1159920014 M * daniel_hozac so if you run mkdir /dev/cpuset/casa82it; echo 0 > /dev/cpuset/casa82it/cpus; echo $$ > /dev/cpuset/casa82it/tasks, nothing returns an error? 1159920018 M * ntrs_ Bertl, can I set per CPU sched parameters with vsched? Or, do you have some special tool for that? 1159920046 Q * meandtheshell Quit: exit (0); 1159920055 M * daniel_hozac ah yes, Bertl's vsched should be able to as well. 1159920066 Q * Blissex Ping timeout: 480 seconds 1159920075 M * Bertl ntrs_: latest vsched can do that 1159920076 M * ntrs_ I was able to execute that and no errors 1159920086 M * ntrs_ is 0.3 the latest one? 1159920111 M * Bertl seems so :) 1159920117 M * ntrs_ daniel_hozac, nothing returned an error 1159920122 M * daniel_hozac ntrs_: really? 1159920130 M * ntrs_ yes 1159920134 M * daniel_hozac how very strange... 1159920174 M * daniel_hozac i really don't see how that would happen then. 1159920181 M * ntrs_ the vsched I have here is the one from util-vserver 1159920241 M * ntrs_ daniel_hozac, I don't know either but I guess cpusets are not ready for production yet. 1159920252 M * daniel_hozac ntrs_: hmm, in scripts/vserver.functions:addtoCPUSET, could you try removing 2>/dev/null from the cat command? 1159920253 M * ntrs_ Bertl, where is the 0.3 version? 1159920258 M * daniel_hozac ntrs_: i'm using them in production already :) 1159920269 M * daniel_hozac i've been using them for months. 1159920285 M * ntrs_ ok, maybe they don't work on every system and on every setup. 1159920313 M * Bertl ntrs_: in the usual place: http://vserver.13thfloor.at/Experimental/TOOLS/vsched-0.03.tar.bz2 1159920332 M * ntrs_ daniel_hozac, where is scripts/vserver.functions 1159920333 M * ntrs_ ? 1159920344 M * daniel_hozac ntrs_: oh, /usr/lib*/util-vserver/vserver.functions 1159920396 M * ntrs_ daniel_hozac, removed 1159920404 M * daniel_hozac and if you start it now? 1159920430 M * daniel_hozac still just the warning? 1159920475 M * ntrs_ yes, still the warning 1159920493 M * ntrs_ # mkdir /dev/cpuset/casa82it; echo 0 > /dev/cpuset/casa82it/cpus; echo $$ > /dev/cpuset/casa82it/tasks# 1159920553 M * daniel_hozac and nothing else? 1159920563 M * ntrs_ yes 1159920600 M * daniel_hozac i don't get how that works. 1159920629 M * daniel_hozac everything that would cause the warning to be printed would output an error. 1159920843 M * ntrs_ daniel_hozac, there must be some special case where that does not happen 1159920871 M * daniel_hozac well, that function is very small, so it's quite easy to tell there are none... 1159920947 M * ntrs_ ok then. 1159920978 M * ntrs_ So, we have at least one case (host/guest) where cpusets do not work properly. 1159920987 M * ntrs_ for whatever reason 1159921030 M * ntrs_ Bertl, would it be a good test to try the vsched instead of cpusets? Or are cpusets more likely to resolve the mysql issue? 1159921034 M * daniel_hozac well, they workforme, and you really should be getting an error if you removed the 2>/dev/null from the cat command. 1159921050 M * ntrs_ daniel_hozac, I did remove it, believe me. 1159921115 M * daniel_hozac well, you have the function. what would cause it to output the warning, but no error message? 1159921117 M * ntrs_ this is what I have now 1159921119 M * ntrs_ read cpuset < "$f"/name 1159921119 M * ntrs_ test -e "$f"/nocreate || { 1159921119 M * ntrs_ test -d /dev/cpuset/"$cpuset" || mkdir /dev/cpuset/"$cpuset" || configured=1 1159921119 M * ntrs_ for i in cpus mems cpu_exclusive mem_exclusive virtualized; do 1159921119 M * ntrs_ if test -e "$f"/"$i"; then 1159921120 M * ntrs_ cat "$f"/"$i" >/dev/cpuset/"$cpuset"/"$i" || { 1159921122 M * ntrs_ configured=1 1159921124 M * ntrs_ break 1159921126 M * ntrs_ } 1159921132 M * ntrs_ fi 1159921134 M * ntrs_ done 1159921136 M * ntrs_ } 1159921142 M * daniel_hozac paste.linux-vserver.org, please ;) 1159921191 M * daniel_hozac that looks fine. 1159921254 M * ntrs_ paste.linux-vserver.org for 10 lines? really? 1159921263 M * Bertl ntrs_: the vsched will not help here, as it is a probability mechanism 1159921264 M * daniel_hozac anything over 3. 1159921280 M * ntrs_ Bertl, ok, I will not try vsched. 1159921289 M * ntrs_ daniel_hozac, oh, come on. 1159921360 M * ntrs_ Bertl, we will have to wait some better time to try to debug the mysql issue as cpusets just don't work on the host. 1159921454 M * matled ntrs_: you can use this as a paste script :) 1159921460 M * matled curl -fsSD - -F 'content=<-' http://pbot.rmdir.de/ | grep '^Location: ' 1159921607 M * Bertl ntrs_: what's the problem with cpu sets? 1159921628 M * ntrs_ they just do not work and daniel_hozac is out of ideas currently, so we are stuck. 1159921652 M * daniel_hozac not out of ideas, out of possibilities :) 1159921659 M * ntrs_ read the 50-100 lines above, the discussion between me and daniel_hozac 1159921675 M * ntrs_ Ok, sorry, he is out of possibilities. 1159921689 M * Bertl have you double checked the tools? 1159921695 M * Bertl I mean the cpuset utilities? 1159921701 M * daniel_hozac there are none. 1159921711 M * daniel_hozac it's managed by a pseudo filesystem. 1159921718 M * daniel_hozac s/by/through/ 1159921719 M * ntrs_ the mount is there 1159921722 M * Bertl hmm, so that changed too? 1159921728 M * ntrs_ cpuset /dev/cpuset cpuset rw 0 0 1159921735 M * ntrs_ that is from cat /proc/mounts 1159921738 M * Bertl I remember pcreate and friends 1159921764 M * daniel_hozac well, it's been a pseudo filesystem at least since baggins wrote the original patch :) 1159921814 M * Bertl okay, and it works for you daniel_hozac? 1159921828 M * daniel_hozac yep. 1159921843 M * daniel_hozac even works for ntrs_ when executing the commands manually. 1159921880 M * Bertl ntrs_: what util-vserver version do you have there? did you remove the old tools? 1159921914 M * ntrs_ # vserver --version 1159921914 M * ntrs_ vserver 0.30.211-rc2 -- manages the state of vservers 1159921914 M * ntrs_ This program is part of util-vserver 0.30.211-rc2 1159921927 M * ntrs_ was installed over the old tools from an rpm package I created 1159921990 M * Bertl over means with -i or -U ? 1159922007 M * ntrs_ -U 1159922061 M * Bertl have we tried a --debug run yet? 1159922080 M * ntrs_ no 1159922084 M * Bertl maybe upload that to paste.linux-vserver.org for further investigations 1159922093 M * ntrs_ ok 1159922102 M * ntrs_ a vserver restart with --debug? 1159922133 M * daniel_hozac vserver stop && vserver --debug start is probably better. 1159922152 M * ntrs_ ok 1159922194 J * oezgueng ~oezgueng@p54A23B33.dip0.t-ipconnect.de 1159922202 M * Bertl welcome oezgueng! 1159922212 M * oezgueng ahoi ;] 1159922279 M * oezgueng hat jemand das openVCP Webinterface am laufen? 1159922330 M * Bertl channel language is english ... 1159922332 M * doener english please, otherwise 90% of the channel won't understand you ;) 1159922353 M * Bertl oezgueng: yes, as far as I know some folks use/test it 1159922361 M * daniel_hozac do you really think it's as high as 90%? sometimes i feel like the only one not speaking german :) 1159922387 M * oezgueng ups, i'm sorry, have anyone has someone openvcp the webinterface to run? 1159922393 M * Nei there really seems to be a large number of german speaking people floating round here~~ 1159922415 M * ntrs_ Bertl, daniel_hozac http://paste.linux-vserver.org/431 1159922424 M * oezgueng ah Bertl, i have problem with the Daemon from openVCP, he says, the port is already in use, which is not correct 1159922439 M * daniel_hozac ntrs_: aaaah, this is why i had mems specified :) 1159922451 M * daniel_hozac ntrs_: you have to set mems and cpus. 1159922457 M * doener oezgueng: did you check in context 1? 1159922464 M * oezgueng http://paste.linux-vserver.org/432 1159922468 M * oezgueng whaT? 1159922505 M * daniel_hozac ntrs_: i get an error though, if i remove it. 1159922515 M * doener oezgueng: chcontext --xid 1 netstat -tunlp 1159922525 M * doener check in that output that the port is unused 1159922543 M * oezgueng doener: http://paste.linux-vserver.org/433 1159922578 M * doener woah, _nothing_ is listening? 1159922586 M * ntrs_ daniel_hozac, set it to 0? 1159922588 M * oezgueng exaktly 1159922588 M * daniel_hozac yeah, that seems very strange. 1159922591 M * daniel_hozac ntrs_: yep. 1159922599 M * doener oezgueng: could you get me an strace of the failing daemon? 1159922608 M * oezgueng strace openvcpd? 1159922621 M * oezgueng which command? 1159922622 M * doener strace -fF openvcpd 1159922630 M * oezgueng moment ;) 1159922662 M * oezgueng hmm its many many lines 1159922680 M * oezgueng http://paste.linux-vserver.org/434 <- the last lines 1159922689 M * ntrs_ daniel_hozac, ok no warning now 1159922713 M * ntrs_ daniel_hozac, I guess there were other possibilities :) 1159922722 M * ntrs_ it is working fine now 1159922733 M * ntrs_ I can see that sched says 0 0 0 0 for cpu1 1159922743 M * ntrs_ only cpu0 has values greater than 0 1159922750 M * ntrs_ Now we wait. 1159922760 M * daniel_hozac ntrs_: well, only if you consider bash not outputting an error on echo > a possibility. 1159922782 M * Bertl excellent work, daniel_hozac, ntrs_! 1159922783 M * ntrs_ no, I meant that mems was required 1159922793 M * ntrs_ Thanks Bertl 1159922817 M * ntrs_ so, at least we now know that mems is required. 1159922823 M * daniel_hozac ntrs_: yes, but a missing mems outputs /usr/lib/util-vserver/vserver.functions: line 726: echo: write error: No space left on device 1159922838 J * Piet hiddenserv@tor.noreply.org 1159922858 M * ntrs_ daniel_hozac, I don't know much about bash scripting. I do believe you. 1159922871 M * daniel_hozac it's not a bash scripting thing, it's a kernel thing :) 1159922885 M * ntrs_ but you also have to believe me. everything I said above was true. There was no error output. 1159922890 M * ntrs_ I know it's strange. 1159922893 M * doener Bertl: hm, does that sa_data in oezgueng's paste look sane to you? 1159922924 M * daniel_hozac doener: and why would it be AF_UNSPEC? 1159922949 M * Bertl doener: nope, also the UNSPEC is strange 1159922988 M * doener yeah 1159923009 M * doener oezgueng: maybe your missing some config and the daemon is stupid about such cases? 1159923021 M * Bertl but I don't know much about the openvcpd ... 1159923031 M * oezgueng doener: http://paste.linux-vserver.org/436 1159923137 M * doener hm, their svn access is pretty broken... checks out an "interesting" trunk directory with no real content 1159923151 M * ntrs_ daniel_hozac, if I want to use cpu 1 do I still set mems to 0? 1159923172 M * daniel_hozac ntrs_: yes, unless you have multiple memory nodes. 1159923186 M * ntrs_ I don't even know what memory nodes are. 1159923190 M * daniel_hozac AFAIK there's no real relation between the memory nodes and the CPUs. 1159923224 M * Bertl well, that is mainly for NUMA archs 1159923242 M * Bertl a dual AMD for example would have memory per CPU 1159923255 M * Bertl memory on the 'other' cpu can be considered far memory 1159923267 M * Bertl and will occur a certain overhead 1159923281 M * daniel_hozac ah. 1159923307 M * Bertl sparc machines with 16 cpus are even more fun :) 1159923404 M * oezgueng http://paste.linux-vserver.org/439 hmm 1159923459 M * doener 1024 CPU SGI Titanic^WItanium boxes are probably even more fun ;) 1159923490 M * romke lol 1159923597 M * doener I just had to think about the amount of NUMA-related comments from Christoph Lameter on lkml I stumbled upon (he works at sgi) 1159923637 M * Hurga sgi still exists? 1159923668 M * doener isn't the new super computer in either Aachen built by sgi? 1159923675 M * doener s/either/ 1159923772 J * meandtheshell ~markus@85-124-36-225.dynamic.xdsl-line.inode.at 1159923778 M * daniel_hozac oezgueng: is one of your IP addresses 192.168.1.3? 1159923796 M * oezgueng no i hase edited to the ip from the hostmaschine 1159923890 M * doener hm, and maybe s/Aachen/Munich/... *memory is broken* 1159923931 M * oezgueng daniel_hozac:http://paste.linux-vserver.org/436 1159923976 M * doener oezgueng: did you compile the daemon yourself? 1159924029 M * oezgueng doener: jepp 1159924153 M * doener oezgueng: in main.c, line 151 could you add this line? 1159924165 M * oezgueng which lnie? 1159924168 M * oezgueng *line 1159924168 M * doener server_addr.sin_family = AF_INET; 1159924172 M * oezgueng mom 1159924190 M * doener right below the server_socket = socket(PF_INET...) call 1159924220 M * oezgueng http://paste.linux-vserver.org/441 <- so? 1159924234 M * doener yep 1159924303 M * oezgueng and now? i have compiled 1159924311 M * doener does it still fail? 1159924324 M * oezgueng openvcpd[9710]: Port already in use 1159924344 M * doener ok, could you do an strace again? strace -e bind -fF openvcpd 1159924378 M * oezgueng http://paste.linux-vserver.org/442 1159924406 M * daniel_hozac now why is it trying to bind 192.168.1.3:6000? 1159924435 M * oezgueng IP=193.254.203.105 1159924436 M * oezgueng Port=1331 1159924441 M * oezgueng its in the config , not 192.168.1.3 1159924456 M * daniel_hozac are you sure it's using that config? could you upload the entire trace somewhere? 1159924495 M * oezgueng http://paste.linux-vserver.org/443 <- 1159924547 M * doener oezgueng: are you sure that it uses the one in the cwd, not sth. like /etc/openvcpd.conf? 1159924601 M * doener "Port already in use" is a stupid message nevertheless... 1159924608 M * oezgueng server1:~/vServerHost/openvcpd-0.2-rc2# ln -s openvcpd.conf /etc/openvcpd.conf 1159924609 M * oezgueng xD 1159924692 M * doener #ifndef OPENVCP_CONFIG 1159924692 M * doener #define "/etc/openvcpd.conf" 1159924694 M * doener #endif 1159924700 M * doener hu? 1159924722 M * oezgueng ? 1159924724 M * doener in config.h... is it just me or does that look broken? 1159924762 M * oezgueng #ifndef OPENVCP_CONFIG 1159924762 M * oezgueng #define "/etc/openvcpd.conf" 1159924762 M * oezgueng #endif 1159924777 M * Bertl yes, that looks broken :) 1159924790 M * Bertl probably wants to become: 1159924793 M * doener oezgueng: add OPENVCP_CONFIG between #define and the path 1159924798 M * Bertl #ifndef OPENVCP_CONFIG 1159924805 M * Bertl #define OPENVCP_CONFIG "/etc/openvcpd.conf" 1159924807 M * oezgueng wha? o.O 1159924807 M * doener then do a full rebuild, incl. configure (just to be sure) 1159924812 M * Bertl #endif 1159924852 M * doener Bertl: does cpp recognize stuff that does not start at column 1 at all? 1159924881 M * doener I have no idea, but when I've seen indentation of cpp stuff, the # stayed at col 1 1159924883 M * oezgueng i have a create symbolic link to the config ?o.O 1159924895 M * Bertl doener: I'd say so 1159924914 M * doener oezgueng: in config.h change the broken part, just like Bertl suggested 1159924915 M * Bertl doener: whitespace is stripped very early 1159924920 M * doener then do a full rebuild 1159924933 M * doener and finally put your config into /etc if you didn't do so already 1159924945 M * doener (a symlink in /etc to some random path doesn't look sane) 1159924960 M * oezgueng how now, don't understand so completely, what I am to make? 1159925008 M * doener Bertl: could you try to explain it? 1159925072 M * oezgueng waaaa my pizza o.O 1159925126 M * Bertl oezgueng: basically change the following line: 1159925135 M * Bertl #define "/etc/openvcpd.conf" 1159925138 M * Bertl to this: 1159925148 M * Bertl #define OPENVCP_CONFIG "/etc/openvcpd.conf" 1159925158 M * Bertl then recompile 1159925260 Q * duckx Ping timeout: 480 seconds 1159925690 M * oezgueng ok mom 1159925779 M * oezgueng aaaaaaaaaaaaaaaaaaaaaaah 1159925891 Q * meandtheshell Quit: exit (0); 1159926067 J * _node node@c-69-143-148-254.hsd1.md.comcast.net 1159926508 M * doener cryptronic: hm, how does your svn repository work? trunk is empty but seems to have two branches in it. Is there any "manual" for your repository layout? 1159926581 M * oezgueng tcp 0 0 server1.hosterplus.:x11 *:* LISTEN 15681/openvcpd 1159926583 M * oezgueng what ist x11? 1159926600 M * oezgueng Host1 193.254.203.105 0.00 GB Online No 1159926600 M * oezgueng 0.2 rc2 1159926600 M * oezgueng 1159926668 J * duckx ~Duck@tox.dyndns.org 1159927175 M * Bertl x11 is the name for the port (6000) 1159927238 M * oezgueng hmm ok 1159927863 Q * FireEgl Read error: Connection reset by peer 1159928046 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1159928057 J * ensc ~irc-ensc@p54B4D708.dip.t-dialin.net 1159928381 Q * Hurga Remote host closed the connection 1159928515 Q * s0undt3ch Ping timeout: 480 seconds 1159928889 J * s0undt3ch ~s0undt3ch@bl7-242-159.dsl.telepac.pt 1159930307 J * Johnnie ~jdlewis@jdlewis.org 1159930814 Q * Johnnie Quit: G'bye! 1159930926 Q * Piet Quit: :tiuQ 1159930938 Q * patulo 1159931365 J * Aiken__ ~james@tooax6-210.dialup.optusnet.com.au 1159931691 Q * Aiken Ping timeout: 480 seconds 1159934880 Q * _node Ping timeout: 480 seconds 1159934975 M * Nei is port 6001 named x12 then :>? 1159934991 Q * oezgueng Ping timeout: 480 seconds 1159939193 J * Johnnie ~jdlewis@216.75.24.218 1159940406 Q * Johnnie Remote host closed the connection 1159940543 M * eyck why would it? it's called x11-2 1159943097 Q * essobi_ Read error: Connection reset by peer 1159943612 M * phedny linux-vserver.org offline? 1159943731 M * phedny Nei: look at /etc/services, there are all names ;) 1159943977 Q * virtuoso Ping timeout: 480 seconds 1159944299 M * Bertl phedny: hmm? 1159944345 M * Zaki morning Bertl :) 1159944361 M * Bertl hey Zaki! I'm almost off to bed :) 1159944368 M * Zaki oops 1159944373 M * Zaki you didn't sleep last night? 1159944386 M * Bertl I didn't sleep yet :) 1159944421 M * Zaki oh hmm 1159944453 M * phedny Bertl: "Firefox can't find the server at www.linux-vserver.org." 1159944505 M * Bertl works fine here .. checking nameservices now 1159944547 M * matled running ejabberd -detach from vserver exec/suexec/start fails, ejabberd does not seem to manage to detach, using a small wrapper program doing fork+setsid helps, is there any other way to do this I miss at the moment? and would it perhaps make sense to add an setsid to exec/suexec/start/..? 1159944555 M * phedny ;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 32858 1159944563 M * phedny ;; AUTHORITY SECTION: 1159944563 M * phedny org. 60 IN SOA tld1.ultradns.net. domadmin.ultradns.net. 2006207083 1800 900 604800 86400 1159944585 M * Bertl well, it seems that all but two nameservers are broken :( 1159944641 M * Bertl and the two main nameservers are not updated by jacques 1159944677 M * Bertl mugwump: IIRC, you have some contact to jacques besides the email (he doesn't read)? 1159944885 M * phedny also, it seems the vserver-build(8) man page is not available on Ubuntu, while vserver(8) en eg. vserver-copy(8) are 1159945269 J * Johnnie ~jdlewis@216.75.24.218 1159945430 M * Loki|muh the ubuntu packet is broken afaik 1159945453 J * virtuoso ~s0t0na@shisha.spb.ru 1159945488 M * Loki|muh phedny: https://launchpad.net/distros/ubuntu/+source/util-vserver/+bug/50545 seems that they do not care 1159945538 Q * Johnnie 1159945653 M * derjohn2 is there a default limit that limits the disk to 30 GB ? 1159945700 M * derjohn2 my guest seems full (hdv1) but the hdd has 450 GB free ......, df -h within the guest says 30 of 30 GB used 1159945729 M * Loki|muh quota-stuff? 1159945789 J * Johnnie ~jdlewis@216.75.24.218 1159945830 M * Loki|muh ah, they use the newer versionfor edgy 1159945834 M * Bertl maybe somebody can check the following, as I'm obviously too tired to understand it ... 1159945847 M * Loki|muh too bad that it is broken for it, too 1159945937 M * Bertl http://paste.linux-vserver.org/444 1159945995 M * Bertl I'm finally off to bed now .. have a nice one everyone! 1159946002 N * Bertl Bertl_zZ 1159946161 M * Loki|muh hmmm who is Jacques Gelinas? 1159946413 Q * Johnnie Remote host closed the connection 1159946559 M * Zaki Bertl_zZ, g'nite 1159946666 M * sid3windr I can't since the dns is bork :p 1159946693 M * sid3windr Bertl_zZ: if you need an extra nameserver let me know :) 1159947176 J * Johnnie ~jdlewis@216.75.24.218 1159947198 M * nayco_work hello, all ! 1159947560 J * dna_ ~naucki@p54BCE1A7.dip.t-dialin.net 1159947787 Q * ntrs_ Ping timeout: 480 seconds 1159948203 Q * shedi Quit: Leaving 1159948763 M * matti Morning. 1159948824 Q * Aiken__ Ping timeout: 480 seconds 1159949181 Q * MrX Remote host closed the connection 1159949699 M * mugwump Loki|muh: Jacques is the original inventor of early ip jail and vserver patch, as well as the 'vserver' userland utilities 1159949729 M * mugwump Bertl_zZ: I once had his postal address, but that was in Easyspace days :) 1159950442 J * dna___ ~naucki@p54BCF429.dip.t-dialin.net 1159950841 Q * dna_ Ping timeout: 480 seconds 1159951161 M * Dimmu mmm, why does getxattr in vserver sometimes returns EINVAL ? 1159951167 M * Dimmu on nfs that is 1159951253 M * Dimmu on physical server it returns ENODATA on the nfs volumes with attribute support, and ENOTSUP on nfs volumes that don't have attribute support 1159951263 M * Dimmu within vserver it sometimes returns ENODATA, and sometimes EINVAL :) 1159951291 M * Dimmu or ENOTSUP :) 1159951564 J * yarihm ~yarihm@whitehead2.nine.ch 1159951589 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1159952140 J * lilalinux ~plasma@dslb-084-058-197-179.pools.arcor-ip.net 1159952471 Q * nayco_work Quit: "gonna install Mdv 2007" 1159952759 J * meandtheshell ~markus@85-124-36-88.dynamic.xdsl-line.inode.at 1159953404 J * MrX ~urk@219.95.24.141 1159953946 M * Loki|muh hmmm, thats nice, vserver guest are working without packages initscripts and sysvinit 1159954035 M * Loki|muh ups, a purge of initscripts was not a very good idea *g* 1159955051 Q * glut Read error: Connection reset by peer 1159955327 J * shedi ~siggi@dsl-149-109-85.hive.is 1159955502 Q * mire Quit: Leaving 1159955699 M * Borg- Loki|muh: sure :) my rc starts just inetd :> and everything works 1159956147 M * Loki|muh hehe 1159957139 J * mire ~mire@85.222.167.76 1159957813 M * meandtheshell hi there - is www.linux-vserver.org down? 1159957818 M * meandtheshell hm ... 1159957837 M * Loki|muh yeah, there is a problem with the dns 1159957844 M * Loki|muh 09:10 < Bertl> maybe somebody can check the following, as I'm obviously too tired to understand it ... 1159957848 M * Loki|muh |>09:12 < Bertl> http://paste.linux-vserver.org/444 1159957848 M * Loki|muh |>09:13 < Bertl> I'm finally off to bed now .. have a nice one everyone! 1159957851 M * meandtheshell I see 1159958302 M * Borg- WTF?! 1159958304 M * Borg- domain expired? 1159958367 M * Borg- ah.. not.. just seriours DNS problems/ 1159958612 J * prae ~Benjamin@5-63.206-83.static-ip.oleane.fr 1159959176 M * daniel_hozac phedny: there is no vserver-build(8) man page. 1159959202 M * phedny then I think vserver(8) should be updated 1159959203 M * phedny SEE ALSO vserver-build(8), vserver-copy(8), vserver-stat(8), vtop(8), vps(8), vpstree(8), chcontext(8), chbind(8), reducecap(8) 1159959281 M * daniel_hozac _all_ of the man pages are at best just outdated. 1159959339 J * Ben81 ~Ben81@tipi0e.lri.fr 1159959342 M * phedny I just needed them as the site is not available ;) 1159959368 M * daniel_hozac vserver - build --help has all the info you need, no? 1159959401 M * phedny yeah, I did find it later on 1159959421 M * phedny and I succeeded building the vserver, but it was a little trail-and-error and remembering from last time ;) 1159959804 J * yang ~yang@cpe-213-157-253-172.dynamic.amis.net 1159959846 M * phreak`` hrm daniel_hozac Bertl_zZ: generic question, is there any known weirdness with FIFO's within a vps ? 1159959872 M * daniel_hozac there shouldn't be. 1159959939 M * phreak`` because I'm currently trying baselayout-1.13 within a vps and it should _just_ work (with some hacks/fixes), but it seems to suffer the same issues as FreeBSD. 1159959952 M * daniel_hozac hmm? 1159960104 M * phreak`` daniel_hozac: http://sources.gentoo.org/viewcvs.py/baselayout/trunk/sh/rc-services.sh?view=markup -> wait_service() { (should be ~ line 396). 1159960135 M * phreak`` but if you say, there are no known issues, then I'll need to look deeper into the code :) 1159960183 M * daniel_hozac that looks fine to me, looks a lot like what util-vserver is using :) 1159960353 M * daniel_hozac so does that wait too long or not long enough? 1159960845 M * phreak`` without the "fix" (just a '|| is_vps_sys') it doesn't wait long enough, and you'll see the timeout msg (from util-vserver that is) with some init-scripts. with the fix, you see the timeout but without anything (just the message from util-vserver without any content) (see https://bugs.gentoo.org/show_bug.cgi?id=149897) 1159962294 M * Dimmu So I decided to dump NFS to run my vservers on :) 1159962628 J * SEAwolfx ~deepsea@p54946E3C.dip.t-dialin.net 1159962716 J * Wido ~wido@kantoor-amersfoort.pcextreme.nl 1159962719 M * Wido hello 1159962753 M * Wido i am testing around a little with the pre and post scripts 1159962757 M * Wido like prepre-start 1159962774 M * Wido but they don't work as i would like to 1159962793 M * SEAwolfx hi, in the vserver-context the amount of processes seems to be limited. the error message " /etc/profile:57: fork failed: resource temporarily unavailable" occures, when for instance too much bashes are started. where in the vserver-config is the limit settable (like with 'ulimit -u')? 1159962841 M * Wido in /etc/vservers//limits 1159962845 M * Wido see: http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1159962855 M * Wido "Possible resources are cpu, fsize, data, stack, core, rss, nproc, nofile, memlock, as and locks. This configuration will be honored for kernel 2.6 only." 1159962912 M * Wido now, my problem is 1159962917 M * SEAwolfx Wido: will check this, thanks so far. 1159962919 M * Wido i have a prepre-start 1159962921 M * Wido #!/bin/sh 1159962921 M * Wido DEV="/dev/disk/by-path/pci-platform-scsi-0:0:0:6-part1" 1159962921 M * Wido mount "$DEV" /var/lib/vservers/reseller14 1159962930 M * Wido this mounts the correct iSCSI target 1159962945 M * Wido when i run vserver start 1159962946 M * Wido i het 1159962947 M * Wido i get 1159962952 M * Wido "No command given; use '--help' for more information. 1159962952 M * Wido An error occured while executing the vserver startup sequence; when 1159962952 M * Wido there are no other messages, it is very likely that the init-script 1159962952 M * Wido () failed. 1159963022 M * SEAwolfx Wido: As this file isnt existant in our vserver-config, is there a default limit to nproc? 1159963049 M * Wido you should create this directory 1159963050 M * Wido and files 1159963075 M * SEAwolfx Wido: ok, will check this too.. 1159963908 M * Borg- hmm 1159963920 M * Borg- [root@test1] iptables -n -L 1159963920 M * Borg- iptables v1.3.5: can't initialize iptables table `filter': Table does not exist (do you need to insmod?) 1159963924 M * Borg- Perhaps iptables or your kernel needs to be upgraded. 1159963924 M * Borg- what could be wrong? 1159963930 M * Wido inside a VPS 1159963935 M * Wido VServer? 1159963938 M * Borg- yeah. Guest 1159963942 M * Wido won't work 1159963944 M * Borg- no host it works 1159963955 M * Borg- ah.. so no iptables on guest? I'll have to do all stuff in host? 1159963955 M * Wido you should firewall on the host 1159963959 M * Borg- oki 1159964264 J * bogues bogues@172.20.149.50 1159964436 M * Wido there are no people with experience with prepre-start scripts? 1159964441 M * Wido ? 1159964560 P * bogues 1159964955 M * daniel_hozac Wido: why aren't you using the guest's fstab for that? 1159964972 M * Wido mount it as /? 1159964978 M * Wido in the fstab? 1159964980 M * daniel_hozac yes. 1159964996 M * Wido i didn't know that was possible 1159965060 M * Wido well, that raises another error 1159965064 M * Wido "vlogin: openpty(): Permission denied" 1159965110 M * daniel_hozac you'll have to specify ,dev in the options for the root mount. 1159965120 M * Wido ok 1159965188 M * phedny does vserver start boot the 2 or 3 runlevel? 1159965197 M * daniel_hozac 3 by default. 1159965214 M * phedny okay 1159965230 M * Wido daniel, adding dev to the fstab solves the error, but raises the next one 1159965232 M * phedny and does vserver stop switch to 0 or 6, since the guest makes messages like "Rebooting..." 1159965237 M * daniel_hozac 6. 1159965240 M * Wido in the fstab i specified ext3 as the filesystem 1159965240 M * Dimmu daniel_hozac: I got rid of that acl related issues by replacing NFS by FiberChannel ;) 1159965247 M * Wido wich it is 1159965252 M * daniel_hozac Dimmu: there you go then :) 1159965258 M * Wido Dimmu, i am also switching from NFS to iSCSI 1159965265 M * Dimmu NFS is giving me to much issues :) 1159965267 M * Wido NFS gives to much troubles 1159965279 M * daniel_hozac phreak``: so you're sure something is writing to it? you don't get any errors anywhere? 1159965281 M * Dimmu locking, EINVAL on getxattr syscall 1159965287 M * Wido correct 1159965296 M * Dimmu FC made all those problems go away :) 1159965303 M * Wido iSCSI also did :) 1159965316 M * Wido but now i want to automate the mount 1159965320 M * Wido wich does not work :P 1159965335 M * Dimmu I mount an FC volume per physical server outside the virtuals 1159965349 M * Wido all my guest have their own volume 1159965357 M * Wido so i can mount them on another host 1159965360 M * Wido and start them there 1159965371 M * Wido my /etc/vservers is still shared through NFS 1159965374 M * Dimmu No need here :) 1159965376 M * Dimmu ohyeah 1159965382 M * Dimmu my physical servers rootfs is still nfs :) 1159965387 M * Wido so i can make a failover with heartbeat 1159965435 M * daniel_hozac Wido: you can't do failover if you have all the guests on the same volume? 1159965445 M * Wido all guest have their own volume 1159965460 M * Wido wich i mount under /var/lib/vservers/guest 1159965468 M * Wido replace guest by the name of the guest 1159965471 M * daniel_hozac which means you can't use hashification to save RAM and disk :) 1159965485 M * Wido those are not my biggest issues 1159965491 M * Wido i want high availability 1159965498 M * Wido and failover 1159965504 M * Wido fast recovery 1159965528 M * Wido but still, why won't a mount in the prepre-start work? 1159965582 M * daniel_hozac it would, it's just cleaner to use the infrastructure already present, IMHO... 1159965592 M * Wido ok 1159965598 M * Wido ah, stupid me 1159965604 M * Wido i used hard as a mount option 1159965606 M * Wido ofcourse 1159965611 M * Wido only for NFS :/ 1159965619 M * Wido have been working to much with NFS lately 1159965640 M * Dimmu I have been irritated by NFS too much lately :) 1159965655 M * Wido well, NFS is not the biggest issue 1159965660 M * Wido it is the Linux kernel 1159965704 M * Dimmu Linux's NFS implementation yes :) 1159965714 M * Wido BSD is much better 1159965716 M * Dimmu client aswell server :) 1159965718 M * Wido and Solaris 1159965729 M * Wido What did you use as NFS server? 1159965760 M * Dimmu Linux for starters, but it gave me too much issue, so I tried or old rusty Netapp F720, it solved some issues, but not all 1159965775 M * Wido i used a EMC Cellera 1159965780 M * Wido works fine 1159965787 M * Wido but when you start pushing the kernel 1159965790 M * Wido with a lot of locks 1159965793 M * Dimmu jup 1159965797 M * Wido you get weird problemes 1159965799 M * Dimmu locks are my biggest iessue :) 1159965802 M * Wido no real errors 1159965809 M * Wido but weird system hangups 1159965809 M * Dimmu samba in a virtual server just behaves weird 1159965814 M * Dimmu on NFS 1159965817 M * Wido jes 1159965818 M * Wido yes 1159965834 M * Wido daniel_hozac, where does this come from? "No command given; use '--help' for more information." 1159965836 M * Curus samba on NFS can't really work well 1159965845 M * Wido vshelper? 1159965850 M * daniel_hozac Wido: probably from vcontext. 1159965862 M * Wido i modified my fstab 1159965864 M * daniel_hozac it most likely means you don't have a /etc/{rc,init}.d/rc script. 1159965864 M * Dimmu Samba's biggest issue on NFS is that it doesn't use NFS supported locks on it's tdb files 1159965873 M * daniel_hozac (if you get it during start) 1159965876 M * Wido starting goes well 1159965885 M * Wido stopping raises that 1159965892 M * Wido "A timeout occured while waiting for the vserver to finish and it will" 1159965913 M * Wido for the record, my scripts dir is gone 1159965933 M * phedny why does this line appear after stopping a vserver? 1159965940 M * phedny /usr/lib/util-vserver/vserver.stop: line 85: 17957 Gedood "${NICE_CMD[@]}" ${USE_VNAMESPACE:+$_VNAMESPACE --enter "$S_CONTEXT" -- } $_VCONTEXT $SILENT_OPT --migrate --chroot --xid "$S_CONTEXT" -- "${INITCMD_STOP[@]}" 1159965948 M * phreak`` daniel_hozac: nothing at all (not on the tty nor in the log), thats why I'm that much curious about it (well except the util-vserver error-msg that in the bug) 1159965952 M * Wido phedny, heey Nederlander ;) 1159965962 M * daniel_hozac Wido: if you put the root mount in fstab, and you're using util-vserver 0.30.210, that's a known issue that should be fixed. 1159965966 M * phedny Wido: Ubuntu has this LANG and LANGUAGE env set :) 1159965968 M * Dimmu There are more dutchies here ;) 1159965984 M * Wido daniel_hozac, ok 1159965993 M * daniel_hozac phedny: most likely it means your stop script is killing itself with force. 1159965995 M * Wido hmm, not a really nice shutdown for your guest 1159966011 M * daniel_hozac Wido: vserver ... enter doesn't work either, right? 1159966011 M * phedny Wido: no, it seems PuTTY sets them, when I'm on the machine itself everything is English 1159966069 M * Wido didn't test daniel 1159966090 M * Wido root@reseller-pool-11:/etc/vservers/reseller14# vserver reseller14 enter 1159966090 M * Wido No command given; use '--help' for more information. 1159966090 M * Wido root@reseller-pool-11:/etc/vservers/reseller14# 1159966098 M * daniel_hozac so it's the same bug. 1159966102 M * Wido ok 1159966102 M * phedny daniel_hozac: point is some vservers do give this message, others don't, so my question is: why is it happening? 1159966113 M * daniel_hozac phedny: same distro? 1159966164 M * phedny I used Debian and Ubuntu, don't exactly know which one is used by which vserver 1159966187 M * Wido daniel_hozac, it does not affect the guest when it is running? 1159966198 M * daniel_hozac Wido: no, just stopping and entering is impossible. 1159966203 M * Wido ok 1159966216 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/namespace.patch is the patch that fixes it. 1159966242 M * daniel_hozac (something like patch -d /usr/sbin -p1 < namespace.patch should patch your installed copy) 1159966262 M * daniel_hozac oh, that's a broken version actually. 1159966279 M * Wido it patches vcontext, right? 1159966287 M * Wido so i would have to recompile my tools? 1159966289 M * daniel_hozac no, it patches the vserver script. 1159966294 M * Wido ok 1159966295 M * phedny hmm, now every vserver gives this message after stopping :) 1159966310 M * daniel_hozac phedny: i've only seen that on Debian guests, so that seems right :) 1159966334 M * phedny okay, is it an error in some way or is it something I can disregard? 1159966345 M * Wido daniel, it works 1159966347 M * daniel_hozac nah, it's not a problem. 1159966351 M * Wido it fixed the problem :) 1159966354 M * daniel_hozac Wido: but now restart is broken! ;) 1159966361 M * Wido well, i never use restart 1159966374 M * daniel_hozac anyways, 0.30.211 will include the proper fix. 1159966379 M * daniel_hozac (already does) 1159966382 M * Wido ok 1159966387 M * Wido eta on the release? 1159966437 M * daniel_hozac i really want it out before this week is over, but it depends largely on what my testing of -rc3 shows... 1159966444 M * Wido ok 1159966445 M * Wido super! 1159966636 M * ex is there some problem with linux-vserver.org domain? 1159966658 M * ex $ host -t ns linux-vserver.org -> Host linux-vserver.org not found: 3(NXDOMAIN) 1159966733 M * daniel_hozac yep, seems DNS is semi-broken. 1159966786 M * daniel_hozac phreak``: is there any way i can test this? 1159966956 M * daniel_hozac my simple mkfifo test; cat test & echo asdf > test works fine. 1159967171 Q * shedi Ping timeout: 480 seconds 1159967269 M * phreak`` daniel_hozac: no clue. Will talk to the baselayout maintainer (hopefully he has an idea) 1159967293 M * daniel_hozac would be interesting if we could get a more minimal test-case... 1159967575 M * daniel_hozac so how do i update to this new baselayout? 1159967635 M * phreak`` daniel_hozac: see /msg 1159967640 M * phreak`` (about the test-case) 1159967648 M * daniel_hozac hmm. 1159967661 M * daniel_hozac that seems like a really strange usage to me. 1159967682 M * daniel_hozac cat ... in one and then echo in the other seems like the more sensible approach... 1159967696 J * shedi ~siggi@dsl-149-109-85.hive.is 1159967736 M * phreak`` daniel_hozac: and on how to update to the new baselayout: unmask the new baselayout (FEATURES=-collision-protect emerge -b sys-apps/baselayout), emerge it, then unmerge baselayout-vserver and then remerge the new again w/ emerge -K baselayout (to ensure everything is there) 1159967760 M * daniel_hozac ok, thanks. 1159967778 M * daniel_hozac is there any particular reason the baselayout usage of fifos is so strange? 1159967811 M * daniel_hozac but it seems to work fine here, regardless. 1159967865 M * daniel_hozac will test the baselayout. 1159967986 M * phreak`` daniel_hozac: I don't see a particular reason, but maybe he has. no idea :) 1159968275 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/util-vserver-0.30.211-rc3.tar.bz2 i think all the bugs are fixed, so if some more people than myself could test it, that would be great :) 1159968288 Q * Ben81 Quit: Leaving 1159968475 M * meandtheshell Hi folks - I'd like to test another ftp command line client. I'm using ncftp nor for a while. Which command line ftp clients other than ncftp are used by Linux-VServer people? /me waiting for inspirations :) 1159968501 M * meandtheshell s/nor/now/ 1159968505 M * daniel_hozac i only ever use sftp :) 1159968513 M * meandtheshell daniel_hozac: hihi 1159968514 M * daniel_hozac and even that is very rare. 1159968583 M * meandtheshell daniel_hozac: well - let's see - maybe one guy comes up with some sort of "all live changing" command line client for us :) 1159968588 M * phedny is it doable to put a mysql server in a single guest, install the client libraries in several guests and hardlink the /var/run/mysqld/mysqld.sock socket? 1159968604 M * phedny or should i let the guest use tcp/ip to connect to mysql server? 1159968634 M * phedny since unix domain sockets have lesser overhead then tcp connections 1159968645 M * daniel_hozac it's possible to use the UNIX socket, but it's certainly easier to use TCP/IP. 1159969361 Q * SEAwolfx Remote host closed the connection 1159970194 M * daniel_hozac phreak``: hmm, so if you try what UberLord said, it doesn't work? 1159970218 J * nayco_work ~nayco@proxy2.laroche.univ-nantes.fr 1159970224 M * nayco_work hello again, all. 1159970283 M * nayco_work Hey, http://ftp.linux-vserver.org/ has no DNS records !? 1159970314 M * nayco_work Do I need to wait for DNS convergence ? 1159970317 M * daniel_hozac nayco_work: the DNS seems to be broken. 1159970329 M * nayco_work oh, dear. 1159970342 M * nayco_work how can I join the ftp server, so ? 1159970358 M * daniel_hozac try helios.dev.croup.de. 1159970366 M * nayco_work is it a mirror ? 1159970465 M * nayco_work daniel_hozac: it works, thanks 1159970490 M * Wido daniel_hozac, little update 1159970494 M * Wido your patch works fine 1159970502 M * Wido did not have any troubles at all 1159970551 M * daniel_hozac yep, it should, only restart/condrestart will break. 1159970561 M * Wido i noticed 1159970565 M * daniel_hozac nayco_work: that's the real name of the server that runs it. 1159970576 M * Wido but i thought you might like some feedback 1159970587 M * daniel_hozac Wido: 0.30.211-rc3 has the proper fix and a whole lot more. 1159970598 M * Wido ok :) 1159970613 M * nayco_work daniel_hozac: Ok, thx ;-) 1159970758 M * daniel_hozac Wido: yes thanks, feedback is always welcome :) 1159971327 J * ntrs_ ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1159971327 Q * ntrs Read error: Connection reset by peer 1159971756 Q * tokkee Remote host closed the connection 1159971757 J * tokkee tokkee@casella.verplant.org 1159972086 M * nayco_work hey guys, is it possible to convert dynamic xid vservers to static xid ones ? 1159972151 M * daniel_hozac sure, vserver guest stop; echo > /etc/vservers//context; vserver guest start 1159972775 J * Nigromante ~username@197.Red-80-35-167.staticIP.rima-tde.net 1159972964 M * nayco_work daniel_hozac: Ok, thanx, that's what I was looking for in the flower page, but I searched "xid" instead of "context" ;-) ! 1159973132 Q * derjohn2 Ping timeout: 480 seconds 1159973132 J * derjohn2 ~aj@dslb-084-059-019-127.pools.arcor-ip.net 1159973148 M * derjohn "vxW: !!! limit: f490b048[DENTRY,22] = 523 on exit" -> what kind of limit did i hit? 1159973247 M * doener the bug limit 1159973254 M * nayco_work :-D 1159973269 M * derjohn doener, what ? 1159973273 M * derjohn I did hit a bug =? 1159973296 M * derjohn maximum number of bugs reached? 1159973302 M * doener yeah, and he's probably dead now, murderer! *lol* 1159973320 M * derjohn rc19 ... 1159973327 M * derjohn might be :) 1159973331 P * Nigromante Abandonando 1159973345 M * doener the message just says that there were accounted dentries left although the context died 1159973348 M * derjohn means: upgrading to rc35 solves the issue= 1159973349 M * derjohn ? 1159973378 M * derjohn hm, i didnt activate a dentry limit on this machine 1159973388 M * derjohn so we some implicit value set ? 1159973395 M * doener accounting is always done? 1159973452 M * doener in /proc/virtual/foo/limit you can always see the current usage 1159973466 M * doener no matter whether a limit is set or not 1159973481 Q * node Remote host closed the connection 1159973611 M * derjohn what means dentry type (?) 22 ? 1159973616 M * derjohn = 523 ? 1159973746 M * derjohn in /proc/virtual/foo/limit "DENT:" means "DENTRY" ? 1159973870 M * doener 22 is just the numeric index of the dentry limit 1159973880 M * doener and DENT probably means DENTRY, yes 1159973899 M * doener 523 is still number of dentries that the kernel thinks are still in use 1159973971 M * nayco_work O, I forgot : What is the range for static XIDs, please ? 1159974169 M * nayco_work Or where can I find the inof ? 1159974251 M * nayco_work info 1159974699 M * derjohn nayco_work, here the inof: XID 0 and 1 are for "internal use". You can use 2 to about 49K ... 1159974719 M * derjohn nayco_work, should be somewhere in the FAQ. LART ME for not having sorted them yet ... 1159974940 M * phreak`` daniel_hozac: haven't tried it within a vps yet. only on my laptop (was at work earlier) 1159974970 J * restill ~restill@c-24-11-195-139.hsd1.mi.comcast.net 1159975053 M * nayco_work derjohn: Thank you :-) ! Because this does not appear in the nearly empty "man vserver", nor the great flower page. Well, I played it safe, I just used the lower byte of the guests' IP adress as XID ;-) 1159975143 N * restill Rich_Estill 1159975151 M * doener nayco_work: 49152 IIRC 1159975157 M * doener (first dynamic xid) 1159975203 M * doener you could try to just run "chcontext true", if you didn't use a dynamic context yet, you'll get the lowest one 1159975321 Q * yarihm Quit: Leaving 1159975526 M * nayco_work doener: I tried, this is true. Ok, so the most important is that the lower bound i 2 :-) 1159975586 M * phedny what is VSZ and RSS? 1159975656 J * dwindsor ~dwindsor@stanford.columbia.tresys.com 1159975659 N * dwindsor node 1159975700 M * nayco_work I rebooted all this stuff, and static context work fine. Good new too : I put the "vrsetup xxxxxx" command in /etc/vservers//scripts/prepre-start , and I now have XFS quota working directly from boot. Ok, entering pre-production phase ;-) ... 1159976159 J * Piet hiddenserv@tor.noreply.org 1159976504 N * Bertl_zZ Bertl 1159976508 M * Bertl morning folks! 1159976564 M * doener morning Bertl 1159976594 M * Rich_Estill yo, morning. 1159976632 M * doener *LOL* I'm just reading http://lkml.org/lkml/2006/10/3/395 1159976639 M * doener nice coincidence 1159976672 M * nayco_work Hello, Bertl 1159976902 M * nayco_work doener: have you read the mail's signature ? 1159976983 M * phreak`` morning Bertl, doener, nayco_work, Rich_Estill, and probably everyone awake at this time of the day :P 1159976999 M * doener nayco_work: yep 1159977003 M * Bertl hey Rich_Estill?! 1159977043 M * Bertl Rich_Estill: new here or just in disguise? 1159977266 M * Rich_Estill aka restill 1159977314 M * Rich_Estill I helped daniel w/ some 64 bit compiling in the past. Then I dissappeared b/c I got wrapped up in my own project. 1159977387 M * Bertl ah, ic. so you are back now! excellent! 1159977420 J * matled_ ~matled@85.131.246.184 1159977487 Q * matled Read error: Connection reset by peer 1159977487 N * matled_ matled 1159977514 M * Rich_Estill Yup. vserver has been working great as a test bed for my distributed computing project. 1159977892 M * matti Hi Bertl. 1159977906 M * Rich_Estill What IDE do you use? I have been using Eclipse, but for Java. 1159977943 M * doener vim 1159977955 M * mnemoc vim+ctags 1159977964 M * doener +cscope here 1159977968 M * mnemoc :) 1159978000 M * matti vim +cscope 1159978051 M * Rich_Estill So I may not last long here reverting back to Emacs. 1159978074 M * Bertl okay, off for dinner now, back shortly ... 1159978078 N * Bertl Bertl_oO 1159978256 J * bonbons ~bonbons@83.222.36.111 1159978648 M * matti Bertl_oO: Enjoy. 1159979518 Q * derjohn Quit: by(t)e 1159980115 M * daniel_hozac ensc: just FYI, i'll change the bug report address to the mailing list. 1159980195 M * ensc daniel_hozac: good idea... ;) (although there were not very much mails in the past) 1159980207 M * daniel_hozac yeah, i figured it wasn't used much. 1159980249 M * daniel_hozac have you had a chance to look at any of the changes? 1159980252 M * ensc most about the failed 'make check' on debian 1159980290 M * daniel_hozac the vhashify test? or the ones micah filed? 1159980308 M * ensc the hashify related ones 1159980324 M * ensc hashcalc* 1159980333 M * daniel_hozac ah, right. i couldn't remember the name :) 1159980371 M * ensc svn.linux-vserver.org can not be resolved? 1159980385 A * micah looks around when his name is said 1159980386 M * daniel_hozac there are some DNS problems, try helios.dev.croup.de instead. 1159980447 N * Bertl_oO Bertl 1159980459 M * Bertl back now 1159980538 M * Bertl okay, as it seems we have a problem with linux-vserver.org 1159980579 M * Bertl I'm not really sure _what_ the problem is, but the result is that the tld servers do not provide glue to our nameservers 1159980603 M * Bertl if somebody has ideas/suggestions/hints ... please let me know 1159980617 M * daniel_hozac still no word from jacques? 1159980623 M * Bertl nope :( 1159980631 M * phedny Domain ID:D101535640-LROR 1159980631 M * phedny Domain Name:LINUX-VSERVER.ORG 1159980631 M * phedny Created On:03-Oct-2003 20:42:41 UTC 1159980631 M * phedny Last Updated On:04-Oct-2006 03:58:34 UTC 1159980631 M * phedny Expiration Date:03-Oct-2007 20:42:41 UTC 1159980669 M * phedny Name Server:DNS.UNIGICIEL.COM 1159980669 M * phedny Name Server:DNS.SOLUCORP.QC.CA 1159980669 M * phedny Name Server:NS1.13THFLOOR.AT 1159980676 M * phedny <-- is that correct? 1159980686 M * Bertl well, no 1159980698 M * Bertl problem is, only the last one is in my administration 1159980707 M * Bertl the other two are in jacques' realm 1159980719 M * Bertl and they do not accept updates from my server anymore 1159980721 P * Rich_Estill Leaving 1159980731 J * Rich_Estill ~restill@c-24-11-195-139.hsd1.mi.comcast.net 1159980755 M * Bertl but there is a different record http://paste.linux-vserver.org/444 1159980773 M * phedny unaccessible ;) 1159980779 M * Rich_Estill I noticed the website. At work here the local DNS cached the IP and seems to work OK, but not from other machines. 1159980801 M * Bertl yeah, folks can use linux-vserver.net in the meanwhile 1159980819 M * daniel_hozac has been working fine for me :) 1159980822 M * phedny it redirects to linux-vserver.org 1159980835 M * Rich_Estill other than it redirects to .org 1159981207 M * Bertl is somebody here located in Canada? 1159981254 N * hardwire hardwire-afk 1159981258 N * hardwire-afk hardwire 1159981364 M * Rich_Estill close 1159981427 M * Bertl hmm, maybe somebody could try to reach jacques via phone 1159981449 M * Bertl last time we tried to reach him via email, he didn't check it for more than 6 months ... 1159981469 Q * prae Quit: Quitte 1159981647 M * Loki|muh whow 1159981971 J * derjohn ~derjohn@80.69.37.19 1159982245 Q * shedi Quit: Leaving 1159983387 M * goblin If some of the host's partitions are mounted in a vserver too (for example, with mount --bind, but not only), is there a way to display a user's quota from the host's context inside a guest? :-) 1159983554 M * goblin even though I add an entry to /etc/mtab with a proper filesystem, and cat /proc/mounts shows it mounted, the quota utility doesn't seem to find any filesystems with quota 1159983573 M * goblin the only gotcha is that the mounted device itself doesn't exist in guest context :-) 1159983738 M * daniel_hozac i believe the device is what's required to perform the calls quota requires. 1159984030 M * goblin hmm 1159984064 M * goblin I think that's not true, as it doesn't have any suids set on my host system 1159984071 M * goblin the device isn't readable to anyone but root 1159984079 M * goblin and it still can fetch the information somehow 1159984095 M * goblin I tried to strace it, but it didn't open anything but mtab, libs, and locale files 1159984121 M * goblin I suspect it does it via some syscall 1159984349 M * goblin yeah, it calls quotactl() 1159984578 M * goblin Hm. You're right, actually. ;-) 1159984615 M * goblin even though the device isn't readable, calling quotactl on it works. :-} 1159984709 M * goblin and the guest vserver gets EPERM when trying to call it 1159984761 M * goblin so maybe adding it to ccapabilities would make it... but I still don't want to have those device files in a guest system :-/ 1159984778 M * daniel_hozac that's what vroot is for. 1159984797 M * daniel_hozac a proxy device allowing only the calls required to make quota work. 1159984905 M * goblin like that described here?: http://66.102.9.104/search?q=cache:j1KqBEOEtFsJ:linux-vserver.org/Standard%2Bnon-shared%2Bquota+quota+vserver&hl=en&ct=clnk&cd=9&client=opera 1159984997 M * goblin Ohh.. I guess I understood only when I read for the second time :-) 1159984998 M * daniel_hozac yep. 1159985004 M * goblin It's cool! thanks a lot! :-D 1159985050 M * goblin one more thing: is it possible to just allow checking the quotas, and not modifying them? 1159985182 M * Bertl yep 1159985224 M * Bertl if you put the quota outside the guest and do not give the quota capability, they will apply (and can be checked somewhat) but not adjusted 1159985267 M * goblin cool 1159985301 M * goblin so, just create a bunch of vroot devices corresponding to guest's devices, but don't put quota_ctl into ccapabilities, right? 1159985374 M * Bertl if you do not want to use the quota tools _inside_ the guest, then you can skip the vroot device too 1159985406 M * goblin I want to use them inside the guest, but only for checking user's quota 1159985430 M * goblin so that users know what they have 1159985545 M * Piet Bertl: http://www.dnsreport.com/tools/dnsreport.ch?domain=linux-vserver.org 1159985568 A * micah looks at Piet 1159985578 M * Piet :) 1159985578 M * Bertl Piet: well, nothing new there :( 1159985617 M * Piet :( 1159985805 M * Piet i guess that explains the google cache urls in the backlog 1159985843 M * goblin thanks very much for solving my problem :-) 1159985848 M * Piet seems to have been some time ago that the A record got lost... http://www.dnsstuff.com/tools/ispdns.ch?name=linux-vserver.org&type=A 1159985924 M * doener hm, SOA at unigiciel and solucorp either 1159985943 M * Bertl that's the problem, I guess 1159985966 M * Bertl both of jacques nameservers stopped updating/serving the domain 1159986053 M * Bertl wht I do not understand is, why the tld does not use the third nameserver 1159986057 M * Bertl *what 1159986087 M * Bertl maybe it is a problem on my side (with the nameserver setup) but it seemed all fine a few weeks ago (when I last checked) 1159986105 A * ay found openvz while waiting for vserver.org. 1159986113 M * ay Anyone tried it? How does it compare to vserver? 1159986194 M * Bertl very similar, based on the commercial Virtuozzo product from swsoft, a lot more instrusive, not as lightweight as Linux-VServer 1159986274 M * Piet $ host DNS.SOLUCORP.QC.CA 1159986274 M * Piet DNS.SOLUCORP.QC.CA has address 207.253.4.241 1159986276 M * Piet $ host DNS.UNIGICIEL.COM 1159986276 M * Piet DNS.UNIGICIEL.COM has address 207.253.4.2 1159986289 M * Piet same network? 1159986307 M * Bertl probably 1159986425 M * ay Bertl: So something in between Xen and vserver. 1159986469 M * Bertl yes, more virtualization, less isolation 1159986502 M * ay Is it any good? What is the good/bad issues? 1159986590 M * Bertl they use a completely virtualized network stack (which adds overhead but gives xen like networking) 1159986793 M * ay I like the easy networking of vserver, maybe except the 127.0.0.1 issues. 1159986800 M * Piet maybe you could temporarily remove the redirection from http://85.10.237.61 to http://linux-vserver.org and arrange that the site can be browsed throught he ip address 1159986819 M * Bertl ay: we will address that in the near future (plans for this are already there) 1159986872 M * Bertl Piet: let me check if I can hack Hollows wiki ... 1159987195 J * kir_home tis-b58b90@213.152.157.70 1159987484 M * Borg- heh 1159987506 M * Borg- I can provide backup DNS or sth ;) 1159987587 M * Bertl Borg-: we have 6 different nameservers 1159987600 M * Borg- so why domain doesnt work now? 1159987602 M * Bertl it doesn't help if the tld points to the wrong ones 1159987616 M * Borg- ah.. you were in middle of update? 1159987672 M * Borg- 20 200 ms 210 ms 201 ms dns.croesus.com [207.253.4.2] 1159987673 M * Borg- ;] 1159987739 M * Bertl nope, Jacques Gelinas (who maintains linux-vserver.org and the two main dns servers) seems unresponsive for some time, and now it did become an issue 1159987781 M * ay Tried phoning him? 1159987782 M * ay Registrant Phone:+1.4505339904 1159987804 A * ay can provide free dns master/slaves as well. 1159987816 M * ay From www.linpro.no 1159987898 M * Bertl Canada is a little expensive from Austria ... was looking for someone a little closer 1159988052 M * mnemoc Bertl: http://www.voipbuster.com/en/index.html, SIP, free for fixed lines to canada 1159988109 M * Rich_Estill I wish I could. I am so close, but even from the US it is expensive. All I have is a cell phone. 1159988179 M * doener hm, looks like 4.5 cent/minute from here, but I need to decipher the tariff to be sure 1159988196 M * mnemoc in fact they also give free calls to austria.... i have not found the "trick" of that provider yet 1159988209 M * matled http://teltarif.de/db/res.html?019x=0&kurz=ja&ve=1&takt=61&von=Analoganschluss&ziel=Kanada 1159988618 Q * Piet Ping timeout: 480 seconds 1159989019 M * Rich_Estill What email do you have for him? 1159989053 M * Bertl several, the most recent seems to be the solucorp though 1159989335 M * independence is there some other place to fetch the patches? 1159989347 M * independence the website seems to be down as you seem to have noticed :) 1159989361 M * Bertl in the meantime (hope it works) http://linux-vserver.at/Welcome_to_Linux-VServer.org 1159989404 M * doener works here 1159989428 M * Bertl the .net domain is unfortunately a tricky forward not administrated by me either 1159989470 M * Bertl I guess I have to send some emails to various maintainers (of those domains) to get them to fix up their entries, once and for all 1159989550 M * independence is there somewhere else i can get this file: patch-2.6.17.13-vs2.0.2.1.diff.bz2 1159989559 M * daniel_hozac ftp://helios.dev.croup.de 1159989635 M * independence thanks 1159989645 M * daniel_hozac and also at http://vserver.13thfloor.at/Experimental/ as usual. 1159989652 M * Bertl hmm? that patch should be where it always was 1159989666 Q * michal` Ping timeout: 480 seconds 1159989802 M * Borg- wuaha ;) www.dnsreport.com added some new tests.. nice 1159989990 J * michal` ~michal@www.rsbac.org 1159990340 M * independence is the patch for +grsecurity also available at some other site? 1159990357 M * independence couldn't find it at the links you gave me 1159990359 M * Borg- I can give you one if you want ;> im mirror boy ;) 1159990390 M * independence please do :) 1159990447 M * Borg- patch-2.6.17.11-vs2.0.2-grsec2.1.9.diff.bz2 <- this one? 1159990482 M * independence nah, 2.6.17.13 1159990505 M * independence patch-2.6.17.13-vs2.0.2.1-grsec2.1.9.diff 1159990519 M * Borg- hmm.. sorry.. didnt mirriored it then. 1159990524 M * Borg- was it on official site? or as beta? 1159990537 M * Borg- I usualy dont mirror alpha/beta/rc stuff 1159990539 M * independence it's official 1159990548 M * independence it's on the first page and marked as stable atleast 1159990553 J * samuel ~samuel@d154-20-170-153.bchsia.telus.net 1159990561 M * samuel hi 1159990582 M * matled independence: http://igit.ath.cx:2080/matled/patch-2.6.17.13-vs2.0.2.1-grsec2.1.9.diff.bz2 1159990591 M * daniel_hozac independence: hmm, i was under the impression the people pages were on the ftp as well. 1159990640 M * daniel_hozac independence: but http://helios.dev.croup.de/~harry/ seems to be accessible. 1159990648 M * Borg- indeed.. 1159990665 M * Borg- it must appeard while ago :) 1159990681 M * matled you could also try adding 85.10.237.61 linux-vserver.org people.linux-vserver.org etc to /etc/hosts 1159990692 M * daniel_hozac yep. 1159990718 M * independence I got it, thanks again 1159990762 M * Bertl http://people.linux-vserver.at/~doener/ 1159990776 M * Bertl just added that entry too ... 1159990798 M * hardwire I love you guys 1159990833 M * daniel_hozac Bertl: do you know if there's a reason people isn't available through FTP? 1159990859 M * Bertl no idea 1159990940 N * samuel samueltc 1159990942 M * hardwire because you can't people use the people transport protocol 1159990947 M * hardwire aka feet 1159991018 M * daniel_hozac ah, because you can't just symlink to public_html. 1159991197 Q * Wido 1159991531 M * hardwire bind mount it 1159991531 M * hardwire :) 1159991556 N * hardwire P 1159991559 M * P oops 1159991561 N * P hardwire 1159991609 M * daniel_hozac yep, but i don't have root access, plus that's kind of ugly ;) 1159991615 M * matti Bertl: :) 1159991623 M * Bertl hardwire: let's keep it low for today, we have a hard day with the dns not working ... 1159991661 M * hardwire what am I a two year old to you! 1159991678 M * samueltc hey, how can I gather network statistic for each vserver? (they are all bound to virtual intercafe on dummy0) 1159991679 M * hardwire <- easily offended - often hyperactive 1159991691 M * matti Bertl: How are you? 1159991719 M * daniel_hozac samueltc: iptables works fine, there's also some stats in /proc/virtual//cacct 1159991774 M * samueltc daniel_hozac: ahh, nice 1159991819 M * matti Hi daniel_hozac. 1159991821 M * samueltc where can I get information about what are these numbers/columns 1159991828 M * samueltc (cacct) 1159991831 M * daniel_hozac hey matti. 1159991835 M * Bertl matti: well, I have seen better days ... seems I'm starting with a flu 1159991862 M * daniel_hozac http://oldwiki.linux-vserver.org/HowTo+Read+ProcFS 1159991864 M * samueltc http://oldwiki.linux-vserver.org/HowTo+Read+ProcFS 1159991871 M * samueltc ehh thanks 1159991900 M * samueltc we are back in the 70s, using /etc/hosts :) 1159991903 J * shedi ~siggi@inferno.lhi.is 1159991921 M * matti Bertl: :/ 1159991931 Q * nayco_work Ping timeout: 480 seconds 1159991935 M * mnemoc http://66.249.93.104/search?q=cache:VVhi2eMCpsUJ:linux-vserver.org/HowTo%2BRead%2BProcFS+vserver+how+to+read+procfs&hl=es&gl=cl&ct=clnk&cd=1 <--- google cache rules :) 1159991947 J * comfrey ~comfrey@h-64-105-215-75.sttnwaho.covad.net 1159992006 M * matti Bertl: I hope you will get better tomorrow. 1159992030 Q * lilalinux Remote host closed the connection 1159992061 M * Borg- hmm 1159992070 M * Borg- [root@test1] cat limit | grep SOCK 1159992070 M * Borg- SOCK: 8 8 -1 0 1159992075 M * Borg- is that sth I should worry about? 1159992086 M * daniel_hozac that you're using 8 sockets? why would you? 1159992109 M * Borg- no ;) what are those columns means? I guess -1 is a limit (means inf) right? 1159992116 M * Borg- first one is current, second is max ? 1159992127 M * Borg- (max as max opened ever) 1159992143 M * daniel_hozac yep. 1159992151 M * daniel_hozac last is number of hits on the limit. 1159992165 M * Borg- ah :) then nothing to worry about, thx 1159992169 M * bonbons Bertl: regarding the DNS trouble, did you see this in `whois linux-vserver.org`: Status:CLIENT HOLD 1159992203 M * hardwire thats no bueno 1159992255 M * bonbons exactly, and possible related: Expiration Date:03-Oct-2007 20:42:41 UTC <-- might be payment issue 1159992271 M * Borg- bonbons: heh? we have year 2006 1159992328 M * bonbons I know, but the point there is 3 October in the evening, and the registrar may have payed registry for the coming year but put domain on holder because himself did not get money yet 1159992358 M * Borg- hmm.. 1159992358 M * Borg- this is bad ten 1159992360 M * Borg- then.. even 1159992361 M * Bertl yes, I assume Jack is gone again (no idea why) and he didn't bother (or wasn't able to) transfer the domain 1159992413 M * Borg- who have access to apache here? maybe someone could add ServerAlias to httpd.conf and bind extra backup domain ? 1159992417 M * daniel_hozac Bertl: could you mirror the entire .org zone on .at? (i.e. svn too, at least) 1159992438 M * Bertl I guess so .. give me a few minutes 1159992477 M * bonbons would be really useful to get that fixed... having a domain-grabber take it would be really anoying 1159992630 M * bonbons look at `whois -h whois.gandi.net`, there it says expiration: yesterday 1159992661 M * Borg- aha.. so payment issue indeed 1159992911 M * independence I'm having some troubles with identd in a vserver guest, I tried to explain my problem here: http://forums.gentoo.org/viewtopic-p-3624086.html#3624086 I would really appreciate it if someone could help me :) 1159992933 M * Rich_Estill I have faith that in a day or two, the money will transfer and we will be back in service. 1159992988 M * Borg- independence: what identd are you using? 1159992992 M * independence oidentd 1159993003 M * Borg- both IPv6 and IPv4 ? 1159993009 M * independence no, just IPv4 1159993018 M * Borg- and do you need oidentd functionality? 1159993022 M * Borg- (NAT for example) 1159993026 M * independence yes 1159993033 M * independence exactly 1159993040 M * Borg- huh? guest does NAT? 1159993052 M * independence no 1159993053 M * independence the host 1159993066 M * Borg- so oident should work on host 1159993076 M * independence yes, it works fine on the host 1159993085 M * independence but not on the guest 1159993090 M * Borg- why you want to move it to guest? 1159993113 Q * bonbons Quit: Leaving 1159993129 M * daniel_hozac because you should run a minimal amount of services on the host? 1159993136 M * daniel_hozac and you have your users in the guest? 1159993143 M * Borg- daniel_hozac: NORMAL services.. not special ones 1159993157 M * Borg- I dont consider IDENTD server as normal service really 1159993168 M * Borg- and AFAIK guest cannot do NAT, right? 1159993183 M * Borg- so how guest can read /proc from host to get NAT -> IDENT ? 1159993185 M * independence uh, wait 1159993190 M * Borg- (connection-track) 1159993196 M * independence I need to run oidentd on both host and guest 1159993201 M * independence It doesn't work otherwise? 1159993207 M * Borg- yes 1159993217 M * independence okay, good 1159993223 M * Borg- provive ident for NAT user.. on ne IP.. and ident for user of guest (other IP) 1159993224 M * independence that's what I'm doing now 1159993241 M * Borg- limit host identd to bind just one IP 1159993265 M * independence how does that help? 1159993284 M * independence the requests are forwarded to the guest, but the guest's oidentd can't resolve them 1159993284 M * Borg- running service on host doing bind(*) will bind it on *ALL* IPs 1159993287 M * Borg- even on guests one's 1159993292 M * Borg- and you dont want that probably 1159993305 M * independence yes, but I run the oidentd on the guest on another port 1159993317 M * Borg- heh.. isnt that.. a bit.. too complicated 1159993324 M * Borg- ? 1159993335 M * Borg- or you have many IPs? 1159993336 M * independence no, it wasn't hard at all.. 1159993337 M * FaUl well - ident on another port is quite useless, isn't it? 1159993347 M * independence the host runs on port 113 1159993353 M * independence so the internet doesn't notice 1159993355 M * Borg- FaUl: not really.. if he does forwarding oident -> oident 1159993360 M * independence the guest runs ident on port 114 1159993360 M * FaUl Borg-: ah, ok 1159993377 M * FaUl havn't read backlog, just comming into the discussion 1159993384 A * FaUl hides again behind his stone 1159993393 M * FaUl ;-) 1159993400 M * Borg- independence: question.. how oident on host forwards IDENT request? 1159993407 M * independence But is there some limitation in vserver that would make lookups of connections in the guest impossible or something? 1159993414 M * Borg- based on IP? or based on result (I didnt found on NAT table nor host passwd) 1159993424 M * daniel_hozac independence: no, e.g. netstat works fine. 1159993431 M * independence Borg-: um, not really sure actually :) 1159993436 M * independence can I check somehow? 1159993441 M * daniel_hozac source? :) 1159993451 M * Borg- independence: dunno.. I dont user that bloatware really ;))) 1159993466 M * independence you think oidentd is bloatware? :P 1159993494 M * Borg- independence: notices ";))))" ... but I dont like it really.. too big.. for my needs 1159993508 M * Borg- and I dont like their aproach to oldschool 1159993532 M * independence Okay, is there some other identd software that does NAT? 1159993540 M * Borg- nope 1159993544 M * independence doh 1159993548 M * daniel_hozac easier to just get it working, no? 1159993587 M * Borg- independence: why you want to user have ident anyway? 1159993602 M * Borg- I like to have difference.. LAN user.. UNIX shell account user in my case 1159993623 M * independence the guest is used for IRC, and some servers require valid ident responses 1159993632 M * Borg- ident nowadays is just checked for IRC.. and some tcpd wrappers.. (rary used) 1159993668 M * Borg- independence: and your users are too lazy to user shell account or irc proxy? ;] 1159993709 M * independence the vserver is supposed to be the shell account/irc proxy :P 1159993741 M * Borg- huh.. so why bothering w/ oident and NAT then? I dont get you really 1159993767 M * independence the vserver guest is used for shell account hosting 1159993773 M * independence people run screen+irssi in there 1159993776 M * independence they need identd 1159993792 M * Borg- yes 1159993807 M * Borg- but oident is big.. he can provide IDENT for NATed users too 1159993810 M * Borg- do you want that too? 1159993814 M * Borg- [borg@cube] cat /etc/identd.conf 1159993814 M * Borg- max-conns 16 1159993814 M * Borg- timeout 30 1159993814 M * Borg- bind * 1159993814 M * Borg- #bind ::* 1159993817 M * Borg- ;) 1159993840 M * independence the users on the vserver guest is nated? 1159993873 M * Borg- okey.. maybe im tired.. or dunno 1159993902 M * independence I can't run the identd software directly on the vserver guest, because it's behind nat 1159993907 M * Borg- we are talking about UNIX shell users on guest that they are natted? 1159993926 M * Borg- ah.. so you have one IP.. public 1159993932 M * Borg- and guests are on private IP? 1159993937 M * independence yes 1159993944 M * Curus Whenever I create a vserver guest, the host downs the network interface that the guest uses 1159993970 M * Borg- independence: holy necro.. its not nice then ;) 1159993987 M * Borg- and you cant do what you want probably.. 1159994009 M * independence it's impossible? 1159994012 M * matled you could probably get all information needed on the host 1159994024 M * Borg- im aware of that... but im thinking 1159994026 M * Curus It's so odd, I can't see a reason why vserver build would do that 1159994048 M * daniel_hozac Curus: what --interface are you specifying? 1159994080 M * Curus bond0.136 last time I tried. Each guest gets its own VLAN interface. 1159994092 M * Borg- independence: example... 192.168.0.1:1024 -> [1.2.3.4:2054] -> irc.server:6667 1159994100 M * Curus After vserver build, bond0.136 is gone 1159994106 M * Borg- irc server will connect to 1.2.3.4:113 and ask like this: 2054 , 6667 1159994109 M * daniel_hozac that's expected. 1159994123 M * Borg- and guest will answer 2054 , 6667 : NO-USER 1159994128 M * daniel_hozac VLANs will be created/removed by util-vserver unless you specify novlandev. 1159994131 M * Borg- because user uses 1024 on guest 1159994141 M * Borg- you need some smart stuff to solve that 1159994145 M * independence Borg-: Um, something like that yes :) 1159994154 M * matled take a identd of your choice and modify it :) 1159994158 M * Curus Oh, I'll go specify that 1159994164 M * Curus Why does it do that? 1159994174 M * Borg- matled: he really need to add ident-proxy software 1159994177 M * Curus Or rather, when would that be useful behaviour? 1159994187 M * daniel_hozac same reason it brings up/down the IP addresses, i suppose. 1159994217 M * daniel_hozac the behaviour has been changed in util-vserver 0.30.211, so you have to set vlandev to get it to do that. 1159994239 M * independence Havn't anyone done this before? modify ident software seems a bit hard.. 1159994243 M * Borg- matled: ident-proxy gets ident request.. lookups /proc/conntrack.. sends modified IDENT request to identd on specified guest.. and then forwards the answer 1159994244 M * Curus So it will be brought back up when I start the vserver. Nifty, actually, except keepalived hates me 1159994252 M * Borg- possible :) just someone needs to write sth like that ;) 1159994268 M * Borg- independence: its not even hard :) if you know C 1159994291 M * independence Well.. I've written some C, but I prefer python :) 1159994314 M * Borg- python will burn on heavy load I guess 1159994319 M * matled identd is not that complicated 1159994338 M * Borg- matled: sure .. I wrote my own identd server (im using it all the time) 1159994344 M * matled I would not expect too much load on an identd.. 1159994354 M * Borg- works both on FreeBSD and Linux (Linux IPv6 is not yet implemented.. im too lazy..;) 1159994369 M * Borg- matled: hehe.. depends how many users do you have on network :) 1159994372 M * independence not very much load no 1159994373 M * matled sure 1159994375 M * Borg- I know networks w/ 3000 users.. 1159994382 M * matled but if this is only for irc.. 1159994384 M * Borg- amount of hits on port 113 is noticeable 1159994501 M * Bertl okay, I'm off to bed ... have a good one everyone! 1159994511 N * Bertl Bertl_zZ 1159994516 M * Borg- hmm quite late indeed 1159994562 M * Curus Can I set novlandev by adding --novlandev to the vserver build command line? The docs aren't so clear 1159994570 M * daniel_hozac no, you can't set it when building. 1159994586 M * Curus But it's when building that my interface goes down... 1159994603 M * daniel_hozac you could upgrade to util-vserver 0.30.211-rc3 then. 1159994610 M * Borg- Curus: very weird 1159994618 M * Borg- show me the vserver build 1159994711 M * Curus /usr/sbin/vserver peterskov build -m yum --hostname=peterskov.vpbx.cust --interface bond0.136:10.0.49.164 -- -d fc5 1159994778 M * Borg- bond ? whats that? and why .136 ? 1159994786 M * Curus bonding interface 1159994791 M * Curus 136 is the VLAN 1159994818 M * Borg- hmm. looks fine then.. dunno really what could be wrong. 1159994828 M * Curus daniel_hozac explained it 1159994881 M * daniel_hozac http://people.linux-vserver.at/~dhozac/p/uv/experimental/util-vserver-0.30.211-rc3.tar.bz2 if you feel it's important enough. 1159994921 M * Curus Oh its important enough 1159995239 Q * Rich_Estill Quit: Leaving 1159995579 M * Borg- oh.. everything works fine via linux-vserver.at 1159996758 J * FCOJ ~mordur@dsl-201-4.hive.is 1159997045 Q * ag- 1159997169 J * ag- ~ag@caladan.roxor.cx 1159997315 Q * ag- 1159997351 J * ag- ~ag@caladan.roxor.cx 1159997532 M * Curus It may be a good idea to add /usr/lib/vserver/vserver-build.fai to util-vserver-build.list 1159997539 M * Curus In rc3 1159997542 M * daniel_hozac i know, it's fixed in HEAD ;) 1159997543 Q * ag- 1159997573 M * daniel_hozac (contrib/manifest.dat.pathsubst is the correct file though) 1159997581 J * ag- ~ag@caladan.roxor.cx 1159997624 M * Curus Thanks 1159998647 Q * ruskie Read error: Connection reset by peer 1159998688 J * ruskie ~ruskie@ruskie.user.oftc.net 1159998751 Q * samueltc Ping timeout: 480 seconds 1159999261 Q * cdrx Quit: Leaving 1159999914 M * Curus rc3 works nicely and doesn't down the interface 1159999923 M * Curus It does remove the IP address though 1159999936 M * daniel_hozac yep, you'll still have to set nodev to avoid that. 1159999946 M * Curus And can I do that in the build? 1159999949 M * daniel_hozac nope. 1159999984 M * Curus So there's nothing I can put on the vserver build command line which will keep vserver from removing the IP? 1159999993 M * daniel_hozac nope. 1160000004 M * Curus Ok 1160000012 M * daniel_hozac you could "forget" to specify --interface and add the directory/contents afterwards though. 1160000025 M * Curus Ah, that'll be the way forward then 1160000032 M * Curus Thanks 1160000082 Q * wenchien Ping timeout: 480 seconds 1160000269 Q * dna___ Quit: Verlassend 1160000521 J * Aiken ~james@tooax6-146.dialup.optusnet.com.au 1160000863 J * _tso_ ~root@196-019.adsl.pool.ew.hu 1160000977 Q * _tso_ 1160001463 M * ntrs_ daniel_hozac, what is new/fixed in 211-rc3? 1160001468 M * daniel_hozac from -rc2? 1160001472 M * ntrs_ yes 1160001560 M * daniel_hozac quite a few things, it seems. 1160001571 M * daniel_hozac for instance -rc3 should run on older 2.0 kernels as well. 1160001581 M * daniel_hozac the fai build method was added. 1160001593 M * daniel_hozac FC6 should now be supported once it hits the streets. 1160001638 M * daniel_hozac and some secure-mount fixing. 1160001646 M * daniel_hozac (those are the most serious fixes) 1160001821 Q * are|lunch Ping timeout: 480 seconds 1160001872 M * daniel_hozac oh, and vserver ... rpm/apt-get/apt-cache/apt-config should work again. 1160001884 M * harry /proc/virtual//cacct 1160001887 M * daniel_hozac (i don't know if anyone uses that though...) 1160001891 M * harry ==> what's the last column? 1160001923 M * daniel_hozac errors. 1160001951 M * harry ? 1160001964 M * harry awtsj 1160001976 M * harry INET: 952014/276155829 1880142/1212289154 7590/157956753 1160001976 M * daniel_hozac huh? 1160001979 M * harry then this is bad? 1160001998 M * daniel_hozac i don't really know what is considered an error. 1160001999 M * harry 157 meg bad data 1160002013 M * harry errors should be 0 :) 1160002342 M * hardwire oops 1160003575 Q * weasel Killed (services.oftc.net (Too many invalid passwords)) 1160003610 J * weasel weasel@asteria.debian.or.at 1160003686 Q * kir_home Ping timeout: 480 seconds 1160004479 Q * meandtheshell Quit: exit (0);