1116202681 M * DaCa Doener`: it looks like Bertl_zZ is having a confrontation with real life :) 1116202697 M * Doener` indeed :) 1116203659 Q * monrad Read error: Operation timed out 1116204494 M * Martin Doener`: Yea, I'm here but just about to leave 1116204505 M * Martin Thanks for the head's up on Bertl 1116204534 M * Martin If he drops past any time soon can you tell him I've done a little more testing on sparc64 and got to a point I think I need his help. 1116204550 M * Martin I'll drop past later this week. 1116204608 M * Doener` ok. i've seen that you made notes of your tests on the wiki, so maybe he can already solve this ;) 1116204671 M * Doener` short question, do you get the bus error after the first SYS_267 call? 1116204827 M * Doener` uhm, nevermind again... i've been reading the wrong function's code... ;) 1116204882 M * Martin Doener`: yes. 1116204895 M * Martin Having a look at the wiki code it appears a bit scrambled 1116204904 M * Martin the plain text seems to make more sens 1116204912 M * Martin most was copied and pasted from command line 1116205242 J * eXplasm2 explasm@p549FEF46.dip.t-dialin.net 1116205652 Q * Martin Quit: using sirc version 2.211+KSIRC/1.1 1116205683 Q * eXplasm Ping timeout: 480 seconds 1116206256 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com 1116206356 J * lilo_ debian-tor@lilo.usercloak.oftc.net 1116206431 Q * lilo Ping timeout: 480 seconds 1116207841 J * terr ~gilles@ip-213-49-77-210.dsl.scarlet.be 1116207893 P * terr 1116211730 Q * daniel_hozac Ping timeout: 480 seconds 1116213350 Q * ciphernaut Quit: 1116213565 N * Bertl_zZ Bertl 1116213572 M * Bertl back now ;) 1116217517 Q * Shuri Quit: 1116219035 J * ciphernaut ~a@61.88.18.130 1116219732 N * lilo_ lilo 1116219871 Q * ntrs Ping timeout: 480 seconds 1116223907 M * SEAwolfx good morning 1116225497 Q * sukria Remote host closed the connection 1116225511 M * Bertl morning SEAwolfx! 1116225524 M * SEAwolfx ;) 1116225765 M * SEAwolfx does masquerading work with vserver or is snat/dnat required? 1116225861 M * SEAwolfx @ Bertl 1116226020 M * Bertl well, masquerading works quite fine, if you _have_ something to masquerade 1116226062 M * SEAwolfx o. 1116226065 M * SEAwolfx k. 1116226074 M * Bertl if the question means: do I need masquerading to allow a vserver guest with a private ip access to the internet, then the answer is no, use snat instead ;) 1116226098 M * SEAwolfx ahhhh .. 1116226104 M * Bertl (i.e. nothing to forward -> nothing to masquerade ;) 1116226107 M * SEAwolfx doh 1116226144 M * SEAwolfx but snat/dnat needs a static ip. 1116226177 M * Bertl so? 1116226274 M * SEAwolfx ok .. if you have a static ip snat is prefered. 1116226332 M * Bertl it's not a question of preference, it's a question of which chains are used and what targets apply ... 1116226347 M * Bertl (and what you actually want to do ;) 1116226360 M * SEAwolfx true, true ;) 1116226380 J * daniel_hozac ~daniel@h56n2fls32o829.telia.com 1116226765 J * sukria ~sukria@213.223.184.201 1116226922 J * erwan_ho ~erwan@konilope.dyndns.org 1116226949 Q * erwan_ho Quit: 1116226952 J * erwan_ho ~erwan@konilope.dyndns.org 1116227324 Q * atsab Ping timeout: 480 seconds 1116227729 M * SEAwolfx yeah, snat is my friend. 1116227799 Q * cryo_ Quit: ðÏËÉÄÁÀ 1116228288 J * cryo ~say@212.86.243.154 1116228291 Q * cryo Quit: 1116228305 J * cryo ~say@212.86.243.154 1116228629 Q * erwan_ho Remote host closed the connection 1116230291 J * prae ~prae@ezoffice.mandriva.com 1116230691 M * Bertl k, folks ... have a nice day ... I'm off to bed now ;) 1116230697 N * Bertl Bertl_zZ 1116230894 J * jsambrook ~jsambrook@aelfric.plus.com 1116230913 J * atsab ~as@lotes.vtu.lt 1116230935 P * jsambrook 1116231185 Q * sukria Remote host closed the connection 1116231391 Q * SEAwolfx Remote host closed the connection 1116231400 J * SEAwolfx ~mike@128.30.2.236 1116232007 J * sukria ~sukria@213.223.184.193 1116232199 J * erwan_taf ~erwan@81.80.43.77 1116232332 Q * rs Quit: rs 1116234235 J * rs ~rs@staff.lycos.fr 1116234272 J * aba__ ~aba@sol.turmzimmer.net 1116234342 Q * aba Read error: Connection reset by peer 1116234378 N * aba__ aba 1116235159 Q * aba Ping timeout: 480 seconds 1116235340 J * aba ~aba@sol.turmzimmer.net 1116235779 P * erwan_taf Leaving 1116235999 J * hws123 ~Werner@p54A17AD8.dip.t-dialin.net 1116236000 M * hws123 hi 1116236005 M * hws123 good morning 1116236318 M * hws123 I am the guy trying to set up MySQL inside a vserver however it does not work the way it is supposed to be ("Can't start server : Bind on unix socket: Permission denied"). I tried to bind MySQL to the ip address of my vserver and disabled networking but that did not change anything. Any ideas on this? 1116238133 J * mplx ~mplx@85.124.48.74 1116238231 M * Eyck it should be able to bind on unix socket, maybe the socket does not exist? 1116238289 M * hws123 Do I just have to touch that file? 1116238344 M * hws123 No, just touching the file does not work 1116238346 M * hws123 hhmmm 1116238352 M * mplx hello... i've got a quick question... is it ok when chkrootkit reports "SIGNIVISIBLE Adore found" inside a vserver? also chkproc -v -v doesn't work like expected 1116238525 M * hws123 Take a look at http://la-samhna.de/library/rootkits/detect.html to detect Adore 1116238540 M * mplx ok, thanks 1116238625 M * mplx i thought it might be related to the vserver as it might check for things which are not allowed inside a vserver 1116238673 M * hws123 could be 1116238684 M * hws123 but just to be sure, use kern_check.c to detect adore 1116238757 M * mplx kern_check doesn't work inside a vserver... 1116238759 M * mplx "You need superuser privileges to use this program, because only the superuser can read from /dev/kmem." 1116238780 M * hws123 ah ok 1116238782 M * hws123 hhmm 1116238803 M * mplx i guess it will be the same problem with chkrootkit... 1116238807 M * hws123 yes 1116238845 M * mplx i've tested inside 3 vservers, and all report that Adore thing.... i just wanted to be sure it's related to vserver 1116238882 M * mplx maybe someone can have a look at their vserver and run a chkrootkit -x lkm ? 1116238896 M * hws123 ah ok, maybe you could open up a bug report? 1116238898 M * hws123 hold on 1116239156 M * hws123 I ran chkrootkit inside a vserver and it reported nothing 1116239224 M * mplx ah, i see.... ok .... so now i'm worried.... thanks for testing it at your server 1116239240 M * mplx which vserver version do you use? 1116239265 M * mplx i'm using 1.2.10 1116239274 M * mplx kernel 2.4.29 1116239282 M * hws123 2.6.11.6-grsec-vs1.9.5 1116239313 M * mplx hm... 1116239315 M * Doener` hws123: i expect your MySQL issue to be grsec related, please check your host's logs 1116239342 M * Doener` btw, unix domain sockets have nothing to do with the networking stuff ;) (just in case that you wondered about that ;) 1116239349 M * hws123 mplx: util-vserver 0.30.205-r1 1116239380 M * mplx util-vserver 0.30.206-3 1116239398 M * mplx i don't think the util is related to this 1116239493 J * wschalk_ ~Werner@p54A13BB8.dip.t-dialin.net 1116239518 M * wschalk_ it is probably not related 1116239679 M * wschalk_ doener: I don't have grsec enabled at all 1116239697 M * wschalk_ inside my first vserver mysql is just fine 1116239781 Q * hws123 Read error: Operation timed out 1116239790 M * Doener` ok... what does thie give you? cat /proc/virtual//status | tail -n4 | head -n3 1116239999 M * wschalk_ Flags: 0000000202000015, BCaps: ffffffffd44c04ff, CCaps: 0000000000000101 1116239999 Q * atsab Read error: Connection reset by peer 1116240029 M * wschalk_ That's what I did on my host for the vserver in question (context #9) 1116240054 J * atsab ~as@lotes.vtu.lt 1116240130 M * Doener` caps look fine, need to check which flags are those.. (though i don't know of any flag that could affect that...) 1116240191 M * wschalk_ Doener: What is the meaning of those three values? 1116240229 M * wschalk_ As I said I do have a dummy vserver which works fine but MySQL in a copied one does not 1116240232 M * Doener` Flags, well the flags for that context, like fakeinit for example, i.e. they modify the virtualization behaviour 1116240245 M * Doener` BCaps, the usual POSIX capabilities 1116240280 M * Doener` CCaps, context capabilities, like for example icmp raw (which is a limited CAP_NET_RAW that only allows those sockets needed for 'ping') 1116240427 M * wschalk_ My flags are lock and nproc 1116240515 M * Doener` and INFO_INIT, HIDE_NETIF and STATE_INIT ;) 1116240535 M * wschalk_ ok :-) 1116240919 M * wschalk_ Any more ideas? 1116240966 M * Doener` i'm reading the code regarding STATE_INIT atm as that looks suspicious to me 1116240985 M * Doener` but first, it's time for lunch... back in a few minutes 1116240988 N * Doener` Doener|gone 1116241080 M * wschalk_ ok 1116242088 N * Doener|gone Doener 1116242096 M * Doener back now... 1116242156 M * wschalk_ ok 1116242330 M * Doener hm, is that a legacy config server? 1116242335 M * Doener s/server/vserver/ 1116242339 M * wschalk_ What do you mean? 1116242342 J * JohnL ~JohnL@nsnoc.plus.com 1116242374 M * Doener is the vserver using a single file for configuration or the directory-tree with multiple files? 1116242495 M * wschalk_ Directory tree with multiple files 1116242595 M * Doener hm, could you make the output of "vserver --debug start" available? 1116242750 M * wschalk_ Of course hold on 1116242785 M * Doener the 'already in use' error occurs because you touched the mysqld.sock file. it may not exist when you start mysql 1116243003 M * wschalk_ http://pastebin.com/285064 1116243008 M * wschalk_ I deleted it also 1116243016 M * wschalk_ it is not there in the first place 1116243034 M * Doener yep, just wanted to clarify it ;) 1116243135 M * wschalk_ Doener: Ok it is online: http://pastebin.com/285064 1116243280 M * Doener do the flags look the same again? 1116243367 M * wschalk_ yes 1116243460 M * Doener creating a test vserver now... 1116243478 M * wschalk_ ok 1116243623 M * Doener hmm... you touched the file as root, right? 1116243650 M * Doener if so, please try it again, but as some regular user 1116243806 M * wschalk_ :-) 1116243811 M * wschalk_ it works fine now?! 1116243813 M * wschalk_ strange 1116243824 M * wschalk_ I had to give /var/run/mysql to the mysql user 1116243851 M * Doener hm, well, the installation should have done this... 1116243905 M * Doener otherwise i had expected that you had done a "chmod 000 /path/to/vservers/" 1116244020 M * wschalk_ I suppose I am trying to copy a vserver with "cp" but I should use "rsync" to preserve the permissions?! 1116244102 M * Doener hm, cp -a should do as well... 1116244121 M * Doener but cp -R (without -p) won't work 1116244176 M * wschalk_ ah ok, that was my problem then 1116244209 M * wschalk_ thanks a lot 1116244213 M * Doener you're welcome 1116244269 M * Doener opinions on vserver copying vary... Bertl uses dump/restore IIRC, some folks use rsync, i'm happy with cp -a... but i copy a vserver once in a year, so... 1116244310 M * wschalk_ I suppose rsync is just fine 1116244321 M * wschalk_ But rsync is slower than cp 1116245756 M * mep_ freevps - vserver. whats the advantage of vserver? 1116246302 M * Eyck hmm, I don't think rsync is slower, it falls back to cp, so it should be exactly the same in speed 1116246341 M * Doener non rh-specific, vanilla kernel support (incl. 2.6), with 2.6 also: uts virt., load average virt., token bucket scheduler, probably more... i just took a look at the freevps website, i don't know much about it 1116246374 M * wschalk_ eyvk: I checked it, rsync was 15 minutes per vserver, cp was 12 1116246481 M * Doener minutes? how big are those? 1116246594 M * Doener i just copied a ~200MB vserver in about 20 seconds... 1116246636 M * Doener (using cp) 1116246642 M * wschalk_ really? they are about 600mb inside a vmware on a 1.6 ghz p4 mobile 1116246657 M * Doener ah ok, i guess vmware explains it... 1116246743 M * wschalk_ Yeah, it makes everything terribly slow 1116246772 M * wschalk_ by the way the major difference between uml and vserver is that vserver still uses only one kernel, isn't it? 1116246937 M * Doener uml provides an arch for the kernel, while vserver modifies the kernel to present individual views on different areas (f.e. /proc) 1116247035 M * wschalk_ ok 1116247146 M * wschalk_ everything is ok with my vserver now - thanks a lot guys. 1116247318 J * ntrs ruzin@Dardeene-68.188.50.87.charter-stl.com 1116247881 M * wschalk_ ok thanks a lot guys, I gotta go - all the best to you guys! 1116247886 M * wschalk_ thanks a lot again and bye 1116247889 Q * wschalk_ Quit: Leaving 1116248139 J * jsambrook ~jsambrook@aelfric.plus.com 1116248149 P * jsambrook 1116248319 M * yarihm i got that right, there is no way to have a kernel-nfsd inside a vserver, no? 1116248376 M * yarihm and according to FSH, the vservers should be under /var/lib/vservers? 1116248761 Q * sukria Remote host closed the connection 1116248891 J * knoppix_ ~knoppix@dsl-213-023-133-094.arcor-ip.net 1116252927 J * orionpanda ~nospam@netblock-66-245-252-180.dslextreme.com 1116253253 J * sukria ~sukria@213.223.184.201 1116256356 Q * ntrs Ping timeout: 480 seconds 1116256527 M * yarihm hey guy ... anyone awake? i'd need some info on how to unify vservers on debian. i've googled somewhat but didn't find something concrete, though i remember having read something about it some time ago. on my system there is a binary vunify installed under /usr/lib/util-vserver ... no manpage though. gonna check the docs 1116256628 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com 1116256688 M * mep_ DaCa alive? 1116256695 M * mep_ need support for his kernel patches 1116256700 M * mep_ seems to be broken 1116256728 M * mep_ when i use his patch: http://vserver.limehouse.org/patch-2.6.11.9-grsec2.1.5-vs2.0rc1.diff.bz2 1116256747 M * mep_ i got this error: 1116256748 M * mep_ arch/i386/kernel/asm-offsets.c: In function `capable_nolog': 1116256749 M * mep_ arch/i386/kernel/asm-offsets.c:66: error: parse error at end of input 1116256749 M * mep_ include/linux/sched.h:1061: warning: unused variable `child_reaper' 1116256855 Q * ntrs Quit: 1116256976 J * DuckMaster ~duckx@195.75.27.158 1116258631 M * albeiro mep_: show me that file 1116258680 M * mep_ the patch?= 1116258684 M * albeiro no 1116258692 M * albeiro asm-offsets.c 1116258696 M * mep_ wait 1116258701 M * albeiro that one on which compilations fails 1116258742 Q * mplx Remote host closed the connection 1116258815 M * mep_ so whats the ebst way to supply you the file? 1116258840 M * albeiro whatever you want, do not have any http server to put it on there ? 1116258917 M * mep_ http://rafb.net/paste/results/40XvTA83.html 1116259063 M * albeiro not that there would be this function error is in 1116259161 M * mep_ CHK include/linux/version.h 1116259161 M * mep_ CC arch/i386/kernel/asm-offsets.s 1116259161 M * mep_ arch/i386/kernel/asm-offsets.c: In function `capable_nolog': 1116259161 M * mep_ arch/i386/kernel/asm-offsets.c:66: error: parse error at end of input 1116259161 M * mep_ include/linux/sched.h:1061: warning: unused variable `child_reaper' 1116259163 M * mep_ make[1]: *** [arch/i386/kernel/asm-offsets.s] Error 1 1116259163 M * mep_ make: *** [arch/i386/kernel/asm-offsets.s] Error 2 1116259254 M * albeiro try put lines 61-62 into one line 1116259263 M * albeiro it may be the problem 1116259312 J * mep__ mep@p5091D6E1.dip.t-dialin.net 1116259321 Q * mep__ Remote host closed the connection 1116259361 J * mep mep@p5091D6E1.dip.t-dialin.net 1116259366 M * mep sry discon 1116259369 M * mep albeiro try put lines 61-62 into one line 1116259371 M * mep one moment 1116259396 M * albeiro just glue them toogether, your compiler might get confused there 1116259426 M * mep nope doesn't help 1116259468 M * albeiro so your just have to wait for patch author :] 1116259474 M * albeiro s/your/you/ 1116259482 M * mep narf 1116259488 M * mep ok thx anyway for the support .) 1116259527 Q * eXplasm2 Quit: Verlassend 1116259537 J * eXplasm explasm@p549FEF46.dip.t-dialin.net 1116259540 Q * eXplasm Quit: 1116259568 J * eXplasm explasm@p549FEF46.dip.t-dialin.net 1116259647 M * DaCa very strange, as I testcompiled it with allyesconfig 1116259663 M * mep daca :D 1116259738 Q * mep_ Ping timeout: 480 seconds 1116259761 M * DaCa mep_: can you email your .config to danny@limehouse.org? I can't look at it now as I have to leave for a concert, maybe I can look at it tonight, otherwise tomorrow evening 1116259828 M * mep arg :( 1116259834 M * mep yes ok 1116260602 M * yarihm could anyone point me to a doc that describes how to unify vservers? i searched but didn't find anything yet. vunify says that it is not configured (i must have missed something there, didn't find any manpage or so) 1116260610 Q * JohnL Quit: Trillian (http://www.ceruleanstudios.com 1116260900 Q * prae Quit: Client exiting 1116261792 Q * DuckMaster Quit: Leaving 1116263764 Q * infowolfe Quit: Leaving 1116264569 Q * rs Quit: rs 1116264675 J * erwan_ho ~erwan@konilope.dyndns.org 1116265323 J * infowolfe infowolfe@66-230-103-130-cdsl-rb1.nwc.acsalaska.net 1116267257 Q * erwan_ho Remote host closed the connection 1116267406 J * hillct ~hillct@client200-5.dsl.intrex.net 1116267465 M * Doener DaCa: the problem does only show up without CONFIG_SECURITY, as the code is #ifdef'ed 1116267761 J * janra janra@paradox.homeip.net 1116267959 M * Doener mep: around? 1116268048 M * Doener mep: in include/linux/sched.h, after line 1011 add: 1116268052 M * Doener return 0; 1116268053 M * Doener } 1116268067 Q * duckx Read error: Operation timed out 1116268154 M * Doener got to go now... 1116268158 N * Doener Doener|gone 1116268175 M * hillct gah! 1116268181 M * hillct damnit. timing is everything 1116268305 M * eXplasm hey Doener :) whats with your vserver webinterface? you have said (long long time ago) that it could go under an GPL/Opensource licence, any news? 1116268323 M * eXplasm what an english xD 1116268735 J * monrad ~monrad@0x535b06c0.ronxx3.adsl-dhcp.tele.dk 1116268983 J * duckx ~Duck@mna75-1-81-57-39-234.fbx.proxad.net 1116269047 J * muadda ~muadda@gar31-1-82-234-50-167.fbx.proxad.net 1116269104 P * hillct 1116269912 J * mef ~mef@pcp09895218pcs.ewndsr01.nj.comcast.net 1116270583 Q * Johnsie Quit: G'bye! 1116271048 M * mep Doener big thx! 1116271097 N * Bertl_zZ Bertl 1116271106 M * Bertl evening folks! 1116271113 M * mep hi bertl! 1116271621 M * Eyck evening! 1116271707 N * Doener|gone Doener 1116271714 M * Doener back now.. 1116271723 M * Doener mep: you're welcome 1116271735 M * yarihm does any of you have any pointers on why after having created /etc/vservers//apps/vunify vserver status gives wrong information? (i.e. says it's down) 1116271740 M * yarihm oh, hi Bertl 1116271777 M * Doener eXplasm: nothing yet... i need 48hour-days ;) 1116271799 M * Doener or even better: 60hour days, IIRC enrico already has those ;) 1116271827 M * yarihm i wanted to unify the servers on my debian testing machine, i read the alpha-utils page and followed the instructions there. during systemstart the vservers come up, but i can't manage them with these dirs in /etc/vservers anymore? how come? 1116271835 M * Bertl evening Doener! 1116271843 Q * SEAwolfx Ping timeout: 480 seconds 1116271845 M * Doener hey Bertl! 1116271903 M * Doener Bertl: back in your timezone? ;) 1116271916 J * Johnsie ~john@acs-24-154-32-12.zoominternet.net 1116272035 M * mep great doener works now :) 1116272046 M * mep but i get the error only on my dual cpu machine 1116272078 M * Bertl Doener: yeah, just got up ... trying to put myself together now ;) 1116272088 M * Doener different config i guess... as i said, it only occurs when CONFIG_SECURITY is off 1116272106 M * yarihm come on ... is it a stupid error i made (that wouldn't be surprising)? any pointers, please? 1116272141 A * Bertl .o( still suffering from a friends birthday party ...) 1116272154 M * Doener yarihm: i would love to help, but i've no idea about the unification stuff ;) 1116272164 M * yarihm thus not able to answer stupid questions ,) 1116272201 M * yarihm Doener, lucky you :) i even started reading the source code until i found the website, there is no manpage for vunify yet 1116272290 M * yarihm Bertl, are you in a state yet where you could answer questions about unification? mkdir -p /etc/vservers//apps/vunify lead to some strange errors here ... 1116272415 J * matta ~matta@69.93.28.254 1116272464 M * matta Bertl: around? 1116272509 M * Bertl nope, I'm a-square ;) 1116272521 M * matta wi rs 1116272524 M * Bertl evening matta! 1116272532 M * matta good afternoon! 1116272535 M * matta question about ng... 1116272544 M * matta i haven't really been following, super busy lately 1116272561 M * matta with your plan, what is used as the gateway for the vserver? 1116272569 M * Bertl Doener: CONFIG_SECURITY is it a grsec issue? 1116272591 M * Doener no, a patch issue... 1116272596 M * Bertl matta: well, you use the 'gateway' the vserver will require 1116272618 M * matta well, let's say vs is 192.168.1.2/24, gateway is 192.168.1.1 1116272621 M * Bertl i.e. vserver guest has 10.0.0.2 as ip, and gateway 10.0.0.254 (maybe)? 1116272627 M * matta is the default gw inside the vs set the gw to .1 ? 1116272636 M * Bertl sure ... 1116272651 M * matta I suppose with proxy arp that could be made to use an alias on the host as the gateway 1116272659 J * Doener` ~doener@p54877A01.dip.t-dialin.net 1116272671 M * Bertl matta: well, with proper arp inside, yes 1116272691 M * matta you'd need to add a static route in the vs and then add as gateway 1116272694 M * matta ok, that is cool then 1116272702 M * matta i was worried it wasn't going to work that way 1116272721 M * Doener` http://www.13thfloor.at/~doener/broken.c 1116272725 M * matta (when maintaining hundreds of vservers across dozens of subnets it's difficult to keep track of subnets/gateways) 1116272738 M * Doener` that's the broken code fragment... 1116272762 M * Doener` it's missing a "return 0; }" 1116272770 M * matta so it's easy to just use the host server as the 'gateway' and have the host do proxy arp 1116272796 Q * Doener Read error: Operation timed out 1116272841 M * Bertl matta: hmm, why proxy arp? 1116272853 M * Bertl (didn't get this part yet ;) 1116272869 M * yarihm hmm ... Bertl, should i ask later about this vunify-thing? i followed http://linux-vserver.org/alpha+util-vserver and have util-vserver 0.30.204 installed 1116272888 M * Bertl yarihm: update to 0.30.207, you have vhashify there ... 1116272907 M * yarihm which is better? 1116272909 M * Bertl (doesn't require a template server) 1116272926 M * yarihm ok, that's true ... 1116272930 M * matta Bertl: well, some method in which the host acts as a router 1116272940 M * matta and not a bridge 1116272955 M * Bertl why do you want that? 1116272958 M * matta but not _exactly_ like a router... you don't want to have to have the upstream router route subnet's directly to the host 1116272974 M * matta Bertl: like I stated for simple network management with thousands of vservers and hundreds of subnets 1116273007 M * Bertl I don't see why you would like the overhead of routing ... 1116273018 M * matta it's what virtuozzo does and what I had to do with xen 1116273060 M * Bertl emphasis on _had to do_ 1116273081 M * matta well, didn't have to do... but for any sort of sanity, I did 1116273098 M * Bertl look, for the folks who absolutely desire to do that, we will add an userspace tool which 'copies' from one tun device to the other ;) 1116273119 M * Bertl everything else is done via virtual shared network interfaces 1116273122 M * matta oh, that is possible? 1116273124 M * Bertl (for performance reasons) 1116273148 M * matta i think the way you say you are doing it I can do the same thing I did with xen 1116273155 M * matta seems similar 1116273226 M * Bertl well, ngnet is based on _sharing_ the interface ... 1116273241 J * Johnny ~john@acs-24-154-32-12.zoominternet.net 1116273251 M * Bertl so what you do is, you tell the system, take interface eth0, make a shared copy as en0 in xid=100 and be done 1116273277 M * Bertl packets leaving en0 inside the vserver, will _leave_ the eth0 interface if permitted 1116273283 M * matta right 1116273299 M * Bertl no routing involved except for the 'normal' host routing inside the guest 1116273300 M * matta there will be ip access granted specifically to devices? 1116273320 M * matta so it's acting like a bridge 1116273325 Q * Johnny Quit: 1116273339 M * Bertl matta: like a bridge with no bridging .. ;) 1116274024 M * yarihm Bertl, i upgraded to 0.30.207 as you suggested. the issue with vserver foo status displaying the wrong status is still there and vserver foo hashify tells me "Failed to initialize unification for vserver" 1116274102 J * Martin ~martin@host-138-38-231-33.nat.bath.ac.uk 1116274108 M * Martin Evenin' 1116274113 M * Doener` hi Martin 1116274119 M * yarihm there are quite strange things (to me) ... vserver foo status says that the vserver is stopped even though it is running (aliases are up, the init-script don't print any errors). when trying vserver foo start it says: line 623: pushd: /etc/vservers/foo/vdir: No such file or directory ... 1116274125 M * Bertl yarihm: wrong status? please elaborate! 1116274129 M * Doener` Martin: could you recompile the tools with fscompat support? 1116274144 M * Doener` (on your sparc64 that is) 1116274148 M * Martin Will do in a bit. 1116274155 M * Martin Have to get food first 1116274157 M * Bertl evening Martin! you have troubles with sparc64? 1116274167 M * Martin Was just dropping past to see who was about later on tonight 1116274170 M * Doener` ok, no problem, just wanted to make sure that i don't forget to ask/tell you ;) 1116274177 M * Martin Bertl: yes and no 1116274186 M * Martin Doener`: OK - where do I set it? 1116274192 M * yarihm i could offer a sparc64, too if someone needed access ... an E3000 1116274197 M * Bertl was working here in a few minutes without big trouble ... 1116274199 M * Martin Bertl: Most of it works, just got a few problems 1116274226 M * Bertl yarihm: sounds good, had to return the one I had here last week :/ 1116274233 M * Martin yarihm: That could be useful, would help catch when I'm being stupid 1116274255 M * Martin Bertl / Doener: Will you guys be about in an hour or so? Must eat. 1116274272 M * Doener` not sure, but probably 1116274277 M * yarihm sure ... connection is not too cool (2mbit down, 400KBit up), but i'll be glad to provide shell if it is of any use 1116274294 M * yarihm the machine is debian though ... dunno whether this will be appreciated :) 1116274303 M * Doener` ./configure --enable-apis=compat,v11,v13,net ... 1116274315 M * Martin yarihm: good to find another Debian / SPARC / vserver user 1116274328 M * yarihm Martin, running debian, too? 1116274329 M * Doener` (that gives the default minus fscompat) 1116274349 M * Martin yarihm: Yep. :-) 1116274364 M * yarihm the sarge-installer made me crazy ... this /boot/etc -> /boot hack-shit got me really angry 1116274399 M * Martin Doener`: OK, and I should compile with fscompat ON 1116274412 M * Martin yarihm: I haven't had too many issue with it so far 1116274424 M * Doener` no, with fscompat OFF, i.e. with the above configure option 1116274455 M * Doener` ah, i wrote "with" above... sorry, should have been "without" 1116274455 M * yarihm hmm ... here the partitioner didn't work as i expected it and as i'm normally not using a /boot partition it blew a few times :( ... but after some time 1116274457 M * Martin Doener`: OK, your initial message said with fscompat support. Will turn if OFF and recompile 1116274458 M * Bertl Martin: yes, will be around ... 1116274471 M * Martin Bertl: thanks 1116274483 M * yarihm Bertl, is this a known thing with this wrong status? 1116274508 M * Martin yarihm: Yea, partitioner seems a little odd but I manually partition everything anyway so it doesn't effect me much. 1116274515 M * Martin OK all: back in a bit. 1116274524 M * yarihm Bertl, i mean if i move /etc/vservers/foo to /etc/vservers/foo.bak things work as expected again 1116274543 M * yarihm l8er Martin 1116274787 J * jdi jdi@c135097.adsl.hansenet.de 1116274861 M * Bertl welcome jdi! 1116274871 M * jdi hi 1116274880 M * Bertl gcc? 1116274884 M * jdi yep 1116274899 M * jdi givme a minute 1116274907 M * Bertl no problem ;) 1116275021 M * Doener` hm? 1116275201 M * Bertl Doener`: http://l4x.org/k/ 1116275381 M * jdi Bertl: we should go over the results with 'your' toolchain 1116275394 M * Bertl Jan and I are somewhat joining forces to improve kernel compile tests ;) 1116275406 M * Doener` ah, ok 1116275448 M * Bertl jdi: jup, but maybe we should make a new channel, or move to private? what do you think folks? 1116275450 M * yarihm hmm ... ok then. gotta go, tomorrow university-week begins ... cu guys 1116275456 M * yarihm thanx for the help 1116275461 M * yarihm n8 everyone 1116275463 M * Bertl cya! 1116275466 M * Doener` good night yarihm 1116275470 Q * yarihm Quit: Leaving 1116275502 M * Bertl Doener`: you're interested in fixing up toolchains? 1116275513 M * jdi Bertl: i don't care 1116275529 M * Doener` Bertl: that would mean doing what? 1116275552 M * Bertl that would mean that I would create a channel instead of going private ;) 1116275577 M * albeiro heya :] what's wrong with toolchain ? 1116275610 M * Bertl okay, a new channel it is ;) 1116275622 M * Doener` i'm more interested in what 'fixing up toolchains' would (probably) involve... ;) i need some way to decide whether i'm up to that... 1116275636 M * Bertl #toolchain (feel free to join there ;) 1116277822 J * Tbery ~tb@pha-84-242-95-4.nat.karneval.cz 1116277827 M * Tbery HI 1116277885 M * Tbery Bertl, are you here?? 1116277913 M * Bertl yes, I am?! 1116277997 M * albeiro He is here and there ;p 1116278055 M * Bertl thanks god, not everywhere ;) 1116278133 M * albeiro Bertl: if i can use your oops hunting knowledge a bit, have you tried it on ppc ? 1116278156 M * Bertl yup ... as I own a TI PB ... 1116278166 M * albeiro i got verid oops to debug, but it does not contain Code: line, so it is almost impossible (or very hard) to do it. 1116278224 M * albeiro zeus.polsl.gliwice.pl/~albeiro/oops if you could look, ksymoops tells me interesting things, but without Code: i have no idea about shadow internals of this function ;p 1116278231 M * Tbery Im trying install on old tar sources..biuld up new vserver 1116278239 M * Tbery on kernel 2.4.29 1116278255 M * Tbery I compiled sources and make server 1116278354 M * Tbery where I have /etc/vservers/newservers config?? 1116278367 M * Tbery vserver borg1 build -m debootstrap --hostname borg1.shell4u.com --context 66 -d sarge -m ftp://ftp.cz.debian.org/debian 1116278374 M * Tbery I maked it like this.. 1116278377 M * albeiro gezz... 2.6.11.10 released 1116278395 M * albeiro root in pktcdvd 1116278399 M * albeiro root in raw device 1116278442 M * Martin Doener`: disabling fstools seems to have fixed the bus error, am doing a "from the top" test and will report back 1116278515 M * Doener` great :) 1116279049 M * Bertl albeiro: h,, why not use addr2line? 1116279152 M * albeiro because i have no idea about its existence ? 1116279155 M * albeiro ;] 1116279178 M * albeiro i am fairly new to oops debuging, this is my first time i am doing it entirely alone 1116279273 M * Bertl okay, you got an idea now, no? 1116279314 M * albeiro i hope so, will find that script, convert information which i have and will know which line it is 1116279337 M * albeiro already figured out how to read ksymoops output 1116279351 M * albeiro thx for huge tip :] 1116279390 Q * Tbery Quit: Ukonèuji 1116279392 M * albeiro zeus.polsl.gliwice.pl/~albeiro/oops2 <- my ksymoops output, a "bit" strange ;] 1116279719 M * Bertl is that kernel booted? 1116279729 M * albeiro yes, it is 1116279754 M * albeiro with every reasonable debuging function enabled 1116279873 M * Bertl what's that strange __crc stuff? 1116279887 M * Bertl is that some rsbac feature? 1116279899 M * albeiro not at all, i have no idea where it comes from 1116279918 M * Doener` Martin: it doesn't make sense to me though... the disabling of fscompat was more or less a wild guess (i just knew where it bailed out). AFAICT it should use the v13 call anyways... but there's A LOT of preprocessor stuff going on there and i feel quite lost ;) 1116279919 M * albeiro there are some funny crc modules in the kernel, that might be it 1116279954 M * Bertl no, looks to me like 'unused' module versioning ... 1116279976 M * Bertl __crc___dev_get_by_index is definitely no crc function ;) 1116279984 M * albeiro hehehe, right 1116280028 M * Martin Doener`: Yea - that was my feeling. I /thought/ it was v13 as well FWIW 1116280267 M * Martin Bertl / Doener - does either of you have some time to give me some pointers? Have documented where I'm up to on linux-vserver.org/sparc64 but could do with some expert guidance 1116280455 M * Doener` eek, don't use "Doener" and "expert" in the same sentence ;) 1116280459 Q * muadda Quit: Leaving 1116280481 M * Bertl hmm, he didn't actually ;) 1116280499 M * Martin :-D 1116280516 M * Bertl so what are the issues with sparc64? 1116280527 M * Martin Well I'd be greatful of any assistance, regardsless of what level of expertise it's owner thinks it has 1116280537 M * Bertl did you build the ultra up-to-date dietlibc yet? 1116280546 M * Doener` regarding the startup problem, let me create a code snippet for Bertl to (hopefully) explain me the code the preprocessor generates, i'll then look into that... 1116280579 M * Martin I think I've got the latest one - I've been using your wiki as a note book :-D 1116280596 M * mep is there a way to test a kernel in runtime? 1116280609 M * Bertl mep: test what? 1116280615 M * mep that it work 1116280625 M * mep i think i killed my dual cpu box again 1116280626 M * Bertl Martin: well, 0.28 + the required patches 1116280636 M * Martin Bertl: yep, think so 1116280641 M * Bertl mep: what aspects would you like to test? 1116280644 M * mep newwest vserver+gr securitypatch 1116280650 M * mep that it can boot 1116280667 M * mep i have only remot4e axx zu der box 1116280671 M * Bertl you mean, test if it will boot, without booting it? 1116280678 M * mep yes ;) 1116280685 M * Martin Boot it in a local emulator 1116280691 M * Bertl you can use an emulator for that 1116280693 M * mep narf boot don't come back 1116280697 M * Martin snap! 1116280700 M * mep nice reinstalla gain DAMN IT 1116280702 M * Bertl mep: like bochs 1116280704 M * mep *box 1116280731 M * Bertl mep: but usually it helps if you specify panic=30 or so 1116280739 M * mep what means that? 1116280747 M * Bertl (at command line, that will reboot the machine after a kernel panic) 1116280756 M * mep and boots what? 1116280770 M * Bertl if you use a one time boot option (with lilo or grub) 1116280779 M * Bertl then it will reboot with the old kernel 1116280797 M * mep hmm thats sounds good 1116280802 M * mep you are longer on this night 1116280811 M * mep i must wait for the one with the confixx axx 1116280828 M * Bertl probably .. don't know yet 1116280838 M * mep kk 1116281013 J * monski ~monrad@213083190130.sonofon.dk 1116281209 M * Bertl evening monski! 1116281233 M * Bertl Martin: okay, after building 0.30.207, what does the testme.sh say? 1116281289 M * Martin Bertl: testme.sh seems to say all clear - I'll just run it again. Paste to the channel? 1116281308 M * Martin Everything succeeds 1116281319 M * Martin OK, have test case for the lockfile bug 1116281346 M * Bertl Martin: no, if it says all fine, then the interface works ... 1116281390 M * Doener` Bertl: the bug is in the tools and/or dietlibc... 1116281423 A * Bertl doesn't even know _what_ doesn't work yet ;) 1116281430 M * Martin lockfile 1116281446 Q * monrad Ping timeout: 480 seconds 1116281475 M * Martin A test case: mkfifo /tmp/fifo ; /usr/local/lib/util-vserver/lockfile /tmp/lock /tmp/fifo ; cat /tmp/fifo (from other shell) 1116281484 M * Doener` if fscompat and v13 apis are enabled, the code to choose one doesn't seem to work, returns with -1 and the following perror gets an unaligned string causing a SIGBUS 1116281511 M * Doener` that's what made the setattr fail... for the lockfile stuff i expect something similar 1116281516 M * Martin Oops - I'm working on a different bug then 1116281542 M * Martin subsequent runs through a bus error immeidately 1116281556 M * Doener` Martin: i just explained your previous bug, since saying 'disable fscompat' isn't really a bugfix ;) 1116281572 M * Bertl probably the same alignment issues as with the hostname (the ones I fixed) 1116281608 M * Bertl obviously alignment was done 'with good hope' isntead of testing ;) 1116281635 M * Martin Doener`: Oh, OK 1116281663 M * Martin From looking at the strace of the current one I'm guessing it's similar 1116281721 Q * knoppix_ Quit: Verlassend 1116281729 M * Martin Checking the source I recon it's perror again. Hmmm... time ot look at dietlibc... 1116281963 M * Martin Given that perror in dietlibc is fairly simple what's the betting that it's an alignment bug in write? 1116282266 N * monski monrad 1116282270 M * monrad evening Beave 1116282274 M * monrad Bertl i mean 1116282865 J * rs ~rs@imhotep.rhapsodyk.net 1116282870 M * Bertl evening rs! 1116282890 M * rs evening bertl! 1116282917 M * rs lycos vds is now running 2.0-rc1 :) 1116282928 M * rs since last friday 1116282950 M * Bertl excellent, everything working fine so far? 1116282964 M * rs seems to be ok, just got some vservers that lost their network connectivity 1116282967 M * rs dunno why 1116282980 M * rs maybe it's unrelated, the IP was just unbound 1116282983 M * Bertl interesting, no primary/secondary issue? 1116282993 M * rs no, all IP are secondaries 1116283036 M * DaCa Doener`: tnx for looking at it, this was indeed a major fsckup from my side, I remember me pondering if I hadn't lost a return somewhere. 1116283054 M * Doener` DaCa: no problem 1116283100 M * Doener` did you check if capable_nolog() should also get the vx_something check that capable() has? 1116283132 M * DaCa well, on the positive side, one learns the best from his errors 1116283174 M * Doener` ACK! 1116283177 M * DaCa I don't think so, I'll load it on my laptop and dig deeper into it tomorrow on my commute 1116283219 M * Doener` well, you know that stuff ;) i just noticed that when i saw the code 1116283261 M * DaCa I don't pretend that I know the stuff :p 1116283283 M * Doener` soon, you will ;) 1116283403 M * DaCa yes, I am working on it, I also got Robert Love's linux kernel development book which I am going thru atm 1116283665 M * Martin OK, there definately appears to be something up with perror() in dietlibc 1116283678 M * Martin What's the viability of linking with glibc? 1116283710 M * Bertl you can 'try' but it has resolver issues ... 1116283774 M * Doener` http://archives.linux-vserver.org/200504/0101.html 1116283782 M * Martin But if it's only for vserver then surely that shouldn't be too much of a problem 1116284384 M * Bertl okay, off for now .. back later ... 1116284389 M * Doener` cya 1116284390 N * Bertl Bertl_oO 1116284648 M * mep arg bertl :( 1116284745 M * albeiro hm, addr2line is gives me strange results. kernel coding is twisted ;p 1116284778 M * mep albeiro we got the error ;) 1116284803 M * albeiro i saw it, strange it was 1116284885 M * mep yes 1116284894 M * mep but the server can't boot with the kernel :( 1116284905 M * mep seems to be a long night... 1116284977 M * Doener` Martin: do you have a strace of the failing save_ctxinfo? 1116285098 M * Martin Ummm... not sure I've run save_ctxinfo 1116285156 M * Martin Can do if you need one though 1116285171 M * mep fuck again kernel update!? 1116285193 M * Martin Don't seem to have a 64 bit strace so args seem to be a little scrambled but mostly OK 1116285213 M * Doener` Martin: hm, save_ctxinfo was the proces getting a SIGBUS, wasn't it? 1116285350 M * Martin Name sounds familiar but I can't see it in the info I've kept on sparc64 1116285364 M * Doener` save_ctxinfo: /usr/local/sbin/vserver: line 142: 21763 Bus error ${NICE_CMD[@]}..... 1116285387 A * Martin looks down 1116285396 M * Martin OK ignore me - I can't read. 1116285399 M * Doener` the third sigbus thing 1116285418 M * Martin There is definately a bug in perror, I'll see if I can fix it. 1116285436 M * Martin Yes - I see it. 1116285506 M * Martin What args do you want strac'ing? 1116285524 M * Martin If I can use strace as a verb :-) 1116285550 M * Doener` hm, let me figure out a command line... save_ctxinfo will bail out in ctx 0... 1116285868 M * Doener` vcontext --create --silent --xid 12345 strace /usr/lib/util-vserver/save_ctxinfo /etc/vservers/ /bin/true 1116285898 M * Doener` the vserver may not be running, /var/run/vservers/ may not exist, and you may have to adjust paths 1116286189 M * Martin /bin/true 1116286190 M * Martin execve("/usr/local/lib/util-vserver/save_ctxinfo", ["/usr/local/lib/util-vserver/save_ctxinfo", "/usr/local/etc/vservers/frequency", "/bin/true"], [/* 16 vars */]) = 0 1116286190 M * Martin SYS_267(0, 0x3f, 0, 0, 0) = 65573 1116286190 M * Martin write(2, "save_ctxinfo: ", 14save_ctxinfo: ) = 14 1116286190 M * Martin write(2, "vc_get_task_xid()", 17vc_get_task_xid()) = 17 1116286191 M * Martin write(2, ": ", 2: ) = 2 1116286191 M * Martin write(2, "[unknown error]", 15[unknown error]) = 15 1116286193 M * Martin write(2, "\n", 1 1116286193 M * Martin ) = 1 1116286195 M * Martin exit(255) = ? 1116286203 M * Martin The syscall looks familiar. 1116286220 M * Martin I'm going ot have to go for a bit but I'll be back in 15 -> 30 mins. 1116286238 M * Doener` that one returns the kernel vserver api version 1116286262 M * Doener` ok, cya! 1116286262 M * Martin Also have a look at linux-vserver.org/sparc64 as i think I've got a work around for the bus errors. 1116286345 M * mep so whats the ebst way to test a kernel functionality in runtime? 1116286355 M * Doener` hm, that strace was made with the tools linked against dietlibc? 1116286442 M * Doener` mep: the vserver interface (and the tools) can be tested with http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh 1116286464 Q * Eyck Read error: Operation timed out 1116286479 M * mep no my problem is that the patched kernel won't boot on this machine 1116286490 Q * jdi Quit: Leaving 1116286519 M * mep the unpatched works fine 1116286522 M * mep same config 1116286536 M * mep then i patchd it apply the new configs and it won't boot 1116286547 M * mep i ahve no physical axx tahts my problem... 1116286578 J * infowolfe_ infowolfe@66-230-103-130-cdsl-rb1.nwc.acsalaska.net 1116286828 Q * infowolfe Read error: Operation timed out 1116286851 M * Doener` Martin: compile with ./configure --enable-apis=v13,net please... there has to be something wrong with that preprocessor hell... 1116286906 Q * rs orion.oftc.net charm.oftc.net 1116286906 Q * monrad orion.oftc.net charm.oftc.net 1116286906 Q * Johnsie orion.oftc.net charm.oftc.net 1116286906 Q * mef orion.oftc.net charm.oftc.net 1116286906 Q * matta orion.oftc.net charm.oftc.net 1116286906 Q * cryo orion.oftc.net charm.oftc.net 1116286906 Q * Hollow orion.oftc.net charm.oftc.net 1116286906 Q * bro orion.oftc.net charm.oftc.net 1116286906 Q * locksy orion.oftc.net charm.oftc.net 1116286906 Q * matti orion.oftc.net charm.oftc.net 1116286906 Q * kevinp orion.oftc.net charm.oftc.net 1116286906 Q * ndim_ orion.oftc.net charm.oftc.net 1116286906 Q * FaUl orion.oftc.net charm.oftc.net 1116286906 Q * hellekin orion.oftc.net charm.oftc.net 1116286906 Q * Vudumen orion.oftc.net charm.oftc.net 1116286906 Q * flock orion.oftc.net charm.oftc.net 1116286906 Q * gaba orion.oftc.net charm.oftc.net 1116286923 M * Doener` Martin: then, in a seperate location, please do a plain "./configure; make;" (don't install) and: 1116286942 M * Doener` cpp -D HAVE_CONFIG_H -I. -Iensc_wrappers lib/syscall_setiattr.c setiattr_processed.c 1116286955 M * Doener` and make that file available somehow 1116286982 J * rs ~rs@imhotep.rhapsodyk.net 1116286982 J * monrad ~monrad@213083190130.sonofon.dk 1116286982 J * matta ~matta@69.93.28.254 1116286982 J * Johnsie ~john@acs-24-154-32-12.zoominternet.net 1116286982 J * mef ~mef@pcp09895218pcs.ewndsr01.nj.comcast.net 1116286982 J * cryo ~say@212.86.243.154 1116286982 J * Hollow ~Hollow@home.xnull.de 1116286982 J * bro ~vanity@lanparty.lv 1116286982 J * matti matti@linux.gentoo.pl 1116286982 J * locksy ~locksy@mrtg.sisgroup.com.au 1116286982 J * kevinp ~kevinp@ny.webpipe.net 1116286982 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il 1116286982 J * Vudumen vudumen@perverz.hu 1116286982 J * hellekin ~hellekin@v41.ath.cx 1116286982 J * FaUl ~immo@ip88.164.1211G-CUD12K-01.ish.de 1116286982 J * ndim_ hun@helena.bawue.de 1116286982 J * gaba ~gaba@protest.net 1116287046 Q * Hollow Ping timeout: 480 seconds 1116287057 J * Hollow ~Hollow@home.xnull.de 1116287069 M * albeiro hm 1116287100 M * albeiro if i will make unification is there any way to back up "attributes" before backup ? 1116287114 M * albeiro grr 1116287118 M * albeiro before upgrade ;p 1116287399 Q * monrad Quit: Leaving 1116287448 J * monrad ~monrad@213083190130.sonofon.dk