1154390484 M * shedi because I was asking earlier, just for the record, to run a ntp-server within a vserver, an additional loopback interface is needed for the vserver 1154390568 M * shedi with that said, I will rest my head upon my pillow :) 1154390657 N * _jake- jake- 1154390995 Q * chand Quit: chand 1154391186 J * s0undt3ch khljyscx@bl7-241-75.dsl.telepac.pt 1154393327 Q * bonbons Quit: Leaving 1154395810 J * NemesisBLK ~malloc@68-190-228-115.dhcp.wsco.ca.charter.com 1154396544 Q * meandtheshell Quit: bye bye ... 1154397772 Q * NemesisBLK Quit: :^) 1154399338 M * Skram im redoing my kernel on my vserver host 1154399401 M * Skram anyone around? 1154399414 M * daniel_hozac probably. 1154399417 M * Skram so 1154399420 M * Skram i am using 2.6.14 1154399430 M * Skram ill use the new patch.. but are there any compat problems with anything? 1154399476 M * daniel_hozac it depends on how everything is configured. 1154399493 M * Skram uhmm 1154399498 M * Skram what are my choices? :P 1154399548 M * daniel_hozac as long as you enable CONFIG_VSERVER_LEGACY, util-vserver should work correctly. 1154399579 M * Skram anythign special when patching for gentoo? 1154399623 N * starleinn starlein 1154399639 M * daniel_hozac i have no idea. 1154399737 Q * s0undt3ch Ping timeout: 480 seconds 1154400056 M * Skram daniel_hozac: all this has left my mind.. what is the procedure for patching the kenrel 1154400065 M * Skram get a new one, run the patch, then make configs, etc/ 1154400157 M * daniel_hozac http://linux-vserver.org/Step-by-Step+Guide+2.6 1154400185 M * Skram Thanks 1154400293 M * Skram it says 2.0.2-rc27 works on 2.6.17.7.. i am using 2.6.17.4.. will it still work? 1154400309 M * Skram or can i just jump down to the patch for 2.6.16.20? 1154400359 M * daniel_hozac why are you using 2.6.17.4? there have been quite a few things fixed in .7. 1154400380 M * daniel_hozac but yes, except for the Makefile, it should apply. 1154400647 J * gerrit ~gerrit@66.46.95.226 1154400787 M * Skram what do you mean exactlu 1154400850 M * daniel_hozac ? 1154401006 M * Skram uhh 1154401031 M * Skram do the patch change the makefile enough to break it from the minor changes (.4 vs .7) 1154401094 M * daniel_hozac kernel.org uses EXTRAVERSION for the micro-releases, and vserver appends the version there too. 1154401123 M * daniel_hozac obviously patch doesn't have sufficient AI to realize it's just changing a version and do the right thing when the line is different. 1154401622 M * Skram patching file Makefile 1154401622 M * Skram Hunk #1 FAILED at 1. 1154401622 M * Skram 1 out of 1 hunk FAILED -- saving rejects to file Makefile.rej 1154401623 M * Skram yeah 1154401779 M * cehteh the usual suspect :) 1154401833 M * Skram it doesnt really matter though does it.. because it was just changing varibles such as VERSIOn info.. right? 1154401934 M * daniel_hozac you could just add it yourself. 1154402528 M * Skram does that part matter though? 1154402594 M * daniel_hozac what? the version? 1154402611 M * daniel_hozac it's always nice to be able to check what exact revision something is. 1154402653 M * daniel_hozac i still don't get why you'd want to stick with .4, when .7 fixes several memory leaks and one local root. 1154402685 M * Skram its gentoo, and i hate kernels.. i really dont want to mess anything up 1154402695 M * Skram and if i start with a new kernel, i may forget something 1154402695 M * Skram eh 1154402697 M * Skram *heh 1154402896 J * KantankerousKid ~blah@149.9.0.27 1154404292 J * s0undt3ch uczvezd@bl7-242-110.dsl.telepac.pt 1154405632 Q * KantankerousKid 1154406787 J * yarihm ~yarihm@whitehead2.nine.ch 1154408350 J * coocoon ~coocoon@p54A06063.dip.t-dialin.net 1154408392 M * coocoon morning 1154408601 M * abi mooorning 1154409791 Q * s0undt3ch Ping timeout: 480 seconds 1154410681 J * s0undt3ch vvccgy@bl7-242-17.dsl.telepac.pt 1154411175 Q * s0undt3ch Read error: Connection reset by peer 1154412382 J * meandtheshell ~markus@85-125-231-3.dynamic.xdsl-line.inode.at 1154412591 J * click_ click@ti511110a080-1946.bb.online.no 1154412700 Q * click Ping timeout: 480 seconds 1154413215 J * pisco ~pampel@p50878A74.dip0.t-ipconnect.de 1154413319 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154413562 J * Pazzo ~thomas@dialin-225136.rol.raiffeisen.net 1154414221 J * s0undt3ch cpegff@bl7-244-141.dsl.telepac.pt 1154414649 J * dna ~naucki@dialer-181-208.kielnet.net 1154416259 J * schimmi2 ~sts@port-212-202-73-176.dynamic.qsc.de 1154416382 Q * s0undt3ch iridium.oftc.net arion.oftc.net 1154416382 Q * h01ger iridium.oftc.net arion.oftc.net 1154416382 Q * blues iridium.oftc.net arion.oftc.net 1154416382 Q * bubulak iridium.oftc.net arion.oftc.net 1154416382 Q * meebey iridium.oftc.net arion.oftc.net 1154416382 Q * lilo2 iridium.oftc.net arion.oftc.net 1154416382 Q * FireEgl iridium.oftc.net arion.oftc.net 1154416382 Q * locksy iridium.oftc.net arion.oftc.net 1154416382 Q * SNy iridium.oftc.net arion.oftc.net 1154416382 Q * MooingLemur iridium.oftc.net arion.oftc.net 1154416382 Q * dna iridium.oftc.net arion.oftc.net 1154416382 Q * Viper0482 iridium.oftc.net arion.oftc.net 1154416383 Q * meandtheshell iridium.oftc.net arion.oftc.net 1154416383 Q * insomnia1 iridium.oftc.net arion.oftc.net 1154416383 Q * kir iridium.oftc.net arion.oftc.net 1154416383 Q * schimmi iridium.oftc.net arion.oftc.net 1154416383 Q * Zaki[] iridium.oftc.net arion.oftc.net 1154416383 Q * Radiance iridium.oftc.net arion.oftc.net 1154416383 Q * pagano iridium.oftc.net arion.oftc.net 1154416383 Q * michal` iridium.oftc.net arion.oftc.net 1154416383 Q * cemil iridium.oftc.net arion.oftc.net 1154416383 Q * baggins iridium.oftc.net arion.oftc.net 1154416383 Q * BenBen iridium.oftc.net arion.oftc.net 1154416383 Q * Hollow iridium.oftc.net arion.oftc.net 1154416439 J * lilo2 hiddenserv@tor.noreply.org 1154416439 J * FireEgl ~FireEgl@Atlantica.Tcldrop.Com 1154416439 J * locksy ~locksy@mrtg.sisgroup.com.au 1154416439 J * SNy 5c1d4a12a4@bmx-chemnitz.de 1154416439 J * MooingLemur ~troy@shells200.pinchaser.com 1154416580 Q * pisco Ping timeout: 480 seconds 1154416638 J * Radiance 1e03a3e572@halt.1984world.eu 1154416725 J * dna ~naucki@dialer-181-208.kielnet.net 1154416776 J * s0undt3ch cpegff@bl7-244-141.dsl.telepac.pt 1154416776 J * h01ger ~holger@socket.layer-acht.org 1154416776 J * blues blues@blysk.ds.pg.gda.pl 1154416776 J * bubulak ~bubulak@whisky.pendo.sk 1154416776 J * meebey meebey@booster.qnetp.net 1154416832 J * Zaki[] ~Zaki@212.118.105.126 1154416832 J * Hollow ~hollow@2001:a60:f026::1 1154416832 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154416832 J * cemil_ ~cemil@defiant.wavecon.de 1154416832 J * meandtheshell ~markus@85-125-231-3.dynamic.xdsl-line.inode.at 1154416832 J * michal` ~michal@www.rsbac.org 1154416832 J * pagano ~pagano@131.154.5.20 1154416861 J * kir ~kir@swsoft-mipt-nat.sw.ru 1154416924 J * ||Cobra|| ~cob@146.50.22.204 1154417003 J * insomniac ~insomniac@slackware.it 1154417015 J * baggins baggins@kenny.mimuw.edu.pl 1154417019 J * BenBen ~benny@defiant.wavecon.de 1154417417 J * pisco ~pampel@p50878A74.dip0.t-ipconnect.de 1154417455 Q * Aiken Ping timeout: 480 seconds 1154418911 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at 1154420236 Q * coocoon Ping timeout: 480 seconds 1154420620 J * lilo2_ hiddenserv@tor.noreply.org 1154420624 Q * lilo2 Remote host closed the connection 1154421501 Q * shedi Quit: Leaving 1154421908 J * wam ~wam@proxy1.msh.de 1154422140 J * mef ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1154422154 P * mef 1154422439 M * wam Hi, I have two hosts connected via openvpn. both hosts run vservers. But the vservers on Host A are unreachable from the vservers on Host B, because they all don't know about the tun0-interface, which is required for the routing (e.g. "route" shows "*" for tun0-routing entries). Yesterday, derjohn proposed to establish a private subnet between the hosts and re-route over this one (which is bound to an accessable interface, e.g. eth1). But the re-routing does not 1154422439 M * wam work. The first found interface is used. Any ideas? 1154422472 M * wam or did i misunderstood your instructions, derjohn? 1154422491 M * wam or did i misunderstand your instructions, derjohn? 1154422561 M * wam would maybe the bcap "NET_RAW" let me use tun0 for routing? What would be the security-problems when all vservers are only under my control? 1154422754 J * coocoon ~coocoon@p54A068DB.dip.t-dialin.net 1154423672 J * mef ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1154424714 M * wam Is it possible to give a vserver access to a network interface without giving it an ip-address on this interface? 1154424841 Q * mef Remote host closed the connection 1154424946 J * shedi ~siggi@dsl-og-108-50.du.vortex.is 1154425519 J * bonbons ~bonbons@83.222.36.236 1154425784 M * derjohn wam, my hint: try to make it all work _without_ vservers. bind the vservers IPs to eth0 and establish routes. 1154425897 M * derjohn wam, if that works, then take care fore the vservers. to change the interface of a vserver look at /etc/vservers//interface/xxx 1154426322 Q * Viper0482 Remote host closed the connection 1154426419 M * wam derjohn: all works without vservers. The vservers IPs are all bound to eth1. When I route the vpn to the eth1-net, I get the problem. Because in the routing table (also from within the vservers) there is a route to the vpn-net over tun0, which is unknown to the vservers. 1154426447 M * wam that means, the vservers can't route to tun0. 1154426610 M * wam derjohn: as soon as i give them an ip in tun0, they can route... 1154426666 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154426765 N * schimmi2 schimmi 1154426766 M * derjohn wam, hm just trying to think about it :) ... 1154426833 J * chand ~chand@gw.net81-65-27.noos.fr 1154426898 Q * locksy Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1154426917 M * bonbons wam: how do your routes through tun0 look like, maybe they require a give source address 1154427667 M * derjohn bonbons, the src should be the guests's ip 1154427728 M * bonbons derjohn: if a source is defined in the route, then there has to be a route with the guest's ip 1154427927 M * derjohn bonbons, you mean a src for the route via tun0? IMO currently the doesnt use policy (src based) routing, and thus the guest will fire a packet with SRC and DST and hand that over to the kernel. nor? the kernel should know the route - the guest doesnt have to "know" the particular interface (i.e. ip ro get shows a route over "if7" within the guest, but that should be no problem) 1154427979 M * derjohn bonbons, wam: from the host directly a ping -i dst-server seems to work if I understood correctly. 1154428043 M * bonbons derjohn: looks like kernel creates some such policy-routes automatically when firingup some interfaces 1154428126 M * derjohn bonbons: this bitch ;) 1154428156 Q * Viper0482 Remote host closed the connection 1154428167 M * bonbons just saw them in my routing table, and nevert set them myself! I have one such on my dummy interface and one on my tun interface 1154428179 M * derjohn bonbons, but anyway: i dint see why it wouldn't route an arbitrary packet on the host/kernel 1154428204 M * derjohn bonbons, maybe it get special on virtual interfaces like dummy and tun? 1154428237 A * derjohn now *afk* bbl ! 1154428259 M * bonbons don't know what exact rules kernel is using, but having src-routes can be imagined as rule to avoid "garbage" src-ips getting out there 1154428297 M * bonbons you wouldn't want to send packets with a src IP for which you have no route back 1154428339 M * derjohn bonbons, i would bet that it gets problematic to auto-magically discover the the src address if a packet is constructed from localhost 1154428361 M * derjohn but if the src addr is "set hard" ... 1154428383 M * derjohn hm, a vserver guest relioes on the kernel, so it can be regarded as local ... 1154428384 M * bonbons derjohn: if no src is specified, the kernel tries to find the best match, but vserver limits the possible choice! 1154428399 M * derjohn but : /me now away ... bbl ! 1154428433 M * bonbons without vserver restriction the kernel would probably have chosen the righ src 1154428579 M * derjohn bonbons, so a workaround would be a policy based route . nor 1154428580 M * derjohn ? 1154428587 M * derjohn but: /me now away 1154428629 M * bonbons I would say no policy on the route 1154428704 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154429280 Q * meandtheshell Quit: bye bye ... 1154429574 N * lilo2_ lilo2 1154429918 J * renihs ~penguin@83-65-34-34.arsenal.xdsl-line.inode.at 1154430063 Q * chand Quit: chand 1154430305 M * wam derjohn: sorry - was away a while. To your question 34 minutes ago (ping -i dst-server) - yes, this works. 1154430793 J * mire ~mire@68-166-222-85.COOL.ADSL.VLine.verat.net 1154430938 Q * yarihm Quit: Leaving 1154430998 M * wam derjohn: bonbons: Thanks for that discussion above. This gave me a new idea - and I solved the problem. The routing is running ok - but the source is the problem. As the vserver doesn't evaluate the route to tun0, it doesn't know that the source should be in its internal network. So it sends to tun0 with an official IP, but never gets replies. When I use "ping -I " it works. Any idea how I can correct this behavour? 1154431039 M * wam Better said: When I use "ping -I " INSIDE the vserver, it works. 1154431085 M * bj can someone give me a hint howto limit max cpu usage of a vserver ? 1154431301 Q * coocoon Remote host closed the connection 1154431310 M * wam bj: http://linux-vserver.org/Scheduler+Parameters 1154431525 M * bj wam: thx :) 1154431597 J * Milf ~Miranda@ipsio245.ipsi.fraunhofer.de 1154431626 Q * Viper0482 Remote host closed the connection 1154431900 M * bonbons wam: so you have two addresses for your guest? Assigning the addresses the other way round would eventually solve your issue 1154431987 M * wam bonbons: good point 1154432005 A * wam goes playing with the order of interfaces 1154432008 M * bonbons that is, first assign internal address to guest, then the global. 1154432053 M * bonbons right now I trying to understand that area of kernel code, IPv6 was easier than IPv4 seems to be :) 1154432097 M * wam bonbons: with the changed order, it works. 1154432117 M * wam bonbons: i do not want to understand why this is implemented in the way it is. 1154432133 M * wam bonbons: I'm simply waiting for ngnet and banging my head against the desk some time. 1154432145 M * bonbons so it's the code I was reading again and again these days that has it'seffect here, many many reference to first IP of the guest 1154432165 N * Bertl_oO Bertl 1154432169 M * Bertl morning folks! 1154432170 M * phedny hi Bertl 1154432173 M * wam hi Bertl 1154432179 M * bonbons morning Bertl 1154432191 M * wam bonbons: the reason seems to be that the guest can't know which src to use when the interface is not available to it. 1154432212 M * Bertl bonbons, wam: well, not unexpected ... there are many cases in the networking stack where the 'primary' ip is used ... unfortunately we can not guarantee that the 'primary' ip is part of the guest :) 1154432214 M * wam bonbons: so it has to make a choice. But if this is reproducably the FIRST defined interface, i can live with that. 1154432244 M * Bertl wam: yes, the first ip is the 'primary' for the guest 1154432275 M * wam Bertl: so this is expected behavour and not by random. So I don't have to bother ;) 1154432367 M * bonbons Bertl: that's a part that makes my game with address ranges a nightmare in IPv4 :) 1154432436 M * bonbons especially as in case of ranges it's reasonable at all to assume that the first address in range is even assigned to any device on the host 1154432512 M * bonbons would eventually be best to go the IPv6 way for src address selection (scan through all addresses and take the best we can get from those assigned to the guest) 1154432781 M * Bertl bonbons: I think we should keep handling the 'primary' ip special 1154432804 M * Bertl and we might also want to re-activate the 'broadcast' ip for this purpose 1154432830 M * Bertl for the ranges, I think we should definitely move the 'business' logic into userspace 1154432842 M * bonbons just that there isn't a primary in case we put n possible addresses as a range 1154432847 M * Bertl i.e. the kernel gets ranges and ip loaded with special masks ... 1154432856 M * Bertl or flags 1154432969 M * bonbons off to lunch 1154432971 Q * bj Ping timeout: 480 seconds 1154432981 M * Bertl derjohn: ping? 1154433021 J * Onceuponatime lol@212.251.16.157 1154433045 M * Milf ... www.ubuntu.de ... Can someone point me to a downloadlink for Ubuntu images to burn an install CD from? 1154433055 M * Bertl welcome Onceuponatime! 1154433162 M * phedny Milf: x86? 1154433168 J * bj ~bj@insanefactory.com 1154433188 M * Milf phedny: yes x86 1154433208 M * phedny ftp://ftp.fu-berlin.de/linux/ubuntu/releases/6.06/ <-- that one has all by the way ;) 1154433248 M * Milf Thanks. 1154433254 M * phedny desktop has graphical install, alternate has more options (like LVM) 1154433272 M * Milf Thanks 1154433277 M * phedny np 1154433283 A * phedny kinda Ubuntu-fan lately :) 1154433312 J * coocoon ~coocoon@p54A05E87.dip.t-dialin.net 1154433411 M * Milf Hmmm, now to gues if Zeng used desktop or alternate ... 1154433428 M * phedny Zeng? 1154433446 M * Milf The Guy who wrote http://linux-vserver.org/MoreUbuntu 1154433470 M * Milf I'm the guy who's supposed to make an article for c't from it. 1154433470 M * phedny doesn't matter 1154433475 M * phedny desktop system can be build from both 1154433483 M * phedny C'T? 1154433495 M * Milf Hmmm, I'll go with desktop then. 1154433499 M * phedny I know someone in .nl that translates and writes articles for C'T 1154433508 M * phedny translates German articles that-is 1154433508 M * Milf A popular german computer magazine 1154433565 M * Milf Must be a nice job to get paid for reading what others read for fun :) 1154433572 M * phedny hehe 1154433596 M * phedny do you know any Dutch collegues? 1154433655 M * Milf colleagues or colleges? 1154433668 M * phedny colleagues 1154433688 M * Milf Nope, sorry, don't have the pleasure of working with any dutch people 1154433736 A * Milf will add Holland to his list of countries good for cycling that have yet to be visited. 1154433751 M * phedny you like cycling? 1154433780 M * Milf a lot 1154433794 M * Milf more if I could get my seating problems in order 1154433801 M * phedny :) 1154433805 M * phedny understandable 1154433865 M * Milf Yeah I'm thinking of packing it in for a week and taking my new bike on a little stroll. 1154433876 A * phedny likes cycling too, but my friends prefer things like swimming 1154433876 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154433889 M * Milf I was thinking of Scotland, but Holland now seems a bit more reachable as I'm going spontaneously 1154433913 A * phedny lives a little more than an hour from the German border ;) 1154433928 M * phedny by car that is 1154433938 M * Milf I was gonna ask :) 1154433940 M * phedny although bike shouldn't take much longer 1154433947 M * phedny you know Nijmegen (Nimwegen?) 1154433966 M * Milf Heard of it. Furthest I ever got was Maasticht 1154433973 A * Milf adds the missing 'r' 1154434079 M * phedny you live in the Southern part of Germany? 1154434098 M * Milf Darmstadt, smack in the center. 1154434149 M * Milf It would be two to three days to the Dutch border I guess. 1154434165 M * phedny okay 1154434172 M * Milf By bike I mean. 1154434173 M * phedny well, the name Darmstadt sound very familiar 1154434180 M * Milf South of Frankfurt. 1154434260 M * phedny I'm looking at a map to find some place near it I might remember better 1154434333 M * Milf Thanks for that download-link. I got a pretty fast connection. Four minutes for a complete ISO is nice :) 1154434423 M * phedny I guess we just passed it some time when we drove to Austria 1154434446 Q * glut Read error: Connection reset by peer 1154434620 Q * mire Remote host closed the connection 1154434639 J * glut glut@no.suid.pl 1154434657 M * Milf Yeah seems likely. 1154435137 J * SNy_ 31b763fc31@bmx-chemnitz.de 1154435137 Q * SNy Read error: Connection reset by peer 1154436059 M * sid3windr 2.4 no longer supported? 1154436484 M * wam Bertl: bonbons: derjohn: now, that I use the internal IP as the primary on all vservers, my Bind-running vserver uses this to access the internet.... :/ 1154436532 M * bonbons wam: you should be able to tell bind to use the public IP for that 1154436561 M * bonbons just tell it which address to use for that purpose (same scenario as if you had two public IPs for it) 1154436667 M * wam bonbons: as far as i can see, all vservers try to access from the local net for DNS queries (10.112.0.11 -> 80.237.xxx.xxx DNS Standard query A altavista.com) 1154436726 M * wam bonbons: I can't tell every application to resolve via public IP. 1154436741 M * bonbons wam: don't know the name of the bind option for this, but check how to define source address for it's outgoing requests 1154436773 M * wam bonbons: it would probarbly be no problem for bind. But for every app that resolves something? 1154436813 M * bonbons the other apps should ask the server(s) in /etc/resolv.conf 1154436833 M * wam bonbons: they do - the external servers from src 1154436885 M * wam bonbons: even when I try telnet to port 80, it looks like this: 10.112.0.9 -> 70.70.70.70 TCP 59196 > www [SYN] Seq=0 Ack=0 Win=5840 Len=0 MSS=1460 TSV=862534 TSER=0 WS=2 1154436948 M * bonbons wam: grr that's really not the best way of choosing source address! 1154436954 M * wam it looks like the primary interface is ALLWAYS the src address. 1154437014 M * bonbons looks so 1154437027 M * wam is there ANY working alternative with two nets for a vserver? 1154437098 M * wam I mean, why doesn't a vserver just use the routing table or leave the choice for a src-address to the host? 1154437127 M * bonbons I don't know... would be worth improving. Have to ask Bertl what special-cases he would like to have/keep in relation to the 'first' address 1154437151 M * wam yeah - maybe there's a plan behind that all which I can't see... 1154437209 M * bonbons I know of one optimisation in cases where there is just a single address, but for the rest... 1154437256 J * romke ~romke@procyon.romke.net 1154437256 J * pisc1 ~pampel@p50878A74.dip0.t-ipconnect.de 1154437595 N * SNy_ SNy 1154438500 Q * gerrit Ping timeout: 480 seconds 1154438769 M * Bertl wam: what was the reason to make the internal (private?) ip the primary in the first place? 1154438849 M * bonbons Bertl: that was to get the right address used for internal traffic (to get it through a tun-vpn tunnel) 1154438947 M * Bertl could it be that wam actually has a multiple gateway setup and no source routing? 1154438989 M * Bertl because if there _is_ a definitive route with a local source ip (assigned to the guest) this should be chosen 1154439000 Q * pisc1 Ping timeout: 480 seconds 1154439001 J * pisc1 ~pampel@p5087A50E.dip0.t-ipconnect.de 1154439016 M * wam Bertl: 1. this was the reason. I want to communicate via tun0 to another host. But tun0 isn't known by the vservers. 1154439045 M * wam Bertl: I only have one default gw. The other nets are defined by / in the routing table. 1154439157 M * wam Bertl: The correct src-ip is not choosen in my case. but maybe I'm overseeing something. I thought the reason was that the vservers don't see the interface in "ip r s" or "route". Because, as soon as I add a interface tun0 to the vserver (even with an invalid IP), all works. 1154439260 Q * pisco Ping timeout: 480 seconds 1154439272 Q * mountie Remote host closed the connection 1154439780 M * Bertl wam, you definitely need to add the tun0 ip to the guest config 1154439798 M * Bertl otherwise the guest will not be able to handle that ip at all 1154439825 M * wam Bertl: i wanted to avoid that. But if there's no other way, I'll have to provide it to all the guests that require access to the tunnel. 1154439852 M * Bertl no way to avoid that, as Linux-VServer's networking is _ip_ based :) 1154439945 J * KantankerousKid ~blah@tor-irc.dnsbl.oftc.net 1154440013 M * wam Bertl: ok, which device and ip should I add? The same the host has? 1154440063 M * wam at least this works ;) 1154440132 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com 1154440409 Q * KantankerousKid 1154440461 M * Bertl wam: well, if you _want_ to use the same ip as the host has, then add that one 1154440481 M * Bertl wam: if you want to use a different one, assigned to tun0 (or whatever) assign that one 1154441327 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1154441421 N * Onceuponatime JohnGreek 1154441494 Q * Roey 1154441504 M * wam date 1154441508 M * wam oops ;) 1154441517 J * click click@ti511110a080-4165.bb.online.no 1154441607 M * Skram eeeks 1154441611 M * Skram updating kernel! 1154441624 M * Skram anyone confirm 2.17.7 and 2.0.2-rc27 work well? 1154441630 Q * click_ Ping timeout: 480 seconds 1154441783 M * Skram Bertl: you do all the kernel stuff, right? 1154442146 M * Skram Bertl: around? 1154442198 Q * coocoon Quit: KVIrc 3.2.0 'Realia' 1154442395 M * Bertl Skram: not all of it, but most ... 1154442410 M * Skram so.. 1154442426 M * Skram if I want a stable install, what kernel, vserver patch should I use? 1154442440 M * Skram according to the site, 2.17.7/2.0.2-rc27 1154442449 M * Bertl latest kernel and stable prerelease 1154442464 M * Skram so 2.0.1? 1154442478 M * Bertl so in this case 2.6.17.7 and 2.0.2-rc27 1154442482 M * Skram okay 1154442486 M * Skram then thats what I am doing 1154442524 M * Skram just let me get this straight, libvserver is Hollow and the gentoo'ers thing, and is not required for normal operations 1154442539 M * Skram the standard "vserver" bins come with util-vserver which is at 0.30.210, right? 1154442545 Q * Viper0482 Quit: one day, i'll find this peer guy and then i'll reset his connection!! 1154442545 J * coocoon ~coocoon@p54A06D14.dip.t-dialin.net 1154442563 M * Hollow libvserver has nothing to do with gentoo 1154442571 J * Viper0482 ~Viper0482@p549774E7.dip.t-dialin.net 1154442571 M * Bertl Skram: yes, that is util-vserver (0.30.210 + some patches from daniel) 1154442581 M * Skram Hollow: Okay, sorry 1154442596 M * Skram vserver-utils is though.. or atleast it isnt required for a base install 1154442602 M * Skram Bertl: Right on 1154442617 M * Hollow well, it is a replacement for util-vserver, and can be used in every distro, not just gentoo 1154442636 M * Skram right on 1154443006 M * renihs hmm? just noticed vserver-sources 2.1 have been removed from gentoo portage (only 2.0 is there) 1154443022 M * Skram and it uses 2.6.15, right? 1154443024 M * Skram last time i checked 1154443033 M * Hollow the devel has been moved to our overlay 1154443035 A * waldi .o0( gent00? harhar ... ) 1154443044 M * Skram im just doing kernel.org and patching it, to be safe and make sure I have a stable box 1154443050 M * Hollow http://overlays.gentoo.org/proj/vps 1154443214 M * Skram uhh... what should i see in .config to make sure my kernel was patched, i searched for "VSERVER" and nothing showed up! 1154443335 M * bonbons Skram: read top of Makefile and check for EXTRAVERSION 1154443352 M * Skram ".7" 1154443396 M * bonbons then you still miss the vserver patch 1154443400 M * Skram weird 1154443482 M * Skram worked this time 1154443488 M * Skram EXTRAVERSION = .7-vs2.0.2-rc27 1154443489 M * Skram NAME=Crazed Snow-Weasel 1154443507 M * renihs thx Hollow 1154443518 M * bonbons Skram: fine, now make oldconfig (or whatever config mode you like) 1154443547 M * Skram bonbons: cant i just copy the old .config and run make menuconfig? 1154443559 M * Skram inux-2.6.17.7-vs2.02-rc27 # cp ../linux-2.6.14-vserver-2.0.1/.config . 1154443571 M * Hollow renihs: you can also use the overlay with layman if you have it installed 1154443571 M * bonbons you can, oldconfig has advantage to tell you whatchanged 1154443588 M * Skram bonbons: okay, so copy the .config and then run make oldconfig? 1154443594 Q * Milf Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org 1154443602 M * Loki|muh yeah 1154443611 M * Loki|muh thats the way i do upgrades 1154443615 M * Skram right on 1154443619 M * Skram bonbons: thats the best way? 1154443619 M * bonbons yep, then you will be told anything that the new kernel has which previous one did not have 1154443627 M * Skram okay 1154443680 M * bonbons way more efficient than running through all menuconfig to find what may have been added 1154443686 M * Skram right 1154443796 M * Skram sorry for the ignoracne, but is ip_tables the same as iptables? 1154443801 M * Skram Netfilter Xtables support (required for ip_tables) (NETFILTER_XTABLES) [N/m/y/?] (NEW) 1154443963 M * bonbons Skram: that's a dangerous area in recent kernels (since march or so) 1154443973 M * Skram hmm 1154443978 M * bonbons do you want IPv6 stateful or not? 1154443985 M * Skram not reall 1154443987 M * Skram no need right now 1154444063 M * bonbons and no special need to IPv6 either? If so you can mostly take the defaults, eventually checking back later using menuconfig 1154444073 M * Skram right 1154444087 M * derjohn Bertl, Thanks ;) 1154444096 M * Skram person who did the kernel last time built in wireless support 1154444108 M * Skram i dont even have physical access to the box, why do we need wireless support 1154444110 M * Skram heh 1154444115 M * Bertl derjohn: np! 1154444177 M * derjohn Bertl, hm, about the "shared caches (buffers?): this only applies, if at leat two clients share some kind of directory (unification?) ? 1154444273 M * Bertl derjohn: well, yes and no, depends on what cache we are talking about, e.g. the filesystem caches and kernel structures are shared too 1154444295 M * Bertl derjohn: think meta structures ... of course, with shared files and such (unification) it is even more 1154444302 M * derjohn what kernel structues are caches and in common (shared) use? 1154444312 M * derjohn *cached 1154444318 M * Skram now for makemenuconfig :) 1154444350 M * Bertl derjohn: for example a filesystem consists of inode bitmaps (metablocks) and such stuff 1154444374 M * Bertl derjohn: or just take the device writeback which will be able to put requests together into the elevator 1154444414 M * derjohn Bertl, so the kernel may cache a part of the bitmap index that may contain information for guest A and (by random) for guest B, too? 1154444583 M * Bertl yep, precisely, and it _will_ definitely share superblock information if they reside on the same filesystem 1154444969 J * meandtheshell ~markus@85-124-207-188.dynamic.xdsl-line.inode.at 1154445025 Q * pisc1 Quit: Download Gaim: http://gaim.sourceforge.net/ 1154445049 J * pisco ~pampel@p5087A50E.dip0.t-ipconnect.de 1154445456 Q * wam Quit: Verlassend 1154445532 J * murdoc ~anything@70.169.163.34 1154445541 M * Bertl welcome murdoc! 1154445659 Q * ||Cobra|| Remote host closed the connection 1154446157 Q * JohnGreek 1154446277 Q * Viper0482 Quit: one day, i'll find this peer guy and then i'll reset his connection!! 1154446343 M * coocoon Bertl: hey bertl now i have installed gentoo on the sparc, but u do not need it at all right 1154446775 Q * renihs Quit: Leaving 1154447388 N * Belu_zZz Belu 1154448146 Q * tso Ping timeout: 480 seconds 1154448389 Q * Pazzo Quit: Ex-Chat 1154448884 J * tatiane ~tatiane@201009043189.user.veloxzone.com.br 1154448898 Q * tatiane 1154450322 J * stefani stefani@D-128-208-208-118.dhcp4.washington.edu 1154451378 J * alexx ~alexx@proxy.ikse.net 1154451562 M * blues re 1154451576 M * blues Bertl: I think that I know was my problem was... 1154451621 M * blues Bertl: it seems that I needed that: http://www.ussg.iu.edu/hypermail/linux/kernel/0601.1/1817.html 1154451640 M * blues Bertl: tnx for you analysis - it helped a lot :D 1154452310 P * murdoc 1154452458 Q * alexx Quit: Ex-Chat 1154454044 Q * sladen Ping timeout: 480 seconds 1154454173 Q * shedi Quit: Leaving 1154454187 J * Pazzo ~thomas@dialin-225136.rol.raiffeisen.net 1154454210 J * sladen paul@starsky.19inch.net 1154456815 Q * lilalinux_ Remote host closed the connection 1154456958 P * stefani parting (is such sweet sorrow) 1154457111 Q * schimmi Quit: Verlassend 1154457482 Q * Pazzo Quit: ... 1154457910 J * x03 ~x03@10001268031.0000030345.acesso.oni.pt 1154457915 M * x03 hi all 1154458675 M * daniel_hozac hello 1154459385 J * Navin ~Navin@206.61.46.5 1154459439 P * Navin 1154461604 A * Belu is away (iŽll be back later...) 1154461604 N * Belu Belu_zZz 1154462512 J * Pazzo ~thomas@host130-250-static.72-81-b.business.telecomitalia.it 1154463174 Q * michal` Ping timeout: 480 seconds 1154463683 J * michal` ~michal@www.rsbac.org 1154464181 Q * zkbrsnie 1154464536 J * Aiken ~james@tooax6-096.dialup.optusnet.com.au 1154465653 Q * Pazzo Quit: Ex-Chat 1154466131 J * mire ~mire@29-167-222-85.COOL.ADSL.VLine.verat.net 1154466149 Q * mire 1154466221 Q * dna Quit: Verlassend 1154468038 Q * bonbons Quit: Leaving 1154468955 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1154469916 J * stefani ~stefani@tsipoor.banerian.org 1154470324 M * Bertl well, I'm actually back now as it seems :( 1154470357 M * daniel_hozac what happened? 1154470400 M * Bertl coocoon: fortunately it seems that we have several sparcs available atm, so I do not _need_ it right now ... but that doesn't mean that it would not be appreciated if you could test on 'yours' too :) 1154470430 M * Bertl daniel_hozac: had some troubles with the 'new' networking, but I see a light at the end of the tunnel :) 1154470449 M * coocoon bertl: cool bertl i am back right now at this moment 1154470452 M * coocoon ;-= 1154470464 M * daniel_hozac hehe, that's good. 1154470480 M * coocoon bertl: no prob i will test it if there is somenthing to test 1154470518 M * Bertl coocoon: but I will get back to 'your' sparc sooner or later, I guess ... 1154470541 M * Bertl coocoon: so keep it warm :) and install the latest Linux-VServer kernels there 1154470564 M * coocoon i have installed the latest 2.6.17.7 patch 1154470568 M * coocoon all works fine 1154470591 M * Bertl excellent, so time for 2.1.1-rc27 then :) 1154470635 M * Bertl together with util-vserver and/or vserver-utils you should be able to have Linux-VServer up and running in no time ... 1154470653 M * coocoon bertl: have tried to install with debian guests but i do not know exactly how to do this because of the guests where not for sparc, thought that the installation routine will recognise the arch 1154470676 M * coocoon yes tomorrow i will install util-vserver and vserver-utils 1154470711 M * Bertl you can specify the arch on the command line IIRC, but by default it should be smart enough to get the same guests as your system is 1154470797 M * coocoon yes i thought so, hm will try it next time with debootstrap, it is also available @ gentoo 1154470811 M * coocoon so i will get a lot of guests i hope 1154470871 M * coocoon bertl: exactly i use patch-2.6.17.7-vs2.1.1-rc27 1154470881 M * daniel_hozac util-vserver will download debootstrap if it's not available. 1154470895 M * Bertl coocoon: excellent! 1154470897 M * coocoon ah ok 1154470906 M * coocoon yum is there also avalable 1154470921 M * coocoon i mean at gentoo 1154472385 J * gerrit gerrit@66.46.95.227 1154472411 M * Bertl wb gerrit! 1154472745 M * daniel_hozac Bertl: i've shown you http://daniel.hozac.com/vserver/delta-mountstats-fix02.diff already, right? 1154472979 M * daniel_hozac i'm going on vacation tomorrow for 10 days, so i'm making sure my patches are ok. :) 1154473431 J * chand ~chand@gw.net81-65-27.noos.fr 1154473690 A * MooingLemur yawns like a wookie. 1154474029 Q * chand Quit: chand 1154474055 P * stefani I'm Parting (the water) 1154475045 M * Bertl daniel_hozac: hmm, no, haven't seen that one (yet) 1154475073 M * Bertl daniel_hozac: if possible, send me a list of all aptches you consider important (via email) 1154475095 M * Bertl daniel_hozac: and make sure that you enjoy your vacation :) 1154475096 M * daniel_hozac that and http://daniel.hozac.com/vserver/delta-cow-feat04.diff are the only patches i have. 1154475107 M * daniel_hozac i'm sure i will :) 1154475166 M * daniel_hozac though COW would also need fs/namei.c:permission fixed to not return EACCES on write. 1154475242 M * daniel_hozac (but doener was working on auditing all of the IS_IMMUTABLE calls) 1154475510 M * Bertl ah, good idea ... 1154476114 M * coocoon good night 1154476119 Q * coocoon Quit: KVIrc 3.2.0 'Realia' 1154476591 M * Skram tarring a running vps is fine, right? 1154476598 M * Skram im pretty sure :) 1154476601 M * Skram O.o 1154476647 M * cehteh eh 1154476704 M * cehteh you need to be sure that tar does the right with special files (ok usually it does) and it depends on the programms which running inside 1154476726 M * Skram "does the right with special files"? 1154476728 M * cehteh taring a database which doesnt support such kinds of hot backups likely breaks 1154476734 M * Skram right.. 1154476743 M * daniel_hozac plus you'll probably have to clean up the guest to get it to start up properly later. 1154476749 J * KantankerousKid ~blah@tor-irc.dnsbl.oftc.net 1154476755 M * Skram daniel_hozac: what do you mean? 1154476762 M * cehteh yes 1154476774 M * Skram cehteh: mysql supports such.. do you know? 1154476796 M * cehteh there are some things to handle ... exclude /proc and /dev cleaning runfiles and so on