1230595507 Q * dowdle Remote host closed the connection 1230595998 J * darrob ~darrob@tor-irc.dnsbl.oftc.net 1230596128 Q * MooingLemur Read error: Operation timed out 1230596164 Q * duckx Ping timeout: 480 seconds 1230596174 J * duckx ~Duck@81.57.39.234 1230596350 J * Moo ~troy@shells195.pinchaser.com 1230596649 J * yarihm ~yarihm@77-56-182-18.dclient.hispeed.ch 1230597246 Q * yarihm Quit: Leaving 1230597471 Q * darrob Quit: WeeChat 0.2.6 1230597604 Q * Moo Read error: Operation timed out 1230597735 J * darrob ~darrob@tor-irc.dnsbl.oftc.net 1230597816 J * Moo ~troy@shells195.pinchaser.com 1230599609 Q * muxueqz Remote host closed the connection 1230600835 J * muxueqz ~muxueqz@123.116.202.157 1230601452 J * arthur ~arthur@pan.madism.org 1230602573 Q * bzed Remote host closed the connection 1230602596 J * bzed ~bzed@devel.recluse.de 1230604143 Q * FloodServ synthon.oftc.net services.oftc.net 1230604912 Q * muxueqz Remote host closed the connection 1230604920 J * muxueqz ~muxueqz@123.116.202.157 1230605516 J * balbir_ ~balbir@122.167.180.7 1230606104 Q * balbir_ Read error: Operation timed out 1230606296 Q * muxueqz Remote host closed the connection 1230606307 J * muxueqz ~muxueqz@123.116.202.157 1230607424 Q * muxueqz Remote host closed the connection 1230607433 J * muxueqz ~muxueqz@123.116.202.157 1230608114 Q * nou Ping timeout: 480 seconds 1230608398 Q * nenolod Ping timeout: 480 seconds 1230608739 J * nenolod nenolod@petrie.dereferenced.org 1230611801 M * Bertl off to bed now .. have a good one everyone! 1230611807 J * takeru ~takeru@nttkyo888227.tkyo.nt.ftth.ppp.infoweb.ne.jp 1230611808 N * Bertl Bertl_zZ 1230614063 J * balbir_ ~balbir@122.167.180.7 1230614357 Q * muxueqz Read error: Connection reset by peer 1230614413 J * muxueqz ~muxueqz@123.116.202.157 1230614649 Q * balbir_ Ping timeout: 480 seconds 1230616846 M * Supaplex m00 1230616984 Q * muxueqz Remote host closed the connection 1230617000 J * muxueqz ~muxueqz@123.116.202.157 1230618130 Q * takeru Quit: takeru 1230619469 J * balbir_ ~balbir@59.145.136.1 1230620216 J * takeru ~takeru@nttkyo888227.tkyo.nt.ftth.ppp.infoweb.ne.jp 1230620549 Q * muxueqz Remote host closed the connection 1230620557 J * muxueqz ~muxueqz@123.116.202.157 1230621153 J * doener_ ~doener@i577B8876.versanet.de 1230621253 Q * doener Ping timeout: 480 seconds 1230621324 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1230621337 Q * takeru Quit: takeru 1230621421 J * takeru ~takeru@nttkyo888227.tkyo.nt.ftth.ppp.infoweb.ne.jp 1230623021 J * ktwilight_ ~ktwilight@26.109-66-87.adsl-dyn.isp.belgacom.be 1230623289 Q * ktwilight__ Ping timeout: 480 seconds 1230624147 J * FloodServ services@services.oftc.net 1230625431 J * darrob2 debian-tor@tor-irc.dnsbl.oftc.net 1230625494 Q * darrob Ping timeout: 480 seconds 1230626654 J * davidkarban ~david@193.85.217.71 1230627669 Q * muxueqz Quit: Leaving. 1230627822 J * muxueqz ~muxueqz@123.115.140.157 1230628918 Q * esa Quit: Coyote finally caught me 1230629043 Q * muxueqz Quit: Leaving. 1230629192 Q * takeru Quit: takeru 1230629809 J * geb ~geb@AOrleans-151-1-72-104.w90-21.abo.wanadoo.fr 1230629872 J * dna ~dna@p54BCF5F1.dip.t-dialin.net 1230630173 M * geb hi 1230630957 M * Supaplex greetage 1230631674 Q * _nono_ Ping timeout: 480 seconds 1230631748 Q * pmenier Quit: Konversation terminated! 1230631964 Q * dna Ping timeout: 480 seconds 1230632410 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1230632519 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1230633016 J * _nono_ ~gomes@libation.ircam.fr 1230634109 M * yang Is it possible to use a ntpd on the guest itself, or is it always getting time from host only? 1230635494 M * pflanze yang: there is a kernel option (iirc) to have separate time in the guest 1230635513 M * pflanze (not sure why that would be useful though) 1230635589 M * yang just a question, becouse i had ntpd running on host and it couldn't set the time 1230635605 M * yang it automatically set it on all guests when running ntpd on host 1230635874 J * dna ~dna@p54BCD53F.dip.t-dialin.net 1230635969 Q * _nono_ Ping timeout: 480 seconds 1230636472 J * _nono_ ~gomes@libation.ircam.fr 1230636546 Q * dna Quit: Verlassend 1230636920 M * derjohn yang, if ntp cant set time on the host, i cant set it in the guest, too ... 1230636961 M * derjohn timekeeping in the guest costs some performance, as for "everything" has be calculated an offset 1230637029 Q * balbir_ Ping timeout: 480 seconds 1230637433 M * geb i am not sure that i understand that you want do but you can have an ntpd in a guest with SYS_TIME caps ( http://linux-vserver.org/Capabilities_and_Flags ) but i don't know any solution to have a different time in the guests and in the host 1230637470 N * Bertl_zZ Bertl 1230637474 M * Bertl morning folks! 1230637494 M * geb hi Bertl 1230637535 M * Bertl well, usually you want the guests to be in sync swith the host (which in turn is usually synced to a time source), as you still can have different time zones per guest 1230637573 M * Bertl but indeed there is a kernel option and flag, which allows to have a completely different time in a guest, even one with no relation to the host time at all 1230637619 M * Bertl still it doesn't make much sense to run ntpd in that guest, as there is only an offset recorded to the host time (so if that one is synced, the offset will remain constant) 1230637644 M * geb VIRT_TIME ? (just found it on the page) 1230637667 M * Bertl yep 1230637714 M * geb never used it 1230637754 M * Bertl doesn't make much sense, unless you have one guest living in the past or future :) 1230637798 M * geb coul be usefull for testing 1230637828 M * geb sending mail from futur/past to see howto filtering work for example 1230637832 M * geb is VIRT_UPTIME activated by default (the doc say no but it looks like in one of my installs) ? 1230637848 M * Bertl what doc? 1230637953 M * geb http://linux-vserver.org/util-vserver:Capabilities_and_Flags 1230638088 M * Bertl well, it says, in addition to the kernel default 1230638323 Q * Aiken Remote host closed the connection 1230638821 J * nou Chaton@2001:6f8:328:bbc:6666:6667:: 1230639144 Q * ztakeru Ping timeout: 480 seconds 1230639283 J * ztakeru ~ztakeru@s35.GtokyoFL11.vectant.ne.jp 1230640594 J * kir ~kir@swsoft-msk-nat.sw.ru 1230642154 M * Bertl daniel_hozac: did you have a look at the devcgroup stuff yet? from the description it looks like it is very similar to the device mapping stuff, except for the mapping part :) 1230642619 Q * bibabu Quit: Coyote finally caught me 1230642623 J * bibabu bip@vserv2.de 1230642737 J * SlackLnx ~Lee@bl7-130-51.dsl.telepac.pt 1230642903 Q * gnuk Quit: NoFeature 1230643244 Q * eyck_ Ping timeout: 480 seconds 1230643755 J * taruti taruti@aoi.yi.org 1230643805 M * taruti Is there a way to have the guests use dhcp to get IP addresses? i.e. have the host run dhcpd on a dummy interface and each guest ask for an IP over that? 1230643856 M * Bertl doesn't make much sense, as the guest(s) will use the same mac address 1230643887 M * Bertl but if you plan to put a dhcpd on the host, why not instead assign the IPs to the guests directly? 1230643935 M * Bertl (i.e. you have to put them into the dhcpd config anyway, so putting them into the guest config should be the same work, minus the rest of the dhcpd setup :) 1230643964 M * taruti Bertl: I want to avoid allocating static addresses for guests and keeping book of them. 1230643998 M * taruti Bertl: I have some scripts that set up vserver instances for testing and delete some of them occasionally. So maintaining addresses by hand sounds like pain. 1230644025 M * Bertl what addresses would the dhcpd hand out? 1230644047 J * hparker ~hparker@2001:470:1f0f:32c:212:f0ff:fe0f:6f86 1230644141 M * taruti Bertl: 172.19.19.0/24 1230644189 M * Bertl so, your script already assigns context IDs to the guests, I presume, why not use them for the IP too? 1230644236 M * taruti Bertl: currently it just omits --context from the vserver build command line. (Is there an issue that they should be managed manually?) 1230644268 M * Bertl well, depending on the util-vserver version, you might run into issues with that 1230644284 M * Bertl i.e. you should assign a specific context id to the guests 1230644337 M * Bertl but I guess you have at least some scheme to number/name the guests already, no? 1230644360 M * Bertl I mean, it doesn't make much sense to have guests with random names/ids, does it? 1230644377 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1230644426 M * taruti Bertl: actually it does :) 1230644459 M * Bertl okay, then just pick any IP randomly, check if it is used, and if so, repeat again 1230644483 M * Bertl (same for the context ids) 1230644510 M * taruti so rewrite the configuration before the vserver is started to get a free ip that does not conflict with already running ones? 1230644538 M * Bertl usually at creation time, but yes, that's the basic idea 1230644578 M * Bertl if you want to make that a little more efficient, keep two lists, one which has the assigned IPs and another one which has the free ones 1230644590 M * Bertl (and move the IPs between those lists) 1230644672 M * taruti that is kind of hacky, but I guess it can work 1230644695 M * Bertl it is - kind of - what dhcpd does :) 1230644765 M * taruti yes. although dhcp would also take care of e.g. nameserver information etc 1230644780 M * Bertl which is not relevant for the guests 1230644797 M * Bertl i.e. you'll have it set once for each, and it stays static 1230644836 M * taruti not really. 1230644845 M * Bertl how so? 1230644896 M * taruti e.g. if I am moving a guest from host A to host B they might use different name servers. (Of course it is possible to have each host just run a caching nameserver that the guests can use) 1230644896 M * Bertl if you configure a dhcpd address pool, you can only specify 'one' set of nameservers for all addresses 1230644932 M * Bertl you can simply wipe out/overwrite the resolv.conf for all guests in that case 1230644985 M * taruti just using dhcp would be cleaner that overwriting lots of files. 1230645004 M * Bertl well, you can do that too, but you have to hack that together :) 1230645021 M * taruti any tips where to start? :) 1230645043 M * Bertl you need to do the dhcp request on the host for each guest, and you need to specify an unique guest/client identifier so that you get a new address from your dhcpd 1230645111 M * Bertl the results you get (the assigned data) needs to be stored in the proper places (guest config, resolv.conf), the IP to be assigned, and then the guest can be started 1230645156 M * Bertl basically the same you have to do without dhcp(d) involvement, just that you put the IP management into the dhcpd 1230645176 M * taruti the guests couldn't do that by normal initscripts? (of course getting an unique MAC might need hacking) 1230645196 M * Bertl nope, as the guest is not allowed to mess with the network stack 1230645230 M * Bertl and dhcp is below the IP layer (naturally) and would require virtual interfaces (with mac addresses) to work 1230645236 M * taruti point. and thus unable to send dhcp 1230645266 M * Bertl which in turn, would require to do complete network stack virtualzation and bridging or something like that on all virtual interfaces 1230645305 M * Bertl (a lot of overhead and useless indirection just for the fun of having dhcp request an IP every 2 minutes :) 1230645503 M * taruti I think I'll just go with the file route instead. 1230645580 M * Bertl if you are not necessarily fixed to those IPs, you could correlate guest context ID and IP in some simple way 1230645610 M * Bertl e.g. use 172.19.x.y where x:y is the context id 1230645704 M * taruti yes, that could work. 1230645710 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com 1230647051 J * eyck yOwwdz9Y@nat05.nowanet.pl 1230647850 J * yarihm ~yarihm@hg-public-dock-76-dhcp.ethz.ch 1230648269 M * Bertl nap attack ... bbl 1230648276 N * Bertl Bertl_zZ 1230648299 P * taruti 1230650496 Q * m_o_d Read error: Connection reset by peer 1230651183 J * xdr ~xdr@gote2.247.cust.blixtvik.net 1230651561 J * xdr_ ~xdr@gote2.247.cust.blixtvik.net 1230651623 Q * xdr_ Remote host closed the connection 1230651669 Q * xdr Ping timeout: 480 seconds 1230652159 J * xdr ~xdr@gote2.247.cust.blixtvik.net 1230652409 J * xdr_ ~xdr@gote2.247.cust.blixtvik.net 1230652446 Q * xdr_ 1230652715 J * dowdle ~dowdle@scott.coe.montana.edu 1230652943 Q * davidkarban Quit: Ex-Chat 1230653682 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1230653768 J * hparker ~hparker@linux.homershut.net 1230654333 Q * hparker Quit: Quit 1230654339 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fe60:79d4 1230654496 J * esa ~esa@ip-87-238-2-45.static.adsl.cheapnet.it 1230654792 M * derjohn Bertl_zZ, daniel_hozac : did you notice the4 container freezer in 2.6.28 ? Any plans you make vserver guests fridge-ready ? 1230654901 M * daniel_hozac if it's in the kernel already, i don't see what we'd have to do. 1230655099 M * derjohn make the userspace tools aware of it, like "vserver xyz freeze". Or maybe something in /proc ..... 1230655132 M * derjohn the next step would be to move those frozen containers to other hosts .... 1230655152 M * derjohn *make moveable 1230655277 M * daniel_hozac well, it'd just be a echo FREEZE > /dev/cgroup//freezer.state, if i recall the API correctly. 1230655297 M * daniel_hozac but sure, we could put that in the utils. 1230655454 M * derjohn ah, there was already a "guest" in .... /dev ?! wow ... I think I lost track of the technology changes .... 1230655489 M * daniel_hozac if you mkdir /etc/vservers/.defaults/cgroup, yes. 1230655520 M * daniel_hozac the util-vserver initscript will mount it, and starting the guest will put it in its own cgroup. 1230656696 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1230656836 J * Piet ~piet@asteria.debian.or.at 1230657930 J * darrob ~darrob@tor-irc.dnsbl.oftc.net 1230658304 Q * darrob2 Ping timeout: 480 seconds 1230658717 J * Aiken ~Aiken@ppp118-208-75-168.lns1.bne4.internode.on.net 1230659175 J * cga ~weechat@94.36.82.206 1230660928 Q * kir Quit: Leaving. 1230661001 N * Bertl_zZ Bertl 1230661008 M * Bertl back now ... 1230661053 M * Bertl derjohn: if you try stuff like the freezer, please let us know ... 1230661091 M * Bertl (how it goes, I mean :) 1230661280 Q * pmenier Ping timeout: 480 seconds 1230661396 Q * pflanze Quit: Leaving 1230661637 Q * jrdnyquist Quit: Leaving 1230662696 Q * Piet Remote host closed the connection 1230664026 Q * SlackLnx Quit: I'll Be Back 1230664286 Q * geb Quit: Quitte 1230664572 Q * gnuk Quit: NoFeature 1230665750 Q * yarihm Quit: Leaving 1230666524 M * welldone hey 1230666547 M * welldone i would like to start dpkg on a vserver 1230666639 M * welldone i tryed "vsomething dpkg -- vservname -- -L perl", but this do not work 1230666654 M * welldone can someone tell me where i'm wrong ? 1230666677 M * derjohn Bertl, daniel_hozac : Is there a devel for .28 in sight? (The container freeze went just recently into mainline) 1230666718 M * derjohn vsomething --debug ? 1230666734 M * welldone tryed 1230666746 M * welldone couple of things are printed, then it "exit 255" 1230666820 M * daniel_hozac welldone: vserver exec dpkg -L perl 1230666847 M * welldone woot, that works 1230666850 M * welldone thanks again. =) 1230666879 M * derjohn then .... Usage: /usr/sbin/vsomething [--quiet|-q] [--debug] [--] at pos 1 ? 1230666930 M * daniel_hozac it's supposed to be. 1230666935 M * derjohn is not even [] (optional) 1230666942 M * daniel_hozac exactly. 1230666951 M * derjohn ahhhh# 1230666958 M * derjohn you used vserver not vsomething 1230666969 M * welldone :) 1230666994 M * daniel_hozac vsomething is for when you want to run something on a lot of guests. 1230667019 M * daniel_hozac e.g., vsomething vserver -- --all -- exec apt-get update will run apt-get update in all your guests. 1230667024 M * derjohn that the * after 1230667025 M * derjohn k. 1230667048 M * derjohn ? 1230667062 M * derjohn so, = vserver in that case ? 1230667080 M * daniel_hozac yes. 1230667084 M * Bertl derjohn: can't take that long, the Linux-VServer folks are usually quite fast :) 1230667129 M * derjohn in what way is that better than "for i in $(echo vserver1 vserver2); do vserver $i foo; done" ? 1230667144 M * derjohn Bertl, i've heard about that :) 1230667179 M * daniel_hozac derjohn: vsomething understands --all, --running, etc. 1230667198 M * derjohn ok, thats a point ... 1230667199 M * daniel_hozac it's a very simple script. 1230667201 M * welldone yeah, this is why i tryed to use it 1230667220 M * daniel_hozac for loops are obviously a valid replacement. 1230667221 M * welldone and thanks for the example, daniel_hozac 1230667227 M * derjohn ahhh vsomething: Bourne-Again shell script text executable 1230667243 M * derjohn so, thats wrapper more or less .... 1230667280 M * daniel_hozac most of the utils are bash wrappers... 1230667302 M * daniel_hozac vserver too. 1230667322 M * derjohn ah, I just checked that ;) 1230667495 M * derjohn thx, flks ;) 1230668579 J * jrdnyquist ~jrdnyquis@slayer.caro.net 1230668803 Q * ard Quit: kernel upgrade 1230669872 Q * hparker Quit: Quit 1230670084 J * ard ~ard@shell2.kwaak.net 1230670927 J * hparker ~hparker@2001:470:1f0f:32c:212:f0ff:fe0f:6f86 1230671353 J * darrob2 ~darrob@tor-irc.dnsbl.oftc.net 1230671399 Q * darrob Ping timeout: 480 seconds 1230673406 Q * bonbons Quit: Leaving 1230674943 J * yarihm ~yarihm@77-56-182-18.dclient.hispeed.ch 1230677187 J * kwowt ~quote@pomoc.ircnet.com 1230677188 M * kwowt hi 1230677194 M * Bertl hey 1230677205 M * kwowt i'm having some problems starting oidentd on guest 1230677232 M * Bertl what kind of problems? 1230677251 M * kwowt it doesnt even show up in processes 1230677254 M * kwowt like its not started 1230677266 M * kwowt Starting ident daemon: oidentd. 1230677267 M * Bertl maybe it is not started? :) 1230677294 M * kwowt well 1230677306 M * kwowt /var/log/messages doesnt say anything 1230677308 M * kwowt no errors 1230677315 M * kwowt or am i looking in the wrong place 1230677323 M * Bertl just means that it dies without logging anything useful, no? 1230677342 M * kwowt right 1230677345 M * Bertl try to start it manually, maybe with strace -fF 1230677416 M * kwowt hm 1230677424 M * kwowt can i paste you the output? 1230677430 M * daniel_hozac that's what pastebins are for. 1230677436 M * Bertl yes, but on some pastebin 1230677445 M * kwowt ofcourse 1230677487 M * kwowt http://rafb.net/p/MJZlug50.html 1230677489 M * kwowt this is the most of it 1230677492 M * kwowt the last part atleast 1230677551 M * Bertl bind(6, {sa_family=AF_INET, sin_port=htons(113), sin_addr=inet_addr("0.0.0.0")}, 16) = -1 EADDRINUSE (Address already in use) 1230677571 M * kwowt why is it binding to 0.0.0.0? 1230677584 M * Bertl that you have to ask the oidentd developers 1230677603 M * kwowt but this worked before i restarted the vps 1230677609 M * Bertl but in general that should be fine, as the guest is limited to guest IPs anyway 1230677629 M * Bertl unless you have one running on the host, binding to 0.0.0.0 :) 1230677709 M * kwowt idetnd? 1230677713 M * kwowt no 1230677742 M * Bertl well, looks like something is bound there (i.e. it overlaps with the guest IPs) 1230677760 M * daniel_hozac could be because you've assigned the same IP to two guests, and the other one is running identd. 1230677772 M * Bertl check with lsof -ni on guest and host (and maybe in the spectator context too) 1230677860 M * kwowt the same ip to two guests? 1230677898 M * Bertl would be a possibility ... yes 1230678131 M * kwowt oh 1230678132 M * kwowt thats it 1230678144 M * kwowt two ips are the same on both guests 1230678147 M * kwowt forgot to remove them 1230678157 M * Bertl hehe, daniel_hozac is good on guessing that stuff :) 1230678640 M * kwowt thanks 1230678640 M * kwowt ;) 1230678845 Q * Aiken synthon.oftc.net saturn.oftc.net 1230678845 Q * TimLyth synthon.oftc.net saturn.oftc.net 1230678859 J * TimLyth ~tux@202.134.227.227 1230678862 J * Aiken ~Aiken@ppp118-208-75-168.lns1.bne4.internode.on.net 1230679422 Q * cga Quit: WeeChat 0.2.6 1230679718 J * friendly ~friendly@ppp118-208-186-36.lns10.mel4.internode.on.net