1170460936 P * jbailey Ex-Chat 1170461436 Q * mcp Read error: Connection reset by peer 1170461445 J * mcp ~hightower@wolk-project.de 1170461638 N * ebiederm ebiederm_away 1170462294 Q * Aiken Quit: Leaving 1170462903 Q * gerrit Ping timeout: 480 seconds 1170464768 M * baldy however maybe openvcp brings an update 1170464875 M * daniel_hozac sure. prodding the right people might make a new version come about. 1170464885 Q * rpc Ping timeout: 480 seconds 1170464893 M * daniel_hozac (who the right people are, i have no idea to be honest) 1170465634 M * baldy daniel_hozac: did u know a image mirror for vserver images? 1170465658 M * baldy when not... 1170465667 M * baldy maybe i will maintain a site 1170465673 M * daniel_hozac there (used?) to be a section on the wiki, but IMHO it's easier/safer to just build your own guests. 1170465717 M * baldy debian guest us quite easy but what about centos/redhut/ubuntu and so on 1170465728 M * daniel_hozac they're all easy. 1170465745 M * daniel_hozac only problematic ones are OpenSuSE and Mandriva, AFAIK. 1170465761 M * baldy opensuse works fne 1170465767 M * baldy 10.0 1170465769 M * daniel_hozac (because nobody has put in the time to get them working, basically) 1170465791 A * baldy had a 10.0image wich works 1170465807 M * daniel_hozac i believe i tried 10.2. 1170465820 M * daniel_hozac while it's installable, the result is horrible. 1170465831 M * baldy hehe 1170465843 M * daniel_hozac the service dependencies pretty much preclude you from running anything short of a complete system. 1170465866 A * baldy think that i will maintain image mirror with guest system... 1170465879 M * baldy traffic and space will be no problem 1170465948 M * xp_prg I hate to ask but my cpan in my vserver appears to not be working well, it keeps saying: Writing Makefile for File::Temp -- NOT OK 1170465948 M * daniel_hozac IMHO documenting the procedure to build said images would be more valuable. 1170465958 M * xp_prg how could it not be ok to write the Makefile? 1170465975 M * daniel_hozac you tell us. 1170465989 M * daniel_hozac is there some form of verbose switch? 1170465999 M * xp_prg not that I know of 1170466116 M * daniel_hozac try strace then 1170466180 M * daniel_hozac strace -fF -o trace command... and upload trace somewhere. 1170466539 M * xp_prg daniel_hozac I want to try something else real quick, I want to a make a debian vserver since that is what I am running on, I am not clear exactly how to do that, can you assist? 1170466567 M * daniel_hozac vserver ... build --help shows a debootstrap build method. 1170466608 M * daniel_hozac so, vserver build -m debootstrap ... -- -d should work. 1170466641 M * xp_prg how do I know what version of debian I am running? 1170466696 M * daniel_hozac uh, cat /etc/debian?release maybe. 1170466714 M * xp_prg cat /etc/debian_version 4.0 1170466752 M * daniel_hozac so... etch, or maybe sid. 1170467146 M * baldy xp_prg: 1170467153 M * baldy vserver test.de build -n test --hostname test --rootdir /home/tester --netdev eth1 --interface test=123.123.12.123 --context 50 -m debootstrap -- -d sarge for example 1170467156 M * baldy ;) 1170467195 M * daniel_hozac test.de -n test? that sounds like it's gonna cause pain. 1170467232 M * daniel_hozac if it works it's probably more by accident than anything else. 1170467322 M * xp_prg vserver debian build --help 1170467330 M * xp_prg that didn't work for me, did I enter that command wrong? 1170467343 M * daniel_hozac why didn't it work? 1170467359 M * xp_prg vserver debian build --help Usage: vserver-build -m -n --force * --rootdir --pkgbase [--] * Options: --force ... remove/rename already existing vservers with the same name 1170467368 M * xp_prg it just spit out all the different ways to run vserver 1170467383 M * daniel_hozac to run vserver_-build_, yes. 1170467402 M * baldy daniel_hozac: it works hehe 1170467404 M * xp_prg perhaps I am confused, what is the command to create a debian vserver on debian? 1170467407 M * baldy just a test... 1170467425 M * baldy xp_prg: as i wrote 1170467436 M * daniel_hozac baldy: just out of curiousity, what does the guest end up being named? test.de or test? 1170467448 M * baldy mom 1170467452 M * daniel_hozac the latter, right? 1170467512 M * xp_prg oh ok, I saw your example thanks 1170467521 M * daniel_hozac about guest images, http://linux-vserver.org/Downloads#Guest_images has a list. 1170467662 M * baldy test:~# 1170467665 M * baldy is the name 1170467733 M * daniel_hozac not the hostname, the name of it. i.e. what you use to vserver ... {enter,start,stop,exec,etc} it. 1170467752 M * baldy test 1170467753 M * baldy ;) 1170467770 M * baldy andrew:~# vserver-stat 1170467770 M * baldy CTX PROC VSZ RSS userTIME sysTIME UPTIME NAME 1170467770 M * baldy 0 60 59.2M 16.3M 1h06m36 5m42s26 7h46m32 root server 1170467770 M * baldy 50 5 11.1M 2.8M 0m00s00 0m00s00 4m12s19 test 1170467772 M * daniel_hozac ok. 1170467797 M * daniel_hozac i guess i might disallow multiple specifications of -n in future versions... 1170467902 M * xp_prg vserver centos4 build -m yum --context 8000 --hostname pricelist --interface pricelist=eth1:10.0.0.1/24 -- -d centos4 1170467913 M * xp_prg that is what I netered before could I not just put: 1170467953 M * xp_prg well I am confused, why is there a -n after the build in your command? 1170467962 M * daniel_hozac there shouldn't be. 1170468101 M * xp_prg vserver deb1 build -m test --context 8001 --hostname pricelist2 --interface pricelist2=eth1:172.16.1.6/24 -- -d etch Unknown build-method 'test' 1170468120 M * xp_prg what is wrong with my command? 1170468151 M * daniel_hozac -m is supposed to be the build method. 1170468164 M * daniel_hozac test, as it says, is not a valid build method. 1170468170 M * xp_prg what build method should I use? 1170468186 M * daniel_hozac for Debian guests? debootstrap. 1170468670 M * xp_prg hmm... says the ar command is not found 1170468677 M * xp_prg anyone know what debian package contains the program ar? 1170469108 M * baldy apt-get install debootstrap 1170469251 M * xp_prg ok it is working, I just put the url in the uri file 1170469302 M * xp_prg once again I appreciate your help greatly :) 1170469806 Q * FireEgl Remote host closed the connection 1170469856 J * gerrit ~gerrit@ip67-88-206-98.z206-88-67.customer.algx.net 1170470670 J * ensc ~irc-ensc@p54B4E070.dip.t-dialin.net 1170471254 J * k421k1 ~k421k1@201.236.31.71 1170471272 M * xp_prg everything is working now thanks! 1170472124 Q * gerrit Quit: Client exiting 1170474379 Q * k421k1 1170479579 N * Bertl_zZ Bertl 1170479586 M * Bertl morning folks! 1170482084 J * Aiken ~james@ppp220-70.lns2.bne1.internode.on.net 1170482106 M * Bertl hey Aiken! still around? 1170482190 M * Aiken yep 1170482218 M * Bertl got time for an alpha userspace test? 1170482243 M * Bertl (not necessarily right now :) 1170482246 M * Aiken yeh 1170482257 M * Aiken I just have to hook the alpha up 1170482280 M * Bertl thing is, I have rewritten the syscall_shiny stuff, to cope with more gcc versions 1170482302 M * Bertl now we basically need to retest on all archs (if it produces the proper syscalls) 1170482623 M * Aiken it is hooked up and running 1170482657 M * Bertl okay, do you have a source tree for util-vserver at hand? 1170482665 M * Bertl one you compiled there? 1170482712 M * Bertl http://vserver.13thfloor.at/Experimental/SYSCALL/syscall_shiny14b.h 1170482757 M * Bertl you want to copy that over lib/syscall-alternative.h IIRC 1170482766 M * Aiken I'll have to compile a new tree 1170482778 M * Bertl and make sure that the configuration uses the alternative syscalls 1170482790 M * Bertl okay, you can get 0.30.213-pre1 here: 1170482920 M * Aiken I can see pre1 pre3 pre4 pre5 rc1 1170482939 M * Bertl ah, great, I couldn't find any of them 1170482947 M * Bertl I'm still lost with the places :/ 1170482957 M * Bertl take whatever you consider latest :) 1170482974 M * Aiken I am looking in http://people.linux-vserver.org/~dhozac/p/uv/experimental/ 1170483007 M * Bertl great! tx! 1170483025 M * Bertl rc1 should be fine and actually is what I meant 1170483033 M * Aiken :) 1170483725 M * Aiken syscall(2) invocation: alternative 1170483725 M * Aiken vserver(2) syscall#: 428/glibc 1170483746 M * Bertl okay, looks good 1170483750 M * Bertl does it work? 1170484416 M * Aiken can not change context: migrate kernel feature missing and 'compat' API disabled: Function not implemented 1170484419 M * Aiken that is new to me 1170484444 M * Bertl sec, let me check the syscall number 1170484487 M * Bertl hmm, 428 would be okay, so we have a problem there 1170484505 M * Bertl sec, let me prepare a vcmd to test with that 1170484517 M * Aiken 2.6.19.1-vs2.3.0.4 1170484548 M * Bertl aah, could it be that you have some legacy stuff enabled? 1170484555 M * Bertl (in the kernel) 1170484573 M * Bertl if so, try to reconfigure with --enable-apis=NOLEGACY 1170484629 M * Aiken http://paste.linux-vserver.org/1022 1170484663 M * Bertl yeah, please add the config above and rebuild 1170484696 M * Bertl you should see a bunch APIs listed with that option on the final overview 1170484706 M * Bertl ATM, you probably have just 3 or 2 1170485014 M * Bertl here is another one to test with: http://vserver.13thfloor.at/Experimental/TOOLS/vcmd-0.08-pre2.tar.bz2 1170485024 M * Bertl just unpack and do 'make' 1170485042 M * Bertl should give you the vcmd tool without any error/warning messages 1170485090 M * Bertl ./vcmd should then show your VCI version 1170485282 M * Aiken vserver: ret = 0x00000000 (0) 1170485302 M * Bertl ah, try with -V 1170485345 M * Aiken version: 2000:aaf8 1170485373 M * Bertl hmm, that is interesting ... not correct, but interesting ... 1170485399 M * Bertl could you run that through strace -fF -o vcmd.strace please? 1170485476 M * Aiken http://paste.linux-vserver.org/1023 1170485647 M * Bertl tx 1170485719 M * Bertl funny, must be an userspace bug, checking now 1170485728 M * Bertl the strace shows the proper values 1170485739 M * Bertl 131330 = 0x20102 1170485922 M * Bertl but for the syscall part, this looks fine 1170485977 M * Bertl yeah, shame on me, I still define the vserver syscall as int 1170485995 M * Aiken a earlier vcmd I had installed gives version: 0002:0102 1170485995 M * Bertl please change the last line in vserver.h 1170486013 M * Bertl from syscall3(int, vserver, uint32_t, cmd, uint32_t, id, void *, data) to 1170486024 M * Bertl _syscall3(long, vserver, uint32_t, cmd, uint32_t, id, void *, data); 1170486057 M * Bertl if that doesn't help, I'll ahve to check the asm output 1170486130 M * Bertl (you will have to do a make clean/make 1170486229 M * Aiken version: 2000:aaf8 1170486235 M * Aiken #define __NR_vserver 236 1170486236 M * Aiken #endif 1170486236 M * Aiken static inline 1170486236 M * Aiken _syscall3(long, vserver, uint32_t, cmd, uint32_t, id, void *, data); 1170486258 M * Bertl looks good, okay, let's try the following: 1170486267 M * Aiken I also have vcmd 0.06 installed and it gives version: 0002:0102 1170486279 M * Bertl yes, that's what I'd expect 1170486307 M * Bertl let me verify that my vcmd does the proper thing on x86 first ... 1170486402 M * Aiken (root@fred) ./vcmd -V 1170486402 M * Aiken version: 0002:0102 1170486402 M * Aiken vserver: ret = 0x00020102 (131330) 1170486407 M * Aiken that is on my x86 1170486414 M * Bertl yes, that looks good and is the same I get here 1170486529 J * rpc ~no@aij186.neoplus.adsl.tpnet.pl 1170486562 M * Bertl Aiken: okay, let's do: 'gcc -O2 -Wall -Ivserver -S -o vcmd.s vcmd.c' 1170486571 M * Bertl and upload the vcmd.s please 1170486582 M * Bertl btw, did util-vserver compile/work? 1170486590 M * Bertl welcome rpc! 1170486596 M * rpc hello Bertl :) 1170486611 M * rpc thanks very much 1170486704 M * rpc here comes my silly question: how bad is it to try and run vserver on 2.6.18-3 that comes patched already (Etch) ? 1170486729 M * Bertl well, it is buggy, the bug is known, and fixed since mid december 1170486752 M * rpc ok definetly not worth trying then :) 1170486760 M * Bertl so I'd suggest to use 2.6.18-4, which is there for a longer time now, but 'just' not integrated 1170486785 M * Bertl and I'd also suggest to add a 'me too' to the debian bugtracker 1170486785 M * rpc i'm just a little worried because i'm going to reinstall the machine with amd64 kernel 1170486804 M * rpc he he that's right 1170486826 M * Bertl I think before 100 people complained, debian will not fix it :) 1170486845 M * rpc lol 1170486865 M * rpc and should i expect any problems when patching 64-bit kernel? 1170486882 M * rpc basically i'm not sure if it's better to run vserver on 32 or 64-bit debian 1170486884 M * Bertl nope, works quite fine here on several machines (with mainline 2.6.19.2 though) 1170486894 M * Aiken Bertl http://www.users.on.net/~boddingt/vcmd.s 1170486899 M * rpc oh hmm 1170486901 M * Bertl yes, 64bit is definitely faster 1170486913 M * Bertl Aiken: tx 1170486927 M * rpc so it's not like many other apps that tend to slow down on 64-bit because they are prepared for 32 env. 1170486936 M * rpc and you recommend 64-bit then 1170486958 M * Bertl nope, you can run 32 and 64 bit guests on 64bit Linux-VServer kernels side by side 1170486973 M * rpc this sounds really good in my ears :) 1170486978 M * Bertl still 32bit guests will run faster on a 64bit kernel 1170486986 M * rpc oh they will? 1170486988 M * Bertl (compared to 32bit kernel) 1170487005 M * Bertl of course, 64bit guests will run even faster on 64 bit kernels :) 1170487016 M * rpc this is great, getting 64-bit Etch then even though i had some doubts before 1170487051 M * Bertl don't know about etch, but it works fine with mandriva here, but AFAICR, debian 64bit is fine too# 1170487079 M * rpc this is going to be a beautiful morning then :) 1170487095 M * rpc i do appreciate your kind help Herbert 1170487101 M * Bertl just if you are going to compile the kernel yourself 1170487114 M * rpc yes that's what i'm going to do as i have no other choice 1170487120 M * Bertl try first with a vanilla kernel and cut it down to the essential stuff 1170487131 M * rpc mhm that's right 1170487139 M * Bertl i.e. remove all those drivers and options you don't want/need 1170487162 M * Bertl once that works for you, make a copy (e.g. cp -la) and apply the Linux-VServer path 1170487165 M * Bertl *patch 1170487167 M * rpc it will take sometime for sure but it well worth it, even though i have never done amd64 kernel before, especially on dual core 1170487182 M * Bertl then do 'make oldconfig' and answer the vserver specific questions 1170487187 M * rpc 2.0.2.1 1170487202 M * Bertl I'd suggest to go for 2.2.0-rc10 1170487205 M * rpc hmm so nice of you - i would never figure that out myself most likely 1170487213 M * rpc so it's ok to use rc10 1170487224 M * Bertl yep, definitely, it's the upcoming stable branch 1170487241 M * rpc so very close to stable 1170487244 M * rpc ok 1170487246 M * Bertl we are spending most of our time making sure that it is rock solid :) 1170487261 M * rpc he he aight 1170487295 M * rpc thank you for all the help - was lucky to meet you today :) 1170487320 M * Bertl you're welcome! 1170487329 M * Bertl feel free to hang around ... 1170487349 M * rpc i will do, promise ;) 1170487387 M * rpc just got to prepare some bnc to make it easier in the future 1170487969 M * Aiken vcmd 0.07-pre2 -> version: 0002:0102 1170488040 J * hubabuba ~hubert@dslb-088-064-004-244.pools.arcor-ip.net 1170488040 M * Bertl Aiken: please try to remove the $(DIET) from the Makefile 1170488049 M * Bertl welcome hubabuba! 1170488060 M * hubabuba hello 1170488209 M * Aiken no change 1170488240 M * hubabuba i'm having some minor problems with my vserver, 1170488257 M * Bertl Aiken: the interesting thing is, strace shows that the return code is correct, the assembler code looks like it should give the proper result, nevertheless the values are strange 1170488264 M * Bertl hubabuba: let's hear ... 1170488284 M * hubabuba is there a cflag that lets me establish tcp listening sockets ? 1170488317 M * Bertl hmm, no, you should always be able to do so :) 1170488365 M * Bertl let's take one step back, and describe your issues :) 1170488389 M * hubabuba oh my. maybe i should ask wikipedia what tcp listening socket are :> 1170488399 M * Bertl Aiken: did/does util-vserver work now? 1170488441 M * Bertl hubabuba: can't hurt, but basically this is a server socket ... 1170488461 M * Bertl (i.e. some point where you can connect from 'outside') 1170488510 M * Aiken no 1170488516 M * Aiken vps ax -> vc_new_s_context(): [unknown error] 1170488549 M * Bertl okay, so we have a problem with the syscall code there ... 1170488580 M * Bertl I assume a missing clobber or so ... let me check 1170488677 M * Aiken for something to do I added a 2nd printf at line 1072 with the & FFFF missing 1170488693 M * Aiken I get different output depending whether that printf is commented out or not 1170488708 M * Bertl yes, I assume we clobber some registers 1170488721 M * Bertl I'm just checking that the arguments are in order 1170488795 M * Bertl Aiken: in the syscall_shiny14b.h inside vcmd 1170488812 M * Bertl move to the alpha section, and change the following line: 1170488825 M * Bertl #define__sysc_clobber \ 1170488843 M * Bertl #define__sysc_clobber __sysc_regs, \ 1170488860 M * Bertl if that doesn't bail out with a compiler error 1170489049 M * Aiken vserver.h:47: error: asm-specifier for variable `__sc_a1' conflicts with asm clobber list 1170489052 M * Aiken and friends 1170489113 M * Bertl hmm, okay, kind of expected that, have to refine it 1170489263 M * Aiken #CFLAGS += -g -Wall -O2 -Ivserver 1170489264 M * Aiken CFLAGS = -Ivserver 1170489271 M * Aiken version: 0002:0102 1170489272 M * Aiken vserver: ret = 0x00020102 (131330) 1170489283 M * Bertl yep, without optimization it works ... 1170489383 M * Aiken it compiles a lot quicker as well 1170489420 M * Bertl naturally :) 1170489509 M * Aiken 8.2 seconds compared to 1 min 13s 1170490452 M * Bertl okay, I have a version with proper clobbering, I hope :) 1170490610 M * Aiken is it ready to download/ 1170490611 M * Aiken ? 1170490623 M * Bertl just a second, fixing up breakage for other archs ... 1170490664 M * Wonka what will happen if someone boots a vserver kernel without ipv6 support, and has ipv6 addresses configured for vservers? util-vserver 0.30.212 are installed. 1170490707 M * Bertl hmm, the guest do not support ipv6 without the patches 1170490722 M * Bertl on the host, the same will happen as with normal linux 1170491051 M * Bertl Aiken: http://vserver.13thfloor.at/Experimental/SYSCALL/syscall_shiny14c.h 1170491140 M * Bertl ah, I guess while this is a nice 'feature' it will not fix anything 1170491169 M * Bertl the problem is probably with the pic register, but let's see ... maybe it helps :) 1170491224 M * Aiken compiling 1170491245 M * Bertl I'll add the pic stuff in the meantime :) 1170491274 M * Aiken version: 2000:ab10 1170491274 M * Aiken vserver: ret = 0x00000000 (0) 1170491648 M * Aiken Bertl we are about to head into town while there is still a little day light left 1170491665 M * Bertl okay, please do so, and thanks for the help 1170491666 M * Aiken maybe /msg me any url to download and I can try it when I get home tonight 1170491677 M * Bertl will have something to test for later/tomorrow 1170493061 M * jjpacheco hey everyone 1170493065 M * jjpacheco hey Bertl! 1170493187 M * Wonka Bertl: but there won't be failures starting the vservers, nor? 1170493236 M * Bertl as the ipv6 patch is not in mainline yet 1170493308 J * bonbons ~bonbons@83.222.37.103 1170493321 M * Bertl okay, have to run now ... cya later folks! 1170493325 N * Bertl Bertl_oO 1170496072 J * duckx ~Duck@tox.dyndns.org 1170496207 Q * Aiken Quit: Leaving 1170498254 Q * ebiederm_away Quit: Leaving 1170498326 J * yarihm ~yarihm@84-75-123-221.dclient.hispeed.ch 1170499968 J * dna ~naucki@27-217-dsl.kielnet.net 1170500090 P * hubabuba linux-vserver 1170501966 M * harry can someone change the topic for 2.2.0-rc10 (the grsec one :)) 1170501979 M * harry it's hell to change 1 thingy in irssi 1170502021 T * harry http://linux-vserver.org/ | latest stable 2.0.2.1, 2.0.3-rc1, 2.2.0-rc10, devel 2.1.1.7.1, 2.3.0.8, stable+grsec 2.0.2.1, 2.2.0-rc10, devel+grsec 2.1.1 | util-vserver-0.30.212 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the Wiki, and we'll forget about the minute ;) 1170502027 M * FaUl harry: it isn't! just do /t and edit 1170502028 M * harry there... did it anyway :) 1170502042 M * harry FaUl: lol, i didn't know that 1170502044 M * harry it is!!!! 1170502050 M * harry i always did: /topic 1170502055 M * harry then copy 1 line 1170502066 M * harry copy part of second, edit, copy rest of line, copy last line 1170502069 M * harry ==> hell :) 1170502084 M * FaUl yes, just do /topic ;-) 1170502085 M * harry tnx 1170502101 M * harry so: patch for grsec updated too 1170502159 T * harry http://linux-vserver.org/ | latest stable 2.0.2.1, 2.0.3-rc1, 2.2.0-rc10, devel 2.1.1.7.1, 2.3.0.8, stable+grsec 2.0.2.1, 2.2.0-rc10 | util-vserver-0.30.212 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the Wiki, and we'll forget about the minute ;) 1170502181 M * harry there... no devel updates for grsec stuff... why would you want that anyways ;) 1170503011 Q * m`m`h Ping timeout: 480 seconds 1170503026 J * Piet hiddenserv@tor.noreply.org 1170503372 J * shedi ~siggi@ftth-237-144.hive.is 1170504457 J * m`m`h ~simba@deb30.mgts.by 1170504528 M * matti harry: Hello! 1170504540 M * harry hey 1170504553 M * matti How are you? 1170504564 M * harry very good... but hungry 1170504569 M * matti Oh. 1170504570 M * harry don't know what to eat 1170504572 M * matti Funny, me too ;] 1170504575 M * matti Me too! 1170504576 M * matti ;] 1170504579 M * matti Any ideas? 1170504590 M * matti Well, I can give you coffee. 1170504591 M * matti ;] 1170504592 M * harry kebab, but it's closed 1170504596 M * harry had coffee , tnx ;) 1170504597 M * matti Kebab? 1170504602 M * matti :) 1170504614 M * harry french fries or so... 1170504620 A * matti is a vegetarian. 1170504622 M * harry but... i don't know :) 1170504624 M * harry lol 1170504626 M * matti ;] 1170504631 M * matti What? 1170504632 M * matti :> 1170504642 M * matti You didn't seen an alien before? ;p 1170504644 M * matti ;D 1170504656 M * harry i eat meat because the animals ask for it 1170504667 M * harry why are you a vegetarian? 1170504676 M * harry it's not natural 1170504683 M * harry it's a self-imposed handicap 1170504689 A * matti pokes harry with soft cushions! 1170504724 M * matti harry: I had this kind of conversation here before ;p 1170504735 M * matti harry: And it become a flame-ware after few minutes ;] 1170504735 M * matti ;p 1170504736 M * harry probably 1170504745 A * harry against vegetarians 1170504750 M * matti Heheh. 1170504756 A * matti runs... 1170504769 M * matti harry: I am not a vegan. 1170504772 M * harry our digestion system is DESIGNED to eat meat 1170504778 M * matti harry: I just don't eat red meat. 1170504786 M * matti harry: I know, believe me I know. 1170504786 M * harry you eat chicken/fish? 1170504790 M * matti Fishes. 1170504800 M * harry so you don't eat it because you don't like it? 1170504817 M * harry that's imho the only acceptable reason to be a vegetarian 1170504909 Q * id23 Remote host closed the connection 1170505034 M * matti ;] 1170506433 Q * FaUl Ping timeout: 480 seconds 1170506980 Q * yarihm Quit: Leaving 1170507911 Q * m`m`h Ping timeout: 480 seconds 1170508055 J * baldy_ 1000@83.223.66.130 1170508170 Q * baldy Ping timeout: 480 seconds 1170510539 Q * shedi Quit: Leaving 1170510785 J * immo ~immo@ip238.164.1211G-CUD12K-01.ish.de 1170510787 N * immo FaUl 1170510787 J * shedi ~siggi@ftth-237-144.hive.is 1170510789 M * FaUl huhu 1170510818 M * FaUl i'm just experiencing http://paste.linux-vserver.org/1024 <- any idea what to do against that? 1170510908 M * daniel_hozac humm, interesting 1170510922 M * daniel_hozac what were you doing to trigger that? 1170510924 M * FaUl it panics afterward 1170510927 M * FaUl booting 1170510928 M * FaUl ;-) 1170510938 M * daniel_hozac the host? 1170510949 M * FaUl well - the first time i experienced the problem i did a /rehash on the ircd 1170510961 M * FaUl and now it appears if the ircd starts i guess 1170510974 M * daniel_hozac do you still have the kernel tree around? 1170510982 M * FaUl because it happens while starting the vserver just after a minute 1170511011 M * FaUl yes, a second please 1170511049 M * FaUl i have to reboot the box again 1170511098 M * daniel_hozac could we get addr2line -e vmlinux c043aae8 c01648c3 c015da51 c015dc18 c016745d c012b247 c019b8e5 c016f5b3 c016f663 c01a02e9 c016de97 c016506f c016786c c015dc75 c015dd5e c0281b55 c015de74 c015dedc 1170511168 M * FaUl mompl 1170511414 M * FaUl argh 1170511430 M * FaUl where was that 'start this server on boot automatically flag again'? 1170511442 M * FaUl i've just very short time to delete it ;-) 1170511585 M * PowerKe /etc/vservers//apps/init/mark 1170511714 M * FaUl ok, so another reboot 1170511990 M * FaUl seems that this has worked so far 1170512177 M * FaUl http://paste.linux-vserver.org/1026 1170512195 M * FaUl daniel_hozac: *highlight* ;-) 1170512223 M * Hollow FYI, our web/svn/ftp server is now monitored with nagios :) 1170512223 J * m`m`h ~simba@deb30.mgts.by 1170512242 M * FaUl Hollow: good idea ;-) 1170512313 J * dlezcano ~dlezcano@AToulouse-252-1-58-239.w83-200.abo.wanadoo.fr 1170512386 M * daniel_hozac FaUl: could you get the first line of the oops too? 1170512402 M * daniel_hozac there should be a line right above that section. 1170512579 M * daniel_hozac something like Unable to handle kernel NULL pointer dereference... 1170512600 M * daniel_hozac (assumption, could just be an invalid pointer too but that would be weirder, i guess) 1170512672 M * FaUl mom, i look what i can find in the logs 1170512868 M * FaUl daniel_hozac: looks not to good 1170512965 M * daniel_hozac ? 1170512978 M * daniel_hozac no oops in the logs, or that line is missing? 1170513148 M * FaUl no oops in the logs 1170513149 M * FaUl ;-( 1170513154 M * daniel_hozac okay. 1170513177 M * daniel_hozac but this is completely reproducable for you? 1170513185 M * FaUl yes 1170513195 M * FaUl i rebootet that box a couple of times 1170513201 M * daniel_hozac hmm. 1170513205 M * FaUl and it started to appear when i did the /rehash 1170513224 M * FaUl it stopped when i started to not start the ircd again ;-) 1170513249 M * daniel_hozac this looks like an old bug that has reappeared... 1170513271 M * daniel_hozac and it appears to be present in 2.2.0-rc10. 1170513296 M * daniel_hozac i'll try to reproduce it, shouldn't be too hard... 1170513331 M * FaUl fine ;-) 1170513464 M * daniel_hozac yep, got it. 1170513490 M * daniel_hozac would you be able to test a fix? 1170513504 M * daniel_hozac (btw, it would probably be a good idea to move to 2.2) 1170513748 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-cowbl-fix01.diff or something like it should do it, i'm test building now... 1170513806 Q * Piet Remote host closed the connection 1170513856 J * Piet hiddenserv@tor.noreply.org 1170513886 M * FaUl mh, the mayor problem is that this is a production-environment-box so it's bad to test to much on it 1170513899 M * daniel_hozac np, i'll test it locally as well. 1170513918 M * FaUl very well 1170513931 M * FaUl is it realy a god idea to move to 2.2? 1170513955 M * daniel_hozac 2.2 is the next stable branch... 1170513977 M * daniel_hozac it's basically 2.1 - untested features + fixes (as we haven't updated 2.1 for a while) 1170514002 M * FaUl ok 1170514009 M * FaUl how stable is 2.3 anyway? 1170514042 M * daniel_hozac i haven't had a chance to use 2.3 yet. 1170514045 M * FaUl ok 1170514062 M * FaUl what are the untested features of 2.1? 1170514069 M * daniel_hozac quota hashes and the device mapper. 1170514102 M * FaUl mh 1170514115 M * FaUl i don't use this anyway 1170514143 M * daniel_hozac the patch fixes the problem here. 1170514161 M * FaUl ok, so i guess I should give it a try 1170514173 M * FaUl thanks for responding so fastly ;-) 1170514182 M * daniel_hozac that's what IRC is for, no? :) 1170514198 M * FaUl yes 1170514252 M * daniel_hozac Bertl_oO: http://people.linux-vserver.org/~dhozac/p/k/delta-cowbl-fix01.diff test-case: mkdir test; echo 1 > test/1; setattr --iunlink test/1; ln test/1 test/2; setattr --immutable test; echo 2 > test/2 1170514309 M * daniel_hozac (for regular users it's even easier to trigger, just needs a directory they don't have access to) 1170514326 M * daniel_hozac s/access/write access/ 1170514338 M * FaUl urgs 1170514351 M * FaUl so i should apply it against 2.2, nor? 1170514365 M * daniel_hozac should apply against 2.1 as well, if you want to use that. 1170514373 M * FaUl yes, it does 1170514380 M * FaUl but if you say 2.2 is recommend anyway 1170514383 M * FaUl i'll use 2.2 1170514384 M * FaUl ;-) 1170514400 M * daniel_hozac it is. i think after this, it's just cleanup and we'll have a release... 1170514986 M * FaUl maybe we should remove the developement-patch from the main-page 1170515816 M * daniel_hozac it will probably be labeled as obsolete when 2.2.0 is released, as IIRC it's getting dropped then. 1170515830 M * daniel_hozac (though in practice i guess it's already been dropped) 1170516804 M * FaUl so, rebooting 1170517228 M * FaUl ok, does not panic anymore since now 1170517286 M * daniel_hozac good, thanks! 1170517296 M * FaUl i have to thank you ;-) 1170518435 M * daniel_hozac FaUl: could you prod Bertl_oO about it when he gets back? i have to leave and don't get back until late tomorrow. 1170518611 Q * comfrey_ Server closed connection 1170518627 J * comfrey ~comfrey@70.91.185.84 1170519575 M * FaUl daniel_hozac: i'll dowhen i see him 1170519591 Q * FaUl Quit: leaving 1170519891 Q * mnemoc Server closed connection 1170519902 J * mnemoc ~amery@kilo105.server4you.de 1170521310 J * yarihm ~yarihm@84-74-16-225.dclient.hispeed.ch 1170521798 M * pflanze Hello. How can I view all network connections and/or listening sockets under vserver 2.2.0(-rc8.7) ? 1170521816 M * pflanze The way I did it on 2.0 doesn't work anymore. 1170521855 M * pflanze which has been "chcontext --silent --ctx 1 {normal command to show the stuff}" 1170521946 M * pflanze I guess I have to use ncontext now? 1170522963 J * tso ~tso@238-048.adsl.pool.ew.hu 1170522965 M * tso hi 1170524457 Q * s0undt3ch Server closed connection 1170524468 J * s0undt3ch ~s0undt3ch@80.69.34.154 1170525022 Q * Piet Quit: Piet 1170526102 J * FaUl immo@shell.chaostreff-dortmund.de 1170526103 M * FaUl huhu 1170529523 Q * kir Ping timeout: 480 seconds 1170531786 J * DoomStar ~asd@ppp85-141-128-165.pppoe.mtu-net.ru 1170531860 M * DoomStar Hello! I've "2.6.17.11-vs2.0.2-rc31" kernel installed a long time. But tow minutes ago I found the exploit :( 1170531880 M * DoomStar Is it ok if I post here a link to exploit? 1170531935 M * DoomStar you can access whole hosts file system after exploiting 1170531978 M * DoomStar how to fix it? Is this exploit fixed in latest stable 2.0.2.1 ?? 1170532000 M * DoomStar I was hacked today :,-( 1170532028 M * DoomStar HALLLOOO... Any body home? 1170532114 M * Radiance no 1170532129 M * DoomStar hi 1170532150 M * Radiance why not upgrade ? 1170532159 M * Radiance perhaps the issue has been solved 1170532178 M * DoomStar I'll try 1170532208 M * Radiance what is the link ? 1170532216 M * DoomStar Radiance, do you have the latest stable installed? 1170532224 M * DoomStar can you test it? 1170532230 M * Radiance not atm 1170532233 M * Radiance but show us the link ? 1170532346 M * Radiance it says it's fixed as from 1.25 1170532352 M * Radiance but you mean it still works on the current ? 1170532354 M * DoomStar I think it is a little dangerous to post it here :( 1170532372 J * FireEgl ~FireEgl@adsl-63-199-202.bhm.bellsouth.net 1170532407 M * Radiance if it does work then one needs at least to be root 1170532456 M * DoomStar link is http://securityvulns.ru/docs5733.html 1170532485 M * Radiance but to be correct, you tested it on the current stable ? 1170532498 M * Radiance looks like a classic escape from chroot 1170532501 M * DoomStar I tested it on 2.6.17.11-vs2.0.2-rc31 1170532521 M * Radiance perhaps the barrier is not set ? 1170532527 M * Radiance what is the dir of your vserver ? 1170532554 M * DoomStar drwxr-xr-x 4 root root 4096 Jan 29 18:43 vservers 1170532572 M * DoomStar hmm i think i should chmod 000 vserver 1170532586 M * Hollow no 1170532590 M * Hollow you just need the barrier 1170532592 M * Radiance can you try this and test it again: /usr/sbin/setattr --barrier /vserver/ 1170532608 M * DoomStar 1 moment 1170532625 M * DoomStar # /usr/sbin/setattr --barrier /vserver/ 1170532625 M * DoomStar lstat(): No such file or directory 1170532626 M * DoomStar ^( 1170532633 M * Radiance if you remove the World permissions from vserver dir itself then the vserver won't work (i experienced this :) 1170532644 M * Radiance add a 's' :) 1170532652 M * Radiance vservers instead of vserver 1170532657 M * DoomStar sorry ) 1170532743 M * DoomStar it worked again, but now using exploit I can see every thing excluding /vservers dir 1170532749 M * DoomStar I'll show now 1170532787 M * DoomStar # vserver hl enter 1170532788 M * DoomStar host124 / # /root/a.out 1170532788 M * DoomStar mkdir baz: File exists 1170532788 M * DoomStar Exploit seems to work. =) 1170532788 M * DoomStar sh-3.00# ls /vservers/ 1170532788 M * DoomStar ls: /vservers/: Permission denied 1170532800 M * DoomStar sh-3.00# ls /home/ 1170532801 M * DoomStar doomstar ftp 1170532815 M * DoomStar sh-3.00# ls /root/ 1170532815 M * DoomStar Ent_Popup_zip.asp 1170532815 M * DoomStar Microsoft Windows Server 2003.iso 1170532815 M * DoomStar Readme.txt 1170532838 M * Radiance can you stop the vserver, set the barrier again just to be sure and start the vserver 1170532870 M * DoomStar # ls /vservers/ 1170532870 M * DoomStar hl vds1 vds11 vds2 vds4 vds6 vds8 1170532883 M * Hollow i'm not sure if there was a bug fixed, but you should try with 2.2.0_rc8 at least 1170532887 M * Hollow rc10 1170532888 M * DoomStar I had to set barier on /vservers/ or on /vservers/hl/ 1170532889 M * DoomStar ? 1170532895 M * Hollow on /vservers 1170533024 M * DoomStar tryed restarting vserver - still works 1170533026 M * DoomStar ( 1170533064 M * DoomStar I'll look now on isntalling 2.2 1170533116 M * DoomStar can somebody test this exploit on 2.2 ?? 1170533165 M * Radiance hmm 1170533165 M * Radiance lemme see and test the EBIL myself 1170533275 M * Radiance doesn't work 1170533290 M * Radiance i just tried it on vs2.0.2 1170533366 M * PowerKe Al it does on my system is dropping me in a shell at the root of the guest (although it says Exploit seems to work) 1170533378 M * PowerKe (2.6.14-vs2.0.1) 1170533442 M * DoomStar on my system it drops me in the root of host 1170533444 M * Radiance same here 1170533454 M * Radiance it does never leave the vserver 1170533474 M * DoomStar but i heve 2.0.2-rc31 1170533486 Q * baldy_ Remote host closed the connection 1170533501 M * DoomStar may be you use grsec ? 1170533558 M * Radiance i use grsec yes but chroot jail restrictions are not enabled, the rest is 1170533570 M * Radiance PowerKe, you use grsec ? 1170533575 M * PowerKe nop 1170533579 M * Radiance ok 1170533592 M * Radiance DoomStar, can you do chmod o-rwx on /vservers 1170533602 M * Radiance although it shouldn't matter 1170533621 M * PowerKe (gentoo-sources, non-hardened) 1170533626 J * shuri ~shuri@hq01.electronicbox.net 1170533687 M * Radiance DoomStar, so something is wrong in your setup 1170534047 Q * shuri Quit: Quitte 1170534071 J * shuri ~shuri@hq01.electronicbox.net 1170534402 J * ]OmqlaGPm ~hollow@styx.xnull.de 1170534428 J * Loki|muh_ loki@satanix.de 1170534738 Q * Loki|muh Ping timeout: 480 seconds 1170534738 N * Loki|muh_ Loki|muh 1170534763 Q * nox Ping timeout: 480 seconds 1170534768 Q * mEDI_S Ping timeout: 480 seconds 1170534778 Q * Hollow Ping timeout: 480 seconds 1170534841 J * mEDI_S ~medi@snipah.com 1170536113 J * nox ~nox@static.88-198-17-175.clients.your-server.de 1170536261 N * ]OmqlaGPm Hollow 1170536417 M * DoomStar I'm back here 1170536436 M * DoomStar I updated my kernel and vserver patch 1170536457 M * DoomStar but my guests dont want to run ^( 1170536467 M * DoomStar vserver hl start 1170536468 M * DoomStar mount: none already mounted or /vservers/hl/dev/pts busy 1170536468 M * DoomStar vsched: non-numeric value specified for '--priority_bias' 1170536468 M * DoomStar An error occured while executing the vserver startup sequence; when 1170536468 M * DoomStar there are no other messages, it is very likely that the init-script 1170536468 M * DoomStar (/sbin/init) failed. 1170536499 M * DoomStar mount: none already mounted or /vservers/hl/dev/pts busy - that is my think from pre-start 1170536516 M * DoomStar what is this - "vsched: non-numeric value specified for '--priority_bias'" ?? 1170536577 M * DoomStar sloved it by delting my /etc/vservers/hl/schedule 1170536918 Q * mire Ping timeout: 480 seconds 1170537046 M * Radiance did you mess with the permissions btw ? 1170537204 J * kir ~kir@swsoft-mipt-nat.sw.ru 1170537502 M * DoomStar I discovered my problem with this exploit 1170537513 M * DoomStar lrwxrwxrwx 1 root root 12 Jun 15 2006 hl -> /mnt/sda3/hl 1170537513 M * DoomStar lrwxrwxrwx 1 root root 14 Jun 12 2006 mail -> /mnt/sda3/main 1170537513 M * DoomStar lrwxrwxrwx 1 root root 14 Jun 16 2006 vds1 -> /mnt/sda6/vds1 1170537513 M * DoomStar lrwxrwxrwx 1 root root 15 Nov 14 01:56 vds10 -> /mnt/sda6/vds10 1170537513 M * DoomStar lrwxrwxrwx 1 root root 15 Nov 16 19:46 vds11 -> /mnt/sda6/vds11 1170537514 M * DoomStar lrwxrwxrwx 1 root root 15 Jan 22 01:27 vds12 -> /mnt/sdb3/vds12 1170537514 M * DoomStar lrwxrwxrwx 1 root root 14 Jun 16 2006 vds2 -> /mnt/sda6/vds2 1170537516 M * DoomStar lrwxrwxrwx 1 root root 14 Nov 1 01:11 vds3 -> /mnt/sda6/vds3 1170537516 M * DoomStar drwxr-xr-x 18 root root 4096 Dec 6 13:10 vds4 1170537518 M * DoomStar lrwxrwxrwx 1 root root 14 Jun 16 2006 vds5 -> /mnt/sda6/vds5 1170537518 M * DoomStar lrwxrwxrwx 1 root root 14 Jan 29 18:43 vds6 -> /mnt/sdd1/vds6 1170537520 M * DoomStar lrwxrwxrwx 1 root root 14 Jan 23 20:44 vds7 -> /mnt/sdb3/vds7 1170537520 M * DoomStar lrwxrwxrwx 1 root root 14 Jun 16 2006 vds8 -> /mnt/sda6/vds8 1170537525 M * DoomStar that's my /vservers/ dir listing ) 1170537541 M * DoomStar my guests are not located in /vservers/ 1170537572 M * DoomStar i setted --barier on /mnt/sda3 and overs and exploit doesn't work any more 1170537580 M * cehteh DoomStar: use bind mounts at least 1170537588 J * mire ~mire@27-167-222-85.adsl.verat.net 1170537620 M * cehteh well and you loose much vserver benefits when you dont share the vservers on one partition 1170537640 M * DoomStar what benefits? 1170537643 M * cehteh you cant unify them that way 1170537659 M * cehteh means each vserver needs quite more ram 1170537677 M * cehteh and also more HD space (which isnt prolly a problem) 1170537684 M * DoomStar mm.. how do you unifiy them ? 1170537713 M * cehteh hashify / immutable links ... read the vserver docs 1170537735 N * Bertl_oO Bertl 1170537740 M * FaUl hey Bertl 1170537741 M * Bertl eveing folks! 1170537747 M * CHTEKK hi Bertl 1170537762 M * Radiance hi Bertl 1170537764 M * cehteh imo i would suggest you rather to aggregate the HD's in a fast and redundant raid and then put them on one partition 1170537769 M * CHTEKK great, just the person I was searching for :) what is vavavoom in the struct vcmd_sched_info ? 1170537778 M * cehteh evening Bertl 1170537799 M * DoomStar thanks cehteh 1170537804 M * FaUl Bertl: daniel-hazoc fixed a problem in current developement that made my kernel panic 1170537876 M * FaUl so i suggest to commit http://people.linux-vserver.org/~dhozac/p/k/delta-cowbl-fix01.diff 1170537903 J * DreamerC_ ~dreamerc@125-225-96-10.dynamic.hinet.net 1170538007 M * Bertl FaUl: no worries, I look at all the patches daniel_hozac provides :) 1170538099 Q * DreamerC_ 1170538137 J * DreamerC_ ~dreamerc@125-225-96-10.dynamic.hinet.net 1170538192 M * FaUl Bertl: yes, daniel just wanted me to point you onto the patch because he has no time 1170538200 M * FaUl it seems to be a quite critical problem anyway 1170538264 M * CHTEKK Bertl, could you tell me what vavavoom in the vcmd_sched_info struct is? :) and is it actually possible to get some % figures of actual CPU utilization from the data in that struct? 1170538308 Q * DreamerC Ping timeout: 480 seconds 1170538402 M * Bertl FaUl, CHTEKK: sure, no problem 1170538432 M * Bertl CHTEKK: you know that there are two Linux-VServer specific scheduler extensions? 1170538450 M * FaUl Bertl: how stable is 2.3 right now? May I use it on my home-server right now? 1170538471 M * Bertl FaUl: I think so, it is used in production in certain places 1170538480 M * FaUl perfect 1170538487 M * FaUl i'll try when i update the system anyway 1170538492 M * Bertl FaUl: but if you do not need the stuff from 2.3 and want stability, 2.2 is the way to go 1170538503 M * FaUl Bertl: i'd like to have ipv6-support 1170538512 M * Bertl okay :) 1170538519 M * FaUl what are the other differences on 2.3? 1170538533 M * FaUl you realy have to learn to write a changelog ;-) 1170538543 M * Bertl all the 'experimental' stuff we had in 2.1 was basically moved to 2.3 1170538547 M * FaUl it would be to nice 1170538552 M * Bertl while the stable stuff got into 2.2 1170538558 M * FaUl what was the experimental stuff? 1170538573 Q * mire Ping timeout: 480 seconds 1170538582 M * Bertl e.g. the quota patches, the new network isolation 1170538597 N * DreamerC_ DreamerC 1170538599 M * Bertl also 2.3 is currently lagging a little behind 2.2 1170538621 M * Bertl so you might want to add the delta between last 2.3 release and current 2.2 1170538633 M * Bertl e.g. the fix from daniel :) 1170538636 M * FaUl what do these quota-patches do? and was the network isolation? ;-) 1170538642 M * CHTEKK Bertl, yeah the priority scheduler and the hard one, right? or you talking about something else? 1170538651 M * Bertl CHTEKK: precisely 1170538672 M * Bertl CHTEKK: while the latter puts the tasks on a hold queue, the former 'adjusts' the priority 1170538678 M * CHTEKK yup 1170538688 M * CHTEKK I'm particularly interested in the hard one 1170538697 M * Bertl CHTEKK: the adjustment was called vavavoom by Sam.V. 1170538729 M * CHTEKK ah ok, so it's the adjustement to priority for the priority sched, k thanks! 1170538737 M * Bertl I kept that name because it sounded like fun :) 1170538751 M * Bertl (and for sure it is something to talk about :) 1170538794 Q * shuri Quit: Quitte 1170538797 M * Bertl FaUl: per context quota on a shared partition (well almost) and the network isolation provides the 'often requested' lo per guest 1170538798 M * CHTEKK and so info that would be around when I'm using the hard sched are the three msec's and the token_usec 1170538807 M * CHTEKK token_uses is in microseconds? 1170538823 M * Bertl yes, that's what it means, the number of usec per token 1170538887 M * CHTEKK hmmm ok, any way to use those values or an other way entirely to get in the end something that says: context X is using Y % of CPU 1 and so on? 1170538901 M * Bertl definitely 1170538934 M * Bertl you know that there are timer ticks at 'regular' intervals? 1170538958 M * CHTEKK yup, the Hz stuff in the kernel iirc, configurable to 100, 250 or 1000 1170539004 M * Bertl correct, the timer interval is identical to our elementary token interval unit 1170539018 M * FaUl Bertl: very cool 1170539049 M * Bertl CHTEKK: and you get the user/sys/hold time in msec 1170539101 M * CHTEKK yeah I figured that with some maths on those three and probably the usec value, knowing the HZ value, I can get a % of cpu usage 1170539102 M * Bertl CHTEKK: which basically tells you how many miliseconds were spent in user time and system time on behalf of the context 1170539137 M * Bertl CHTEKK: you do not really need the HZ value for any of those calculations anymore, that's what the token_usec is for 1170539180 M * Bertl FaUl: ipv6 will be soon integrated into 2.3, so you can almost add that to the list :) 1170539189 Q * dna Quit: Verlassend 1170539217 M * FaUl Bertl: excellent *mrBurnsgesture* 1170539240 J * mire ~mire@113-167-222-85.adsl.verat.net 1170539253 M * Bertl CHTEKK: so if you are just interested in the % spent on that context, collect the user/sys times and add them up over all cpus 1170539285 M * Bertl (or show them per cpu, they will match the percentage on a timeline) 1170539291 M * CHTEKK actually I'm interesetd in that per-cpu 1170539296 M * CHTEKK yup k 1170539337 M * CHTEKK but is just summing them enough? pretty sure that doesn't give me a value I can immediately use as % ? 1170539361 M * Bertl you sample at an interval T 1170539408 M * Bertl you get numbers like U0,U1,U2,....,Un, S0,S1,S2,...,Sn 1170539437 M * Bertl now lets assume you had intervals T01, T12, T23, ... T(n-1)n 1170539457 M * Bertl then the percentage for each interval will be: 1170539490 M * Bertl (Un-U(n-1))/T(n-1)n 1170539523 M * Bertl similar for S and the sum (depends on if you want to calculate user and system time separately or not) 1170539529 M * CHTEKK ok assuming I sample at known intervals, but I wanted this to work immediately, smt like "vcpu " and I immediately get "34%" back... I could probably though do the measurement for 3-4 extremely tiny intervals and then average that out 1170539576 M * Bertl CHTEKK: the shorter you make the intervals, the more random values you'll get 1170539602 M * CHTEKK yeah, how long is a "good" interval? 1170539602 M * Bertl percentage of CPU usage at infinite intervals are 0 or 100% :) 1170539625 M * Bertl either the cpu is executing the context or not :) 1170539652 M * Bertl you can see similar with top for processes ... 1170539668 M * Bertl the longer the interval is, the better (and more realistic) is the result 1170539735 M * CHTEKK would something like 5 measurements, each 500ms apart (total of 2s "measure time") be possible to give useable results? 1170539764 M * CHTEKK cause I'd like to not go over 1.5-2 seconds of total time for the measurements, since this has to be relatively immediate 1170539778 M * Bertl if you display the result every 2 secons, you can simply do a measurement at the beginning and one at the end 1170539783 M * Bertl *seconds 1170539809 M * Bertl if you start with short intervals, you will get 'jumping' values at the beginning 1170539821 M * Bertl which stabilize over time (i.e. after 2 seconds) 1170539871 M * CHTEKK hmmm ok I'll try with measure, wait 1 sec, measure, wait 1 sec, measure, and then calculate the percentage using those three values 1170539877 M * Bertl of course, you know another time/value pair 1170539904 M * Bertl the time at which the context was started, where it had zero user/sys msecs :) 1170539926 M * Bertl so if you are interested in a percentage over the entire lifetime 1170539934 M * Bertl you can calculate that immediately too 1170539974 M * CHTEKK yup, but I'm only interested in percentage in that moment atm 1170539981 M * CHTEKK thanks a lot for your help! :) 1170539992 M * Bertl you're welcome! 1170541647 M * Hollow ah, i see.. scheduler discussion :) 1170541656 M * Hollow the formulas in the wiki could need some checking 1170541695 M * Bertl yeah, still on my todo :) 1170541855 Q * mire Ping timeout: 480 seconds 1170542593 J * mire ~mire@209-166-222-85.adsl.verat.net 1170542829 Q * DoomStar 1170544450 Q * duckx Read error: Connection reset by peer 1170544467 Q * mire Ping timeout: 480 seconds 1170544634 J * duckx ~Duck@tox.dyndns.org 1170544911 J * Aiken ~james@ppp220-70.lns2.bne1.internode.on.net 1170545059 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1170545069 M * Bertl hey Aiken! 1170545069 J * ensc ~irc-ensc@p54B4E09B.dip.t-dialin.net 1170545128 J * mire ~mire@171-167-222-85.adsl.verat.net 1170545375 M * Aiken good morning 1170545390 M * Aiken anything to test? 1170545397 M * Bertl how was the town visit yesterday? 1170545409 M * Bertl yeah, I should have something shortly 1170545433 M * Aiken the trip into town was ok 1170545436 M * Bertl the bug is trivial, but I ended up rewriting larger parts before I found it :) 1170545445 M * Aiken not s good as this morning, just got home from breakfast 1170545674 M * Bertl how is the weather? 1170545692 M * Aiken dry 1170545712 M * Aiken for once hoping it does not rain today because of a bbq tonight 1170545738 M * Bertl ah, nice ... 1170546237 J * CokE-cOKe ~VhYaN@125.162.53.121 1170546266 M * Bertl welcome CokE-cOKe! 1170546278 M * CokE-cOKe tq Bertl 1170546501 Q * bonbons Quit: Leaving 1170546504 P * CokE-cOKe 1170546991 M * Bertl Aiken: let's see if my assumption was right regarding the bug in shiny14b 1170547017 M * Bertl Aiken: you see the two 'mov' instructions in the ALPHA section? 1170547017 M * Aiken ok 1170547045 M * Bertl "mov %0,$0" and "mov %1,$19" 1170547056 M * Aiken yes 1170547064 M * Bertl let's try to just exchange the arguments, i.e. make that 1170547083 M * Bertl "mov $0,%0" and ""mov $19,%1" :) 1170547176 J * dna ~naucki@27-217-dsl.kielnet.net