1146355253 Q * ntrs Ping timeout: 480 seconds 1146355427 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1146355913 Q * ntrs Ping timeout: 480 seconds 1146356071 M * _coocoon_ ok good night thanx to all 1146356075 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia' 1146356192 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1146362315 M * doener anyway... off to bed, good night! 1146362327 M * doener uhm, wrong window, but good night anyway *g* 1146363486 Q * doener Ping timeout: 480 seconds 1146364410 J * doener ~doener@i5387CB23.versanet.de 1146365141 M * Bertl good idea .. off to bed here too 1146365152 M * Bertl night everyone .. have a good one! 1146365157 N * Bertl Bertl_zZ 1146366499 Q * Zaki[] Ping timeout: 480 seconds 1146369760 Q * wenchien Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1146371868 J * jesse_ ~wenchien@221-169-69-23.adsl.static.seed.net.tw 1146371894 N * jesse_ wenchien 1146374061 P * Skram 1146374061 J * Skram ~mark@admins.sentiensystems.net 1146385657 J * bonbons ~bonbons@83.222.39.134 1146391417 J * _coocoon_ ~coocoon@p54A05D3B.dip.t-dialin.net 1146391447 M * _coocoon_ hello 1146391538 J * Viper0482 ~Viper0482@p5497502D.dip.t-dialin.net 1146393243 Q * FireEgl Quit: Bye... 1146394229 J * kir_home ~kir@ffm-nat.sw-soft.de 1146394852 Q * softi42 Ping timeout: 480 seconds 1146395247 Q * shedi Quit: Leaving 1146395947 Q * kir_home Ping timeout: 480 seconds 1146396001 J * softi42 ~softi@p549D7E2E.dip.t-dialin.net 1146396045 J * FireEgl Atlantica@Atlantica.Tcldrop.US 1146401094 J * shedi ~siggi@inferno.lhi.is 1146401755 Q * ntrs Read error: Connection reset by peer 1146401760 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1146404381 J * Dr4g ~Dr4g@82-40-40-135.cable.ubr06.uddi.blueyonder.co.uk 1146407430 J * _are_ are@dslb-084-057-149-202.pools.arcor-ip.net 1146407432 M * _are_ hi 1146407925 J * Hollow ~hollow@home.xnull.de 1146408863 N * Bertl_zZ Bertl 1146408871 M * Bertl morning folks! 1146408886 M * _coocoon_ hello bertl 1146410643 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia' 1146411003 J * American_Pie IRCSohbet@195.87.162.223 1146411047 P * American_Pie 1146411397 Q * ntrs Read error: Connection reset by peer 1146411560 Q * bonbons Quit: Leaving 1146411561 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1146412436 J * yarihm ~yarihm@217-162-112-98.dclient.hispeed.ch 1146412480 M * Skram morning, Bertl 1146412509 M * Bertl morning Skram! 1146412588 Q * yarihm Quit: 1146412675 M * Skram :) 1146412699 M * Skram Okay, am now looking for a good place to advertise my company's services.. was thinking Slashdot 1146412725 J * _webus_ ~djweb@ip-83-149-3-141.nwgsm.ru 1146412810 J * Zaki ~Zaki@212.118.99.6 1146412854 M * _webus_ you speak about Virtual Server in Linux ??? 1146412870 M * _webus_ i don't understand it, i'm from Russia! Help me 1146413173 M * Skram I webus :) 1146413179 M * Bertl _webus_: what kind of help do yu need? 1146413181 M * Skram *Hi 1146413182 M * Bertl *you 1146413203 M * Skram http://linux-vserver.org/ has a good description. 1146413217 M * _webus_ Bertl: help about SAT+GPRS+Traffic counter 1146413253 M * Skram for a virtual private server? 1146413275 M * Skram this has nothing to do with Cell-phone b/w / traffic / etc. (GPRS) 1146413366 Q * _webus_ Quit: ðÏËÉÄÁÀ 1146413476 M * Bertl off for dinner now .. back later ... 1146413483 M * Skram See ya 1146413503 N * Bertl Bertl_oO 1146414717 N * Bertl_oO Bertl 1146414721 M * Bertl back 1146415152 J * hillct ~H@client200-5.dsl.intrex.net 1146415178 M * hillct good afternoon all 1146415196 M * Bertl welcome hillct! 1146415201 M * hillct I'm still fighting with reigning in my vservers 1146415224 M * hillct the processes causing a problem seem to be apache 1146415248 M * hillct which, when it runs out of memory, it continues ot try and spawn processes 1146415250 M * hillct I guess 1146415270 M * hillct this then drives up CPU usage 1146415285 M * hillct and comes close to killing the host 1146415296 M * Bertl hmm, CPU limits? 1146415319 M * hillct so I'm trying to work out ahat things should be limited how, and with what relative percentages 1146415335 M * hillct Bertl: yes, and presumably process count limits 1146415367 M * hillct I was hoping you or someone else may have guidance on percentages and counts 1146415388 M * hillct or is it esentially a trial and error sort of thing 1146415398 M * Bertl there is a wiki howto IIRC 1146415409 M * hillct yeah 1146415416 M * rmoriz hm 1146415419 M * hillct I believe I've read most if not all of the material 1146415428 M * hillct I may have missed something though 1146415435 M * rmoriz that "per option one new configuration file" scheme is messy :/ 1146415449 Q * Viper0482 Ping timeout: 480 seconds 1146415454 M * hillct I understand how to set the limits. I'm just trying to work out what might be the proepr limits 1146415475 M * Bertl rmoriz: well, that's where the opinions go apart ... I guess you cannot satisfy everyone ... 1146415487 M * hillct I like that structure 1146415493 M * hillct it took a while to get used to 1146415494 M * rmoriz Bertl: yes. it was not meant offensive 1146415496 M * hillct but it grows on you 1146415508 M * Bertl rmoriz: what I wonder is that the folks complaining about the new tree based config, do not write some xml to tree (or vice versa) tool 1146415523 M * rmoriz hmmm :) 1146415527 M * hillct I said some months ago that I would do exactly that 1146415535 M * hillct time hasn't allowed me to do so 1146415543 M * hillct but it would theoretically be easy to do 1146415619 M * daniel_hozac didn't someone say on the mailing list that there would be scripts for that coming soon? 1146415623 M * rmoriz how can i allow traceroute (net raw i think?) with util-vserver 1.31 and vs2.1.x ? 1146415640 M * hillct Bertl: I read the memory alocation, resource alocation and CAPS pages. Is there a HOWTO that suggests reasonable settings, that I missed? 1146415646 M * Bertl http://linux-vserver.org/vsched+explained 1146415745 M * rmoriz NET_RAW in bcapabilities seem not to work 1146415765 M * rmoriz raw_icmp in ccapabilities neither 1146415779 M * Bertl hmm? objection! 1146415803 M * intrigeri rmoriz, you can use tracepath 1146415803 M * Bertl rmoriz: CAP_NET_RAW should work, raw_icmp is not sufficient 1146415823 M * Bertl traceroute insists on opening unlimited raw sockets 1146415823 M * rmoriz echo "CAP_NET_RAW" >> bcapabilities ? 1146415827 M * intrigeri rmoriz, that works inside a vserver and does not need any non-default capability iirc 1146415900 J * _coocoon_ ~coocoon@p54A07557.dip.t-dialin.net 1146416011 J * Viper0482 ~Viper0482@p54977205.dip.t-dialin.net 1146416039 M * rmoriz traceroute: ifaddrlist: SIOCGIFADDR: eth1: Cannot assign requested address :/ 1146416082 M * Bertl what address does it try to assign` 1146416210 M * rmoriz i've "dev1ffm01" in "name" 1146416231 M * rmoriz the ip for the vserver is bound to eth1 in the host 1146416260 M * rmoriz traceroute -i eth1:dev1ffm01 -s $realsourceip target.com does not work :( 1146416339 M * Bertl rmoriz: what kernel/patch versions? 1146416391 M * hillct are dlimits implemented in 2.01-pre2 ? 1146416417 M * rmoriz 2.6.16.9 with vs2.1.1-rc17 1146416431 M * hillct not related to my current problem but this is what happens when I start persing the flower page 1146416441 M * Bertl hillct: I'd say so, see the feature amtrix for details 1146416459 M * hillct feature matrix... URL? 1146416465 M * Bertl rmoriz: okay, could you upload an strace -fF traceroute target.com 1146416466 A * hillct growses 1146416470 M * hillct browses 1146416483 M * rmoriz strace? hmm.. hopefully it does not crash this time ;) 1146416497 M * Bertl x86_64 ? 1146416500 M * rmoriz yes 1146416506 M * Bertl okay, then no strace 1146416517 M * Bertl this seems to be an unresolved issue in mainline 1146416518 M * rmoriz oh well it worked 1146416683 M * rmoriz http://codernaut.org/strace1.txt 1146416771 M * Bertl 217.172.51.211 is assigned to your guest? 1146416783 M * rmoriz yes 1146416789 M * rmoriz i can ssh from outside for example 1146416796 M * rmoriz i can ping in and out 1146416802 M * rmoriz but traceroute seems not to work 1146416815 M * rmoriz traceroute -V 1146416815 M * rmoriz Version 1.4a12 1146416886 M * Bertl hmm, I assume it tries to get the addr of eth1 instead 1146416916 M * Bertl could be something we might be able to improve in the future 1146416942 M * rmoriz iirc this worked with 2.4 kernels and the older vs? 1146416977 M * Bertl could be, but I would not bet on it 1146416999 M * Bertl I think it really depends on which device and what ip gets where 1146417053 M * rmoriz traceroute-nanog works 1146417067 M * rmoriz traceroute.lbl doesn't 1146417099 M * Bertl tracepath should work always 1146417114 M * rmoriz :) 1146417119 M * rmoriz that's tcp only i guess? 1146417121 M * Bertl could you make a static version of your failing traceroute? 1146417140 M * Bertl (for testing purposes) 1146417150 M * rmoriz hm it's just debian's default one 1146417412 Q * lilalinux Ping timeout: 480 seconds 1146417943 J * lilalinux ~plasma@dslb-084-058-239-067.pools.arcor-ip.net 1146418300 M * rmoriz Bertl: http://codernaut.org/traceroute.tar.bz2 1146418343 M * Bertl rmoriz: tx! 1146418511 M * rmoriz i just dig gcc -static is that what you need? 1146418601 M * Bertl yep 1146420414 M * ray6 doesn't look like I'll be making it to the linuxtag :/ 1146420470 M * Bertl :( 1146420515 M * rmoriz ray6: didn you have an image of the wtc on your site? 1146420574 M * ray6 rmoritz: yeah, I had.. peace on earth when browsed with IPv4, broken down when browsed with v6 :) 1146420623 M * rmoriz heh 1146420626 M * ray6 rmoritz: I guess you mean this: http://www.ray.net/wtc.html (thats me in front of the WTC one day before Y2K :) 1146420641 M * rmoriz yes, but rmoriz without t ;) 1146420673 M * ray6 oh, I add little mistakes to proof I handtype the nicks instead of using unpersonal nickname completion :) 1146420681 M * Bertl rmoriz: you can buy one for 10? 1146420718 M * rmoriz Bertl: that tv show still has airtime in .at? ;) 1146420739 M * Bertl no, but I still remember a lot of useless stuf :) 1146420788 Q * rmoriz Quit: Reconnecting 1146420793 J * rmoriz ~roland@moriz.name 1146420803 M * rmoriz re 1146421136 M * Bertl wb rmoriz! 1146421607 M * rmoriz had some dns troubles a few days ago while switching from bind to powerdns and finally a day later to mydns ;) 1146422129 J * doownek ~doownek@emma.ipv6.doownek.org 1146422394 M * hillct I'm trying to understand a value reported by vserver-stat 1146422396 M * hillct CTX PROC VSZ RSS userTIME sysTIME UPTIME NAME 1146422396 M * hillct 0 95 15.9+ 232.2M 21h46m36 2h38m25 1d03h49 root server 1146422396 M * hillct 5 30 1.8G 633.8M 45m38s60 9m27s32 2h05m25 hulver.com 1146422396 M * hillct 10 46 15.9+ 1G 29m44s30 5m09s00 2h06m21 stdscoop 1146422396 M * hillct 20 40 1.8G 204.7M 1m41s58 0m47s65 23h44m50 mostly-harmless.ca 1146422415 M * nebuchadnezzar which one ? 1146422420 M * Bertl welcome doownek! 1146422421 M * hillct what does VSZ of 15.9+ indicate? 1146422439 M * Bertl a bug, IMHO :) 1146422443 Q * doownek Remote host closed the connection 1146422455 M * hillct relative to the values which include M or G suffixes 1146422457 M * hillct ah 1146422488 J * doownek ~doownek@emma.ipv6.doownek.org 1146422501 M * hillct so I've been reding and re-reading the vsched page 1146422509 M * nebuchadnezzar wb doownek 1146422556 M * hillct interesting stuff. Can the vsched values be set within the /etc/vservers tree or do I have to set them on the command line? 1146422581 M * Bertl hillct: yes, they can be set in the config 1146422595 M * Bertl but it doesn't support all features available in 2.1.1-rc* 1146422608 M * hillct didn't see them in the flower page 1146422632 M * Bertl it's currently one file which has all 'old' ABI values 1146422645 M * hillct is vsched only available in 2.1.x ? It seems to be in 2.0.x 1146422671 M * hillct ah 1146422749 M * Bertl http://linux-vserver.org/Release+FAQ 1146422790 M * Bertl but we need to extend that soon 1146422826 M * hillct as listed in the matrix, it's 'hard CPU limits' ? 1146422847 Q * doownek Remote host closed the connection 1146422848 M * Bertl yep 1146422854 M * hillct ah 1146422855 M * hillct K 1146422881 J * doownek ~doownek@emma.ipv6.doownek.org 1146422885 M * hillct how can I easily determine whether the kernel was compiled with hard CPU limits? 1146422895 M * Bertl hey doownek! troubles with the connection? 1146422902 M * hillct is that in proc/vitrtual/info or something? 1146422919 M * doownek Bertl: yes a little problem 1146422923 M * Bertl hillct: yes, the VCI information contains that 1146422924 M * doownek hello all 1146422934 M * nebuchadnezzar hello doownek 1146423079 M * hillct Bertl: I guess I'm not looking in the right place 1146423095 M * hillct I was looking at /proc/vitrtual/info 1146423098 M * Bertl use testme.sh 1146423120 M * Bertl it reports the VCI flags, give me that info and I can tell you what is configured 1146423120 M * hillct hmm 1146423129 M * hillct default path? 1146423138 M * hillct /usr/lib? 1146423154 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/ 1146423160 M * hillct oh 1146423164 M * hillct not in the default distro 1146423167 M * hillct K. thanks 1146423168 M * Bertl it's packaged in most distros 1146423183 M * Bertl but I don't know where they put it if at all 1146423197 M * hillct checking 1146423434 M * Bertl rmoriz: that traceroute seems to behave very strange 1146423455 M * Bertl # ifconfig lo 127.0.0.1 1146423455 M * Bertl # ifconfig eth0 10.0.0.2 1146423459 M * Bertl # chbind --ip 10.0.0.2 -- chcontext --xid 666 -- tracepath 10.0.0.1 1146423462 M * Bertl works fine 1146423466 M * Bertl # chbind --ip 10.0.0.2 -- chcontext --xid 666 -- traceroute 10.0.0.1 1146423470 M * Bertl works fine 1146423475 M * Bertl # chbind --ip 10.0.0.2 -- chcontext --xid 666 -- /tmp/traceroute 10.0.0.1 1146423480 M * Bertl (that's the one from you) 1146423484 M * Bertl /tmp/traceroute: 4: Syntax error: "|" unexpected 1146423634 M * Bertl could you try similar on your system? 1146423678 M * hillct Bertl: I clanced at the code and ran testme.sh in vorbose mode but I'm unclear on what output I would expect to see relating to CPU Hard Limiting tests 1146423710 M * Bertl VCI: 0002:0101 273 031101e6 1146423718 M * hillct ah 1146423722 M * Bertl this (or a similar line) is the improtant one 1146423749 M * hillct VCI: 0002:0001 236 03000016 (TbLgnP) 1146423801 M * hillct is it the existance of the line or the values that are meaningful? 1146423807 M * Bertl if you now look into kernel/vserver/vci_config.h 1146423832 M * Bertl you see that VCI_KCBIT_HARDCPU will get 5 1146423849 M * Bertl so it's basically 0x20 in hex 1146423930 M * Bertl (in your case that is not set) 1146423961 M * Bertl when I get around, I extend the testme.sh to decode that in plaintext 1146424014 M * hillct ah 1146424022 M * hillct I see. thanks 1146424030 M * Bertl you're welcome! 1146424054 M * hillct so I guess I need to at least rebuilt, and it's probably worth upgrading 1146424092 M * Bertl well, looking at the recent kernel mainline exploits, probably yes 1146424132 M * ray6 Bertl: hm, which important ones were there? 1146424151 M * hillct yeah, I wanted to hold off to at the same time do a full update to FC5, but first let FC5 release issues get ironed out 1146424181 M * Bertl ray6: well, basically 2.6.16.7-11 1146424250 M * hillct Bertl: rlimits/cpu is also an aspect of CPU Hard scheduling right? 1146424263 M * hillct or is that a different limiting mechanism? 1146424273 M * nebuchadnezzar is it possible to edit http://linux-vserver.org/HowtoCloneVserver to add http://www.asgardr.info/~nebu/vserver/vserver-copy ? 1146424285 M * nebuchadnezzar maybe a new page on the wiki because I'm only on adsl 1146424285 M * Bertl hillct: you don't want that (actually it's not even implemented) 1146424312 M * Bertl hillct: it would limit your guest to soandso many seconds of total cpu time 1146424334 M * Bertl hillct: i.e. 10 seconds and then the guest would be killed of :) 1146424337 M * hillct nebuchadnezzar: I have a tweaks page I put up many montrhs back on somewhat the same topic 1146424361 M * hillct Bertl: K 1146424370 A * hillct avoids that one 1146424686 M * nebuchadnezzar it's seems that the wiki is public editable... 1146424706 M * Bertl nebuchadnezzar: I hope so :) 1146424716 M * hillct It is 1146424727 M * nebuchadnezzar and I ask the permission to someone to edit that page ? 1146424727 M * hillct nebuchadnezzar: I'll add your tool 1146424737 M * nebuchadnezzar hillct: thanks 1146424766 M * Bertl nebuchadnezzar: usually you just edit the page (using preview) and then save it ... 1146424946 M * nebuchadnezzar does someone look at implementing all those scripts in perl ? it can be interessting I think ;-) 1146425028 M * daniel_hozac what happened to mugwump's libvserver perl module? 1146425060 M * daniel_hozac but i guess the utils are complicated enough as is. 1146425072 M * daniel_hozac we relaly don't want to use a write-only language as well :) 1146425077 M * daniel_hozac +spelling 1146425122 M * nebuchadnezzar daniel_hozac: arf, shell script is more write-only, it's script, Perl is a programming language 1146425124 M * nebuchadnezzar :-) 1146425150 M * daniel_hozac shell is far easier to read than Perl... 1146425163 A * nebuchadnezzar do not agree 1146425333 J * knotty ~void@loin.ailleurs.org 1146425996 M * Bertl wb knotty! 1146426010 M * knotty hi Bertl 1146426081 M * Bertl okay, off for medidation .. back later ... 1146426088 N * Bertl Bertl_oO 1146426777 Q * Viper0482 Remote host closed the connection 1146428138 Q * cehteh Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1146428408 J * cehteh foobar@cehteh.homeunix.org 1146429140 J * doener_ ~doener@p54876DE9.dip.t-dialin.net 1146429659 M * _coocoon_ got this message after building a vserver --> chattr: Unpassender IOCTL (I/O-Control) fr das Ger� beim Lesens der Flags in /etc/vservers/.defaults/vdirbase/apache 1146429667 M * _coocoon_ has anyone an idea why 1146429706 M * _coocoon_ kernel 2.6.16.11-vs2.1.1-rc18 1146429725 M * _coocoon_ util-vserver: 0.30.210 1146429755 M * daniel_hozac how did you build the guest? when exactly do you get that message? 1146429793 M * _coocoon_ vserver apache build -m skeleton --hostname apache --context 59 1146429810 M * _coocoon_ during building i think so 1146429906 M * _coocoon_ directly after i fired up the command 1146430184 M * _coocoon_ first time i got this message is with 2.1.1-rc18 1146430406 M * daniel_hozac that command shouldn't run chattr, ever. 1146430454 M * _coocoon_ hm i do not know 1146430525 M * _coocoon_ i got this message on two different systems at my home pc and at the system in the net 1146430629 M * _coocoon_ home --> chattr: Unpassender IOCTL (I/O-Control) für das Gerät beim Lesens der Flags in /etc/vservers/.defaults/vdirbase/apache 1146431108 M * daniel_hozac hmm, it appears it does run chattr... i guess that's a bug. 1146431112 M * daniel_hozac does that make your build fail? 1146431375 M * _coocoon_ no everything is working fine 1146431386 M * _coocoon_ i will reboot with another kernel to check 1146431616 M * _coocoon_ it is the same 2.6.16-vs2.1.1-rc15 --> have disabled legacy kernel api maybe this is the r4eason 1146431711 M * daniel_hozac yes, chattr .t is legacy cruft. 1146431727 M * daniel_hozac i have no idea why the utils uses it at all. 1146432082 J * mkhl ~mkhl@200-148-41-150.dsl.telesp.net.br 1146432239 M * derjohn daniel_hozac, where can I find your magic 12 (?) patches for the utils? 1146432428 Q * shedi Quit: Leaving 1146432705 M * daniel_hozac http://daniel.hozac.com/vserver/util-vserver/ 1146433136 Q * doener_ Read error: Connection reset by peer 1146433214 J * doener_ ~doener@p54876DE9.dip.t-dialin.net 1146434017 N * Bertl_oO Bertl 1146434025 M * Bertl back now ... 1146434073 M * Hollow hey Bertl, uploaded the ppc test results for rc18 1146434159 M * Bertl ah, and? looking good? 1146434674 J * kir_home ~kir@ffm-nat.sw-soft.de 1146434828 M * Bertl wb kir_home! 1146434846 M * Bertl Hollow: almost got a new lo test version ready, are you interested? 1146434847 M * kir_home Bertl, hello 1146434869 M * Hollow Bertl: yep 1146434907 M * nebuchadnezzar Bertl: i am, I can test it on sparc :-) 1146434941 M * Bertl nebuchadnezzar: even better, okay, will have it ready in 30min or so 1146434979 M * nebuchadnezzar ok 1146435000 M * nebuchadnezzar 2.6.17 or 2.6.16 ? 1146435020 M * Bertl 2.6.16.11, didn't verify if the xfs issue in 2.6.17 is already fixed 1146435047 M * nebuchadnezzar ok 1146435681 Q * intrigeri Quit: ERC Version 5.0.4 $Revision: 1.726.2.20 $ (IRC client for Emacs) 1146436602 M * nebuchadnezzar I don't know for XFS in rc3 1146436612 M * nebuchadnezzar the bug is still present in rc2 1146436845 Q * mkhl Quit: 1146437834 M * nebuchadnezzar daniel_hozac: do your remember my problem on sparc with input ? 1146437844 M * nebuchadnezzar which version of util-vservers correct this ? 1146437873 M * nebuchadnezzar doownek have 0.30.210-7bpo2 and the problem is present 1146437927 M * nebuchadnezzar commenting le line with --vlogin in /usr/lib/util-vservers/vserver.function make it works correctly 1146438184 M * Bertl interesting ... 1146438198 M * nebuchadnezzar Bertl: arf, on sparc the context can not be 65535 :-/ 1146438217 M * nebuchadnezzar it works on i386 1146438226 M * Bertl aha, any ideas why? 1146438230 M * nebuchadnezzar no 1146438241 M * Bertl what error do you get when you do, e.g. 1146438251 M * nebuchadnezzar chbind: vc_net_create(): Invalid argument 1146438253 M * Bertl chcontext --xid 65535 -- true 1146438276 M * Bertl ah, for chbind? well, that is fine 1146438282 M * nebuchadnezzar I can not start the vserver 1146438290 M * Bertl -1 was/is used for 'current' 1146438313 M * Bertl but it's unusual that it should only use 16 bit? 1146438335 M * Bertl nebuchadnezzar: do you have debug enabled? 1146438340 M * nebuchadnezzar 16bits is the arbitrary limit in 'I don't remember the header' 1146438347 M * nebuchadnezzar Bertl: yes normally 1146438357 M * Bertl okay, does that include the vserver debug? 1146438370 M * baggins Bertl: hello 1146438376 M * Bertl hey baggins! 1146438386 M * baggins I'm back with bugs ;) 1146438395 M * Bertl great! let's hear! 1146438406 M * nebuchadnezzar do I need to enable something for debug ? 1146438417 M * baggins Bertl: see the list, I hit kernel BUG in kernel/exit.c at 676 1146438437 M * baggins gdb kills my systems 1146438438 M * Bertl nebuchadnezzar: try if you can do: sysctl -w vserver.debug_switch=255 1146438451 M * Bertl baggins: which kernel/patch release? 1146438478 M * Bertl nebuchadnezzar: then try to start your guest, and check the output in syslog/dmesg 1146438481 M * baggins 2.6.16.11, 2.1.1-rc17 with my ugly hack (euid == 0) 1146438530 M * nebuchadnezzar Bertl: it's cryptic to me 1146438533 M * Bertl baggins: hmm, so you got a reaper outside your context 1146438572 M * Bertl baggins: did your init die? 1146438573 M * baggins Bertl: all I did was: vserver test enter ; gdb /bin/ls , run and BUG 1146438614 M * baggins Bertl: no idea, it's dead after BUG so I can't check anything 1146438614 M * Bertl okay, let's make that a debug info and output the pids/xids shall we? 1146438658 M * Bertl nebuchadnezzar: yes, please upload it to paste.linux-vserver.org :) 1146438696 M * nebuchadnezzar the limit is between 4000 and 50000 1146438716 M * Bertl ah, I'd supsect it is 49152 1146438723 M * Bertl which would be quite fine 1146438738 M * Bertl i.e. 2-49151 is static, above is dynamic 1146438776 M * nebuchadnezzar arf, I do not disable dynamic context in that kernel 1146438783 M * Bertl yep 1146438805 M * nebuchadnezzar http://paste.linux-vserver.org/12 1146438867 M * Bertl nebuchadnezzar: yeah, looks fine, you can disable the debug with sysctl -w vserver.debug_switch=0 1146438879 M * nebuchadnezzar it's already done:-) 1146438897 M * nebuchadnezzar ok, so, where is my patch ? :-) 1146438922 M * Bertl nebuchadnezzar: no patch, either disable dynamic contexts or live with the limit :) 1146438938 M * nebuchadnezzar lo patch I say 1146438954 M * Bertl ah, okay, sec, takes a little longer 1146438959 M * nebuchadnezzar ;-) 1146438972 M * Bertl aprox 15min 1146438976 J * s0undt3ch ~s0undt3ch@bl7-245-254.dsl.telepac.pt 1146438982 M * Bertl welcome s0undt3ch! 1146438983 M * s0undt3ch hello ppl 1146438987 M * nebuchadnezzar no pb, my bed can wait, holidays 1146439124 M * s0undt3ch I'm building the vserver kernel on a system that is running a normal kernel(gentoo) 1146439136 M * s0undt3ch i'm accessing it remotely 1146439149 M * Bertl sounds good! 1146439167 M * s0undt3ch think it would be safe to reboot after grub is config'ed to run it? 1146439188 M * Bertl you might want to add panic=60 and enable the one time grub boot feature? 1146439275 M * s0undt3ch k, help me out on that one ;) set panic=60 where? kernel /boot/kernel-2.6.14-gentoo-r2 root=/dev/sda2 panic=60? for example? 1146439288 M * Bertl yep, kernel boot command 1146439317 M * s0undt3ch a what about the one time grub boot feature? 1146439348 M * Bertl grub-set-default, read the grub info page 1146439359 M * Bertl basically you set default saved 1146439367 M * Bertl and change the default to the known good with 1146439371 M * Bertl savedefault 0 1146439379 M * Bertl after each option select 1146439439 M * doener_ Bertl: hm, how do I interpret vx_bcaps and vx_cap_bset? (comments didn't help :/) 1146439463 M * Bertl doener_: you know the 'default' capset which can be configured via sysctl? 1146439471 M * Bertl doener_: (on vanilla) 1146439482 M * s0undt3ch so I've set on top default = 0, and after the vserver kernel I set savedefault 1 (the kernel that runs)? 1146439505 M * doener_ no, but that sounds like either limiting root's caps or raising regular users' caps to that value 1146439506 M * Bertl s0undt3ch: no, the default is an index (i.e. the number of your boot entry) 1146439520 M * Bertl s0undt3ch: at the beginning you specify 'default saved' 1146439531 M * Bertl s0undt3ch: which means, that the default is saved somewhere 1146439550 M * Bertl s0undt3ch: now on position 0 (first entry) you place your known good kernel 1146439566 M * Bertl s0undt3ch: and for each other entry you add savedefault 0 1146439582 M * Bertl s0undt3ch: then you do grub-set-default 1 (e.g.) 1146439591 M * Bertl which will boot entry 1, and save a default for 0 1146439599 M * Bertl so next time it will boot the 0 again 1146439608 M * Bertl doener_: kernel.cap-bound 1146439636 M * Bertl doener_: it is a global limit, inside the kernel that is known as cap_bset (global) 1146439660 M * Bertl doener_: the vx_bcaps basically used that as inital default 1146439697 M * Bertl doener_: the new code remapped this to vx_cap_bset (currently unused default) 1146439716 M * Bertl doener_: and the vx_bcaps is the mask which gets applied on 'test' 1146439724 M * s0undt3ch Bertl: and remove 'default 0'? 1146439735 M * Bertl yep, you replace that with default saved' 1146439741 M * s0undt3ch k 1146439782 M * doener_ Bertl: so root's permission cannot exceed cap_bset, right? 1146439798 M * doener_ s/root's/noone's/ 1146439824 M * Bertl doener_: the default capset is cap_bset 1146439847 M * Bertl changes to that global value will only affect new processes 1146440039 M * Bertl nebuchadnezzar, Hollow: http://vserver.13thfloor.at/Experimental/delta-lo0.04.1.diff 1146440056 M * nebuchadnezzar Bertl: appli on top of vs2.1.1-rc18 ? 1146440069 M * Bertl has one 'known bug', crosstalk via lo is still possible with udp (checking for that now) 1146440076 M * Bertl yes, ontop of rc18 1146440089 M * nebuchadnezzar ok 1146440104 M * s0undt3ch Bertl: so to go over the drill, I've commented out 'default 0' and instead put 'default saved', then the 0 entry is my good kernel, and 1 entry is: 1146440110 M * s0undt3ch title=kernel-2.6.15-vs2.0.1-gentoo-r1 1146440113 M * s0undt3ch root (hd1,0) 1146440118 M * s0undt3ch kernel /boot/kernel-2.6.15-vs2.0.1-gentoo-r1 root=/dev/sda2 panic=60 1146440122 M * s0undt3ch savedefault 0 1146440143 M * s0undt3ch now I'm safe to reboot after 'grub-set-default 1' 1146440146 M * s0undt3ch correct? 1146440150 M * Bertl yes, just check that your grub supports that 1146440167 M * Bertl and that the syntax is correct, I didn't verify it 1146440170 M * s0undt3ch grub (GNU GRUB 0.96) 1146440181 M * nebuchadnezzar Bertl: ok, building 1146440204 M * s0undt3ch Bertl: support what grub-set-default? 1146440217 M * knotty too much suspens nebuchadnezzar 1146440230 M * nebuchadnezzar sure 1146440243 M * Bertl s0undt3ch: yes, that and the savedefault statement 1146440269 M * Bertl s0undt3ch: certain patched versions do not, so if that is the first time you try, better be careful and check twice 1146440274 M * s0undt3ch well I do have the grub-set-default binary 1146440345 M * Bertl which is a good sign :) 1146440354 M * doener_ s0undt3ch: just start the grub shell and enter "savedefault 0" 1146440373 M * doener_ but my gentoo grub supports that syntax 1146440409 M * doener_ patched version (e.g. older debian grub version) had a special "savedefault --once --default=X" syntax and IIRC don't accept "savedefault X" syntax 1146440411 M * s0undt3ch doener_: and grub's info page also mentions that 1146440427 J * are|lunch are@dslb-084-057-131-144.pools.arcor-ip.net 1146440558 M * s0undt3ch trying and praying ;) 1146440651 M * Bertl praying to murphy? :) 1146440725 M * s0undt3ch Linux kermit 2.6.15-vs2.0.1-gentoo-r1 #1 SMP PREEMPT Mon May 1 00:31:14 WEST 2006 i686 Intel(R) Pentium(R) 4 CPU 2.40GHz GenuineIntel GNU/Linux 1146440728 M * s0undt3ch ;) 1146440733 M * s0undt3ch it booted! 1146440735 M * Bertl so seems it worked :) 1146440817 Q * _are_ Ping timeout: 480 seconds 1146440833 M * s0undt3ch k, all I need now I to config it as a vserver and create a new vserver so I can temporarely migrate my server services to this comp(the desktop one) while I do a stage one gentoo vserser on the real server :) 1146440954 J * ntrs_ ~ntrs@dsl-203-81-rt2-bras.wnvl.centurytel.net 1146440979 M * Bertl wb ntrs_! 1146441021 M * s0undt3ch wb ntrs_! ? 1146441035 M * Bertl wb = welcome back :) 1146441090 M * s0undt3ch oh ;) 1146441114 M * s0undt3ch my irssi is ignoring joins/parts/quits :) 1146441116 M * daniel_hozac nebuchadnezzar: the fix for that was a dietlibc patch. 1146441125 M * nebuchadnezzar arf, ok 1146441130 M * daniel_hozac nebuchadnezzar: and then recompiling util-vserver with that dietlibc. 1146441132 M * baggins Bertl: any idea how gdb could cause reaper to be from outside the context? 1146441170 M * Bertl baggins: no idea, it could be a bug too, but I'd assume it messes with the father (which can be outside when you do enter) 1146441183 M * Bertl baggins: you might try to 'enter' via ssh for example 1146441184 M * nebuchadnezzar daniel_hozac: so I should as micah for what is on backports.org, isn't it ? 1146441235 M * baggins Bertl: my users tried the ssh attemp, that's how I got the x8664 oops 1146441273 M * Bertl okay, what about making that BUG_ON a kernel message? 1146441285 M * Bertl baggins: could/would you test that for me? 1146441331 M * nebuchadnezzar arf, I got error in net/dccp, I disable it since I do not use that 1146441342 M * nebuchadnezzar suspence, suspence 1146441345 M * nebuchadnezzar :-/ 1146441345 M * baggins Bertl: of course :) But not now, it's 2am here and i need some sleep ;) 1146441362 N * are|lunch _are_ 1146441363 M * Bertl nebuchadnezzar: good to know, will add that to testing/config 1146441379 M * nebuchadnezzar prblem with number of parameters 1146441379 M * Bertl baggins: okay, contact me when you are ready ... 1146441384 Q * ntrs Ping timeout: 480 seconds 1146441395 M * nebuchadnezzar net/dccp/ipv4.c: In function ??dccp_v4_err??: 1146441395 M * nebuchadnezzar net/dccp/ipv4.c:323: erreur: too few arguments to function ??inet_lookup?? 1146441395 M * nebuchadnezzar net/dccp/ipv4.c: In function ??dccp_v4_hnd_req??: 1146441395 M * nebuchadnezzar net/dccp/ipv4.c:650: erreur: too few arguments to function ??__inet_lookup_established?? 1146441396 M * s0undt3ch doener_: you're running vserver on gentoo? 1146441399 M * nebuchadnezzar net/dccp/ipv4.c: In function ??dccp_v4_rcv??: 1146441401 M * nebuchadnezzar net/dccp/ipv4.c:961: erreur: too few arguments to function ??__inet_lookup?? 1146441405 M * nebuchadnezzar make[3]: *** [net/dccp/ipv4.o] Erreur 1 1146441408 M * nebuchadnezzar 1146441410 M * nebuchadnezzar oops 1146441414 M * nebuchadnezzar sorry 1146441443 J * ntrs ~ntrs@dsl-203-81-rt2-bras.wnvl.centurytel.net 1146441479 Q * ntrs_ Read error: Connection reset by peer 1146441495 M * doener_ s0undt3ch: no, my thinkpad (still) runs gentoo, my desktop box is debian once again... will probably move to some other distro on the laptop as well, but that's yet to decide 1146441510 M * doener_ ... and on the laptop, i have no vservers 1146441533 M * Bertl doener_: why so? 1146441539 M * s0undt3ch oh, was just asking to know which vserver-sources you where running 1146441556 M * doener_ Bertl: why what? 1146441569 M * Bertl no vservers on your laptop? :)