1134432385 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134432591 Q * lonewolff Ping timeout: 480 seconds 1134432645 M * Bertl welcome Johnnie! 1134432652 M * Johnnie Hi, thanks. :) 1134433085 J * lonewolff ~lonewolff@host86-128-0-23.range86-128.btcentralplus.com 1134433424 J * Gerden ~Danger@200-213-138-246-moselli.inetbusiness.com.br 1134433442 P * Gerden 1134433443 M * Bertl welcome Gerden! 1134434633 J * prae ~benjamin@sherpadown.net 1134434942 Q * Johnnie Remote host closed the connection 1134434949 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134435523 M * Bertl welcome prae! 1134435602 Q * Johnnie Quit: G'bye! 1134435633 J * renegade334 ~renegade3@209.144.225.19 1134436125 J * Aiken_ ~james@tooax7-137.dialup.optusnet.com.au 1134436256 M * renegade334 Anyone around? 1134436273 M * Bertl nope, only us chickens ... 1134436279 M * renegade334 mm mm good 1134436281 M * renegade334 i am hungry 1134436282 M * renegade334 haha 1134436325 M * renegade334 I am getting an error with some of the vservers i have 1134436329 M * renegade334 RTNETLINK answers: Cannot assign requested address 1134436345 M * renegade334 The inet connectivity stops i restart the vserver and it says that 1134436347 M * renegade334 but all inet works 1134436378 M * Bertl that means that the ip is already configured differently 1134436391 M * Bertl it's usually a config issues and easily resolved 1134436398 M * renegade334 ok 1134436404 M * Bertl your current kernel/tool version and distro is? 1134436445 M * renegade334 just a moment 1134436473 Q * Aiken Ping timeout: 480 seconds 1134436659 M * renegade334 2.6.12-vs2.0-gentoo-r1 util-vserver-0.30.208-r5 gentoo 1134436682 M * Bertl okay, how did you configure your IPs? 1134436699 M * renegade334 gave each box a static ip 1134436750 M * Bertl ah, you assigned it on the host? 1134436789 M * renegade334 yeah 1134436806 M * renegade334 when i did the vserver-new command 1134436824 M * Bertl hmm ... k, let's do the following: 1134436835 M * Bertl stop all your guests if possible 1134436853 M * renegade334 i am on a guest atm 1134436878 M * Bertl k, then take an example guest 1134436883 M * renegade334 ok 1134436893 M * Bertl stop that one, and check with 'ip addr ls' for the ip 1134436906 M * Bertl it should not show the guest ip 1134436959 M * renegade334 no up 1134436976 M * renegade334 not there sorry 1134437024 M * renegade334 so you are correct it did not show the ip 1134437103 M * Bertl okay, now start the guest, and show me the output 1134437126 M * renegade334 no output 1134437170 M * Bertl now check for the ip again 1134437180 M * renegade334 it is there now 1134437195 M * Bertl okay, check with vserver-stat that the guest is still running 1134437209 M * renegade334 yep 1134437227 M * Bertl okay, now stop it again, and look for error messages 1134437234 M * renegade334 there isnt any 1134437253 M * Bertl okay, so everything is working fine, what was your problem again? 1134437284 M * renegade334 it works, no problem, then out of nowhere the inet conx stops 1134437307 M * renegade334 you go into the guest and it doesnt have any interface 1134437310 M * Bertl ahh, okay, well, that's probably not out of nowhere .... 1134437319 M * renegade334 true 1134437325 M * Bertl that is very likely when you stop the 'first' guest 1134437341 M * Bertl the linux kernel has a strange behaviour ... 1134437351 M * Bertl look at the 'ip addr ls' output again 1134437372 M * Bertl it will list most IPs as secondary (all except one for each subnet) 1134437383 M * renegade334 yep 1134437391 M * Bertl this one is the 'primary' IP 1134437406 M * Bertl and if that IP is removed, the kernel removes all secondary IPs too 1134437466 M * renegade334 didnt stop any 1134437525 M * renegade334 i think i know what the issue is 1134437556 M * renegade334 brb 1134437559 Q * renegade334 Quit: leaving 1134437777 J * renegade334 ~renegade3@209.144.225.19 1134437785 M * Bertl wb! 1134437788 M * renegade334 thanks 1134437811 M * renegade334 not sure if this is it i set the prefix wrong 1134437823 M * renegade334 could that have caused the interface to stop? 1134437830 M * Bertl yup, that could cause issues 1134437840 M * renegade334 ok i think it was htat 1134437841 M * renegade334 that 1134437842 M * Bertl but not a 'sudden stop, IMHO 1134437855 M * renegade334 hmmm 1134437925 M * renegade334 host was set to /29 so i changed them to 29 1134438017 M * renegade334 have you seen interfaces stop like this? 1134438036 M * Bertl no, they just stop when the primary is removed 1134438070 M * renegade334 so you think i would see something in a log? 1134438275 M * Bertl IMHO interfaces do not 'disappear' 1134438297 M * renegade334 have 4 guests and one host 1134438324 M * Bertl in any case the 'change' is caused by userspace 1134438339 M * Bertl maybe you have some fancy network hotplug script 1134438350 M * Bertl which takes the 'main' interface upp and down 1134438355 M * renegade334 interface was up on host and 1 guest, 3 guests down 1134438378 M * Bertl which I would explain by a restart of the _one_ guest 1134438389 M * Bertl which probably was lucky to have the 'primary' ip 1134438587 M * renegade334 when it happens again what should i try? 1134438597 M * Bertl well, you should 'prevent' it 1134438607 M * Bertl that's what I was getting to ... 1134438632 M * Bertl just assign a 'dummy' ip in the same network as your guests are on the host 1134438663 M * Bertl (or if the host uses an IP in the same network, make sure that the IP/prefix is correct) 1134439611 M * renegade334 So if the guests were all set to /24 and the one that was set to primary was reset it could cause the interface of the other 3 to stop 1134439622 M * renegade334 is this a correct thought/statement? 1134439642 M * Bertl no, it _will_ do so :) 1134439647 M * renegade334 OK 1134439651 M * renegade334 so that is what was happening 1134439667 M * Bertl yes, that's why I'm trying to tell you since the beginning :) 1134439692 J * mep_ mep@p5091BCD0.dip0.t-ipconnect.de 1134439758 M * renegade334 sorry didnt catch that 1134439766 M * Bertl np 1134439772 M * renegade334 was readin over what you said and realized it 1134439777 M * renegade334 i am a lil slow 1134439788 M * renegade334 long day at work 1134439970 M * renegade334 i remembered i resarted the box that was 'primary' 1134440133 Q * mep__ Ping timeout: 480 seconds 1134440201 Q * renegade334 Remote host closed the connection 1134442473 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134443134 Q * Johnnie Read error: Connection reset by peer 1134443815 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134445662 J * balbir ~balbir@59.92.129.150 1134447069 M * Bertl welcome balbir! 1134447510 Q * Johnnie Ping timeout: 480 seconds 1134447939 M * balbir Thank you 1134447950 M * monrad Bertl: still in the US? 1134447959 M * Bertl monrad: yup! 1134447979 M * monrad if anyone is hungry i have pancakes ready in 5 min :) 1134447991 M * monrad i do really strange things when i am really tried 1134448055 M * monrad Bertl: where are you overthere 1134448074 M * Bertl Princeton, NJ 1134448127 M * monrad cool i never been to the us 1134448152 M * Bertl been here several times ... it's always kind of fun ... 1134448160 M * monrad but i have to say that i think japan is number one on my list 1134449589 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134450239 Q * Johnnie Read error: Connection reset by peer 1134450494 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134450735 Q * Johnnie Remote host closed the connection 1134451018 J * Johnnie ~jdlewis@acs-24-154-53-16.zoominternet.net 1134451826 J * Smutje`_ ~Smutje@xdsl-87-78-59-250.netcologne.de 1134451935 Q * Smutje` Ping timeout: 480 seconds 1134452630 Q * Johnnie Read error: Connection reset by peer 1134454508 J * Aiken__ ~james@tooax8-075.dialup.optusnet.com.au 1134454531 M * Bertl okay, off to bed now .. back tomorrow! 1134454536 N * Bertl Bertl_zZ 1134454785 N * Aiken__ Aiken 1134454835 Q * Aiken_ Ping timeout: 480 seconds 1134454938 Q * entroposcope Ping timeout: 480 seconds 1134455464 J * Aiken_ ~james@tooax6-097.dialup.optusnet.com.au 1134455489 J * entroposcope ~entroposc@user-0c992og.cable.mindspring.com 1134455790 Q * Aiken Ping timeout: 480 seconds 1134457459 Q * balbir Ping timeout: 480 seconds 1134460124 J * micah ~micah@ool-4355be81.dyn.optonline.net 1134460645 P * Aiken_ Leaving 1134461224 N * Smutje`_ Smutje 1134463401 J * BWare ~bware@office.intouch.net 1134463873 J * balbir ~balbir@59.92.155.63 1134463984 Q * lilo Quit: bbiab 1134464566 J * lilo ~lilo@lilo.usercloak.oftc.net 1134466552 J * lilo_ ~lilo@lilo.usercloak.oftc.net 1134466636 Q * lilo Ping timeout: 480 seconds 1134467440 Q * FireEgl Ping timeout: 480 seconds 1134467831 J * menomc ~amery@200.75.27.72 1134467939 Q * mnemoc Ping timeout: 480 seconds 1134467939 N * menomc mnemoc 1134468224 Q * BWare Ping timeout: 480 seconds 1134468632 J * FireEgl Atlantica@Atlantica.IRCNut.Com 1134468902 J * BWare ~look@office.intouch.net 1134470641 J * Duckx ~duckx@195.75.27.158 1134473274 J * glut glut@no.suid.pl 1134473314 M * glut g'day 1134474970 Q * shedi Quit: Leaving 1134475922 J * Ben81 ~Ben81@tipi0e.lri.fr 1134475933 M * Ben81 hello 1134475973 M * Ben81 have anyone already use the "vserver-copy" with the last vserver version ? 1134476046 M * Ben81 no ? :'-( 1134476271 M * Ben81 ok it look like to be deprecated, the vserver ... -copy is not yet implemented ? 1134480683 J * click click@ti511110a080-0438.bb.online.no 1134481006 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl 1134482750 Q * balbir Ping timeout: 480 seconds 1134483551 J * balbir ~balbir@59.92.152.179 1134483712 J * nurull4h ~nurull4h@85.103.221.76 1134483843 P * nurull4h 1134484193 M * harry Ben81: deprecated? 1134484375 M * Ben81 it doesn't work with the last version 1134484411 M * harry never used vserver-copy, don't even know if it exists 1134484454 M * mnemoc afaik those are not executed directly, those work via `vserver` wrapper 1134484474 J * mef ~mef@targe.CS.Princeton.EDU 1134485044 N * Bertl_zZ Bertl 1134485050 M * Bertl morning folks! 1134485064 M * Bertl mef: got my email? 1134485076 M * mnemoc moin master Bertl 1134485097 M * Bertl good morning mnemoc! 1134486086 M * Bertl okay, moving out, back shortly ... 1134486092 N * Bertl Bertl_oO 1134487087 J * jayeola ~jayeola@host-87-74-46-211.bulldogdsl.com 1134487100 M * jayeola hey chaps 1134487114 N * nokoya nokoyaz 1134487147 N * nokoyaz nokoya 1134487150 M * jayeola any recommended partition methods for vservers? size per vserver appprox 8gb? 1134487232 J * shedi ~siggi@inferno.lhi.is 1134487258 M * Ben81 What is the fastest method to copy a Vserver ? 1134487290 M * mnemoc jayeola: lvm 1134487309 M * mnemoc Ben81: cp -al with CoW facilities on kernel 1134487373 M * Ben81 CoW ? but with this method, i have to copy the conf in /etc, change symbolics links, ... 1134487431 M * jayeola thanks ben! i have 2x40gb hard disks... i was thinking of something like /boot swap then everything wlse in the lvm 1134487451 M * jayeola s,wlse,else 1134487500 M * mnemoc i have swap on lvm too 1134487526 M * jayeola so boot is noton lvm? 1134487663 Q * balbir Quit: Leaving 1134487829 M * ||Cobra|| hi 1134487837 M * ||Cobra|| i have a problem for building 1134487843 M * ||Cobra|| a guest system 1134487857 M * ||Cobra|| i do the simple command : 1134487873 M * ||Cobra|| vserver test build -m yum --hostname=test.example --interface test0=eth0:192.168.0.1/24 -- -d fc4 1134487893 M * ||Cobra|| and i've got this message : pkgcfgbase-dir '/vservers/.pkg' does not exist or is invalid 1134487901 M * ||Cobra|| does anyone have any ideas ? 1134488115 M * daniel_hozac mkdir -p /vservers/.pkg 1134488193 M * mnemoc jayeola: you would need an initrd to have /boot at lvm 1134488209 M * ||Cobra|| thank you daniel 1134488216 M * ||Cobra|| i try it 1134488223 M * daniel_hozac mnemoc: does GRUB read LVM now? 1134488226 M * mnemoc jayeola: and dm-mod kernel module is also a problem on some places, so i have / as a 100M partition, and all the rest on lvm 1134488238 M * mnemoc daniel_hozac: no idea, i use lilo 1134488257 M * daniel_hozac wow, i didn't know lilo still existed :) 1134488353 M * ||Cobra|| daniel_hozac : it works , thx :) 1134488357 M * jayeola thanks guys. i've just installed via "auto partitioin". lvm chosen by default. i'll roll with that for now 1134488361 M * mnemoc daniel_hozac: i don't use machines to play, so i'm more than happy with static loaders :) 1134489266 J * JimmyGulp ~james@ns0.esagroup.co.uk 1134489391 Q * click Read error: Connection reset by peer 1134490107 M * ||Cobra|| does anyone know the short name of the red hat 7.3 1134490108 M * ||Cobra|| ? 1134490115 M * daniel_hozac rh73? 1134490130 M * jayeola cat /etc/redhat-version ? 1134490131 M * ||Cobra|| i was not sure 1134490148 M * ||Cobra|| thx guys 1134490163 M * daniel_hozac why would you still use RHL 7.3 though? 1134490186 J * meandtheshell ~markus@85-124-11-178.dynamic.xdsl-line.inode.at 1134490197 M * daniel_hozac has it received any updates for the past... 3 years? 1134490225 M * ||Cobra|| i have to use compiler 1134490240 M * ||Cobra|| and i can recompiler them only on this platform 1134490249 M * ||Cobra|| recompile* 1134490256 M * daniel_hozac what, exactly? 1134490259 M * ||Cobra|| with an old version of gcc 1134490265 M * daniel_hozac can't you patch it to work with more recent gcc's? 1134490270 M * ||Cobra|| no i cant 1134490290 M * ||Cobra|| i have to use gcc 2.95.x or 2.96 1134490299 M * ||Cobra|| for recompiling my stuff 1134490310 M * daniel_hozac why can't you? 1134490332 M * daniel_hozac i mean, we're at gcc 4.1 now... 2.9* is ancient. 1134490352 M * ||Cobra|| yes i know 1134490371 M * ||Cobra|| i reuse a software which can be compile only by this gcc 1134490378 M * daniel_hozac and why can't you fix it? 1134490389 M * ||Cobra|| and i dont know yet how to fix it 1134490600 M * daniel_hozac well, first step would be to read the errors new gcc's give you ;) 1134490617 M * ||Cobra|| sure 1134490629 M * ||Cobra|| most the errors came from dev libraries 1134490636 M * ||Cobra|| i mean old libraries 1134490671 M * daniel_hozac so you're using ancient APIs as well. 1134490680 M * daniel_hozac doesn't sound unfixable at all. 1134490711 M * ||Cobra|| yes i think so 1134490725 M * ||Cobra|| but the software is huge 1134490740 M * ||Cobra|| and so i dont know where i have to modify 1134490775 M * daniel_hozac well, compiler errors generally give you a file and line number ;) 1134490826 M * harry /dev/mapper/vservervg-harrylv on /vservers/harry type reiserfs (rw,attrs) 1134490847 M * harry vserver1:/vservers# umount harry/ 1134490847 M * harry vserver1:/vservers# rmdir harry/ 1134490847 M * harry rmdir: `harry/': Device or resource busy 1134490851 M * ||Cobra|| yes and most of these errors are in the dev libraries and used functions ^^ 1134490853 J * click click@ti511110a080-3507.bb.online.no 1134490854 M * harry ??? 1134490856 M * daniel_hozac harry: namespaces. 1134490874 M * harry daniel_hozac: what to do about them? 1134490884 M * harry how can i ... ahm... fix it? 1134490892 A * harry never worked with namespaces etc... 1134490900 M * harry gimme a manpage or so!!!!!!!!! ;)))) 1134490913 A * harry wants a rtfm, but i want to have the fm included ;) 1134490918 M * daniel_hozac harry: IIRC soemone posted to the mailing list not too long ago with the same/equivalent problem. 1134490936 M * harry define "not so long ago"? 1134490946 M * harry then i can see if i have to look in my own archives or not :) 1134490980 N * Bertl_oO Bertl 1134490984 M * daniel_hozac my sense of time is really bad :) 1134490996 M * harry :) 1134491030 M * Bertl harry: here you go: http://linux-vserver.org/Namespaces 1134491039 M * harry Bertl: i googled there allready :) 1134491041 M * harry tnx anyway :) 1134491050 M * harry An error occurred while loading http://linux-vserver.org/Namespaces: 1134491057 M * harry let's try again :) 1134491066 P * JimmyGulp Leaving 1134491104 M * Bertl http://archives.linux-vserver.org/200512/0143.html 1134491112 M * harry what happened to the server? 1134491114 M * Bertl (this is the thread daniel was talking about) 1134491131 M * harry can't seem to connect to anything 1134491151 M * Bertl works here ... 1134491163 M * harry what's the ip address? 1134491173 M * harry might have to do something with some dns servers in .be 1134491210 M * harry : 15:57 lois ~/mp3 ;host linux-vserver.org 1134491210 M * harry ;; connection timed out; no servers could be reached 1134491221 M * harry no, i don't have mp3's... :p 1134491229 M * harry just a coincidence :) 1134491295 M * Bertl 207.253.4.250 1134491318 M * harry let's try that :) 1134491331 M * harry hah... i can reach it from home... that will do 1134491416 M * Hollow *bounce* 1134491502 M * Ben81 i'm looking for limit the bandwith of Vserver , can i do that with vserver ? 1134491516 P * meandtheshell 1134491531 M * Ben81 or do i have to use tc ? 1134491569 M * harry hmm... 1134491575 M * harry i can't find it! 1134491581 J * meandtheshell ~markus@85-124-11-178.dynamic.xdsl-line.inode.at 1134491585 M * harry how do i get a list of all namespaces? 1134491597 M * Bertl welcome meandtheshell! 1134491603 M * Bertl harry: vps 1134491625 M * Bertl harry: each task potentially has one 1134491636 P * click [IRSSI] 1134491647 M * harry Bertl: ??????? 1134491656 M * harry PID CONTEXT TTY TIME CMD 1134491657 M * harry 32162 0 MAIN pts/0 00:00:00 bash 1134491657 M * harry 1384 1 ALL_PROC pts/0 00:00:00 vps 1134491657 M * harry 1385 1 ALL_PROC pts/0 00:00:00 ps 1134491665 M * harry i have 2 dirs i want to remove 1134491694 M * harry i stopped the vservers.. 1134491698 M * harry phone... brb 1134491784 M * harry so, i stopped vserver, unmounted the logical volume 1134491790 M * harry then try to remove the dir 1134491795 M * harry but the last one doesn't work 1134491800 M * harry so i remove the vserver 1134491805 M * harry but still... not removable 1134491821 M * harry (and.... phone again... but i'll be back in a jiffie) 1134491822 M * Bertl ~well, did you read the namespace page? 1134491828 M * harry yes 1134491837 M * harry but... 1134491858 M * harry vnamespace: vc_xidopt2xid("tobiaas"): No such file or directory 1134491862 M * harry brb... 1134491895 M * TheSeer wtf..? 1134491923 M * Bertl TheSeer: hmm? 1134491939 M * TheSeer just wondering why someone would "tobiaas" as a xid ;) 1134492019 M * Bertl well, folks do all kind of funny things :) 1134492058 M * harry it's not xid :) 1134492071 A * harry just named the vserver tobiaas 1134492093 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1134492095 M * Roey hi 1134492101 M * Roey wow, this is bigger than that freenode place 1134492135 M * Roey Does VServer work with 2.6.14.3? 1134492141 M * Roey or just 2.6.12.6? 1134492142 M * Bertl yup 1134492146 M * Bertl welcome Roey! 1134492146 M * Roey Bertl: to what? 1134492148 M * Roey hey! 1134492158 M * Bertl you are too fast, the first one :) 1134492160 M * Roey Bertl: listen I'm so happy that this place exists... that other chan was so quiet 1134492172 M * Roey oh, so vserver will patch cleanly to 2.6.14.3 1134492174 M * Bertl that's why this is the main channel :) 1134492174 M * Hollow Roey: which chan? 1134492183 M * Roey Hollow: #vserver on irc.freenode.net 1134492190 A * Hollow goes looking 1134492196 M * harry i want a list of all used xid's 1134492207 M * harry vps gives me ... well... pid stuff pasted above... 1134492212 M * Bertl harry: try vserver-stat 1134492222 M * Roey Bertl: So I'm assuming that 2.6.14.3 is OK with vserver. 1134492226 M * Roey Bertl: I'll patch it then. 1134492234 M * Roey Bertl: also, does vserver integrate OK with Xen? 1134492248 M * Bertl it should, but there is no xen support per se 1134492256 M * FaUl .oO) an vserver-enabled kernel on xen? ) 1134492257 M * harry Bertl: since it's stopped, removed, it doesn't show anything 1134492262 M * Bertl Roey: http://vserver.13thfloor.at/Experimental/patch-2.6.14.3-vs2.0.1-rc5.diff 1134492270 M * Bertl harry: well, then there is none 1134492271 M * Roey oh 1134492276 M * Roey Bertl: wonderful, thanks ;) 1134492280 M * harry Bertl: than why can't i remove the dir? 1134492291 M * Bertl Roey: do you plan to combine it with xen? 1134492302 M * Roey Bertl: yes 1134492305 M * Bertl harry: let's check the permissions (you are with debian?) 1134492308 J * Aike s528254@stuwww.uvt.nl 1134492321 M * Bertl Roey: great! if you need help, just let me know 1134492322 M * Roey Bertl: I want to have vserver available for guests 1134492326 M * Roey Bertl: you've done it? 1134492334 M * Bertl Roey: what? 1134492339 M * Bertl welcome Aike! 1134492339 M * Roey Bertl: combined xen+vserver? 1134492345 M * Bertl Roey: not yet ... 1134492348 M * Roey oh, ok. 1134492352 M * Aike Bertl: hey :) 1134492359 M * Roey Bertl: I dunno if it'll ask me to change stuff manually... I hope not. 1134492366 M * Bertl Roey: I assume you want xen arch support, yes? 1134492374 M * Roey what do you mean? 1134492382 M * Roey Bertl: arch as a sub-arch 1134492384 M * harry isnt there a list, showing me all the namespaces that are in use? 1134492385 M * Roey a Linux arch? 1134492387 M * harry Bertl: yes 1134492393 M * Roey Bertl: if that's what it means ;) 1134492405 M * Roey Bertl: but yes, both for arch and non-arch 1134492412 M * Roey Bertl: (that is, both as a server and as a guest) 1134492436 M * Bertl Roey: you want to put the vserver kernel in domU 1134492445 M * Bertl to put the vserver kernel in domU, right? 1134492462 M * Bertl *sorry* my connection is flakey 1134492498 M * Bertl Aike: need anything? if not, feel free to hang around ... 1134492513 M * Ben81 How can i limit the network bandwith of a Vserver ? 1134492523 M * Bertl Ben81: tc? 1134492552 M * Ben81 yes but it is hard to use it ;) 1134492570 A * harry gotta go now... cya'll later 1134492572 M * Ben81 but if there is no simple way ... i'll use it again :-( 1134492583 M * Aike Bertl: I'm not really looking for anything, I heard Roey talking about vserver in ##xen and it sounds kinda interesing 1134492586 M * Bertl well, what 'simple' way do you envision? 1134492592 M * Roey Bertl: well both--domU and dom0. 1134492595 M * Aike so why not hang around :) 1134492598 M * Roey Bertl: is it that big a deal, you think? 1134492611 M * Roey Aike: ##xen on Freenode? 1134492615 M * Roey Aike: wow, cool :) 1134492627 M * Bertl Roey: well, I'm pretty sure that DomU will work quite fine 1134492631 M * Ben81 maybe a config file of a vserver where i can specified is upload and downbload bandwith :-) 1134492632 M * Roey Bertl: ok 1134492651 M * Bertl Roey: linux-vserver supports many arch, including UML 1134492664 M * Roey ok 1134492667 M * Bertl Roey: so, the xen arch, is 'just' another one 1134492670 M * Roey Bertl: ahhh... 1134492670 M * Aike Roey: are you in both #xen and ##xen? 1134492681 M * Roey Aike: #xen, ##xen, #vserver. 1134492694 M * Roey Aike: #apt-proxy, #debian,#debian-kde,#kde,#kde-devel. 1134492695 A * Aike too 1134492700 M * Bertl Roey: but the arch support has to be implemented (though it's quite simple to do that) 1134492706 M * Aike and #linux-cluster 1134492708 M * Roey and ##unavailable, which was where I got when I tried to /j #vserver 1134492709 M * Roey Aike: :) 1134492722 M * Roey Bertl: you mean manually implemented? 1134492732 M * Bertl yeah, as in 'coded' :) 1134492739 M * Roey Bertl: doh :( 1134492747 M * Roey <--- not up for coding kernel magic 1134492776 M * Ben81 Bertl, i think i'll add QoS to my kernel :-) do you know a good url about tc ? 1134492847 M * Bertl Roey: don't worry, I have some experience there .. 1134492906 M * Roey :) 1134492916 M * Roey (tc? no never heard of it ;) 1134492936 M * Roey Bertl: hmm, this patch doesn't seem to take well to linux-2.6.14.3 1134492944 M * Bertl http://lartc.org/ 1134492964 M * Bertl Roey: which one? 1134492971 M * Roey the 1134492973 M * Roey one sec 1134492981 M * Roey the wrong patch 1134492983 M * Roey I'm sorry 1134492984 M * Roey one sec 1134492987 M * Bertl :) 1134493067 M * Ben81 thx Bertl 1134493087 M * Bertl Ben81: you're welcome! 1134493100 J * stefani ~stefani@superquan.apl.washington.edu 1134493112 M * Bertl morning stefani! 1134493137 M * stefani salut. 1134493176 M * Roey heya stefani 1134493182 M * Roey how do you pronounce your name? 1134493186 M * Roey with the accent on the e or the a? 1134493346 M * daniel_hozac wasn't there a vserver-for-xen patch posted to the mailing list a few months ago? 1134493369 M * Roey hey Bertl, the patch went well 1134493390 M * Roey daniel_hozac: I don't know :( 1134493394 M * Roey daniel_hozac: actually yes. 1134493406 M * Roey daniel_hozac: yes there was, but it is not... stored anywhere... other than on Google. 1134493412 M * daniel_hozac i remember it, because it had some suspicious code in it... 1134493413 M * Bertl * Linux kernel 2.6.11 1134493413 M * Bertl * Xen 2.0-testing from yesterday 1134493422 M * daniel_hozac well, mailing list archives should have it. 1134493424 M * Bertl so yes there was something :) 1134493447 M * Roey Bertl: but it evidently was not important enough; else it'd be linked to from either vserver or xen's web sites. 1134493449 M * Bertl but I guess Roey is aiming at 3.0 1134493452 M * Roey yes 1134493458 A * Roey wants it all 1134493459 M * Roey come to papa 1134493465 M * daniel_hozac ah, yes, of course. 1134493472 A * Roey nods emphatically 1134493476 M * Bertl and of course with the latest vserver release too 1134493480 M * daniel_hozac i guess the patch shouldn't have changed much though? 1134493493 M * Bertl daniel_hozac: no, probably not 1134493494 M * stefani Roey: st?fani 1134493495 M * daniel_hozac basically just the syscall in the tables, IIRC. 1134493500 M * Roey stefani: hehe ok 1134493507 M * daniel_hozac -s 1134493525 M * Roey how is it that vserver changes network code and xen changes network code but this code does not intersect 1134493531 M * Roey *the code changes do not intersect 1134493547 M * Roey does VServer ever touch the hardware?? 1134493555 M * Roey or is it just process-level routing 1134493563 M * Roey that it modifies? 1134493568 M * Bertl neither nor ... networking happens on the host 1134493572 M * Roey ok 1134493576 M * Bertl the guests are 'just' isolated 1134493580 M * Roey ahh ok 1134493595 Q * Ben81 Quit: Leaving 1134493596 M * Roey so a guest cannot have separate routes 1134493598 M * Roey that's what you mean? 1134493611 M * Bertl not with the legacy networking (i.e. not now) 1134493622 M * Roey ah 1134493641 M * Bertl but you can easily set up multiple routing tables to have a 'private' route for each guest 1134493691 M * Roey aye ok 1134493693 M * Roey hmm 1134493698 M * Roey I just hope I'll be able to set this thing up soon then 1134493702 M * Roey for now I'll just use vserver. 1134493734 M * harry could my problem be a vnamespace bug? 1134493748 M * Bertl harry: I doubt that, let's do the following first> 1134493771 M * Bertl harry: ls -lda /path/to/dir/{.,..} 1134493793 M * Bertl and then lsattr -lda /path/to/dir/{.,..} 1134493804 M * harry drwxr-xr-x 2 root root 48 2005-11-23 18:14 tobiaas/ 1134493816 M * harry tobiaas/ --- 1134493837 M * Bertl that's not really what I asked for, is it? 1134493842 M * harry yes it is... 1134493853 M * harry vserver1:/vservers# ls -lad tobiaas/ 1134493854 M * harry drwxr-xr-x 2 root root 48 2005-11-23 18:14 tobiaas/ 1134493854 M * harry vserver1:/vservers# lsattr -lad tobiaas/ 1134493854 M * harry tobiaas/ --- 1134493868 M * harry okay... lad instead of lda :) 1134493869 M * Bertl well, check the lines again (the ones I pasted :) 1134493883 M * harry aaaaaah 1134493884 M * harry sry 1134493910 M * harry vserver1:/vservers# ls -lad tobiaas/{.,..} 1134493910 M * harry drwxr-xr-x 2 root root 48 2005-11-23 18:14 tobiaas/. 1134493910 M * harry drwxr-xr-x 10 root root 264 2005-11-25 10:48 tobiaas/.. 1134493910 M * harry vserver1:/vservers# lsattr -lad tobiaas/{.,..} 1134493910 M * harry tobiaas/. --- 1134493912 M * harry tobiaas/.. --- 1134493945 M * Bertl and you want to remove the dir /tobiaas, yes? 1134493948 M * harry yups 1134493971 M * Bertl what does rmdir tobiaas tell you? 1134493981 M * harry rmdir: `tobiaas/': Device or resource busy 1134493998 M * Bertl that's on the host, right? 1134494001 M * harry yup 1134494028 M * Bertl let's see what 'cat /proc/self/mounts' gives (please upload to pastebin.com) 1134494070 M * harry http://pastebin.com/462445 1134494077 M * Bertl tx 1134494126 M * Bertl harry: hmm, what kernel version do you use? 1134494138 M * harry Linux vserver1 2.6.14.2-vs2.1.0-rc7-harry #2 SMP Thu Nov 17 16:15:24 CET 2005 i686 GNU/Linux 1134494157 A * ||Cobra|| bye bye 1134494163 Q * ||Cobra|| Quit: Leaving 1134494163 M * harry ||Cobra||: cya 1134494167 M * harry too late :0 1134494213 M * Bertl hmm, I'd try to update to rc10 and see if it still persists ... 1134494214 J * click click@ti511110a080-3507.bb.online.no 1134494233 M * Bertl okay, off for a meeting ... back later ... 1134494238 N * Bertl Bertl_oO 1134494248 M * daniel_hozac well, it wouldn't, as you'd have to reboot for the new kernel ;) 1134494251 M * harry cya 1134494257 M * harry Bertl_oO: that means... reboot... 1134494260 M * harry which will have to wait 1134494265 M * harry until i get the new server 1134494273 M * harry (quad dualcore opteron ;)) 1134494290 M * daniel_hozac ... 8-way? 1134494294 A * daniel_hozac drools 1134494308 M * harry daniel_hozac: 32 gig ram, and 4 146GB 15k rpm disks 1134494313 M * harry 4 gigabit network interfaces 1134494323 M * daniel_hozac jesus. 1134494324 M * harry supermachine!!!!!! 1134494326 M * harry allmine :) 1134494327 M * daniel_hozac indeed. 1134494364 M * harry just to run a few (i guess 10... 15 max) vps :) 1134494372 M * harry which do... nothing at all ;) 1134494380 M * daniel_hozac lol 1134494394 M * harry like printserver, ftp server, license server, testservers, mail rewrite servers 1134494395 M * harry etc... 1134494427 Q * prae Quit: Pwet 1134494454 M * daniel_hozac damn. 1134494511 Q * jayeola Quit: leaving 1134494896 N * Bertl_oO Bertl 1134494900 M * Bertl back now ... 1134496406 J * Smutje_ ~Smutje@xdsl-87-78-56-235.netcologne.de 1134496440 M * micah I'm having a network problem inside a guest, I am sure it something I screwed up 1134496458 M * Bertl let's hear 1134496466 M * ag- harry: what's the average cost? ;D 1134496492 M * micah the guest has two interfaces, one a rfc1819 address, and the other a static IP. When I start up the vserver, i can ping the static via the internet, but inside the guest I can't seem to get traffic back 1134496541 M * Bertl micah: proper SNAT entry in POSTROUTING? 1134496555 Q * Smutje Ping timeout: 480 seconds 1134496576 M * micah Bertl: I have a real static IP assigned to the vserver as well, so I shouldn't need to do SNAT 1134496578 M * harry ag-: 23k euro 1134496600 M * Bertl micah: depends on which IP is chosen ... have a look with tcpdump 1134496610 M * Bertl (on the host) 1134496619 M * ag- harry: it seems right, this was what i computed ;) 1134496677 M * micah Bertl: this is an exact copy of a guest from a previous machine, the only difference is I am using 2.6.14-vs2.0.1-rc5 on this machine and patch-2.6.12.4-vs2.0 on the previous 1134496735 M * micah I must be missing something... 1134496811 M * micah Bertl: it looks like the 192.168 address is the address that it is trying to use 1134496820 M * micah based on tcpdump 1134496822 M * Bertl see 1134496866 M * Roey micah: hi. UMD Micah? 1134496872 M * micah What is the criteria for choosing this address? 1134496876 M * micah Roey: nope, different Micah 1134496879 M * Roey oh ok! 1134496944 M * Bertl micah: what does 'ip addr ls' say on the host and guest? 1134496964 M * Bertl micah: it's basically userspace choosing the source ip 1134496977 M * Bertl micah: and it usually choses the first one 1134496995 M * micah I am getting a weird error from the utils when I try to stop this guest: 1134496998 M * micah /usr/sbin/vserver: line 85: 22395 Killed "${NICE_CMD[@]}" ${USE_VNAMESPACE:+$_VNAMESPACE --enter "$S_CONTEXT" -- } $_VCONTEXT $SILENT_OPT --migrate --chroot --xid "$S_CONTEXT" -- "${INITCMD_STOP[@]}" 1134497034 M * Bertl 0.30.209 mainline? 1134497133 M * micah its the 0.30.209 debian package 1134497158 M * micah ip addr ls in the guest shows the two IPs on the two interfaces 1134497175 M * Bertl which one is the first? 1134497192 M * micah the real static IP 1134497202 M * micah 2: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 1134497202 M * micah link/ether 00:30:48:83:5d:4c brd ff:ff:ff:ff:ff:ff 1134497202 M * micah inet 69.90.134.186/32 scope global eth0 1134497202 M * micah 3: eth1: mtu 1500 qdisc pfifo_fast qlen 1000 1134497202 M * micah link/ether 00:30:48:83:5d:4d brd ff:ff:ff:ff:ff:ff 1134497203 M * micah inet 192.168.0.2/32 scope global eth1 1134497330 M * Bertl and routing is configured via what interface? 1134497340 M * micah I have discovered something interesting 1134497354 M * micah in /etc/vservers/guest/interfaces I have a 0 and 1 directory 1134497370 M * micah the 0 directory had the 192.168 configured, and the 1 had the 69.90 configured 1134497383 M * micah if I switch those, so the 0 has 69.90 and the 1 has the 192.168, it works 1134497421 M * micah (does that make sense?) 1134497471 M * Bertl ah, yes, that _makes_ sense 1134497496 M * Bertl the first guest IP is special in that way, that it is used if you do not specify a specific IP 1134497521 M * micah if you do not specify a specific IP how? 1134497656 M * Bertl for example, if a process creates a socket, and does not bind/connect it, and just starts sending out packets 1134497673 M * Bertl then the kernel has to choose an IP for that socket (source) 1134497757 M * micah that makes sense 1134497798 M * micah so it chooses the lowest numbered /etc/vserver/guest/interface/# to do that from? 1134497841 M * Bertl well, after some indirection, yes 1134497855 M * Bertl basically it's the _first_ ip listed in the ninfo 1134497877 M * Roey hey Bertl 1134497881 M * micah this appears to be new behavior in this kernel patch, because the other server has the 192 address as the first interface configured and it doesn't have this problem (also using the debian user space tools) 1134497889 M * Roey yeay! I recompiled 2.6.14.3 to support VServer! Hmm, what next! 1134497893 A * Roey reads... 1134497895 M * Bertl micah: and as the userspace tools set them in the alphabetical order of the dir, that will be the dir with the lowest interface 1134497920 M * micah I quite literally tarred up the config dir and moved it over to the new machine 1134497952 M * Bertl micah: also depends, if there is already a routing entry there and/or SNAT takes care of that (which IMHO would we the proper solution) then it will work 1134498029 T * Bertl http://linux-vserver.org/ | latest stable 2.01, 1.2.10, 1.2.11-rc1, devel 2.1.0-rc10 | util-vserver-0.30.209 | 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 ;) 1134498045 M * Bertl finally 2.01 is released :) enjoy! 1134498086 M * micah oooh! congrats Bertl! 1134498089 M * Kara_ :) 1134498099 M * micah is there an announcement somewhere with changes, etc. that we can start sending out? 1134498141 M * Bertl will be there shortly 1134498709 M * Bertl http://www.13thfloor.at/vserver/s_rel26/v2.01/ 1134498724 M * Bertl (check the overview for the changes) 1134499106 M * Bertl micah: the ML entry is at http://archives.linux-vserver.org/200512/0152.html 1134500099 M * Hollow hellas 2.01 1134500106 M * Hollow any changes? 1134500127 M * Hollow (i mean since rc5) 1134500222 M * Bertl a few cleanups 1134500230 M * Bertl and one fix I lost (for the reiserfs) 1134500797 M * Bertl (all should be available in FOR-2.0.1) 1134500960 P * meandtheshell 1134501098 M * Hollow Bertl: ok, i'm adding it to gentoo cvs now 1134501121 M * Bertl excellent, btw, is there a tools version which should work with the stable release? 1134501139 M * Hollow atm 208 is stable in gentoo 1134501146 M * Bertl (i.e. which you would like to see on the release page) 1134501179 M * Bertl I was thinking about the Hollow-tools whatever they are called now :) 1134501197 M * Hollow hm, they won't work with 2.0.1 i guess 1134501211 M * Bertl that's what I thought too, okay,w e keep that for 2.1.x then 1134501215 M * Hollow yup 1134502001 M * Roey hi 1134502080 M * Roey quick question. I was following this guide: http://www.howtoforge.com/linux_vserver_debian . When it came time to do apt-setup, I selected the 'http' method for completing the setup. It can't connect to the Debian mirror. What should I do? 1134502213 A * Bertl is checking the url ... 1134502271 M * Roey ok 1134502293 M * Bertl hmm, why can't you connect? because a) you are not on the network b) it does not find the name (resolver) or c) it does connect but fail to find the stuff? 1134502295 M * Roey btw I did give it a different network number (192.168.1.10) than the local network (192.168.0.0/24) 1134502305 M * Roey Bertl: ah! the resolver, right. 1134502309 M * Roey Bertl: lemme check 1134502332 M * Bertl also check that the 192.168.1.10 IP is allowed to go online 1134502355 M * Roey ok, ifconfig shows no IP. 1134502357 M * Roey on eth0 at least. 1134502360 M * Bertl simple test would be: ping -c 1 -I 192.168.1.10 www.google.com (on the host) 1134502386 M * Bertl if that doesn't work, you are missing an SNAT entry :) 1134502393 M * Roey hmm 1134502400 M * Roey Bertl: well, it's like how I siad 1134502416 M * Roey ifconfig shows an eth0, but there's no IP assigned to it. 1134502426 M * Roey how does VServer manage network connections? 1134502431 M * Bertl that's fine, use 'ip addr ls' 1134502444 M * Roey are all vservers in their own little virtual lan? 1134502444 M * Bertl ifconfig is just too old to see everything :) 1134502462 M * Bertl nope, they share the host networking 1134502462 M * Roey root@mail:/# ip addr ls 1134502463 M * Roey bash: ip: command not found 1134502487 M * Roey Bertl: ok, so I can set up different vservers with different IPs, all in the same LAN as the master, right? 1134502488 M * Bertl except for isolation (restriction of IP addresses) and some localhost remapping, it's 100% pure linux networking 1134502495 M * Roey ok 1134502508 M * Bertl ip is part of the iproute(2) package 1134502509 M * Roey hmm, can I configure DHCP maybe? 1134502520 M * Roey Bertl: (can't download the package, as I don't have IP working yet ;) 1134502522 M * Bertl the IP is already there, trust me 1134502525 M * Roey ok 1134502526 M * Roey ok ok 1134502542 M * Bertl let's do the following: 1134502557 J * lilo ~lilo@70.230.73.20 1134502563 M * Bertl iptables -t nat -I POSTROUTING -s 192.168.1.10 -j SNAT --to 1134502565 Q * lilo_ Remote host closed the connection 1134502580 M * Bertl where is the IP on the host which can reach the outside 1134502601 M * Bertl again you can test with: ping -c 1 -I www.google.com 1134502613 M * Roey root@mail:/# iptables -t nat -I POSTROUTING -s 192.168.1.10 -j SNAT --to 192.168.5.185 1134502613 M * Roey modprobe: Can't open dependencies file /lib/modules/2.6.14.3-vs2.0.1-rc5/modules.dep (No such file or directory) 1134502613 M * Roey iptables v1.2.11: can't initialize iptables table `nat': Permission denied (you must be root) 1134502613 M * Roey Perhaps iptables or your kernel needs to be upgraded. 1134502620 M * Roey hmm 1134502622 M * Roey 5.134 maybe. 1134502624 M * Bertl not inside the guest, on the host 1134502633 M * Roey nope, not that either 1134502636 M * Roey Bertl: oh :) ok 1134502661 M * Bertl for now, keep in mind: networking is on the host :) 1134502693 M * Roey hey! it worked! 1134502696 M * Roey Bertl: ok 1134502732 M * Roey Bertl: so if I want to ping this vserver from another physical box, I can set a route using the host as the gateway 1134502732 M * Roey yes? 1134502748 M * Roey (if the guest has a different subnet than the host) 1134502830 M * Bertl no 1134502841 M * Roey oh 1134502850 M * Bertl the 'host' is the one sending packets (even if they are created inside a guest) 1134502858 M * Roey ok 1134502866 M * Bertl there is no 'virtual' network or indirection involved 1134502871 M * Roey well there are two question I'm asking, then 1134502889 M * Bertl go ahead 1134502905 M * Roey 1) is it possible to have the guests share a vlan, and to be able to communicate with the Outside (that is, outside of the host) 1134502915 M * Bertl yes 1134502934 M * Roey 2) is it possible to give each guest an IP from the Outside ip space 1134502940 M * Bertl yes, sure 1134502949 M * Roey ok. 1134502972 M * Bertl the linux-vserver networking is solely based on IP addresses 1134502982 M * Roey ok 1134503000 M * Bertl so you 'assign' a few addresses (up to 16) to a guest 1134503003 M * Roey won't it mess my firewall up to see the same mac responding to different IPs? 1134503015 M * Roey er, the same MAC sending with different IPs rather. 1134503018 M * Bertl and the processes inside the guest will be restricted to those IPs 1134503021 M * Roey ok 1134503032 M * Bertl Roey: is that a problem on a linux host with several IPs? 1134503041 M * Roey Bertl: not that I know of ;) 1134503053 M * Bertl see, so it's not a problem here either :) 1134503056 M * Roey like, eth0:0, eth0:1, etc. 1134503063 M * Bertl precisely 1134503074 M * Bertl if you want names for the interfaces, just configure them 1134503087 M * Bertl (ala :1 :2 or :hansi :karli ...) 1134503114 M * Bertl you can do that on the host (and just let the guests use them) or with the util-vserver config at guest startup 1134503122 M * Roey oh 1134503129 M * Roey oh I see. 1134503152 M * Roey so it's not... it's not like the host can have 192.168.0.3 and the guest will have 192.168.0.8 1134503155 M * Bertl if you are creating the config with util-vserver, then the syntax is like this: 1134503170 M * Roey it's more like, the host would have 192.168.0.3 and the guest would be 192.168.0.3:1 1134503182 M * Bertl hmm, no :) 1134503193 M * Roey ah\ 1134503205 M * Bertl look, you are used to xen, (and maybe UML) right? 1134503209 M * Roey nono 1134503215 M * Roey I didn't get to that point ;) 1134503223 M * Bertl okay :) 1134503230 M * Roey but I have used.... stuff... before 1134503234 M * Bertl well, then let me explain it like this: 1134503254 M * Roey (ah, right, I've set up OpenVPN) 1134503258 M * Bertl there are interfaces, like eth0, lo, ppp0, tun0 1134503263 M * Roey ok 1134503277 M * Bertl now those interfaces can hold an almost arbitrary number of IPs 1134503282 M * Roey rignt 1134503285 M * Roey right. 1134503298 M * Bertl _without_ any :0 :1 or so 1134503303 M * Roey ok 1134503312 M * Bertl but, you can also use something 'older' 1134503320 M * Bertl which is called alias 1134503323 M * Roey ok 1134503334 M * Bertl and uses a separate name for each IP 1134503347 M * Roey hmm ok 1134503350 M * Bertl like this: eth0:1 ppp0:karli lo:g027 1134503371 M * Bertl in addition to those aliases, you can also use vlans 1134503376 M * Roey oh... so what I wrote up there (eth0:1) is just a subset of aliasing 1134503380 M * Bertl which use a notation like this: 1134503391 M * Bertl eth0.101, ppp0.42 1134503397 M * Roey hmm ok 1134503412 M * Bertl basically you can also combine those to 1134503413 M * Roey maybe that's why I was consued when I saw eth0.0 notation alongside eth0:1 and thought they're the same 1134503419 M * Bertl eth0.101:hansi 1134503448 M * Bertl the tools basically support all of them 1134503492 M * Bertl ip addr ls (on the host) will paint you a picture of the current setup 1134503575 M * Roey ok 1134503583 M * Roey so when I see the device 'eth0' 1134503587 M * Roey in the guest 1134503587 M * Bertl http://freshmeat.net/ (btw :) 1134503602 M * Roey is this the same 'eth0' as the host's? 1134503608 M * Bertl yep, precisely 1134503613 M * Roey ohhh ok 1134503619 M * Bertl only with a restricted set of IPs\ 1134503623 M * Roey ok 1134503629 M * Roey well can I use ifconfig to set it? 1134503638 M * Roey or that I already have an ip 1134503640 M * Roey ? 1134503641 M * Bertl it is already set 1134503643 M * Roey oh ok 1134503645 M * Roey got it. 1134503650 M * Roey controlled by ip/0/ip 1134503652 M * Roey that file 1134503654 M * Bertl just in a way that ifconfig doesn't see it 1134503658 M * Roey I see now 1134503665 M * Roey see, I want higher security than just chroot 1134503679 M * Bertl jep, that's what you get :) 1134503686 M * Roey BUT 1134503694 M * Bertl that also why you cannot set the ip inside :) 1134503727 M * Roey I still have to be vigilant in setting up postfix so that an attacker cannot write to (the guest's) / 1134503730 M * Roey Bertl: ok 1134503753 M * Roey Bertl: that was also something I was going to ask, if not being able to set the Ip inside has to do with security 1134503781 M * Bertl Linux-VServer increases security, but of course, you do not want a bunch of 'rooted' guests sitting around :) 1134503787 M * Roey aye 1134503795 M * Roey even though, they wouldn't be able to change their IPs. 1134503801 M * Bertl so, you still have to care about security for the guests 1134503811 M * Bertl no, they would not be able to harm the host 1134503811 M * Roey Bertl: see, I didn't know vservers involves replicating the entire filesystem tree 1134503823 M * Bertl actually it doesn't .... 1134503827 M * Roey well 1134503837 M * Roey you're giving each vserver its own space 1134503838 M * Roey right. 1134503846 M * Bertl it's own 'namespace' 1134503865 M * Bertl (which might be also a view into the host's file space) 1134503871 M * Roey but I meant that you essentially have to install a new file tree for each vserver. 1134503883 M * Bertl well, no, wrong again ... 1134503888 M * Roey ok 1134503891 M * Bertl first, you can share them between guests 1134503896 M * Roey oh, ok 1134503903 M * Bertl (in a secure way, which is called unification) 1134503911 M * Roey I'll read up on that :) 1134503930 M * Bertl then, you could easily use /usr or /sbin from the host, if mounted read only 1134503947 M * Roey what about lib 1134503951 M * Bertl (or with the devel releases, or the BME patches, just make ro --bind mounts) 1134503955 M * Roey or for that matter, all of /usr 1134503964 M * Roey see, the only thing that the guests need to change are /var and /etc 1134503965 M * Roey er 1134503967 M * Roey just /var rather 1134503969 M * Bertl should be no problem if done properly 1134503998 M * Bertl the template approach for example, usually has a 600MB distro with 20MB differences for each guest 1134503999 M * Roey Bertl: oh ok, so they're not like virtual machines at all -- guests are mor elike 'partitions' of the same system. 1134504010 M * Bertl exactly! 1134504014 M * Roey heh 1134504015 M * Roey ok 1134504025 M * Roey so there's no 'communication' needed among guests 1134504026 M * Bertl that's why they are complementary to xen (for example) 1134504028 M * Roey right 1134504030 M * Roey right right 1134504059 M * Bertl this 'soft-partitioning' allows for resource sharing and less overhead 1134504076 M * Roey so if I want to use the host's /lib and /usr but force a separate /var, what would I do? Just mount -mode ro,bind the dirs? 1134504091 M * Roey plus I want to be able to apt-get update only one machine 1134504110 M * Bertl well, that's trickier, but possible too 1134504118 M * Bertl first, a few details ... 1134504120 M * Roey ok 1134504146 M * Bertl - did you know? ro,bind does not work in vanilla kernels 1134504163 M * Roey dah! 1134504164 M * Roey :) 1134504164 M * Bertl - there is a patch called BME, which makes it work 1134504175 M * micah debian package with 2.01 stable uploaded 1134504183 M * Bertl micah: thanks a lot! 1134504211 M * Roey you mean, for the purpose of using vservers, you need to apply the BME patch before you can mount ro/bind 1134504212 M * Bertl Roey: second, if you do ro --bind mounts, apt-get will not be able to update them, right? 1134504231 M * Roey well I just want to update the host system 1134504235 M * Bertl Roey: no, for the general purpose of doing ro --bind mounts :) 1134504238 M * Roey oh 1134504246 M * Roey because I use it on my computer at home w/o problems 1134504250 M * Roey but it's a standard debian kernel there. 1134504258 M * Roey (sid) 1134504268 M * Bertl well .. let's try if it work then, shall we? 1134504274 M * Roey ok! :) 1134504275 M * Roey but 1134504277 M * Roey I have to go home 1134504279 M * Roey can we try tomorrow? 1134504281 A * Bertl doubts that debian includes the patches 1134504289 A * Roey joins Bertl in doubt 1134504296 M * Bertl Roey: sure, whenever you like to 1134504302 M * Roey ok 1134504306 M * Roey thanks for being so patient and positive 1134504314 M * Bertl you're welcome! 1134504316 M * Roey I'll be here tomorrow! 1134504318 N * Roey _Roey 1134504495 M * Bertl okay, moving out ... back in a few hours ... 1134504501 N * Bertl Bertl_oO 1134504694 J * infowolfe jthm@66-230-115-123-cdsl-rb1.nwc.acsalaska.net 1134504762 M * micah Bertl_oO, I sent a note to LWN about the announcement 1134506029 Q * flock Ping timeout: 480 seconds 1134506097 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il 1134507206 N * Bertl_oO Bertl 1134507210 M * Bertl back now! 1134507218 M * Bertl micah: thanks a lot! good idea! 1134508826 Q * infowolfe Ping timeout: 480 seconds 1134509001 Q * Pazzo Quit: ... 1134509239 J * jeeves ~jeeves@c-24-11-171-10.hsd1.mi.comcast.net 1134509287 M * Bertl welcome jeeves! 1134509379 M * derjohn hiho! wasn't there a feature which could tag a file with xid 'automatically' when root-of-host (not guest) copies s.th. into the namespace (?) of the guest? 1134509420 M * Bertl hmm, well, I wonder how you would do that ... 1134509436 M * Bertl (I'm talking about the namespace part) 1134509498 M * derjohn maybe 'namespace' ist the wrong word 1134509537 M * derjohn I just wonder if a daemon on host could store files in a vserver's dir .. and that file would be taggeg "auto" 1134509568 M * derjohn how? man I not a lonh haird kernel hacker ;) 1134509577 M * derjohn s/lonh/long/ 1134509595 M * Bertl lol 1134509600 M * Kara_ not? 1134509605 M * derjohn I remember this there being here or on ML 1134509608 J * Aiken ~james@tooax7-160.dialup.optusnet.com.au 1134509609 M * Bertl well, actually it's easier to implement that in userspace 1134509625 M * derjohn Kara_, nope. Pengiuns dont have long hairs ;) 1134509643 M * derjohn Bertl, so, do a chxid after copy? 1134509644 M * Bertl make a slightly modified 'cp' version which just supports specifying a context id 1134509650 M * Kara_ phew. _I_ have long hair :) so don't dare you to call me geek again ^^ 1134509669 M * derjohn Kara_, you won ;( 1134509678 M * Kara_ :) 1134509678 M * derjohn vcp? 1134509711 M * Bertl derjohn: yeah, something like that .. well I'm dreaming of a set of coreutils and similar which are all just context aware 1134509733 M * Bertl find / --type f --xid 27 ... 1134509737 M * derjohn man vcp: a copy like tool that finds out the xid of `pwd` and sets it automagically. 1134509756 M * Bertl well, xid of pwd? 1134509764 M * Bertl that's a little too vague ... 1134509787 M * derjohn Bertl, go up to the barrier and check what's the xid there? 1134509805 M * derjohn Bertl, if you hack it kara_ will test 1134509808 M * Bertl not necessarily correct either 1134509828 M * Bertl the problem is, a guest does not have to have any files which are context tagged at all 1134509834 M * daniel_hozac Bertl: well, we already have find -xid ;) 1134509847 M * Bertl daniel_hozac: yes, indeed, thanks to you! 1134509907 M * derjohn Bertl, yes, but that's true. mine dont have. if it does nozt find any, it sets none. there problem is if you have -for whatever reason- changin xid in the path?? 1134509914 M * derjohn daniel_hozac, vfind? 1134509921 M * derjohn daniel_hozac, or find? 1134509924 M * daniel_hozac find. 1134509957 M * daniel_hozac http://daniel.hozac.com/stuff/findutils-4.1.20-xid-libvserver.patch 1134509961 M * Bertl derjohn: that's why folks are advised to use static contexts 1134510004 M * derjohn Bertl, I do. I want stuff that runs without strange sideeffects. :) 1134510108 M * Bertl then no guest should create funny tagged files in your setup, right? 1134510132 M * derjohn daniel_hozac, chances to get it into std find == 0 ? Or would it be feasable to put it into util and call ut vfind (all tools are prefixed 'v', so....) 1134510178 N * Smutje_ Smutje 1134510198 M * daniel_hozac derjohn: as it requires util-vserver's libvserver, no standard distro build would ever have it. 1134510214 M * daniel_hozac derjohn: so i haven't even submitted it upstream. 1134510228 M * Kara_ i would consider it being usefull though 1134510260 M * Bertl daniel_hozac: that might get easier with hollows implementations 1134510263 M * Bertl no? 1134510272 M * daniel_hozac Bertl: how? 1134510293 M * Bertl well, the few syscall commands could be easily added 1134510294 M * derjohn daniel_hozac, our future: debian sarge|etch|sid|bertl <-- the one with vserver in ;) 1134510308 M * Bertl lol 1134510316 M * daniel_hozac Bertl: util-vserver's library is already really easy to use. 1134510323 M * Bertl okay 1134510472 M * derjohn Bertl, the new stable release ... is it worth to inform heise? or to small? 1134510507 M * Bertl well, it took quite some time to finish ... so go ahead 1134510518 M * daniel_hozac oohh, 2.0.1 got released? changes from -rc5, besides reiserfs fix? 1134510525 M * derjohn besides that i am notware of a changelog 1134510546 M * Bertl http://www.13thfloor.at/vserver/s_rel26/overview 1134510566 M * Bertl http://linux-vserver.org/Changelog26 1134510567 M * derjohn daniel_hozac, same as me. some houres not here ... boom .. new release ;) 1134510572 M * derjohn k 1134510586 M * Bertl daniel_hozac: yes, I did a few cleanups 1134510598 M * daniel_hozac vs201-clean02? 1134510599 M * Bertl mainly white space adjustments 1134510605 M * Bertl yup 1134510638 M * Bertl and the last part of the vroot fix was missing, IIRC 1134510644 M * daniel_hozac ah. 1134510650 M * Bertl but you can get a delta to -rc5 1134510739 M * daniel_hozac where? or isn't it in Experimental yet? :) 1134510779 M * Bertl http://vserver.13thfloor.at/Experimental/delta-2.6.14.3-vs2.0.1-rc5-vs2.01.diff 1134510810 M * daniel_hozac thanks. 1134510840 M * Bertl http://vserver.13thfloor.at/Experimental/delta-2.6.14.3-vs2.0.1-rc5-vs2.01-nowhite.diff 1134510844 M * Bertl you're welcome! 1134510863 M * daniel_hozac any particular reason it's not 2.0.1 rather than 2.01? 1134510922 M * Bertl I thought I'll use the 1.2x notation from 2.4 1134510937 M * Bertl to make it easier for folks to tell them apart ... 1134510981 M * daniel_hozac isn't it 1.2.10 now? :) 1134510993 M * Bertl yeah ... right :) 1134511011 Q * jeeves Quit: Leaving 1134511039 M * Bertl well, the internal version is 2.0.1 anyway 1134511083 M * Bertl so when we get the first complaint, I'll change the urls and download pages :) 1134511084 M * mnemoc Bertl: do you have the changes between 2.0.0 and 2.0.1? :) 1134511090 M * daniel_hozac hehe. 1134511104 M * Bertl mnemoc: sure! 1134511112 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0.1/ 1134511131 M * mnemoc Bertl: thanks! :D 1134511200 M * derjohn Bertl, the changelog ist not very "spannened" to read. just "normal" bigfixes? or is there s.th. pretty cool in you changed and forgot to mention? 1134511306 M * Bertl no, stable is stable ... we have a bunch of changes in devel 1134511978 M * derjohn Bertl, what is 'ioprio'? "My guest is better than your's io preference"? 1134512004 M * Bertl for example, basically it's support for set/get_ioprio() syscalls 1134512034 M * Bertl the devel release also has the cfq I/O scheduler modified to handle contexts in a fair way 1134512050 M * derjohn Bertl, as i said, I'm short haird. When do those syscalls occur? before the diskdriver does io on the dev? 1134512071 M * Bertl no, they occur when you artificially set the priority 1134512091 M * derjohn Bertl, ah, this is a std kernel feature you exploit? 1134512093 M * Bertl they are pretty new, as they require the v3 scheduler (I/O) which was not there in 2.6.13 1134512111 M * Bertl yes, it's a mainline feature 1134512118 M * Bertl nothing vserver specific 1134512159 M * derjohn Bertl, wow a new scheduler? hmm . btw... can you tell me what happened to netlink? it's no compile time option anymore ... 1134512226 M * Bertl CONFIG_NETFILTER_NETLINK ? 1134512283 M * derjohn errrr ... 1134512285 M * derjohn mom 1134512377 M * Bertl okay, it's warm .. I had something to eat ... now ... nap attack! 1134512387 M * Bertl be back later ... have fun! 1134512393 N * Bertl Bertl_zZ 1134512393 M * derjohn eh 1134512395 M * derjohn yo 1134512397 M * derjohn bye 1134512513 M * Kara_ I will be gone for now, too. Goodnight! 1134514467 J * infowolfe ~infowolfe@209-112-212-182-cdsl-rb1.nwc.acsalaska.net 1134515581 Q * micah Quit: Leaving 1134515688 J * micah ~micah@69.90.134.205 1134516202 J * comfrey ~comfrey@peoples.coop 1134516529 Q * micah Quit: leaving 1134516542 J * micah ~micah@69.90.134.205 1134516926 Q * infowolfe Quit: Leaving 1134516968 Q * click Ping timeout: 480 seconds 1134517000 Q * micah Quit: leaving 1134517033 J * micah ~micah@69.90.134.205 1134517034 Q * micah Quit: 1134517046 J * micah ~micah@69.90.134.205 1134517061 Q * micah Quit: 1134517096 J * micah ~micah@69.90.134.205 1134517676 J * click click@ti511110a080-0549.bb.online.no