1160957028 M * micah hmm, I'm close 1160957043 M * micah it seems to die when it tries to add the route 1160957068 M * micah Mon Oct 16 00:02:50 2006 Linux ifconfig failed: shell command exited with error status: 1 1160957081 M * micah i need to make it so it doesn't do that I guess 1160957125 M * micah there we go... needed to remove 'pull' from config 1160957155 M * Bertl I thought, maybe we should add ignore flags for certain ip operations too and/or combine that with the helper in the future 1160957290 M * micah hmm, it seems to have negotiated and appears to be connected, but I cannot move traffic through it 1160957384 M * Bertl hmm? 1160957470 M * micah i think I must add a route on the host server 1160957593 M * micah my host has the tun0 interface configured with the right IP, the vserver has that device configured with that IP, but there is no route on the host to the VPN network, and if I try to add it, I get "SIOCADDRT: Network is unreachable" 1160957668 M * Bertl then you are doing something wrong (regarding routing) 1160957705 Q * Piet Quit: Piet 1160957710 M * Bertl show me the ip route ls and ip addr ls data 1160957728 M * micah # ip route ls 1160957728 M * micah 208.99.202.0/24 dev eth0 proto kernel scope link src 208.99.202.59 1160957728 M * micah 10.0.1.0/24 dev eth2 proto kernel scope link src 10.0.1.3 1160957728 M * micah default via 208.99.202.1 dev eth0 1160957769 M * Bertl okay 1160957772 M * micah http://paste.linux-vserver.org/492 1160957793 M * micah the tun0 is on 10.8.0.6/32 (and you can see no route there in ip route ls) 1160957805 M * Bertl what's the tap device for? 1160957827 M * micah I dont know, I didn't make it 1160957839 M * micah I'd like to remove it, if I can 1160957843 M * Bertl then it's very likely the software created it 1160957860 M * Bertl are you sure that you connected it to tun0? 1160957864 M * micah I think I created it incorrectly 1160957872 M * micah because I ran tunctl with no arguments and it made it 1160957875 M * micah I just did that again: 1160957878 M * micah # tunctl 1160957878 M * micah Set 'tap1' persistent and owned by uid 0 1160957882 M * micah so now I have a tap1 heh 1160957892 M * Bertl and this is a tap not tun device 1160957910 M * Bertl i.e. it is with the -e option (which you usually do not need/want) 1160957914 M * micah yes, I originally did: "tunctl -t tun0" 1160957935 M * Bertl the tun0 is a tap device too 1160957943 M * micah I did a tapctl -d tap0, and tapctl -d tap1 and those are gone now 1160957955 M * micah oh it is? how do you know that? 1160957969 M * Bertl tun devices have not mac 1160957975 M * Bertl -t 1160958030 M * micah ok, I should create the tun0 correctly then 1160958083 M * micah hmm 1160958083 M * Bertl well, I don't know at what layer your software operates 1160958093 M * micah my tunctl has strange options: 1160958093 M * micah my tunctl has strange options:Create: tunctl [-b] [-u owner] [-t device-name] [-f tun-clone-device] 1160958104 M * Bertl most of my tunnels are layer 3 not layer two 1160958106 M * micah not sure how I said that twice... :) 1160958179 M * micah maybe I create the tun0 this way: tunctl -t tun0 -f /dev/net/tun 1160958205 M * Bertl well, just check, if it shows a mac, it's layer 2 1160958277 M * micah yeah, it creates layer2 :P 1160958307 M * Bertl maybe get a more recent tunctl then? 1160958355 M * micah my version comes from uml-utilities 20060323-3 1160958425 M * Bertl let me see where mine comes from :) 1160958478 M * micah the one here is quite old http://www.user-mode-linux.org/cvs/tools/tunctl/ 1160958514 M * Bertl I have a source file for you, it is from one of my rpms ... 1160958600 M * Bertl http://vserver.13thfloor.at/Stuff/tunctl.tar.bz2 1160958665 M * micah i wonder where that comes from 1160958683 M * Bertl no idea, I have it (as it seems) for some while now 1160958696 M * Bertl (must be something like 2 years at least) 1160958775 M * micah ok, with this one, how should I properly create the tun0? 1160958789 M * Bertl tunctl (that's it) 1160958798 M * micah huh, yeah that does it 1160958883 M * micah ok, now I have the tun device, and the IP is added, but I'm still back to where I was 1160958890 M * micah I have no route, and I can't seem to add one 1160958898 M * Bertl it is a tunnel, so PtP 1160958916 M * Bertl i.e. add a far IP to the PtP 1160958932 M * Bertl when you, e.g. ping that, it will go through the tun 1160958951 M * Bertl you can also specify 'generic' device routes, like 1160958961 M * Bertl ip route add default via tun0 1160958970 M * micah hmm 1160959041 M * Bertl hmm, sorry, s/via/dev/ 1160959084 M * micah got it 1160959092 M * micah ip route add 10.8.0.0/24 dev tun0 is what I needed 1160959643 M * micah Bertl: thanks for the help and the tunctl that actually works... :) 1160959657 M * Bertl np :) 1160960584 M * derjohn2 Bertl, I just try tu build a clean patch for 2.6.18.1 1160960613 M * derjohn2 why is my result of diff -NurpP --minimal1.3 MB and yours only about 970 KB ? 1160960638 M * Bertl hmm, did you clean up the kernel tree first? 1160960655 M * Bertl lsdiff will give you some hints about the files involved 1160960660 M * derjohn2 in which way? I simply unpacked the full src from kernel org 1160960698 M * Bertl yeah, but you do 'things' to that tree then :) 1160960730 M * derjohn2 onl y the vserver patch. I didnt touch anything i swear ;) ! 1160960745 M * Bertl well, that created a bunch of .orig files 1160960837 M * derjohn2 only for the rejects .. nor ?? 1160960842 A * derjohn2 looks up 1160960849 M * Bertl no, the rejects are .rej 1160960866 M * Bertl (which are probably there too) 1160960887 M * Bertl basically you want to do a test compile after your 'fixesÄ 1160960896 M * Bertl s/Ä/' 1160960909 M * Bertl and then a 'make distclean' in the kernel tree 1160960927 M * Bertl hmm, 'make mrproper' 1160960934 M * derjohn2 I deleted the rej and ,org after i merged the sstuff 1160960943 M * doener Bertl: aren't they the same? 1160960951 M * derjohn2 but I just see 6 more .origs 1160960954 M * derjohn2 why only 6 ? 1160960958 M * Bertl doener: probably 1160960972 M * Bertl doener: as this is something I do very seldom myself :) 1160960976 M * doener derjohn2: I've seen mysterious .origs some time ago, too 1160960993 M * doener don't remember how they came to life though 1160961023 M * derjohn2 we touch more than 8 "original" files , nor ? 1160961044 M * Bertl I hope so 1160961102 M * derjohn2 930K after distclean.... 1160961111 M * derjohn2 thats reasonable ... 1160961149 M * Bertl yep, looks better 1160961172 Q * ensc Ping timeout: 480 seconds 1160961190 M * derjohn2 BTW: rc40 is only on 13thfloor and not linked on the wiki. intention? 1160961227 M * derjohn2 and: an rc40 for 2.6.18.1 is nit there yet 1160961239 M * Bertl correct 1160961268 M * Bertl no intentions 1160961703 M * Bertl derjohn2: btw, I would appreciate feedback from rc40 test 1160961706 M * Bertl +s 1160961735 M * derjohn2 hm, do you consider it a dangerous release can. ? 1160961735 M * Bertl i.e. with load and such as discussed, and we need to check the new fake init too, I suspect it is not 100% yet 1160961758 M * Bertl not any more 'dangerous' than rc39 :) 1160961771 M * derjohn2 the last i have in production is rc35 1160961782 M * derjohn2 i thought its time to change to a new one ;) 1160961789 M * Bertl on 2.6.18? 1160961795 M * derjohn2 yes 1160961797 M * derjohn2 no 1160961801 M * derjohn2 2.6.18.1 1160961809 M * derjohn2 to be precise 1160961811 M * Bertl I meant the rc36 1160961813 M * Bertl *35 1160961823 M * derjohn2 no, rc35 on 2.6.17.13 1160961838 M * Bertl yep, what I expected, 2.6.18 changed a lot 1160961852 M * derjohn2 then i went on holiday to italy for 6 days. 6 days -> +5 rcs ... :) 1160961881 M * derjohn2 well, Il 1160961917 M * derjohn2 try tomorrow because I have to get up in time and sleep now 1160961926 M * Bertl okay 1160961949 M * derjohn2 is there any particuar to look for? more then fakeinit thoughts ? 1160961969 M * Bertl well, yes, we fixed a lot, and changed even more 1160961986 M * Bertl most of the pid stuff and signalling got slightly changed 1160961995 M * derjohn2 i am even curious to see ocfs on drdb with tagxid ;) 1160962021 M * Bertl yep, seems that we will get gfs2 soon too :) 1160962038 M * derjohn2 do we make use of the biederman patches? or so they do on vanilla with 2.6.19 ? 1160962072 M * derjohn2 I read on heise that there is even some IBM container stuff in vanilla .. 1160962073 M * Bertl .19 will be another change, and we will utilize the patches/changes there 1160962086 M * Bertl uts, ipc, pid 1160962092 M * derjohn2 ah 1160962156 M * derjohn2 but now my eyes arent hardly open .. bertelish: nap attack. I must S1 now. 1160962175 M * derjohn2 bye! 1160962176 M * Bertl have a good one! cya! 1160962998 M * doener S1? I'd try to get S3 at least ;) 1160963044 M * Bertl lol 1160963076 A * doener .oO( S1 == day dreaming? ) 1160964724 Q * dreamind Quit: dreamind 1160964896 Q * Aiken Remote host closed the connection 1160965357 Q * s0undt3ch Ping timeout: 480 seconds 1160965367 Q * dmax Ping timeout: 480 seconds 1160965406 N * DreamerC_ DreamerC 1160965952 J * Aiken ~james@tooax8-063.dialup.optusnet.com.au 1160967052 M * Bertl welcome Aiken! 1160967060 J * ensc ~irc-ensc@p54B4D4B4.dip.t-dialin.net 1160967235 M * Aiken hi 1160967255 M * Bertl TNS! how's going? 1160967261 M * Bertl *LTNS 1160967345 M * Aiken life is good 1160967365 M * Bertl good to hear! how's the weather? 1160967378 M * Aiken 30C 1160967386 M * Aiken wish it rain 1160967395 M * Aiken this area is on water restrictions 1160967400 M * Bertl eek! 1160967430 M * Aiken no watering of gardens or lawns, no washing of cars and so on 1160967451 M * Bertl yeah, we had that a few years ago (in the summer too) 1160967490 M * Bertl but hey, it's only 'comfort' not really a problem ... 1160967519 M * Bertl thinking of all the folks who have no water to drink ... 1160967573 M * Bertl (or not enough at least) 1160967620 M * Aiken not just summer 1160967630 M * Aiken water restrictions since last year 1160967645 M * Aiken the dams that feed Toowoomba are getting very low 1160967666 M * Bertl hmm, at least rain in sight? 1160967693 M * Aiken predicted for later this afternoon 1160967706 A * Aiken looks at bright blue sky and does not believe the prediction 1160967718 M * Bertl hehe 1160967773 M * Bertl here in Austria, you listen to the wheather forecast, and then assume the opposite, this way you're at least 50% right :) 1160967773 M * Aiken just had a bad experience with linux, 1 gig ram, vmplayer with xp as guest, host compiling c++ code, 300 meg or swap swap in use and processes started dying 1160967801 M * Aiken he when they say rain you ignore it 1160967832 M * Bertl OOM killer? 1160967848 M * Aiken nothing wth dmesg and still had > 1 gig of swap to go 1160967862 M * Bertl hmm, the killer should report in dmesg 1160967871 M * Aiken thunderbird, firefox and vmplayer were segfaulting until I rebooted 1160967883 M * Bertl that's strange ... 1160967919 M * Bertl but I remember something similar, about a month ago or so 1160967920 M * Aiken had one though I don't really want to consider, memory. might try booting with memtest later 1160967933 M * Aiken the kernel is 2.6.17-vs2.1.1-rc24 1160967967 M * Bertl after running for 6 or 7 months, my desktop suddenly decided to crash some galeons ... 1160967994 M * Bertl but except for the galeon restart, nothing was required to make it work again 1160968030 M * Aiken the web page looks different 1160968038 M * Aiken have not looked at it for awhile 1160968040 M * Bertl indeed :) 1160968052 M * Bertl we switched to mediawiki 1160968066 M * Bertl (transition still in progress) 1160968099 M * Aiken first thing I noticed is the vserver patches are easier to find :) 1160968117 M * Bertl and the logo? 1160968121 M * Aiken have been wanting to upgrade my machines to 2.6.18 1160968139 M * Aiken did not even notice :( 1160968149 M * Aiken I went straight for the latest 2.1.1 patch 1160968161 M * Aiken I like the look of it 1160968169 M * Bertl well, the .18 patches have been reported working fine, but there are still some minor issues to iron out 1160968179 M * Aiken like what? 1160968190 M * Aiken I was going to start with this thing (xp1600) and the alpha 1160968192 M * Bertl like the fake init, which got broken 1160968235 M * Bertl i.e. it isnt shown correctly yet (probably not even in rc40) 1160968288 M * Aiken I am still using util-vserver 0.30.208 1160968299 M * Bertl ah, we have a new release from daniel too 1160968301 M * Aiken it will auto start/stop any vservers marked as default 1160968313 M * Bertl so 0.30.211 is the new one :) 1160968315 M * Aiken but at shutdown leaves other vservers running 1160968338 M * Bertl I think I read somewhere that this was fixed 1160968357 M * Aiken which means / is not remounted properly because it is still in use which leaves me with the disk not filesystem not umounted cleanly 1160968387 M * Aiken I was wondering if newer version would shutdown all vservers so the filesystem could be cleanly unmounted 1160968411 M * Aiken ok, something else for me to check out 1160968456 M * Aiken my ultra still has solaris 9 on it so have tried anything vserver on that machine yet 1160968489 M * Bertl well, you can 'try' zones on that too :) 1160968502 M * Aiken on 9? 1160968511 M * Aiken I have played with zones on 10 for x86 1160968517 M * Aiken quite liked them 1160968562 M * Aiken my solaris guru uses 10 and zones a lot where he works so I can get info if I want 1160968686 M * Bertl yeah, my SO (has to) works with solaris too ... 1160969134 M * Aiken not sure what to make of solaris, close to enough to what I am used to and use it but enough differences there is a lot I don't know how to do 1160969171 M * Aiken I think the main problem is it is different and I am basically a newbie with it :) 1160969174 M * Bertl yeah, it seems that sun tries to make everything a little different than other unixes 1160969206 M * Bertl not very much, just enough to be annoying :) 1160969437 M * Aiken just tried looking at http://vsmon.revolutionlinux.com/Virtual_server_monitor 1160969446 M * Aiken Bad Gateway 1160969446 M * Aiken The proxy server received an invalid response from an upstream server. 1160969446 M * Aiken 1160969446 M * Aiken Apache-AdvancedExtranetServer/2.0.48 (Mandrake Linux/6mdk) mod_ssl/2.0.48 OpenSSL/0.9.7c PHP/4.3.4 Server at www.revolutionlinux.com Port 80 1160969497 M * Bertl yep, same seems to happen here 1160969773 M * Aiken any possible problems I should know about before I reboot this thing with a shiny new 2.6.18-vs2.1.1? 1160969819 M * Aiken I am backing up my home directory, I just don't feel like for a full system backup to finish. 1160969822 M * Bertl as usual, it can blow away your box .. but it's unlikely :) 1160969876 M * Aiken last major filesystem crash was one drive in a strip lost power and reiserfs got trashed 1160969886 M * Aiken very glad I had a recent system back up that day 1160970446 M * Bertl yeah, apropos reiserfs, have you heard about hans? 1160970493 M * Aiken who hasn't? 1160970519 M * Bertl good question .. I wonder how that will affect reiserfs ... 1160970529 M * Aiken I think any reader of slashdot will know about it 1160970536 M * Aiken don't know 1160970640 M * Aiken I think wait and see and continue using reiserfs in the mean time 1160970656 M * Aiken time for a reboot to see what the new kernel does 1160970682 Q * Aiken Quit: Leaving 1160971481 J * Aiken ~james@tooax8-063.dialup.optusnet.com.au 1160971507 M * Bertl wb Aiken! 1160971587 M * Aiken first impression is not the best, currently using the nv driver instead of nvidia because the nvidia driver I have been using does not work with 2.6.18 1160971643 M * Bertl not _that_ unusual 1160971655 M * Aiken I know, seems to be expected :( 1160971670 M * Bertl i.e. you probably need the latest patch sets (for the proprietary driver) 1160971684 M * Aiken my default guest starts as per normal, can stop and star it with nothing in dmesg 1160971712 M * Bertl hope that is a good sign :) 1160971725 M * Aiken I think it is 1160971730 M * doener Aiken: works fine here (debian package built with module-assistant on sid) 1160971867 M * doener hm, cmake is nice... it works, it's quite convenient, it's fast, produces nice quiet Makefiles and the best: I actually understand how to use it! 1160971874 M * Aiken I'll just put up with not being able to play quake while I download the lastest driver 1160971913 M * Aiken doener you are one up on me, my first experience with cmake was installing qt 4.2 and kde trunk in a vserver 1160971920 M * Aiken still working it out 1160971930 M * Aiken I do like the % done 1160972057 M * doener it took me _hours_ to grasp the basics to write a probably sucky autohell config for a small project... In the same time, I've configured a project 10 times as big, that depends on a bunch of libraries and I've even managed to write a comfy module to detect a library which cmake didn't support ootb 1160972085 M * doener and autohell basics still equal to "what do I need to copy and paste" ;) 1160972925 M * Aiken what version of mktemp is expected with the current util-vserver? 1160972946 M * Bertl one which knows to handle directories IIRC 1160972966 M * Bertl i.e. create temp directories 1160972994 M * Aiken 1st thing I notice going from .208 to .211 tools is the .211 tools don't work with the mktemp I have installed 1160973199 M * Aiken http://paste.linux-vserver.org/493 1160973286 M * Bertl mktemp version 1.5 1160973299 M * Bertl Usage: mktemp [-V] | [-dqtu] [-p prefix] [template] 1160973785 M * Aiken don't ask why, the mktemp that was installed is the openbsd mktemp 1.5 1160973808 M * daniel_hozac you need one that understands -t -d and -p. 1160973844 M * Aiken I now have the other mktemp 1.5 which does 1160973927 M * Bertl excellent! 1160974438 Q * mugwump Ping timeout: 480 seconds 1160974486 M * Aiken Bertl I think you mentioned a reference earlier to shutting down all guests at host shutdown 1160974493 M * Aiken what was the reference? 1160974522 M * daniel_hozac you're gonna have to use a non-default initscript to do that. 1160974546 M * Aiken was wondering 1160974565 M * daniel_hozac the default one will just stop guests marked as default. 1160974670 M * Bertl daniel_hozac: btw, what about the following idea: 1160974715 M * Bertl - have a shutdown/halt (runlevel) script for linux vserver, which does the kill -15/kill -9 for all active contexts? 1160974742 M * Bertl basically as last resort when the machine is going down 1160974769 M * daniel_hozac well, vserver ... stop should run that if it can't stop it. 1160974799 M * Bertl I was more thinking of contexts not related to 'proper' guests 1160974804 M * Aiken would save doing what I just did, copied vservers-default to vserver-others and set MARK for the other guests as others 1160974804 M * daniel_hozac ah. 1160974847 M * Aiken at least now I should get my clean shutdowns with out reiserfs having to replay it's log at startup 1160974855 M * Bertl more like what the halt script for example does, before trying to unmount stuff, just for xids ... 1160974865 M * daniel_hozac yeah. 1160974897 M * daniel_hozac makes sense. 1160975090 M * doener hm, can we read the name of the vserver somewhere in /proc/virtual/$XID? 1160975101 M * daniel_hozac in cvirt, IIRC. 1160975147 M * doener so that could be used to try "vserver $NAME stop" and vkill later, right? 1160975147 M * daniel_hozac hmm, no. 1160975168 M * Bertl well, there is a field for userspace data (65chars) 1160975187 M * doener and IIRC util-vserver sets the vserver name there, doesn't it? 1160975194 M * Bertl IIRC, ensc uses it for storing the path or so 1160975208 M * doener interesting... 1160975236 M * daniel_hozac yep. 1160975291 M * daniel_hozac vuname -g --xid ... context seems to be the only way to get it. 1160975350 J * mugwump ~samv@watts.utsl.gen.nz 1160975453 M * Aiken vuname -g --xid 49156 | cut -f1 -d' ' | sed 's:/etc/vservers/::' 1160975462 M * Aiken egg 1160975630 Q * Aiken Quit: Leaving 1160975873 J * Aiken ~james@tooax8-063.dialup.optusnet.com.au 1160976923 J * bronson_ ~bronson@c-71-198-75-160.hsd1.ca.comcast.net 1160976943 Q * bronson__ Quit: Ex-Chat 1160976986 M * Aiken is there a simple way of getting a list of running contexts beside something like 'vps ax | awk '{print $2}' | sort -u | grep -v CONTEXT' and ignoring contexts 0 and 1? 1160977020 M * doener ls /proc/virtual/[1-9]* 1160977028 M * doener -d 1160977034 M * Bertl *damn* you're fast today :) 1160977064 M * Aiken ls -d /proc/virtual/[1-9]* | sed -e 's:/proc/virtual/::' 1160977109 M * Aiken maybe ls --color=none -d /proc/virtual/[1-9]* 1160977141 M * Bertl color should be disable if you pipe it (or use it otherwise) 1160977153 M * Aiken no 1160977154 M * doener ls -d /proc/virtual/[1-9]* | xargs -n1 basename 1160977186 M * Aiken that gets rid of the colours 1160977198 M * Bertl what about (cd /proc/virtual && ls -d [1-9]*) 1160977217 M * doener too easy ;) 1160977233 M * Aiken propably less expensive than the sed 1160977245 M * Bertl note that the [1-9] still leaves room for thinking :) 1160977305 M * Aiken is vuname -g --xid still the easiest way to get vserver names from the context id? 1160977328 Q * bubulak Quit: leaving 1160977347 M * Bertl not all contexts might have names 1160977377 M * doener but you can at least try to get one and just vkill after the vserver xxx stop run 1160977392 M * Bertl yup 1160977417 J * bubulak ~bubulak@whisky.pendo.sk 1160977423 M * Aiken more versatile than my work around with using MARK=others for the non default guests 1160977606 M * Bertl okay, here for the adventurous folks out there: http://vserver.13thfloor.at/Experimental/delta-lback-feat01.diff 1160977629 M * Bertl it's a cruel hack, but it seems to proof that this can work quite fine 1160977723 M * Bertl here how I tested: http://paste.linux-vserver.org/494 1160977736 M * Bertl basically you have to assign the 'magic' 1160977759 M * Bertl loopback ip as the first one, with normal remapping 1160977814 M * Bertl what I didn't check so far is the isolation, but that should be no problem, as the stack still sees/handles the magic ips 1160977914 M * Bertl okay, I'm off to bed now ... have a nice one everyone! cya! 1160977940 M * doener g'night! 1160978035 N * Bertl Bertl_zZ 1160978057 M * Aiken (root@fred) ls --color=none -d /proc/virtual/[1-9]* | xargs -n1 basename | xargs -n1 vuname -g --xid 1160978058 M * Aiken /etc/vservers/egg Linux egg.bedrock 2.6.18-vs2.1.1-rc38 #1 PREEMPT Mon Oct 16 13:44:35 EST 2006 i686 (none) 1160978058 M * Aiken /etc/vservers/avon Linux avon.bedrock 2.6.18-vs2.1.1-rc38 #1 PREEMPT Mon Oct 16 13:44:35 EST 2006 i686 (none) 1160978063 M * Aiken the --color=none is still needed 1160978135 M * doener do you force colors? 1160978174 M * Aiken yes 1160978178 M * Aiken just had a look 1160978200 M * Aiken ls --color=none -d /proc/virtual/[1-9]* | xargs -n1 basename | xargs -n1 vuname -g --xid | cut -f1 -d' ' | xargs -n1 basename | while read line; do vserver $line stop; done 1160978242 M * Aiken that line just stopped all vservers that I had running 1160978258 M * Aiken which will do what I want at system shutdown 1160978392 M * doener what does vuname actually print? 1160978458 M * doener ah, pathname in front... 1160978488 M * Aiken (root@fred) vuname -g --xid 49163 1160978488 M * Aiken /etc/vservers/avon Linux avon.bedrock 2.6.18-vs2.1.1-rc38 #1 PREEMPT Mon Oct 16 13:44:35 EST 2006 i686 (none) 1160978770 M * Aiken my init.d/vserver-others scrip now contains that line 1160978775 M * Aiken I'll see how it goes 1160978807 M * doener Aiken: xargs -n1 -I '{}' vuname -g --xid '{}' context 1160978812 M * doener saves the cut ;) 1160978852 M * Aiken ok 1160978874 M * doener same thing is possible for "vserver ... stop" 1160978901 M * doener huge xargs-only chain :) 1160978927 M * Aiken not that famillar with xargs, the while done was the easiest way for me 1160978955 M * doener I didn't know the option either, just been curious if it can be made xargs-only ;) 1160978973 M * Aiken I am getting xargs: invalid option -- I 1160979035 M * doener 4.2.28 here, -i might work for you 1160979050 M * doener xargs -n1 -i vuname -g --xid {} context 1160979103 M * Aiken the -i bit works for me 1160979216 M * Aiken http://paste.linux-vserver.org/495 1160979253 M * Aiken apart from the ls it is xargs only 1160979282 M * Aiken the xargs I am using is 4.1.7 1160981391 Q * mugwump Ping timeout: 480 seconds 1160982826 Q * Aiken Ping timeout: 480 seconds 1160983740 Q * besonen Ping timeout: 480 seconds 1160984186 Q * spTim Ping timeout: 480 seconds 1160984471 J * besonen ~besonen@dsl-db.pacinfo.com 1160985110 Q * Johnnie Read error: Connection reset by peer 1160985133 J * Johnnie ~jdlewis@jdlewis.org 1160985357 Q * shedi Quit: Leaving 1160985971 J * meandtheshell ~markus@85-124-232-125.work.xdsl-line.inode.at 1160986235 J * adray ~adray@109.102.broadband6.iol.cz 1160986300 J * dna_ ~naucki@129-204-dsl.kielnet.net 1160986333 M * adray Please could you tell me how to use CoW works in vserver? 1160986358 M * Johnnie CoW? 1160986370 M * adray Copy on write 1160986408 Q * mcp Read error: Connection reset by peer 1160986413 J * mcp ~hightower@wolk-project.de 1160986514 M * doener adray: the files are marked as CoW (immutable + another flag) and certain operations that would cause the file to be changed cause a copy of the file to be created 1160986587 M * doener check cow_break_link and its callers for details 1160986610 M * adray doener: Do I need to hashify my vservers? 1160986689 M * doener you don't have to, but you can ;) (CoW will only come into play for hashified vservers of course) 1160986707 M * doener ... or (manually) unified vserver ;) 1160986750 M * doener note that only 2.1.1-rcX include CoW and for debian you should use a really recent version to avoid problems with dpkg (IIRC these were fixed now) 1160986766 M * doener off for a shower 1160986907 J * Aiken ~james@tooax8-209.dialup.optusnet.com.au 1160987136 Q * mcp Remote host closed the connection 1160987225 J * mcp ~hightower@wolk-project.de 1160987247 Q * michal` Ping timeout: 480 seconds 1160987577 M * adray Ok. In fact, I haven't found what does vhashify and vunify exactly do. I tried to use it. I make the '/etc/vservers//apps/vunify' dirs in each vserver. Then I did 'vserver hashify' for each server. But I can't see any changes. 1160987645 J * michal` ~michal@www.rsbac.org 1160987967 M * doener adray: after restarting the vservers you should see a difference in disk usage 1160987983 M * doener the restart is required to actually free the now removed files 1160988511 Q * Johnnie Remote host closed the connection 1160988583 J * Johnnie ~jdlewis@jdlewis.org 1160988691 Q * adray Quit: Leaving 1160988852 J * Hollow_mobile ~bene@160.45.223.34 1160988856 A * Hollow_mobile waves 1160988926 M * doener hey Hollow_mobile 1160989134 M * Hollow_mobile vpn@uni :D 1160989142 M * Hollow_mobile finally online more than 10 minutes ;) 1160989254 M * doener *lol* 1160989389 M * Hollow_mobile well, on wednesdeay my dsl line will be activated at home 1160989496 J * shedi ~siggi@dsl-149-109-85.hive.is 1160991736 Q * DreamerC Quit: leaving 1160991754 J * DreamerC ~dreamerc@61-224-133-120.dynamic.hinet.net 1160992612 Q * Skram Ping timeout: 480 seconds 1160992747 Q * Hollow_mobile Ping timeout: 480 seconds 1160992827 J * Skram ~Mark@hermes.sentiensystems.com 1160992944 J * cunha ~adray@109.102.broadband6.iol.cz 1160993134 M * cunha Hashify: From my point of view: I have two vservers. One is 'refserver'. I install 'mc' on 'refserver'. Then it could be visible for the other vserver. Is it true? 1160993252 M * cunha In other words: how to install e.g. 'mc' on one server and make it visible for another? 1160993568 M * phedny is there a way to make netstat show all connections for all vservers? 1160993572 M * phedny like vps does for ps? 1160993626 M * doener chcontext --xid 1 netstat 1160993644 M * phedny shows no connections at all 1160993660 M * doener ? 1160993683 M * phedny empty list 1160993732 M * doener was just an expression of confusion ;) 1160993737 M * phedny okay :) 1160993740 M * doener chcontext --xid 1 netstat -tnalp 1160993754 M * phedny Active Internet connections (servers and established) 1160993754 M * phedny Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name 1160993773 M * doener which kernel is that? 1160993783 M * phedny Linux tenerife 2.6.18-vs2.1.1-rc38 #1 SMP Sun Oct 15 16:11:45 CEST 2006 i686 GNU/Linux 1160993861 J * mugwump ~samv@watts.utsl.gen.nz 1160993872 M * cunha Is there any useful HowTo for vunify/vhashify? 1160995019 J * lilalinux ~plasma@dslb-084-058-204-150.pools.arcor-ip.net 1160995077 Q * cunha Quit: Leaving 1160995211 Q * FireEgl Read error: Connection reset by peer 1160995859 M * meandtheshell does anybody know a smarter way counting all the files within a linux kernel source tree than du -a tree/ | wc -l 1160995913 M * matled meandtheshell: find path -type f | wc -l 1160995966 M * matled git ls-tree -r | wc -l :) 1160996049 Q * click Read error: Operation timed out 1160996050 M * meandtheshell matled: :) 1160996316 M * meandtheshell matled: "find path/to/tree -type f | wc -l" is correct because my du -a | wc -l also counts directories - FYI we do have around 1200 directories within the kernel source tree ;) 1160997144 M * harry ahm... what's wrong with: find path/to/tree -type d | wc -l for the number of directories? 1160997323 M * meandtheshell harry: nothing if you want to figure out the number of directories :) I was looking for the number of files ... 1160998096 Q * Aiken Ping timeout: 480 seconds 1160998253 J * Piet hiddenserv@tor.noreply.org 1160998259 J * s0undt3ch ~s0undt3ch@81.193.60.134 1160998274 J * dmax ~semaj@81.193.60.134 1160998511 Q * Johnnie Remote host closed the connection 1160999200 M * morfoh harry: hi, fyi ... I rediffed latest grsec to apply against the recent linux-2.6.18.1 kernel sources. maybe you want to test merging the vserver stuff into it. I posted the patch to the grsec list 1160999214 M * morfoh ;) 1160999521 J * prae ~Benjamin@5-63.206-83.static-ip.oleane.fr 1160999766 Q * fs Read error: Connection reset by peer 1160999769 J * _fs fs@213.178.77.98 1160999777 M * harry morfoh: not before 2.0.2.1 is out for 2.6.18(.1) 1160999790 M * daniel_hozac 2.0.2.1-t8 is out... 1160999805 M * daniel_hozac and if i understand things correctly, 2.0.2.2 will be next. 1160999811 M * harry that's t8 indeed ;) 1160999916 M * morfoh harry: ok ;) 1160999970 M * morfoh harry: just that the proper integration of vserver + grsec within opensde is on my todo 1160999993 M * morfoh mnemoc refused to do that 1160999999 M * harry ah :) 1161000006 M * morfoh and because I maintain grsec 1161000011 M * morfoh I have to do it :) 1161000021 M * harry you maintain grsec? 1161000031 M * harry i thought that was spender and pipacs 1161000032 M * morfoh within opensde ... yes 1161000037 M * harry ah, mkay 1161000229 M * morfoh I mean ... due to mnemoc's great integration work it's very easy to build custom "distributions" with vserver support ... now it's up to me to finish the grsec part of it ;) 1161000260 M * harry hehe 1161000273 M * harry so you want me to merge them, so your work becomes easier too... 1161000278 M * harry i'll do that ... tomorrow? 1161000301 M * harry then it's just some incremental patches to the release of 2.0.2.2 (as daniel_hozac said) 1161000303 M * morfoh harry: I won't ever try to force you :) 1161000324 M * harry hehe... if you force me, i might put a backdoor in there :p 1161000337 M * morfoh :) 1161000340 M * morfoh evil harry :p 1161000349 M * morfoh *spank* *spank* 1161000350 M * harry hehe 1161000363 M * harry /nick ben_dover 1161000365 M * harry ;) 1161000372 M * meandtheshell morfoh: this http://sourceforge.net/projects/opensde opensde? hm ... 1161000405 A * meandtheshell nerver heard about it 1161000408 M * meandtheshell *never 1161000417 M * morfoh meandtheshell: no! 1161000419 M * morfoh :) 1161000438 M * morfoh http://freshmeat.net/projects/opensde/ 1161000448 M * morfoh this one 1161000462 M * meandtheshell morfoh: ahhh ... :) 1161000523 M * morfoh meandtheshell: we have a already existing vserver host distro called "sheba" 1161000553 M * morfoh meandtheshell: and you can build vserver guests as well of course 1161000566 M * morfoh thanks to da mighty mnemoc :) 1161000569 M * meandtheshell morfoh: ok - I see - how does that distro differ from a usual DebianGNU/Linux with Linux-VServer kernel 1161000588 M * morfoh meandtheshell: opensde is no distro at all 1161000599 M * morfoh it's just the build kit 1161000610 M * meandtheshell I know - it was regarding sheba 1161000618 M * morfoh ah 1161000637 M * morfoh meandtheshell: it uses runit instead of sysvinit for example 1161000643 M * meandtheshell ok 1161000657 M * morfoh meandtheshell: it comes with an dialog based small helper to setup vservers 1161000663 M * morfoh restarting stopping etc 1161000669 M * meandtheshell ok - ncurses? 1161000673 M * morfoh yes 1161000677 M * meandtheshell :) 1161000696 M * morfoh it behaves very well 1161000711 M * morfoh in fact all opensde servers are running on a sheba 1161000793 M * meandtheshell rofl http://www.sheba.com/Splash/default.asp 1161000819 M * morfoh yeah :) 1161000840 M * meandtheshell makes happy admins and cat's as well 1161000846 M * morfoh meandtheshell: if mnemoc returns from shopping you can ask him why he called it "sheba" :) 1161000846 M * meandtheshell :) 1161000894 M * morfoh afaik him he had this in mind http://en.wikipedia.org/wiki/Sheba 1161000921 M * morfoh all the "distros" he made are within that historical context 1161000929 M * meandtheshell I see 1161000939 Q * eyck Ping timeout: 480 seconds 1161001069 M * morfoh meandtheshell: http://en.wikipedia.org/wiki/Shedu <-- this is a firewall/gw "distro" made with opensde 1161001084 M * morfoh I mean the name is derived from it 1161001092 M * morfoh even if it is called "sedu" 1161001131 M * morfoh but as I said ... if you need in deepth explanations ... ask mnemoc :) 1161001163 M * morfoh from time to time I can't follow his excourses in history :p 1161001190 J * eyck eyck@ghost.anime.pl 1161001250 J * Rich_Estill ~restill@c-24-11-195-139.hsd1.mi.comcast.net 1161001657 J * Ben81 ~Ben81@tipi0e.lri.fr 1161002363 J * Borg- borg@cube.benet.uu3.net 1161002367 M * Borg- afternoon 1161003283 M * Borg- there was a dood that said he has an Java APP with a lot of threads.. he said that he will test how it works under vserver.. did he said sth here about that? 1161003315 M * Rich_Estill Thats me. So far so good. 1161003359 M * Borg- so everything is ok? :) 1161003361 M * Rich_Estill I am trying to do some reporting now on the performance. 1161003382 M * Borg- great :) im looking forward of that.. thx :> 1161003389 M * Rich_Estill They seemed to start slowly, but ramped up fine and is running good. 1161003416 M * Rich_Estill I have ~ 5000 threads / hour that will live and die 1161003556 M * Rich_Estill I usually have a problem after about 4 days of running getting an error having to do with too many connections database connections open with MySQL. That doesn't have anything to do with vserver. I just need to make sure the database connections close. 1161003726 M * Borg- aha.. 1161003729 M * Borg- sounds good then.. 1161003888 M * Rich_Estill So far so good. It is a distributed app. The server is what I just moved to a vserver. The clients, which are also in java have been running in vservers for months. They use threads too. 1161003956 M * Borg- okey.. I already bugged my developers to fix 127.0.0.1 issue.. 1161003969 M * Borg- and I see progress.. just one component left 1161006157 Q * gdm Read error: Operation timed out 1161007196 Q * mountie Quit: LUNCK! 1161007780 J * gerrit ~gerrit@c-67-160-146-170.hsd1.or.comcast.net 1161008176 N * _fs fs 1161008667 N * Bertl_zZ Bertl 1161008672 M * Bertl morning folks! 1161008675 M * daniel_hozac morning Bertl! 1161008676 M * meandtheshell Bertl: morning! 1161008691 M * Bertl hey, fs around too? 1161008715 M * fs Bertl: yes, at work :) 1161008738 M * Bertl ah, good, did you get to test one of the later rcs regarding the X issue? 1161008811 M * fs not yet, sorry 1161008830 M * fs I've been too busy :( 1161008851 M * Bertl np, just curious 1161008870 M * Bertl okay, off for now ... back a little later ... 1161008875 N * Bertl Bertl_oO 1161008877 M * fs do you plan to backport the changes if it works? 1161008912 M * daniel_hozac to? 1161009092 Q * michal` Ping timeout: 480 seconds 1161009423 J * michal` ~michal@www.rsbac.org 1161011649 M * Borg- uhhh :) great.. everything now works :) 1161011658 M * Borg- 127.0.0.1 -> localhost fixed all problems as expected 1161011686 M * Borg- going home :) cu folks 1161011983 J * gdm ~gdm@www.iteration.org 1161012067 M * Rich_Estill sweet. Glad to hear it works 1161013760 N * Bertl_oO Bertl 1161013765 M * Bertl back now ... 1161014142 Q * shedi Quit: Leaving 1161014174 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1161014192 M * Bertl welcome Roey! 1161014415 Q * bronson_ Quit: Ex-Chat 1161014962 Q * ||Cobra|| Ping timeout: 480 seconds 1161015068 M * Borg- Rich_Estill: yeah ;) java eating up 1GB RAM ;) but thats not a issue ;) for now 1161015126 J * bonbons ~bonbons@83.222.36.111 1161015165 M * Roey Hello Bertl!!! 1161015824 J * FireEgl FireEgl@Sebastian.Atlantica.US 1161016193 M * Bertl wb bonbons! FireEgl! 1161016211 M * bonbons Hey Bertl! 1161016251 M * Bertl you might be interested in the latest 'hack' (proof of concept code) for loopback 1161016280 M * bonbons by the way, my nbd issue tourned out to be low stack space... XFS + NBD + 4kStack => *bang* 1161016294 Q * prae Quit: Quitte 1161016305 M * Bertl ah, explains it ... and I would be interested to know if similar could be done for ipv6 1161016337 M * bonbons second loopback hack? What are the differences regarding previous try? 1161016356 M * Bertl well, it's a completely new approach actually (at least for me :) 1161016366 M * Bertl it works like this: 1161016387 M * Bertl - assign a separate loopback ip to each guest 127.x.x.1 (for now) 1161016394 M * Rich_Estill Borg: What command are you using to get the memory usage? 1161016412 M * Bertl - map each request for 127.0.0.1 to that one (already done if that ip ist the first one) 1161016436 M * Bertl - revmap the 'magic' ip to 127.0.0.1 when it goes to userspace 1161016495 M * bonbons yep that's completely different approach 1161016511 M * Bertl quite simple, patch is here: 1161016531 M * Bertl http://vserver.13thfloor.at/Experimental/delta-lback-feat01.diff 1161016547 M * Bertl should not add overhead if done properly 1161016561 M * Bertl and it passes my nc_server/client tests 1161016578 M * bonbons regarding ipv6 I assume a similar thing could be done, but there we need to take really care of in-kernel handling as kernel does address type detection 1161016608 M * Bertl is it possible to have more than one loopback ip in ipv6? 1161016641 M * Bertl or would a local-to-machine ip suffice? 1161016657 M * bonbons e.g. there is the function that determines what kind of address one has (multicast, unicast, ...) that would at least need to be hacked 1161016659 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl 1161016680 M * Bertl wb ||Cobra||! 1161016697 M * bonbons probably some link-local address on lo should be sufficient, an possibly cleaner as well 1161016715 M * Bertl bonbons: i.c. well would be interesting to see if that works, will continue testing (and of course cleaning up) the hack 1161016799 M * bonbons but here again we need to be cautious, link-local addresses always require interface... 1161016854 M * Bertl well, lo is considered an interface, no? 1161016905 M * bonbons it is, but we need to take care setting interface when we map to that address (when socket is configured + avoid userspace breaking that) 1161016906 Q * Ben81 Quit: Leaving 1161016951 M * Bertl bonbons: hmm, I thought ipv6 loopback would already use lo, no? 1161017008 M * bonbons it uses lo, but ::1 is not a link-local address, it's a host-local address (would need to check if there are some other such addresses) 1161017045 M * Bertl okay, please do so and let me know ... still have to dig into ipv6 myself 1161017067 M * bonbons try pinging some fe80:: address without specifying interface, kernel will reject that 1161017271 M * Borg- Rich_Estill: top 1161017513 M * Rich_Estill Borg: Thats what I was using too, and I am using 1.8Gigs 1161017580 M * Rich_Estill Borg: I have been doing some reading, and large memory sizes for java apps are not unusual. 1161017596 M * Borg- Rich_Estill: VM/SIZE or RSS? 1161017677 M * Rich_Estill vm 1161017701 M * Bertl yeah, java is memory hungry ... bad design ... 1161017739 M * Rich_Estill mainly on my part. My app is designed to have everything in memory 1161017759 M * Rich_Estill RSS = 330764 1161017777 M * Borg- RSS is true memory usage:) 1161017799 M * Borg- yeah.. Java is Memory Monster ;) 1161017810 M * Rich_Estill I am going to try to limit it with the -Xms and -Xmx parameters and see what happens 1161017830 M * Borg- 330 is biiig :) 1161017850 M * Rich_Estill Thats what she said. 1161017882 M * bonbons Bretl: some general question, do you know how application resouces (ro data compiled into binaries) is handled by kernel? (if it's considered cache, or has to be swapped out on low memory) 1161017915 M * Rich_Estill THere isn't anything else running, so I don't think much about it. I would like to game at the same time, so I am going to limit it. 1161017930 M * Bertl bonbons: it is file backed memory 1161017945 M * Bertl i.e. pages will be simply discarded when needed 1161018029 M * bonbons that's what I tought, same principle for memory-mapped files I assume - is it accounted under cache or under what memory status field (/proc/meminfo) 1161018056 M * Bertl yeah, should be cached 1161018076 M * Bertl i.e. the line 'Cached' 1161018121 J * bronson ~bronson@66.160.177.208 1161018130 M * Bertl wb bronson! 1161018362 M * Rich_Estill be back later 1161018378 N * Rich_Estill Rich_Estill_afk 1161018579 M * bonbons Bertl: regarding IPv6 address type detection, function __ipv6_addr_type(...) in net/ipv6/addrconf.c is of interest (for #defines, check include/net/ipv6.h) 1161018693 M * bonbons we could use some space in ::/112 or ::/96 that would match the context-ID and just adjust __ipv6_addr_type() to identify those the same way as loopback 1161018723 M * bonbons this would certainly be the cleanest approach in the sense that we don't have to care about too much of the rest of kernel code 1161018799 M * bonbons at most look for equality checks in regard to ::1 (which should be rather rare; in correlation with netfilter caution is required as for IPv4) 1161018862 M * Bertl okay 1161019326 Q * derjohn2 Ping timeout: 480 seconds 1161019332 J * derjohn2 ~aj@dslb-084-059-015-252.pools.arcor-ip.net 1161019531 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com 1161019564 M * Bertl wb derjohn2! welcome mountie! 1161020632 J * stefani ~stefani@tsipoor.banerian.org 1161020650 M * Bertl wb stefani! 1161020702 M * stefani salut. 1161022601 Q * bubulak Ping timeout: 480 seconds 1161022610 J * click click@ti511110a080-4997.bb.online.no 1161022618 M * Bertl welcome click! 1161022632 Q * michal` Ping timeout: 480 seconds 1161022655 J * bubulak ~bubulak@whisky.pendo.sk 1161022775 M * Borg- click: ;) world is small 1161022789 M * click hehe hi borg :) 1161022807 M * click hi bertl, any news going on? 1161022815 M * Bertl always new stuff :) 1161022834 M * Bertl has been some time IIRC, so what's your last status? 1161022839 M * click haven't tested the vs in quite some time, so i'm kind of in the gray on new feats 1161022929 M * click guess i'll read up a but on the latest docs and changelogs :) 1161022933 M * click *bit 1161022950 M * Bertl can't hurt :) 1161022952 M * click borg: when did you start hanging here then? 1161022966 M * Borg- click: well.. 3 weeks ago :) or so 1161022974 J * michal` ~michal@www.rsbac.org 1161022986 M * click ah, well, welcome :) 1161023013 M * click i've been here a bit longer... enough to become a regular idler :P 1161023018 M * Borg- very nice channel :) too bad not much projects provide so great support via irc 1161023033 M * click i'm kind of always popping by on the oddest of occations 1161023095 Q * doener Quit: Lost terminal 1161023119 M * Borg- okey.. time to upgrade second cacti ;) 1161023177 J * doener ~doener@host.magicwars.de 1161023245 M * click oh, ok. been some time since i've looked at the webpages lately, new design... nice :) 1161023291 M * doener *sigh* another gnu screen died :( 1161023304 M * Borg- doener: what version? inside guest? 1161023338 M * doener inside guest, died while writing to a pty, bad it's a pretty outdated kernel... 1161023375 M * doener Screen version 4.00.02 (FAU) 5-Dec-03 1161023814 M * Borg- hmm.. not good 1161023825 M * Borg- died after some of workign? or after detach/attach? 1161023951 M * Bertl had some hangs here over the years, but never a dying screen so far 1161023957 A * Bertl knocks on wood 1161023982 M * Borg- ;) 1161024012 M * doener well, irssi "hung" (not sure if it was really irssi), created a second window in screen, which then also hung and finally screen completely hung 1161024029 M * Borg- hmm.. 1161024045 M * Borg- happens to me sometimes.. a hung.. but after 60 secs or so.. everything backs to normal 1161024052 M * Borg- but its cygwin-specific in my case 1161024053 M * doener a SIGTERM to screen caused irssi and the bash in the second window to become zombies, had to send a SIGKILL to screen 1161024069 M * Borg- weird. 1161024234 J * Johnnie ~jdlewis@jdlewis.org 1161024248 M * Bertl welcome Johnnie! 1161024264 M * Johnnie Howdy. 1161024524 J * rgl Rui@217.129.151.190 1161024527 M * rgl hellooo 1161024530 M * Bertl welcome rgl! 1161024537 M * rgl hi Bertl :D 1161024610 M * rgl Bertl, maybe you don't remeber me, but did you manage to use that ruby tests cases for the quota stuff? 1161024625 M * Bertl ah, now I remember ... 1161024643 M * Bertl no, we did put the quota stuff on hold again shortly after 1161024658 M * rgl ah ok 1161024668 M * rgl is there a timeline for it? 1161024698 M * rgl or its really really into the bottom of TODO? 1161024718 M * Bertl as you know, my 'priorities' shift with the 'demand' 1161024754 M * rgl demand by whom? 1161024771 M * Bertl so if you want to continue working on that (i.e. testing or so) it automatically starts moving up on my list :) 1161024808 M * rgl I see :) 1161024845 M * doener and the "demand" needs to be "active" for longer than 30 minutes, which seems to be the average time span for quota demand ;) 1161024884 M * Bertl right :) 1161024934 M * rgl I don't get why there is no need for it. how I'm supposed to manage HD space without it? 1161024956 M * Borg- rgl: by your harisma ;))) 1161024958 M * Bertl most folks do not manage harddisk space per user 1161024958 M * rgl by hopping users don't fill /home? 1161024985 M * daniel_hozac i'm using disk limits to limit the guest. 1161024994 M * Bertl and, if you control the host machine, you can use normal quota per guest 1161024998 M * daniel_hozac then the users can duke it out amongst themselves :) 1161025008 M * Borg- daniel_hozac: how? :)) 1161025012 M * Bertl (or have a separate partition per guest) 1161025048 M * daniel_hozac Borg-: dd if=/dev/zero of=$HOME/MYSPAGEMUHAHA bs=8192k? :) 1161025063 M * rgl daniel_hozac, yeah that what I though, using /home in a different partition, then the users can nuke themselfs if they are evil hehe 1161025078 M * rgl daniel_hozac, but again, if they are evil, I'll just nuke them from host hehe 1161025111 M * Borg- daniel_hozac: you create file and mount in via loop? 1161025133 M * Bertl Borg-: lvm/device mapper is a better choice 1161025147 M * doener Borg-: no, he just filled the disk to claim all space for himself with that dd call ;) 1161025160 M * doener was an extension of the "duke it out" joke 1161025163 M * doener ;) 1161025184 M * rgl Bertl, humm didn't known about that, is that on the wiki? 1161025210 M * Borg- ugh ;) 1161025217 M * Bertl rgl: what? 1161025227 M * rgl Bertl, the lvm/device mapperstuff 1161025250 M * Bertl well, quota works quite fine, even in a secure way if you have a partition per guest 1161025251 N * Roey WorkRoey 1161025262 M * Bertl rgl: and yes, that should be on the wiki 1161025313 M * rgl Bertl, humm lvm didn't return anything 1161025348 M * Bertl it's not lvm specific, let me dig out the urls for you 1161025410 M * rgl Bertl, I could find a lvm howto at http://oldwiki.linux-vserver.org/Documentation, is that the "only" doc about it? 1161025427 M * rgl which points to http://www.5dollarwhitebox.org/wiki/index.php/Howtos_Linux-Vserver_With_LVM_And_Quotas 1161025431 M * daniel_hozac for LVM itself you likely won't find any docs. 1161025446 M * daniel_hozac (on linux-vserver.org, that is) 1161025462 M * Bertl http://oldwiki.linux-vserver.org/Standard+non-shared+quota 1161025601 M * rgl ah thx :) 1161026011 M * rgl topic says "latest stable 2.02.1", shouldn't it says "2.0.2-rc31" ? 1161026037 M * Bertl hmm, no? 1161026060 M * rgl oh, I was looking at the "oldwiki.*"! sorry! 1161026065 M * Bertl np 1161026086 M * rgl still, its 2.0.2.1 hehehe 1161026115 M * rgl the new wiki layout is quite nice congrats ;) 1161026128 M * Bertl yeah, 2.02.1 is a synonym for 2.0.2.1 and tx! 1161026212 M * Bertl btw, just finished 'porting' the Happy Users page to the new wiki, if somebody want's to add there, please go ahead (but try to keep the table intact :) 1161026216 M * Bertl http://linux-vserver.org/VServer_Users 1161026333 M * rgl :D 1161026349 M * rgl there is even a Portuguese there heheh 1161026374 M * Borg- K1 Fight Club on EuroSport ;P 1161026391 M * rgl BTW, which linux dist do you guys use? 1161026397 M * Borg- me.. Slackware 1161026454 M * Bertl madriva, debian, and fc here 1161026479 N * Rich_Estill_afk Rich_Estill 1161026511 M * bonbons gentoo here 1161026536 M * rgl alright :) 1161026545 M * rgl at least a debian user heheh 1161026802 J * blueprint foobar@dslb-088-073-102-159.pools.arcor-ip.net 1161026815 M * blueprint hejo 1161026842 M * blueprint I suppose another fool who prefers to ask by reading am i. 1161026869 Q * FireEgl Ping timeout: 480 seconds 1161026917 M * blueprint 2.6.16-2-vserver-amd64-k8 unstable I cannot get there from here, anyone a debian profile? 1161026965 T * blueprint util-vserver-0.30.211 1161026998 M * blueprint hejo 1161027009 M * blueprint blue in berlin, 1161027120 T * doener http://linux-vserver.org/ <- new and shiny | latest stable 2.02.1, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.1.1-rc40, stable+grsec 2.0.2.1 | util-vserver-0.30.211 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 1161027130 M * blueprint I'm game for the wiki action, only, all of my documentation ( 2.6.16-2-vserver-amd64-k8) is a dead end 1161027171 M * blueprint I've rarely written or read so little in the way of useful information. Even if I smell that vserver is 'the way' 1161027471 M * daniel_hozac well, if you want to get help, you'll have to ask a question. 1161027546 M * blueprint ah, it's the old problem of ips/devices 1161027584 M * blueprint I initially had a debian stable install and older util-vserver and was using iptables 1161027615 M * blueprint now I'm attempting the debian way (tm) and am running a sid host trying to install sarge guests 1161027641 M * blueprint vserver build and newvserver (regardless of how I add prefix / name to the confs) 1161027643 M * blueprint work 1161027686 M * blueprint There are two, apparently fundementally different approaches on debian, and I'm lost as to what to do 1161027698 M * blueprint I'm stubborn, it's been three days. 1161027724 M * Bertl sounds interesting ... but maybe you could focus on _the problem_? :) 1161027745 M * blueprint ok. 1161027749 M * Bertl (as far as I got it, everything is working fine :) 1161027764 M * daniel_hozac yeah, that's my understanding too :) 1161027786 M * blueprint except the network. had 2 guests running but the first always greedy 1161027792 A * Bertl .o( hmm, maybe we are not used to this kind of feedback :) 1161027814 M * rgl what do you guys think about intel vt support on the latest pentium d cpu? does vserver support it? 1161027819 M * blueprint I'm always in 'verbose mode' sorry. I'll try to keep it breif 1161027821 M * Bertl greedy means? it blocks services of the second guest? 1161027828 M * blueprint yup 1161027846 M * Bertl that is a good sign that you didn't configure the networking at all 1161027846 M * blueprint vserver build appeared to build proper confs 1161027868 M * Bertl rgl: doesn't give any advantage to OS-Level virtualization as Linux-VServer 1161027875 M * blueprint I tried both with interfaces configured on the host (dummy0 and or eth0) 1161027877 M * daniel_hozac not configuring network at all doesn't make greedy guests, does it? 1161027883 M * Bertl rgl: only useful for para virtualization as Xen 1161027898 M * Bertl daniel_hozac: first IP=0.0.0.0 means allow bind all 1161027906 M * rgl Bertl, I see thx :) 1161027906 M * daniel_hozac Bertl: right, but that's not the default. 1161027915 M * Bertl daniel_hozac: right 1161027955 M * Bertl blueprint: could you upload the contents of the 'interfaces' dir in teh config to paste.linux-vserver.org 1161028010 M * blueprint the interfaces dir of the running guest? sure. but i just added prefix and name since newvserver didn't 1161028022 M * blueprint tar ball it? 1161028052 M * daniel_hozac or something like tail /etc/vservers//interfaces/*/* 1161028064 M * Bertl yeah 1161028111 M * blueprint it's just dev 'eth0', ip '192.168.1.1', prefix '32', name '1' no more 1161028132 M * Bertl okay, that is the first guest, right? 1161028140 M * blueprint yup 1161028147 M * Bertl and the second one? 1161028164 M * blueprint had it at .2 name 2, prefix 32 1161028180 M * blueprint but also with fixed ips (real /29 subnet) 1161028185 M * Bertl and they collide with services? 1161028218 M * Bertl I would suspect that they collided with host services if you had some enabled (like sshd) 1161028220 M * blueprint even when I explicitly tell all services exactly the ip to use (ie apache ports Listen 1.2.3.4:80) 1161028253 M * Bertl that sounds very strange .. what kernel version is that? 1161028258 M * blueprint killed apache on host, listen for sshd to ip address (primary ip, different from fake, 192.168.1.1) 1161028278 M * blueprint followed both approaches, well documeted to the letter, 3 times. 1161028324 M * blueprint looked at ip addr show , list and puzzled, but I'm no routing expert 1161028366 M * Bertl let's get the testme.sh script and upload the output of 1161028369 M * blueprint given inet 88.198.33.200/27 brd 88.198.33.223 scope global eth0 1161028395 M * blueprint I had tried prefix at 32 all the way down to 27 (following a tip somewhere) 1161028395 M * Bertl 'testme.sh ' and 'vserver-info - SYSINFO' to paste.linux-vserver.org 1161028417 M * Bertl nah, 32 is fine, as seems your network config 1161028438 M * Bertl the files dev/ip/prefix files are where exactly? 1161028443 M * blueprint testme checks fine for all 1161028463 M * Bertl I'm interested in the magic lines at the beginning 1161028467 M * blueprint /etc/vservers/some-wiki/interfaces/ 1161028475 M * Bertl ah! 1161028486 M * Bertl you need to make a dir there, i.e. 1161028490 M * blueprint 0/dev, 0/prefix, 0/ip, etc 1161028503 M * Bertl okay, that's fine then 1161028525 M * blueprint I though 0 and dev a bit redundant but then I'm thinking a bit recursively ;) 1161028645 M * blueprint just pasted the SYSINFO 1161028701 M * blueprint hmm just noted that the Rootdir is wrong .. is set with a symlink to /vservers/vserver-name 1161028736 M * blueprint /etc/vservers/.defaults/vdirbase -> /vservers 1161028750 M * Bertl which is the default 1161028799 M * blueprint nope the defaul is /var/lib/vservers or? I have a mount at vservers on another disk... 1161028805 M * rgl does linux-vserver work fine on em64t? 1161028830 A * rgl is thinking on getting a pentium D 1161028938 M * blueprint just pasted ip route show to paste as well 1161028968 M * blueprint obviously I need to add the ip address, I'm just confused about what SHOULD be configured on debian 1161028979 M * blueprint and how 1161029071 M * blueprint I have a feeling it's a debian unstable problem, not vserver, per se 1161029120 M * blueprint I did try a sarge, with menuconfiged custom kernel, but went with the binary in unstable as 'it just' appeared to work :) 1161029228 Q * ruskie Remote host closed the connection 1161029250 M * blueprint Bertl: would more info be useful. 1161029331 M * Bertl sec 1161029336 M * blueprint just posted ifconfig from within guest. sure. 1161029409 J * ruskie ~ruskie@ruskie.user.oftc.net 1161029424 M * Bertl the fact that the guest config points to /vservers, but the info says 28. 1161029425 M * Bertl /var/lib/vservers 1161029440 M * Bertl suggests that you have a mix of different utils installed 1161029449 M * Bertl (which is always kind of problematic) 1161029479 M * Bertl but the ifconfig info looks fine for me, could you show me the following output from within both guests? 1161029498 M * Bertl 'ip addr ls' and 'cat /proc/self/*info' 1161029506 M * Bertl welcome ruskie! 1161029591 M * blueprint I think you're right with the mix (or old conf cruft?) ... i'll past in a sec 1161029724 M * blueprint just pasted the cats ... ip is not present, shall I install? 1161029749 M * blueprint or ifconfig ? 1161029751 M * Bertl yes, that would help (iproute2) 1161029757 M * blueprint ok. 1161029782 M * blueprint of course, I'll have to do it outside, no network :< 1161029831 M * Bertl hmm, wait, your problem is what exactly? 1161029883 M * blueprint shit, not supported (externalize) on debian 1161029915 M * blueprint no network (since upgrading) from either of the vservers now. formerly it was just one greedy 1161029925 M * Bertl sidenote: you want to add a context number too (i.e. just do echo " >/etc/vservers//context" where xid is in the range from 2 to 49151 1161029945 M * Bertl oops echo > /etc/vservers//context 1161029962 M * blueprint funny, context was being automagically generated before, but I noticed it's no longer being done so 1161029980 M * Bertl you have to talk to the debian folks about that 1161029992 M * Bertl by default you specify it with --context on creation 1161030000 Q * Johnnie Remote host closed the connection 1161030001 M * Bertl (similar with the --interface) 1161030004 M * blueprint Bertl: what kind of OS are you running? I 1161030010 M * Bertl Linux :) 1161030021 M * blueprint I did in fact do the context creation flag thang. 1161030034 M * blueprint Linux is a kernel, netbsd is an operating system :) 1161030051 M * blueprint errr, was, was. :) 1161030062 M * rgl Bertl, are you running vserver on em64t? 1161030069 M * Bertl next thing you are going to tell me that Windows(R,TM,*) is an OS too? 1161030104 M * Bertl rgl: I don't possess a 64bit machine yet, but many folks use x86_64 (amd64,em64t) 1161030118 M * blueprint I've never in my life run windows. Started on vax mainframes, to cbm, to amiga, apple, bsd/netbsd, and now 4 years linux 1161030151 M * blueprint I have some old apples and linux machines (and freebsd jail boxes) 1161030169 M * blueprint I prefer debian. 1161030227 M * Bertl debian is fine, only some 'older' debian packages are not 1161030259 M * Bertl the guest setup and ip isolation in your config seems perfectly fine, I doubt that services will collide here 1161030279 M * Bertl it's fine that you cannot reach the internet with private ips though 1161030301 M * Bertl the question now is, do you have more than one public ip for that machine` 1161030306 M * Bertl +? 1161030328 M * rgl Bertl, I see, so I'm ok with it on 64bit :D 1161030338 M * Bertl rgl: you definitely should be :) 1161030363 M * Bertl rgl: same for sparc/mips/arm/alpha/hppa/s390 and others :) 1161030502 M * blueprint Bertl: I could use the 29 subnet ... I have spare ips, but trying that has failed 1161030548 M * blueprint rgl: I did have a 99% functional setup on amd 64 bit with SMP, I did something funky 1161030563 M * Bertl blueprint: you have a /27 network configured right now 1161030572 M * rgl blueprint, which was? 1161030581 M * Bertl blueprint: what do you mean with 29 subnet? 1161030585 M * blueprint X2 4400 1161030606 M * blueprint 8 ripenet meaning 6 hosts which is /29 ? or 1161030627 M * Bertl okay, which ips are assigned to you? 1161030650 M * blueprint ok. primary: 1161030652 M * blueprint 88.198.33.200 1161030654 J * debugger Rui@217.129.151.190 1161030661 M * Bertl welcome debugger! 1161030679 M * debugger Bertl, you got that in auto mode or what? -:D 1161030683 N * debugger rgl_ 1161030693 M * Bertl nope, ah, didn't conenct that to you :) 1161030702 M * blueprint subnet: 88.192.213.161-67 1161030713 M * rgl_ ah ok :) 1161030734 M * Bertl blueprint: so that is a completely different network, which kind of explains why you are having issues 1161030739 M * blueprint I feel like a very poor debbuger now. 1161030753 M * Bertl blueprint: I assume, the network has it's own router too? 1161030757 M * blueprint well, I tried to avoid the different network by just using the primary ip 1161030785 M * Bertl blueprint: yeah, but with a _single_ ip, of course, your services will collide (at least the services shown to the outside) 1161030811 M * blueprint I did try to specify the broadcast and netmasks correctly 1161030829 M * Bertl is there a router within the 88.192.213.160/29 network= 1161030832 M * Bertl s/=/? 1161030851 M * Bertl or where is it routed to/from? 1161030868 M * blueprint there 'must' be ... the ip .161 does resolve to 88.198.33.200 1161030879 M * Bertl uhum? 1161030900 M * Bertl ips _do_not_ resolve to other ips :) 1161030906 M * blueprint that's what I said, 1161030911 M * blueprint but it's a rental :) 1161030937 M * Bertl fine, you have to figure the routing setup first, otherwise nothing will work as expected 1161030937 M * blueprint I know an ip is an ip is an ip. they wind up on the same dev. they resolve to the same mac address if you will 1161030958 M * Bertl okay, let's try a few things ... 1161030963 M * blueprint ok. 1161030989 M * Bertl first, let's change the guest1 interface to ip=88.192.213.161 prefix=29 1161031001 M * Bertl and the guest2 interface to ip=88.192.213.162 prefix=29 1161031008 M * Bertl make sure to shut down the guests before 1161031084 M * blueprint ok 1161031127 Q * rgl Ping timeout: 480 seconds 1161031129 M * daniel_hozac Bertl: is 2.6.18-vs2.0.2.2 in sight? :) 1161031142 M * Bertl blueprint: once the guests are 'up' again, try the following 1161031154 M * Bertl daniel_hozac: yep, working on it ... any patches from your side? 1161031174 M * Bertl blueprint: ping -I 88.192.213.161 www.google.com (on the host) 1161031178 M * daniel_hozac nope, i'm just restarting my rebasing on 2.6.18 efforts. 1161031188 M * Bertl daniel_hozac: and the same with 162 1161031208 M * Bertl *blueprint 1161031245 M * blueprint Bertl: 100% packet loss. 1161031257 M * Bertl so your 'ips' are not routed 1161031277 M * blueprint I have my internal dns setup, had hosts interfaces setup and ONE guest running (https apache) with the 161 1161031284 M * Bertl try to ping the 88.192.213.161 and 162 ips from outside, and check with tcpdump on the host 1161031296 M * blueprint Bertl: k 1161031332 M * blueprint shit. wrong ip, misread, sorry 1161031402 M * blueprint s/192/198/ 1161031411 M * Bertl that looks better :) 1161031437 M * Bertl okay, adjust the ips 1161031451 N * rgl_ rgl 1161031479 M * blueprint rgl: identity crisis! 1161031523 M * rgl blueprint, not at all. my network died, so I had to use other nick 1161031602 J * Johnnie ~jdlewis@jdlewis.org 1161031609 M * blueprint rgl: ach. 1161031611 M * Bertl wb Johnnie! 1161031638 M * blueprint Bertl: ok, 'appears' to ping from outside otherwise not from inside 1161031658 M * Bertl so the 'pings' to google fail, yes? 1161031673 M * blueprint sorry, google ping works for host 2 1161031678 M * blueprint err, guest 2 1161031692 M * Bertl ah, so guest2 but not guest1, right? 1161031701 M * blueprint 88.198.213.162, but not 161, yes 1161031724 M * Bertl okay, so I just assume for now, that 161 is reserved for the router (double check the ips) 1161031736 M * Bertl let's use 163 for guest1 for now 1161031736 M * blueprint 88.198.213.160 / 255.255.255.248 1161031746 M * blueprint ok, sec 1161031800 M * Bertl and let's repeat the ping checks (all of them) 1161031917 M * blueprint shit. bloody fencepost. 1161031928 M * blueprint Bertl: pings in both dirs fine. 1161031930 M * blueprint hmm. 1161031939 M * Bertl okay, now let's see inside a guest 1161031943 M * daniel_hozac Bertl: what is 2.0.2.2 going to be? 2.0.2.1-t8+pidtype+secheduler fix? 1161031984 M * Bertl daniel_hozac: yep, +iosched+ext3-fix+something_I_just_forgot :) 1161031989 M * blueprint both ping fine from within. 1161031994 M * daniel_hozac ext3-fix meaning the warnings? 1161032004 M * Bertl well, it's not just a warning :) 1161032031 M * daniel_hozac right, well, the underlying problem causing the warnings then :) 1161032037 M * Bertl yep 1161032044 M * Bertl it's already fixed here ... 1161032058 M * daniel_hozac ok, i'll wait for 2.0.2.2 then. 1161032063 M * blueprint Bertl: so we know fixed ips specified correctly (sigh, sorry) work. 1161032065 M * Bertl blueprint: so now let's try some services (e.g. apache) on that guests 1161032074 M * blueprint In a jiff 1161032086 M * Bertl blueprint: no ip restrictions are required for the guest services 1161032250 M * blueprint Bertl: I tried setting up iptables with a private network on a dummy interface ... 1161032250 M * Rich_Estill later 1161032252 P * Rich_Estill Leaving 1161032701 J * Aiken ~james@tooax8-110.dialup.optusnet.com.au 1161032713 M * Bertl morning Aiken! 1161032720 M * Aiken hello 1161032740 M * Aiken the final all xargs script seems to work well for shutting down all guest 1161032782 M * blueprint BERTL: i love you guys! 1161032783 M * daniel_hozac /usr/lib*/util-vserver/start-vservers --stop --all should also stop all guests. 1161032808 M * daniel_hozac (i assume i said that earlier though :)) 1161032808 M * Bertl blueprint: let me get the url for donations :) 1161032814 M * blueprint I have 80:443:22 on both instances. and dupvserver works. 1161032870 M * Aiken don't remember that from yesterday 1161032884 M * Aiken tried it now and all guests stopped 1161032886 J * lafeuil ~lafeuil@4be54-2-82-225-0-224.fbx.proxad.net 1161032907 M * daniel_hozac the vkill should only be required for non-guest contexts. 1161032909 M * Bertl welcome lafeuil! 1161032914 M * blueprint Bertl: thanks a bunch. And I assume this only works with dedicated ips? 1161032933 M * Bertl blueprint: well, you _can_ share a single ip between guests too 1161032942 M * blueprint thanks all, I'm blasted but happy. going to have drink. 1161032946 M * lafeuil Bertl: thanks 1161032948 M * blueprint Bertl: how? 1161032949 M * Bertl blueprint: but the problem is that you cannot have port 80 on both of them 1161032965 M * Bertl blueprint: you could for example use 80/81/82 ... 1161032968 M * bonbons blueprint: and you can also nat guests' private IPs if you like 1161032991 M * Bertl yep, ant + private ips would be the choice with a single public one 1161032994 M * Bertl *NAT 1161032996 M * blueprint ah, ok, thought so. I did have success with nats ... 1161033009 M * blueprint DNAT and SNAT are your freinds :) 1161033025 M * Bertl for outgoing connections, the SNAT is sufficient 1161033083 M * blueprint Bertl et al: thanks a bunch ... I'm falling down happy. 1161033093 M * Bertl you're welcome! 1161033101 M * blueprint good night! 1161033105 M * Bertl have fun! 1161033124 M * blueprint thanks! 1161033133 Q * blueprint Remote host closed the connection 1161033176 J * shedi ~siggi@inferno.lhi.is 1161033256 Q * bonbons Quit: Leaving 1161033401 J * mire ~mire@127-166-222-85.adsl.verat.net 1161033409 M * Bertl wb shedi! mire! 1161033420 M * shedi thank you kind sir 1161033519 J * FireEgl FireEgl@Sebastian.Atlantica.US 1161033753 Q * lafeuil Quit: lafeuil 1161034005 Q * mire Quit: Leaving 1161034413 J * mire ~mire@127-166-222-85.adsl.verat.net 1161034482 Q * FireEgl Ping timeout: 480 seconds 1161034790 Q * pusling Remote host closed the connection 1161034808 J * pusling pusling@195.215.29.124 1161034835 M * Bertl wb pusling! 1161034853 M * pusling thx. who killed me? ;) 1161035109 M * gdm hi. quick question 1161035122 M * gdm if individual vservers are limited aprropariately 1161035127 M * gdm adn one has a high load 1161035135 M * gdm does that get reflected on the host? 1161035139 M * gdm or just within the vserver? 1161035155 M * Bertl depends on the kernel version 1161035163 M * Bertl but typically the load is per guest 1161035166 M * daniel_hozac and configuration, etc. :) 1161035176 M * gdm ok, good :) thank you 1161035186 M * gdm Bertl: was the one i created with you for the smp server 1161035204 M * Bertl now if I just could remember that one :) 1161035221 M * gdm hehe, yeah, don't worry 1161035230 M * gdm was vanilla and 2.0.2 patch iirc 1161035235 M * gdm but it's not impt right now 1161035240 M * gdm it's just a general answer 1161035254 M * Bertl in general, with stable, load is per guest 1161035262 Q * dna_ Quit: Verlassend 1161035300 M * Bertl I/O scheduling is not fair per guest (yet) on 2.0.x 1161035360 M * fs Bertl: do you plan to backport the X-sloppyness fix to 2.0.x? 1161035386 M * Bertl yes and no, part of the 'fix' will be in 2.0.x too 1161035412 M * fs at least the sloppyness-fixing part of the fix :) 1161035415 M * Bertl (not sure we will add the complete enter/leave part there) 1161035442 M * Bertl fs: but yes, it should get better in the next stable 1161035450 M * fs great 1161035458 M * fs do you have a schedule for the release? 1161035480 M * Bertl no, but it will be soon (read today/tomorrow) 1161035484 M * fs (background: we need it to stabilize 2.6.18 as release kernel for Debian Etch) 1161035487 M * fs great 1161035526 M * Bertl i.c. well, if you can do some testing tonight/tomorrow, that would definitely help 1161035598 M * fs yes 1161037017 M * gdm hmm, that all sounds cool. thank you for your help Bertl :) 1161037158 M * Bertl no problem! 1161037171 Q * meandtheshell Quit: exit (0); 1161037662 J * FireEgl FireEgl@Sebastian.Atlantica.US 1161037856 J * debugger Rui@217.129.151.190 1161038327 Q * rgl Ping timeout: 480 seconds 1161038800 Q * Piet Quit: Piet 1161039977 Q * debugger Ping timeout: 480 seconds 1161040107 J * lilalinux_ ~plasma@dslb-084-058-209-079.pools.arcor-ip.net 1161040146 J * debugger Rui@217.129.151.190 1161040246 N * debugger rgl 1161040389 P * stefani I'm Parting (the water) 1161040538 Q * lilalinux Ping timeout: 480 seconds 1161042576 M * ntrs Bertl, what is new/fixed in rc40? 1161042652 M * Bertl compared to rc39? 1161042722 M * Bertl minor cleanups in signalling code, some additional warnings, a test fake init (which didn't work out as expected yet) 1161042960 M * rgl Bertl, are you already using git or something alike? :) 1161043029 M * Bertl nope, I started to use it (a few weeks ago) but it turned out as chaos 1161043039 M * Bertl will give it a second try soon though 1161043053 M * rgl ouch! hehe 1161043128 M * rgl I like mercurial a lot, maybe you could try it ;) 1161043149 M * Bertl I know hg ... not very happy with it ... 1161043166 M * rgl oh why? :) 1161043167 M * Bertl what I currently do is very similar to git actually