right away. 03:11:17< pflanze> What does it say? 03:11:52< ensc> netrose: did you remove the 'deb http://debootstrap.invalid/ sarge main' line from sources.list 03:12:41< ensc> netrose: regarding libdb4.2, please try 'apt-get install -f' 03:13:30< ensc> else, everything seems to be fine 03:13:31< netrose> ensc: Yes I did remove that line 03:13:37< netrose> perl: warning: Setting locale failed. 03:13:37< netrose> perl: warning: Please check that your locale settings: 03:13:37< netrose> LANGUAGE = (unset), 03:13:37< netrose> LC_ALL = (unset), 03:13:37< netrose> LANG = "en_US.UTF-8" 03:13:38< netrose> are supported and installed on your system. 03:13:38< netrose> perl: warning: Falling back to the standard locale ("C"). 03:13:45< netrose> This is what happens after 03:13:57< pflanze> That's no problem, just a warning 03:14:15< pflanze> "dpkg-reconfigure locales" and choosing your locale will solve this. 03:14:20< ensc> netrose: either unset LANG, or install something so that Debian understands UTF-8 03:14:31< pflanze> But that's not a real issue. 03:14:34< netrose> dpkg: error processing webmin (--install): 03:14:34< netrose> subprocess post-installation script returned error exit status 1 03:14:34< netrose> Errors were encountered while processing: 03:14:43< pflanze> Ah, that's a problem. 03:14:44< netrose> this is another example just after that 03:14:52< pflanze> failing post-installation scripts are ugly. 03:14:56< netrose> same happens with apache-ssl and many other packages 03:15:17< pflanze> They prevent the installation process from proceeding. 03:15:36< Bertl> why do they fail at all? 03:15:39< pflanze> I think the only thing you can do is trace down why it fails. 03:15:42< netrose> Almost everything fails with post-installation failing 03:15:52< pflanze> Hm, any hint to why? 03:16:09< netrose> Have no idea. Everything fails with the same error 03:16:17< ensc> mmh, can not reproduce it here 03:16:20< netrose> subprocess post-installation script returned error exit status 1 03:16:29< netrose> Installed it with: 03:16:49< pflanze> yes, but doesn't the post-installation script issue a message itself? 03:17:01< netrose> vserver 0000sarge build -m debootstrap -- -d sarge 03:17:08< netrose> No. No message 03:17:30< pflanze> strange. then edit the post-installation script of the first package that fails, 03:17:39< pflanze> they are in /var/lib/dpkg/.. 03:17:54< pflanze> moment.. 03:18:14< pflanze> /var/lib/dpkg/info/packagename.postinst 03:19:00< pflanze> open the file and enable set shell tracing -- what is it again? set -x I thinnk 03:19:11< Bertl> yep 03:19:31< pflanze> then type "apt-get install -f" again 03:19:38< netrose> Ok. I'll try that. I just thought it might be some obvious problem in the script and that it could be fixed fast 03:20:10< pflanze> well maybe it's obvious but if the script doesn't print anything useful we need to help it 03:20:45< netrose> No, I was referring to the vserver build script. 03:21:02< pflanze> maybe sarge is just broken at the moment?.. 03:21:11< pflanze> (hope not) 03:22:14< pflanze> Maybe it's something obvious like /proc not mounted.. 03:22:22< pflanze> (whatever) 03:22:38< Bertl> ... or proc security kicking in ;) 03:23:01< pflanze> yep maybe. I didn't fiddle around with proc security yet. 03:23:14< pflanze> (Hope that doesn't make my host vulnerable) 03:23:33< Bertl> if you are the only user, then no ... 03:23:39< pflanze> I'm not 03:23:59< pflanze> then yes? 03:24:05< Bertl> probably ... 03:24:17< Bertl> what does ls /proc show inside a vserver? 03:24:39< pflanze> well quite much, what do you mean? 03:24:48< Bertl> /msg me 03:25:22< cypromis> where do I find all the possible build methods ? 03:29:30< ensc> netrose: problems seem to be caused by the broken libdb4.2 dependency. Because of that, some final steps in debootstrap will not be executed (e.g. renaming ...REAL, or fixing apt.sources) 03:29:58< ensc> netrose: nothing which can be fixed by util-vserver 03:33:27< netrose> ensc: so this is really a debian problem, right? 03:33:34< ensc> yep 03:33:39< netrose> ok 03:33:44< ensc> netrose: try '-d woody' for the meantime ;) 03:35:21< netrose> Yes, that worked flawlessly 03:37:06>> click [click@gonnamakeyou.com] has joined #vserver 03:39:10< ensc> netrose: I have a better idea. There seems to be a new debootstrap package. Execute 03:39:13< ensc> echo http://ftp.debian.org/debian/pool/main/d/debootstrap/debootstrap_0.2.38.1_i386.deb >/etc/vservers/.defaults/apps/debootstrap/uri 03:39:18< ensc> and try sarge again 03:47:06< Bertl> I guess I'm off to bed now ... 03:47:31< Bertl> ensc: will you be around this weekend, or away till monday? 03:47:54< ensc> latter; not before monday 03:48:18< Bertl> okay, then have a nice weekend, I'll try to draft something for the ipv4 interface ... 03:48:35< Bertl> we'll discuss that on monday or tuesday ... 03:48:43< ensc> ok 03:49:47< Bertl> g'night everyone ... have a nice wossname ... 03:50:06>> Bertl is now known as Bertl_zZ 03:57:00>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has quit [Quit: ] 04:08:28< netrose> ensc: I already did a dist-upgrade and that seems to be going fine, so far. 04:09:02< netrose> What is this with LANGUAGE = (unset), 04:09:02< netrose> LC_ALL = (unset), 04:09:16< netrose> It is quite annoying and pops up all the time. 04:09:30< pflanze> netrose: I already told you.. dpkg-reconfigure locales 04:09:34< cypromis> ok the staff from tax... works 04:09:35< cypromis> :) 04:09:35< ensc> netrose: the new debootstrap works better. E.g. you do not need the REAL renaming 04:10:56< netrose> ensc: I am sure it does work better, but this is almost done already. 04:11:06< netrose> I don't want to start from scratch. 05:16:31>> serving [~serving@213.186.189.12] has quit [Read error: Connection reset by peer] 06:24:25>> ensc [~ircensc@ultra.csn.tu-chemnitz.de] has quit [Ping timeout: 480 seconds] 06:39:27>> pflanze [~chris@gate.wyona.com] has quit [Ping timeout: 480 seconds] 07:12:33>> serving [~serving@213.186.189.12] has joined #vserver 08:06:18>> serving [~serving@213.186.189.12] has quit [Read error: No route to host] 08:06:27>> serving [~serving@213.186.189.12] has joined #vserver 10:04:24>> hiaslboy_zZ is now known as hiaslboy 11:44:17< cereal> moin at all 12:06:40>> cereal is now known as Sebastian 14:47:38>> pflanze [~chris@gate.wyona.com] has joined #vserver 15:25:01>> Doener` [~doener@p5082DFC6.dip.t-dialin.net] has quit [Quit: Leaving] 15:33:02>> tanjix [tanjix@ACB23742.ipt.aol.com] has joined #vserver 15:35:37< tanjix> hi together 15:36:14< Sebastian> hi tanjix 15:37:21>> Doener [~doener@p5082DFC6.dip.t-dialin.net] has joined #vserver 15:38:41< tanjix> Hi Doener :) 15:38:52< Doener> hi 15:39:05< Sebastian> hi Doener ;) 15:55:44< cypromis> how do I make sshd listen on 2 addresses with v_sshd ? 16:07:25< Doener> cypromis: create a file called /etc/vservices/sshd.conf and add a line IP="x.x.x.x y.y.y.y" 16:08:16< cypromis> I mean the sshd on the root server 16:08:24< Doener> so do I 16:08:27< cypromis> ok 16:09:15< cypromis> thx :) 16:09:17< Doener> the name of the file has to be the actual parameter that is handed to vsyswrapper... on debian for example, the init script is called ssh and not sshd, so it would be /etc/vservices/ssh.conf... 16:09:32< cypromis> k 16:10:00< cypromis> currently I run vserver in legacy mode since I could not understand how to build a vserver with the new design 16:13:37< Doener> hmm... you read the alpha tools page in the wiki? 16:16:42< cypromis> where is it 16:16:43< cypromis> ? 16:16:47< cypromis> I feel like a digger 16:16:51< cypromis> dug the whole of yeterday 16:16:51< cypromis> :)) 16:17:06< cypromis> but fixed a patch from tax... for rhel and have it working now on RHEL 16:20:29< cypromis> ok found it 16:27:51< Doener> sorry, have been busy 16:28:45< cypromis> hmmm 16:28:57< cypromis> with the new version I would have to create a framework for RHEL 3.0 16:29:01< cypromis> sinve that is what I use 16:29:02< cypromis> ok 16:29:12< cypromis> is there any way to convert the legacy stuff to the new way ? 16:34:21< Doener> i guess no... at least no way I know... 16:35:00< Doener> except using vi or whatever your favorite editor is, of course... 16:38:45< cypromis> :) 16:39:51< cypromis> hmmm 16:39:56< cypromis> I have a strange error 16:40:07< cypromis> in the root server I can see the process running with vps 16:40:14< cypromis> but in the vserver ps does not show it 16:40:20< cypromis> any capability I have to add ? 16:40:24< Doener> no 16:40:47< cypromis> hmmm 16:40:48< Doener> you're sure that the process is in the context of the vserver? 16:40:53< cypromis> yes 16:40:57< cypromis> I start it inside the vserver 16:41:02< cypromis> and vps also shows it correctly 16:41:24< cypromis> but tax... I don't remember his nick said that there are some proc bugs in his patch fix for RHEL kernels 16:41:27< cypromis> :) 16:41:40< Doener> well, i guess you hit them ;) 16:42:00< Doener> btw: the name's taxcollector ;) 16:43:53< cypromis> ok 17:13:27>> monrad [~monrad@213083190235.sonofon.dk] has joined #vserver 17:32:32>> Sebastian [~cereal@217.20.120.10] has quit [Remote host closed the connection] 17:35:51>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has joined #vserver 17:36:06< Napalm> hello everyone 17:41:15>> Sebastian [~cereal@pD9EAB2BE.dip.t-dialin.net] has joined #vserver 17:41:17>> Bertl_zZ is now known as Bertl 17:41:30< Bertl> morning everyone! 17:43:25< Bertl> cypromis: hmm, is the effect you see reproducible with ryan's patches? 17:43:56< UFOczek> Hi Bertl! 17:44:06< Bertl> hey UFOczek! how are you? 17:44:16< UFOczek> I'm fine, thanx. 17:44:26< UFOczek> i'm still waiting for ... you know :-)))) 17:44:47< Napalm> moring bert 17:45:36< Bertl> UFOczek: did you bug mcp recently? if not, I'd suggest to do so ;) 17:45:48< UFOczek> Bertl: no, i didn't. 17:45:52< Bertl> hi Napalm! what was the reason for the _not_ coming back? 17:45:59< UFOczek> Bertl: i need only wolk+grs+vs 17:46:06< Bertl> Napalm: (of the remote server) 17:46:21< UFOczek> there is avaible wolk+grs, but i have a lot of rej for vs (and i can't handle of it) 17:47:17< Napalm> bert: no idea, i look at it now nad apprently the uptime is 13 hours 17:47:26< Napalm> ie it did come back on 17:48:35< Bertl> UFOczek: IIRC the following statement where made on this list: mcp said, that he will make sure that vs and grsec play nicely together as soon as vs is ported to wolk, and I said (and did) the port to the older wolk, and I will do the port to the 2.6.6 wolk, as soon as it comes out ... 17:49:28< UFOczek> Bertl: and that's why i'm still patient. 17:49:28>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection] 17:49:39< UFOczek> huh! 17:50:01< UFOczek> The latest stable version of the Linux kernel is: 2.6.6 17:50:04< UFOczek> i didn't noticed even. 17:55:21< Napalm> anyone good with redhat? 17:56:31< Napalm> ive just updated my kernel to 2.6.6-vs1.9.0 from 2.4.26-vs1.27 and rpm has stopped working 18:00:34< Napalm> bert: what tools do i now need to compile? 18:00:48< Napalm> bert: util-vserver-0.29.213 ?? 18:01:03< Bertl> 0.29.214 or 0.29.5 is suggested 18:01:14< Bertl> older tools will work too ... 18:01:31>> chaosle [~yvan@port-212-202-168-55.dynamic.qsc.de] has joined #vserver 18:01:49< Bertl> hey chaosle! 18:02:03< chaosle> hi Bertl 18:05:47< cypromis> ryans patches ? 18:05:58< cypromis> remember I am using tax... patches for RHEL 18:06:17< Bertl> yep taxcollector == ryan ... 18:06:27>> shuri [~shushushu@cpu183.adsl.qc.bellglobal.com] has joined #vserver 18:06:35< Bertl> (soory for the confusion) 18:06:43< Bertl> hello shuri! 18:06:49< shuri> hello Bertl 18:07:10< cypromis> lol 18:07:11< cypromis> ok 18:07:15< cypromis> this is with ryans patches 18:07:31< cypromis> so the only problem I have is that it does not show me the pid in ps inside the vserver 18:08:04< Bertl> hmm, could you /msg me the output of a cat /proc//status of such a process? 18:11:57< cypromis> ok 18:12:02< cypromis> postgresql on rhel 18:12:02< cypromis> :) 18:13:54< Bertl> okay, dinner time ... back in 30 mins ... 18:13:58< cypromis> ok 18:14:02>> Bertl is now known as Bertl_oO 18:14:20>> ccooke [~ccooke@spc1-walt1-4-0-cust238.lond.broadband.ntl.com] has quit [Ping timeout: 480 seconds] 18:24:34< cypromis> it's not in proc 18:24:43< cypromis> neither on the root server nor in the vserver 18:24:46< cypromis> now that is more fun 18:24:47< cypromis> :) 18:26:30>> click [click@gonnamakeyou.com] has joined #vserver 18:35:12>> Bertl_oO is now known as Bertl 18:35:38< Bertl> hey click? 18:36:08< Bertl> cypromis: hmm, maybe you should try a vanilla kernel+vs patches ... 18:36:22< cypromis> want work 18:36:31< cypromis> vanilla does not work on RHEL 18:36:43< cypromis> anyway will talk to ryan when he shows up :) 18:36:54< cypromis> but there is one question to you as well 18:37:03< Bertl> go ahead! 18:37:08< cypromis> how can I move a vserver from legacy to the new setup ? 18:37:12< click> bertl: re, working, back in a bit. 18:37:45< Bertl> cypromis: you can create the config for a new one, and just copy over the files ... 18:38:39< cypromis> how ? 18:38:39< cypromis> :) 18:38:50< cypromis> I was using vserver long ago so the legacy system is one I know 18:38:55< Bertl> you#ve read the alpha utils page? 18:38:57< cypromis> but the new is very different :) 18:39:01< cypromis> yes 18:39:14< Bertl> there are some examples to create a new vserver ... 18:39:38< cypromis> ok 18:39:45< cypromis> with the alpha utils 18:39:53< cypromis> how do I create vservers from a sample vserver ? 18:39:58< cypromis> just use vserver-copy ? 18:40:55< Bertl> from the alpha util-vserver page: 18:40:57< Bertl> * a basic skeleton (configuration entries, /dev entries); the command would be 18:41:00< Bertl> # vserver build -m skeleton 18:41:19< Bertl> for complete list see vserver - build --help output 18:43:25< cypromis> ok 18:43:41< cypromis> I built a skeleton server but did not fully understood what it does 18:43:42< cypromis> :) 18:43:46< Bertl> of course you can 'jsut' copy a vserver fwith cp for example . 18:46:59< cypromis> ok I will play and come back later with questions :) 18:47:54< _id> re 18:48:43< Bertl> evening _id! 18:48:44< _id> Bertl, i am running a mysqldb inside a vserver and dbmail-mysql on another one. werks fine =) 18:49:36< Bertl> good to hear ... 18:50:21< _id> still on stable @ the moment 18:50:36< _id> but maybe i can switch next week to a dev version 18:50:47< _id> i like to play with mem limits & co 18:50:49< _id> ;) 18:50:50< Bertl> I have no problem with stable ;) 18:54:44< _id> oki - its now time for the late saturday RPG session - cya laters 18:57:56< Bertl> good game then! 19:10:55>> monrad [~monrad@213083190235.sonofon.dk] has quit [Quit: Leaving] 19:24:46< netrose> cypromis: " vanilla does not work on RHEL" What do you mean by this? What happens if you install a vanilla kernel? 19:25:28< Bertl> I guess a few apps/libs will break because of nptl ... but I consider it not a big issue ... 19:26:08< netrose> I wonder if those are important apps/libs. Do you know which ones are going to break? 19:26:37< Bertl> don't know which libs use nptl ... sorry ... 19:36:54>> _id [~id@pD95193A3.dip.t-dialin.net] has quit [Ping timeout: 480 seconds] 19:46:53>> _id [~id@pD9E6122D.dip.t-dialin.net] has joined #vserver 19:47:46< cypromis> glibc for example 19:47:46< cypromis> lol 20:04:46>> tanjix [tanjix@ACB23742.ipt.aol.com] has quit [Quit: ] 20:29:47>> ccooke [~ccooke@spc1-walt1-4-0-cust238.lond.broadband.ntl.com] has joined #vserver 20:48:38< eyck> that doesen't sound important... 20:48:54< eyck> whole uwoody for example works without glibc. 21:08:50< Bertl> fun: http://www.cs.vu.nl/~ast/brown/followup/ 21:19:59>> tchan [~tchan@c-24-13-81-164.client.comcast.net] has joined #vserver 21:20:16< Bertl> hey tchan! 21:20:24< tchan> hi Bertl 21:34:27< Sebastian> bertl, we have running 20 vserverīs on a host, and now the third time the vserver werenīt availble at their ip, but on the host they are up and running, and ip addr show only 21:34:27< Sebastian> 1: lo: mtu 16436 qdisc noqueue 21:34:29< Sebastian> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 21:34:30< Sebastian> inet 127.0.0.1/8 scope host lo 21:34:32< Sebastian> 2: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 21:34:33< Sebastian> link/ether 00:0d:61:c4:b5:fd brd ff:ff:ff:ff:ff:ff 21:34:35< Sebastian> inet 217.20.114.114/24 brd 217.20.114.255 scope global eth0 21:34:36< Sebastian> inet 217.20.115.24/24 brd 217.20.114.255 scope global eth0:abauer 21:34:38< Sebastian> inet 217.20.115.27/24 brd 217.20.114.255 scope global secondary eth0:nkorna 21:34:38< Sebastian> sp 21:34:39< Sebastian> 3: tunl0@NONE: mtu 1480 qdisc noop 21:34:41< Sebastian> link/ipip 0.0.0.0 brd 0.0.0.0 21:34:42< Sebastian> 4: gre0@NONE: mtu 1476 qdisc noop 21:34:44< Sebastian> link/gre 0.0.0.0 brd 0.0.0.0 21:34:57< Sebastian> what could be the problem 21:35:06< Sebastian> after a reboot all vservers works fine 21:35:22< Sebastian> but then after one or two hours the same happens 21:35:53< Bertl> hmm, please describe your site/network layout 21:36:14< Bertl> (and as usual, kernel/patch version) 21:38:40< Sebastian> 2.4.24-vs1.26 21:38:51< Sebastian> with 0.29.3 21:39:03< Sebastian> but what do you mean with site/network layout? 21:40:10< Bertl> it seems to me that you are using 217.20.114/24 and 217.20.115/24 on eth0 ... 21:40:21< Bertl> and some tunl and gre ??? 21:43:20>> tanjix [tanjix@pD9049C13.dip.t-dialin.net] has joined #vserver 21:43:24< tanjix> hi together 21:43:29< Bertl> hi tanjix! 21:44:43< tanjix> Bertl, could you join #problem please :) 21:45:59< Bertl> what is #problem? 21:46:28< tanjix> a channel here in that irc network 21:46:43< Bertl> hmm, vserver related? 21:46:50< tanjix> yes 21:47:13< Bertl> well, then move over here ... this is 'the' vserver channel after all ... 21:53:22< eyck> hmm, hey, bertl, could you please join #eyck channel.. 21:53:30< eyck> it's a channel here in that irc network.. 21:53:34< eyck> and it's vserver related 21:53:37< eyck> heh 21:53:41< Bertl> sure eyck, I'm on it ... 21:54:21< eyck> hmm, not to mention how bad publicity would be to call 'vserver' a 'problem' 22:01:16< Bertl> hmm, was there, but nobody else ... ;) 22:02:13< eyck> oh my, you really went there? 22:02:46< eyck> hmm, I need to get a life.. 22:02:49< eyck> or maybe not. 22:03:02< eyck> I find life very disturbing in code-producing activities. 22:07:31>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has quit [Quit: ] 22:13:02< Sebastian> ok the problem was, that one vserver get a primary ip and when this vserver was stopped or rebootet all secondary ip adresseīs will be down 22:13:28< Sebastian> the solution was, to set a not used ip adress as the primary ip :) 22:13:33< Sebastian> thx Bertl for help 22:14:16< Bertl> as usual, you are welcome! 22:14:53< Bertl> tanjix: hmm, so what's up? 22:15:20< tanjix> Bertl, it was that you talked with Sebastian :) 22:15:34< Sebastian> :) 22:16:12< Bertl> okay, so you are happy again ... 22:16:32< tanjix> i didnt expect anything else qith your help :) 22:27:59< Bertl> you need something else? 8-) 22:28:48< tanjix> no :) 22:29:02< tanjix> not this time :) 23:09:15>> Khahan [~Filbert@D5E0628B.kabel.telenet.be] has quit [Ping timeout: 480 seconds] 23:14:16< UFOczek> any 'tc' guru here ? 23:14:25< UFOczek> :-) 23:14:54 * Bertl is hiding behind the keyboard .oO( maybe nobody will notice ) 23:15:23< shuri> i notice 23:15:28< UFOczek> (just asked, for fun :)) 23:23:59>> Sebastian [~cereal@pD9EAB2BE.dip.t-dialin.net] has quit [Quit: ] 23:29:27>> tanjix [tanjix@pD9049C13.dip.t-dialin.net] has quit [Quit: ] --- Log closed nie maj 23 00:00:33 2004