1138579299 Q * bwana Read error: Connection reset by peer 1138579355 J * bwana ~bwana@71-82-125-40.dhcp.roch.mn.charter.com 1138579495 N * bwana bwana2oo 1138579508 N * bwana2oo bwana 1138580989 J * shed| ~siggi@inferno.lhi.is 1138581394 Q * shedi Ping timeout: 480 seconds 1138581492 J * SNy_ 8de58f95fb@bmx-chemnitz.de 1138581583 M * Bertl welcome SNy_! 1138581647 Q * SNy Ping timeout: 480 seconds 1138581846 Q * bwana Quit: reboot 1138582013 M * marl_ ok, i have entered /dev/mem into /etc/vservers/.defaults/apps/vprocunhide, and restarted vprocunhide as well as the client vserver, and im still getting operation not permitted when trying to access /dev/mem on X startup :( any pointers as to were im messing things up? 1138582208 M * Bertl well, it could be that simply the required caps are missing 1138582223 M * Bertl best would be to provide an strace -fF of the relevant section 1138582328 M * marl_ ok, REALY daft question, how do i strace the 'relivant section' ? 1138582336 M * marl_ is the any info ont he site to read on that? 1138582350 M * Bertl best do an strace -fF -o X.trace X 1138582367 M * Bertl then look through the really huge file and see where you get an EPERM 1138582389 M * Bertl copy out 20 lines around that place and put it on pastebin.com 1138582416 M * marl_ ok thanks, will try in a few mins 1138582698 M * derjohn Bertl, wiki says: devel vs2.1.0.4 ? 1138582708 M * Bertl wiki is wrong ... 1138582728 M * derjohn Bertl, it on old exp ? 1138582732 M * derjohn *it's 1138582742 M * Bertl all experimental releases should be there 1138582743 A * derjohn tired 1138582767 J * Torvan ~toarvan@24.138.193.90 1138582774 M * Bertl welcome Torvan! 1138582784 M * Torvan HI all 1138582853 Q * Torvan Quit: 1138582886 M * marl_ Bertl, http://pastebin.com/529730 eperm is on line 36 1138582911 M * Bertl derjohn: I told the guy who added those sections, that I really think this is a good idea, but (and that's the point) I do not intent to update this info very often .. so he has to keep them updated somehow ... 1138582928 M * derjohn http://linux-vserver.org/ChangeLogExperimental 1138582929 M * derjohn :) 1138582988 M * Bertl hmm, is this a good idea? 1138583017 M * Bertl derjohn: I mean, are you going to maintain it? 1138583040 M * derjohn Bertl, just started maintaining :) 1138583059 M * Bertl okay, we are at 2.1.0.9 ... ask if you ahve questions ... 1138583100 M * derjohn Bertl, I have, yes, but not this night ... just splitted up the changelogs. 1138583105 M * Bertl marl_: sure that is the 'right' EPERM? 1138583245 M * marl_ lol, sorry, just grabed the first one i saw, didnt realy understand the syntax, try this one http://pastebin.com/529736 1138583248 M * marl_ line 24 1138583285 M * marl_ and this is the trace from starting gdm 1138583313 M * marl_ as gdm is what apears to be failing 1138583352 M * Bertl what capabilities do you give? 1138583428 M * marl_ cat /etc/vservers/test/capabilities has one line consisting off : ALL 1138583486 M * Bertl and? does that work? 1138583502 M * marl_ when starting gdm? no :( 1138583505 M * Bertl what gives 'grep Cap /proc/self/status' 1138583532 M * marl_ from inside trhe vserver? 1138583573 M * marl_ nothing is returned for ether host or client 1138583630 M * marl_ i know im missing something REALY stupid, i just cant find it :( 1138583638 M * derjohn Bertl, bye. I changed the wiki. I will take care for the changelog tomorrow or so. 1138583706 M * marl_ Bertl, is the cap grep is searching for definatly lower case? 1138583846 M * Bertl marl_: hmm? can't follow 1138583874 M * Bertl the grep will give you 3 lines (inside the guest and on the host) 1138583883 M * Bertl we are interested in the three lines from the guest 1138583890 M * marl_ the grep statment u told me to run : grep cap /proc/self/status : should it be grep -i cap /proc/self/status? 1138583894 M * Doener marl_: it's Cap, with uppercase C 1138583904 M * Doener and Bertl actually said that ;) 1138583915 M * Bertl marl_: does your client wrap the case? 1138583976 M * marl_ wrap the case? u mean is it case insensitive in greps normally, no 1138583989 M * marl_ cap lines (stating with Cap) are as follows 1138584006 M * Doener no, he meant if your irc clients converts everything to lowercase 1138584010 M * marl_ CapInh 0000000000000000 1138584025 M * marl_ no, irc dosnt convert 1138584054 M * Bertl ok, keep going :) 1138584059 M * marl_ sorry :( re-read your line, very hard to see the case on this screen :( 1138584080 M * marl_ CapPrm 00000000344c04ff 1138584093 M * marl_ CapEff 00000000344c04ff 1138584096 M * marl_ and thats it 1138584099 M * Bertl doesn't look like 'ALL' is doing anything good 1138584122 M * Bertl not sure it is supposed to do anything 1138584125 M * marl_ should they be all 'f's? 1138584141 M * Bertl assumed that ALL option works, yes 1138584163 M * Bertl or well, at least all but the first (highest one) 1138584181 M * marl_ ah, ok, how do i assign all caps to a vserver? 1138584188 M * Bertl Doener: any thought regarding the ALL stuff? 1138584209 M * Doener just checking util-vserver, at least the testsuite has it, still grepping 1138584232 M * Bertl marl_: in the meantime, what versions :) 1138584242 M * Doener marl_: is there a newline at the end? IIRC util-vserver doesn't like lines without newline at the end 1138584249 Q * SNy_ Remote host closed the connection 1138584254 J * SNy a0491ccf70@bmx-chemnitz.de 1138584271 M * marl_ lol, sorry should have said at start, kernel-2.6.15.1 with util-vserver-0.30.210 1138584319 M * marl_ newline plus one blank after it 1138584357 M * Doener ALL seems to be flags-only, not bcaps 1138584424 M * marl_ ok, so i mucked up there :( do i have to enter all the caps lines seperatly then? 1138584429 M * Bertl marl_: okay, you can list relevant ones one by one, but be careful, a guest with sufficient permissions will take down your network interfaces and happily reboot the system 1138584466 M * marl_ any hints asto wich one would allow access to /proc/mem? 1138584484 M * meebey please document the caps/bcaps to the wiki 1138584495 M * Bertl yeah, please do so :) 1138584517 M * marl_ lol 1138584521 M * meebey so others know which one are required, searching the caps/bcaps can be very annoying :) 1138584564 M * marl_ so there isnt any single page with the info on it at the moment? :( what about within the install tar.gz ? 1138584592 M * meebey CAP_SYS_RAWIO X needs I know 1138584606 A * meebey has vmware running inside vserver ;) 1138584620 M * meebey I should update the working programs page with my stuff 1138584700 M * meebey for some reason I still prefere and use the old config style 1138584710 M * meebey its just one plan file, easy to deploy/copy 1138584716 M * meebey s/plan/plain/ 1138584803 M * Bertl marl_: do you ahve /dev/mem inside the guest? 1138584818 M * marl_ ok, ive done SOMETHING wrong, have entered the CAP_SYS_RAWIO into the cap file and /proc/self/status hasnt changed 1138584839 M * marl_ and yup /dev/mem is inside the client 1138584846 M * Bertl did you restart the guest? 1138584850 M * marl_ yup 1138584854 M * meebey maybe the caps are called different in the new config style? 1138584862 M * meebey I pasted that from old config :-P 1138584869 Q * SNy Ping timeout: 480 seconds 1138584880 M * Doener 02:10:27 marl_ cat /etc/vservers/test/capabilities has one line consisting off : ALL 1138584890 M * Doener that's bcapabilities on 2.6 kernels 1138584897 M * meebey /etc/vservers/*.conf vs /etc/vservers/*/* 1138584909 M * marl_ b**ger :( 1138584915 M * meebey whats the diff between cap and bcaps? 1138584939 M * meebey are they enforced differently? 1138584957 M * Doener AFAIK util-vserver just uses two different files for the two kernel versions... for whatever reasons... 1138584974 M * meebey aha :) 1138584975 M * marl_ :) IT WORKED :) 1138584980 M * marl_ bcaps.... it was :) 1138585047 M * Bertl yes, CAP_SYS_RAWIO is the check 1138585055 M * Bertl (kernel code) 1138585068 M * Bertl will allow the guest to mess with arbitrary ports 1138585068 M * marl_ ok, for information, all i ended up with in bcaps is the CAP_SYS_RAWIO 1138585090 M * Doener Bertl: how did you check that? i have no idea where to look for caps concerning dev nodes 1138585101 M * Bertl static struct file_operations mem_fops 1138585114 M * Bertl (check the .open= 1138585118 M * meebey Doener: I do try & error 1138585128 M * meebey Doener: full list of all caps, and remove one by one ;) 1138585133 M * meebey good that I have apprentices 1138585156 M * Doener Bertl: so $NODENAME_fops is a good thing to search for in general, when looking for dev nodes? 1138585165 M * Doener (unless I do dirty things with udev ;) 1138585249 M * meebey rawio is indeed dangerous, but X doesnt give you a choice 1138585251 M * Doener looks like :) thanks 1138585288 M * Bertl Doener: well, I usually search for "mem" or so 1138585306 M * Bertl which in turn points me to the fops ... 1138585319 M * Doener hm, plain searching for mem led to a 'few' too many results 1138585333 M * Bertl note the double quotes :) 1138585373 M * Doener ah, those are part of the search term 1138585568 M * mugwump Hey guys. I'm back home now, looking at playing with stgit to manage producing some patchsets 1138585596 M * Doener hey mugwump 1138585601 M * Bertl welcome home mugwump! 1138585617 M * mugwump First day back at my desk since 23rd Dec :) 1138585624 J * gerrit ~gerrit@216.65.245.8 1138585631 M * Bertl wb gerrit! 1138585660 A * Hollow yawns 1138585667 M * gerrit Hi Bertl - 35,000 feet and zooming... 1138585697 M * Bertl cool! 1138585722 M * gerrit satellite link and slow, or so it seems but it works. ;) 1138585777 M * gerrit somewhere near the Baffin Islands, which I've never heard of (somewhere near Greenland) 1138585887 M * marl_ gerrit, which direction? 1138585898 M * Hollow Bertl: i have replaced dietlibc's printf functions for vserver-utils.. http://phpfi.com/99161 1138585907 M * gerrit headed from portland to frankfurt 1138585935 M * marl_ well if u pass over west scotland give us a wave :) 1138585945 M * gerrit I'm waving the whole way. ;) 1138585948 M * Bertl Hollow: cool, url for the printf? 1138585978 M * Hollow http://home.xnull.de/work/vserver/vserver-utils/src/libinternal/printf.c 1138586028 M * Hollow it has not all features described on the man page (yet) but i have implemented more than i ever used before ;) 1138586100 M * Bertl yeah, looks nice, somebody should audit it ... 1138586108 M * Hollow definitely :) 1138586146 M * Hollow libowfat is really nice 1138586172 M * Bertl yeah, I've heard so 1138586252 M * Hollow ok, off to bed then.. cu tomorrow 1138586267 J * SNy 696f7054ec@bmx-chemnitz.de 1138586295 M * marl_ is it posible to start a client in a mode other than that in its own inittab file? like booting a kernel with a 1 at the end to enter single suer mode? 1138586365 M * Bertl you can start guests in many different modes 1138586387 M * Bertl check the Flower Page for the relevant config files 1138586430 M * marl_ thanks 1138586456 M * Bertl IIRC runlevel 3 is default for some distros 1138586462 M * Doener hm, those non-blank blank lines distract me... 1138586497 M * meebey Bertl: why is that printf so important to be small? 1138586523 M * Bertl meebey: no idea, but it's nice to have a non-bloated version, no? 1138586529 M * meebey heheheh 1138586560 M * Bertl I'd assume (out of the blue) Hollow got annoyed by the diet messages stating that the file got bloated 1138586582 M * meebey :) 1138586585 M * Bertl and so he decided to combine coding practice with something useful 1138586614 M * meebey annoying messages can change the world, interesting :) 1138586656 M * meebey for embedded stuff its definetly important 1138586758 M * Bertl well, for testing it's useful too 1138587419 J * bwana ~bwana@71-82-125-40.dhcp.roch.mn.charter.com 1138587432 M * Bertl wb bwana! 1138588149 M * bwana howdy Bertl 1138588839 M * Doener good night! 1138588841 Q * Doener Quit: Leaving 1138589925 M * matti Uhh, there's no better thing than good book, indeed. 1138589926 M * matti :) 1138589948 M * Bertl well, I cannot object here :) 1138589950 M * matti romke: Thanks for "Limes Inferior" it is really awesome :) 1138589965 M * matti Bertl: :-) 1138590111 M * matti Bertl: http://en.wikipedia.org/wiki/Limes_inferior 1138590133 M * matti Bertl: Not very good description, but I cannot find better iDn english :) 1138590205 M * matti Hm, and what is that... s/iD[D[Dn/in/ 1138590816 M * locksy is vserver SMP safe? 1138590829 M * mugwump locksy: yes. 1138590868 M * locksy Any known issues with hyper-threading? 1138590960 M * mugwump I doubt it, many people use fairly high end systems with vserver 1138590964 M * mugwump not tried that myself 1138591021 M * bwana so as far as services, dns, mail, ftp, db, httpd, etc., are their any that don't run well/properly as a vserver 1138591120 M * mugwump this is a faq, see the wiki 1138591131 M * mugwump in summary, not really. 1138591182 M * mugwump ntp has problems, samba can be hard, also non-ip daemons like dhcp/rarpd can't be restricted to IPs 1138591203 M * mugwump but almost everything else works fine 1138591780 M * Bertl ntp doesn't make sense in a guest 1138591792 M * mugwump well, except to jail ntp 1138591858 M * mugwump why is fs/xfs/linux-2.6/xfs_sysctl.c modified in 02_xxx_base.diff ? :) 1138592056 M * Bertl which patch set? 1138592092 M * mugwump split-2.6.14.4-vs2.1.0 and split-2.6.14.3-vs2.01 are both that way 1138592418 M * Bertl ah, because the sysctl needs patching 1138592438 M * Bertl it's not xfs specific, it's just for all sysclt entries 1138592742 M * mugwump does 2.1.0 include PID virtualisation? 1138592762 M * mugwump I see find_task_by_real_pid etc 1138592868 M * Bertl no virtualization except for initpid 1138592874 M * mugwump ahh 1138593015 M * mugwump hmm, syslog virtualisation? I see include/vserver/cvirt.h:vx_do_syslog() 1138593020 M * Bertl yep 1138593021 M * mugwump is that for printk() ? 1138593038 M * Bertl well, more for ksyslogd :) 1138593051 M * mugwump right, so you can run ksyslogd within each vserver? 1138593061 M * Bertl yes, just gets an empty log :) 1138593818 Q * emp Ping timeout: 480 seconds 1138593903 J * emp ~emp@70.57.239.35 1138594839 Q * gerrit Ping timeout: 480 seconds 1138595471 J * menomc ~amery@200.75.27.81 1138595579 Q * mnemoc Ping timeout: 480 seconds 1138595579 N * menomc mnemoc 1138595750 J * gerrit ~gerrit@216.65.245.8 1138596297 M * mugwump hey, look what I made with stgit 1138596387 M * mugwump wilber:~/vserver/linux-2.6$ stg export ../split-2.6.14.3-vs2.01-stgit 1138596410 M * mugwump output => http://utsl.gen.nz/vserver/split-2.6.14.3-vs2.01-stgit 1138596441 M * mugwump git repo for that is at http://utsl.gen.nz/vserver/vserver.git/ 1138596451 M * mugwump try with stgit: stgit clone (the URL) 1138596465 M * mugwump with reference to http://wiki.procode.org/cgi-bin/wiki.cgi/StGIT_Tutorial 1138596535 M * Bertl hmm ... 1138596536 A * mugwump goes home for now, having at least achieved *something*, however minor 1138596541 M * Bertl and what's the prpose? 1138596555 M * mugwump it manages layered patches 1138596581 M * Bertl okay, guess you have to explain the difference to me tomorrow or so 1138596590 M * Bertl (the difference to a dir with those patches) 1138596593 M * mugwump stgit was recommended to me by Linus... so, I have pulled in your patches to it to track my efforts to break them up the way Linus wants them 1138596607 M * mugwump I typed one command to generate them 1138596619 M * mugwump and can easily switch between them etc 1138596633 M * Bertl like with quilt? 1138596656 M * mugwump ah, you already use quilt, ok. that was also recommended with a minor deprecation notice :) 1138596666 M * mugwump does quilt use git? 1138596674 M * Bertl well, no, but I know quilt ... 1138596683 M * Bertl akpm used to use it 1138596695 M * Bertl (maybe still uses it :) 1138596727 M * Bertl okay, we are both tired I guess, and we should really delay that ... 1138596769 M * mugwump ok, well, I will be publishing my progress at those places ... 1138596779 M * mugwump feel free to get involved or not as suits 1138596787 M * Bertl okay, great! 1138596790 A * mugwump & 1138596804 Q * gerrit Quit: Client exiting 1138599254 M * Bertl okay, off to bed now .. have a good whatever everyone! 1138599254 Q * bwana Read error: Connection reset by peer 1138599259 J * bwana ~bwana@71-82-125-40.dhcp.roch.mn.charter.com 1138599261 N * Bertl Bertl_zZ 1138602793 J * balbir ~balbir@59.145.136.1 1138603870 J * Smutje ~Smutje@xdsl-87-78-0-11.netcologne.de 1138603979 Q * Smutje_ Ping timeout: 480 seconds 1138607004 Q * bwana Read error: Connection reset by peer 1138607039 J * bwana ~bwana@71-82-125-40.dhcp.roch.mn.charter.com 1138609561 Q * jkl Quit: shutdown 1138611459 Q * _are_ Ping timeout: 480 seconds 1138611878 Q * Aiken Quit: Leaving 1138612613 J * prae ~prae@ezoffice.mandriva.com 1138612785 J * _are_ ~are@62.112.159.81 1138612849 J * meandtheshell ~markus@85-124-37-206.dynamic.xdsl-line.inode.at 1138614237 J * bonbons ~bonbons@83.222.38.150 1138614705 J * s4edi ~siggi@inferno.lhi.is 1138614915 J * shedi ~siggi@inferno.lhi.is 1138615119 Q * shed| Ping timeout: 480 seconds 1138615269 Q * s4edi Ping timeout: 480 seconds 1138615512 J * Rescueraider ~email@CPE-149-167-120-10.qld.bigpond.net.au 1138615515 P * Rescueraider 1138615792 M * Hollow bonbons: around? 1138615808 M * bonbons Hollow: Yup 1138615836 M * Hollow just wanted to let you know that i have commited the source reorganization, just in case your patching around.. 1138615866 M * bonbons Ok, will sync before continuing patching 1138615871 M * Hollow and we depend on dietlibc now, no more glibc support 1138615897 M * Hollow due to the local printf implementation 1138615900 M * Hollow finally 1138615920 M * Hollow fyi: http://phpfi.com/99161 1138615971 M * bonbons oh :) that are real size differences! 1138615994 M * Hollow yep, well.. the local printf function do not implement all conversions, but the most common are implemented 1138616025 M * bonbons was just going to ask what it implements 1138616046 M * Hollow look in src/libinternal/printf.c in the comment above vsnprintf 1138617035 M * bonbons Hollow: doesn't compile fine for me... FMT_LEN undeclared in libinternal/printf.c line 510 (an a few warnings about implicit declarations) 1138617067 M * Hollow you need libowfat 1138617077 M * Hollow reminds to put a check for it in configure.ac 1138617123 M * bonbons where do I get that lib from? 1138617137 M * Hollow http://www.fefe.de/libowfat/ 1138617175 M * Hollow it's a gpl-ported version of djbs c libraries by the author of dietlibc 1138617272 M * bonbons is there an ebuild for it in portage, or do you have one for overlay? 1138617315 M * Hollow it's in portage.. 1138617319 M * Hollow emerge libowfat 1138617373 M * bonbons ups, did not find it because I typed lobowfat (my bad) 1138618246 M * SiD3WiNDR lib "ow, fat!" 1138620650 Q * Hunger xenon.oftc.net iridium.oftc.net 1138620650 Q * monrad_ xenon.oftc.net iridium.oftc.net 1138620650 Q * meebey xenon.oftc.net iridium.oftc.net 1138620650 Q * click xenon.oftc.net iridium.oftc.net 1138620650 Q * bragon xenon.oftc.net iridium.oftc.net 1138620650 Q * Psy0rz_ xenon.oftc.net iridium.oftc.net 1138620650 Q * kilian xenon.oftc.net iridium.oftc.net 1138620650 Q * aba xenon.oftc.net iridium.oftc.net 1138620650 Q * zobel xenon.oftc.net iridium.oftc.net 1138620650 Q * FireEgl xenon.oftc.net iridium.oftc.net 1138620650 Q * shedi xenon.oftc.net iridium.oftc.net 1138620650 Q * bonbons xenon.oftc.net iridium.oftc.net 1138620650 Q * lonewolff xenon.oftc.net iridium.oftc.net 1138620654 Q * derjohn xenon.oftc.net iridium.oftc.net 1138620654 Q * Loki|muh xenon.oftc.net iridium.oftc.net 1138620654 Q * ComplexMind xenon.oftc.net iridium.oftc.net 1138620654 Q * BartVB xenon.oftc.net iridium.oftc.net 1138620654 Q * Vudumen xenon.oftc.net iridium.oftc.net 1138620654 Q * wibble_ xenon.oftc.net iridium.oftc.net 1138620654 Q * Cru xenon.oftc.net iridium.oftc.net 1138620654 Q * entroposcope xenon.oftc.net iridium.oftc.net 1138620654 Q * Bertl_zZ xenon.oftc.net iridium.oftc.net 1138620654 Q * phreak`` xenon.oftc.net iridium.oftc.net 1138620654 Q * meandtheshell xenon.oftc.net iridium.oftc.net 1138620654 Q * SNy xenon.oftc.net iridium.oftc.net 1138620654 Q * michal_ xenon.oftc.net iridium.oftc.net 1138620655 Q * cohan xenon.oftc.net iridium.oftc.net 1138620655 Q * AllenJB xenon.oftc.net iridium.oftc.net 1138620655 Q * mire xenon.oftc.net iridium.oftc.net 1138620655 Q * jgommers xenon.oftc.net iridium.oftc.net 1138620655 Q * SiD3WiNDR xenon.oftc.net iridium.oftc.net 1138620655 Q * samuel_ xenon.oftc.net iridium.oftc.net 1138620655 Q * dhansen xenon.oftc.net iridium.oftc.net 1138620655 Q * micah xenon.oftc.net iridium.oftc.net 1138620655 Q * neofutur_ xenon.oftc.net iridium.oftc.net 1138620655 Q * andrew_ xenon.oftc.net iridium.oftc.net 1138620655 Q * prae xenon.oftc.net iridium.oftc.net 1138620655 Q * balbir xenon.oftc.net iridium.oftc.net 1138620655 Q * daniel_hozac xenon.oftc.net iridium.oftc.net 1138620655 Q * lilo xenon.oftc.net iridium.oftc.net 1138620655 Q * ag- xenon.oftc.net iridium.oftc.net 1138620655 Q * brc_ xenon.oftc.net iridium.oftc.net 1138620655 Q * ntrs__ xenon.oftc.net iridium.oftc.net 1138620655 Q * sladen xenon.oftc.net iridium.oftc.net 1138620655 Q * blizz xenon.oftc.net iridium.oftc.net 1138620655 Q * cryo xenon.oftc.net iridium.oftc.net 1138620655 Q * Roey xenon.oftc.net iridium.oftc.net 1138620655 Q * mountie xenon.oftc.net iridium.oftc.net 1138620655 Q * yang xenon.oftc.net iridium.oftc.net 1138620655 Q * romke xenon.oftc.net iridium.oftc.net 1138620655 Q * mcp xenon.oftc.net iridium.oftc.net 1138620655 Q * waldi xenon.oftc.net iridium.oftc.net 1138620655 Q * wasser xenon.oftc.net iridium.oftc.net 1138620655 Q * Adrinael xenon.oftc.net iridium.oftc.net 1138620655 Q * _Medivh xenon.oftc.net iridium.oftc.net 1138620655 Q * gdm xenon.oftc.net iridium.oftc.net 1138620655 Q * blackfire xenon.oftc.net iridium.oftc.net 1138620655 Q * mugwump xenon.oftc.net iridium.oftc.net 1138620655 Q * LCamel xenon.oftc.net iridium.oftc.net 1138620655 Q * distortion xenon.oftc.net iridium.oftc.net 1138620791 J * andrew_ ~andrew@tnlug.linux.org.tw 1138620791 J * neofutur_ ~neofutur@neofutur.net 1138620791 J * micah ~micah@69.90.134.205 1138620791 J * dhansen ~dave@sprucegoose.sr71.net 1138620791 J * samuel_ ~samuel@levinux.UQAR.UQUEBEC.CA 1138620791 J * SiD3WiNDR luser@bastard-operator.from-hell.be 1138620791 J * phreak`` ~phreak``@styx.xnull.de 1138620791 J * Bertl herbert@212.16.62.52 1138620791 J * entroposcope ~entroposc@user-0c992og.cable.mindspring.com 1138620791 J * Cru ~mindwarp@turbodiesel.e.de.wahlich.com 1138620791 J * wibble_ wibble@vortex.ukshells.co.uk 1138620791 J * jgommers ~virtual@jgommers.vserver01.linit.nl 1138620791 J * Vudumen vudumen@perverz.hu 1138620791 J * FireEgl Atlantica@Atlantica.DollarDNS.Net 1138620791 J * BartVB ~BartVB@84.35.54.120 1138620791 J * kilian kk@projects.verfaction.de 1138620791 J * Psy0rz_ ~psy0rz@lounge.datux.nl 1138620791 J * ComplexMind ~ComplexHo@cpc1-brig3-6-0-cust194.brig.cable.ntl.com 1138620793 J * bragon ~bragon@sd866.sivit.org 1138620793 J * click click@ti511110a080-4706.bb.online.no 1138620793 J * Loki|muh loki@satanix.de 1138620793 J * meebey meebey@booster.qnetp.net 1138620793 J * Hunger Hunger.hu@Hunger.hu 1138620793 J * mire ~mire@150-167-222-85.COOL.ADSL.VLine.verat.net 1138620793 J * AllenJB ~Allen@stuEAED.kent.ac.uk 1138620793 J * zobel zobel@zobel.irc.ftbfs.de 1138620793 J * aba ~aba@eos.turmzimmer.net 1138620793 J * derjohn ~derjohn@80.69.37.19 1138620793 J * cohan ~cohan@koniczek.de 1138620793 J * lonewolff ~lonewolff@host-84-9-143-159.bulldogdsl.com 1138620793 J * monrad_ ~mikkel@213083190131.sonofon.dk 1138620793 J * michal_ ~michal@www.rsbac.org 1138620793 J * SNy 696f7054ec@bmx-chemnitz.de 1138620793 J * meandtheshell ~markus@85-124-37-206.dynamic.xdsl-line.inode.at 1138620793 J * bonbons ~bonbons@83.222.38.150 1138620793 J * shedi ~siggi@inferno.lhi.is 1138620793 J * prae ~prae@ezoffice.mandriva.com 1138620793 J * balbir ~balbir@59.145.136.1 1138620793 J * daniel_hozac ~daniel@c-2d1472d5.010-230-73746f22.cust.bredbandsbolaget.se 1138620793 J * lilo ~lilo@lilo.usercloak.oftc.net 1138620793 J * ag- ag@caladan.roxor.cx 1138620793 J * brc_ bruce@20151222045.user.veloxzone.com.br 1138620793 J * ntrs__ ~ntrs@68-188-50-87.dhcp.stls.mo.charter.com 1138620793 J * sladen paul@starsky.19inch.net 1138620793 J * blizz ~blizz@evilhackerdu.de 1138620793 J * cryo ~say@212.86.233.146 1138620793 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1138620793 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com 1138620793 J * yang ~boni@cpe-213-157-253-172.dynamic.amis.net 1138620793 J * waldi ~waldi@bblank.thinkmo.de 1138620793 J * romke ~romke@acrux.romke.net 1138620793 J * mcp ~hightower@wolk-project.de 1138620793 J * wasser ~wasser@ip86.ipax.at 1138620793 J * Adrinael adrinael@hoasb-ff09dd00-79.dhcp.inet.fi 1138620793 J * _Medivh ck@paradise.by.the.dashboardlight.de 1138620793 J * gdm ~gdm@209.51.169.84 1138620793 J * blackfire blackfire@dp70.internetdsl.tpnet.pl 1138620793 J * mugwump ~samv@watts.utsl.gen.nz 1138620793 J * LCamel luors@chengjeng.iis.sinica.edu.tw 1138620793 J * distortion distortion@junipero.3sheep.com 1138620827 M * Bertl bonbons: --rbind? 1138620829 M * bonbons eg: /proc, /vserver/myvs/proc, /home/projects/vserver-utils/src/proc; the last one is even not possible as that dir does not exist! 1138620893 M * bonbons hmm could really be there's something wrong with rbind 1138620905 M * Bertl why should be something wrong? 1138620944 M * bonbons that's the last call, and reading my debug info the rbind mount has the "incorrect" root 1138621187 M * Bertl hmm? 1138621257 M * bonbons the rbind mount for the guest is done from the "broken" location, checking why 1138622021 M * bonbons Hollow: what is the need for the rbind mount on the guest's root? 1138622257 M * Hollow *shrug* i adapted it from util-vserver, but i guess it is meant to prevent cd ../.. exploits 1138622267 M * Hollow but that's for the barrier anyway 1138622382 M * bonbons hmm, then I will try without (the rbind currently happens to an unlogical target => cwd from where vserver is called!) 1138622422 M * Hollow hm, at this point cwd should be vdir 1138622451 M * bonbons that what I thought, but somehow it's not 1138622459 M * Hollow than it's a bug :D 1138622854 M * bonbons Hollow: from inside the guest looks the same with and without 'vps.mount root' ... 1138622948 M * Hollow hm, yeah.. we can probably skip it completely 1138623131 M * Hollow Bertl: would be nice if we can clear things with regard to breaking out the chroot, what is actually needed to preven it? 1138623278 J * eyck eyck@81.219.64.71 1138623401 M * Bertl Hollow: well yeah, suggestions? 1138623433 M * Bertl I mean, there is a 'root-plug' functionality in mainline (which isn't known/used) 1138623446 M * Bertl There is --rbind which should be sufficient 1138623449 M * Hollow e.g. vx_cleanup_namespace.. does it do anything useful to us? 1138623453 M * Bertl and there is our barrier ... 1138623468 J * dothebart ~willi@xdsl-213-196-247-161.netcologne.de 1138623488 M * Bertl well, the cleanup will remove mounts which are not pinned IIRC 1138623499 M * dothebart re. 1138623516 M * Hollow so, theoretically we could get rid of all mounts of the cloned namespace which are not used by the guest? 1138623521 M * dothebart how do i best let two vservers communicate in private? 1138623529 M * Hollow dothebart: ipv4 1138623541 M * dothebart over loopback? 1138623543 M * Bertl Hollow: yes, but I'm not sure you really want that ... 1138623553 M * Hollow Bertl: why wouldn't i? 1138623562 M * Hollow dothebart: no, over their configured ip adresses 1138623578 M * bonbons Bertl: what would be reasons not to want it? Possibility to add mounts lateron? 1138623584 M * Bertl Hollow: thing is, if you remove the 'host' stuff, you will lose all ability to mount something from there afterwards 1138623590 M * dothebart can't i give both of them 127.0.0.5 for example? 1138623603 M * Hollow well, do we want that? 1138623604 M * dothebart or one .5, one .6? 1138623611 M * Hollow i can still mount from inside, right? 1138623615 M * Hollow with appropriate caps 1138623616 M * bonbons Bertl: this would be solved once we can (bind)mount accross namespaces 1138623620 M * Bertl dothebart: sure you can, but what would be the advantage? 1138623636 M * Bertl bonbons: that is correct ... 1138623640 M * dothebart that i wouldn't need a firewall to block access from outside in ;) 1138623650 M * Hollow take 192.168. 1138623668 M * dothebart because of i want i.e. communicate between mysql and apache for example... 1138623670 M * Bertl dothebart: why? does your machine protect against 127.x.x.x by default? 1138623699 M * dothebart hm, you're right. i would need to add that too. 1138623743 M * Hollow imo, it should be sufficient to set the barrier and chroot, no need for rbind 1138623756 M * Hollow bonbons: did you try the exploit without the rbind? 1138623791 M * dothebart well, but i'd like to share the mysql between several vservers... 1138623816 M * Bertl do that 1138623834 M * bonbons Hollow: don't know the possible exploits... 1138623905 M * dothebart each vserver should than have its own 192.168.x.y1 / y2 address? 1138623925 M * Hollow http://vserver.13thfloor.at/Stuff/rootesc.c 1138623983 M * dothebart hm. the crappy onboard nic disapeared from lspci. 1138624048 M * Bertl Hollow: there is another one too 1138624063 M * Bertl Hollow: and the chroot escape stuff could use a rewrite ... 1138624449 M * Hollow hm, we could include a tool in vserver-utils which can be used from inside to check common flaws 1138624490 M * Hollow Bertl: btw.. i switched vserver-utils completely to dietlibc now, no more glibc support.. 1138624508 M * Hollow it's annoying to fiddle with glibc/dietlibc libs/bins and linkage 1138624520 M * bonbons that would be a useful test/debug-tool 1138624544 M * Hollow yeah, the thing is.. you have to know the flaws ;) 1138624600 M * Bertl Hollow: ad diet, that's fine for me 1138624618 M * Bertl Hollow: we know diet is buggy in some places, but at least it's a constant :) 1138624643 M * Hollow heh 1138624661 M * bonbons without the rbind that exploit works (note, I have no barrier setup) -- rbind made to rbind to / (not cwd which should have been '/') 1138624673 M * Hollow without barrier it works always i guess 1138624699 M * bonbons did not work with the correct --rbind to / 1138624722 M * Hollow no barrier, correct rbind = no breakout? 1138624743 M * bonbons exact 1138624744 M * Bertl Hollow: I would assume so 1138624773 M * Hollow so, we either use rbind or barrier, or both just to be sure 1138624819 M * Bertl the latter one is preferred for now :) 1138624827 M * Hollow barrier? why? 1138624831 M * Bertl motto: better safe than sorry :) 1138624857 M * bonbons with incorrect rbind and no barrier escape works 1138624928 M * Hollow well, since barrier is an extension, and rbind is vanilla, we can probably get rid of the barrier, if rbind is secure 1138624944 M * bonbons one thing left for checking, what vnamespace can still do after correct rbind (if it sees "old" or "new" root) 1138624991 M * bonbons if it sees "new" root, we are at same point as removing the host's namespace copy 1138624999 M * Hollow you should guess ENOSYS or EPERM 1138625009 M * Hollow s/guess/get/ 1138625028 M * Hollow most vserver syscalls do not exist in the guest (enosys) 1138625099 M * Hollow have to leav for school.. back in about one hour 1138625144 M * bonbons bye Hollow, see you later 1138625171 J * id23 ~id@p54A04868.dip0.t-ipconnect.de 1138625179 M * id23 hi #vserver 1138625252 M * SiD3WiNDR hi id23 1138625282 M * bonbons vnamespace enters the namespace, and still sees the host-equivalent root, rbind correct and escape exploit fails 1138625363 M * Bertl welcome id23! 1138625882 J * LiNuX ~LiNuX@85.104.9.218 1138625890 M * Bertl welcome LiNuX! 1138625895 M * LiNuX Slemas. 1138625902 N * LiNuX LiNuX[Yok] 1138625907 M * LiNuX[Yok] Thanx. 1138625929 A * LiNuX[Yok] I will come back! 1138625952 M * Bertl well, do so! :) 1138626041 M * LiNuX[Yok] bye 1138626049 Q * LiNuX[Yok] Quit: 1138626195 M * romke matti: you're welcome ;) 1138626631 M * Bertl okay, off for a while, back later ... 1138626644 N * Bertl Bertl_oO 1138626829 M * bonbons Off too, back in an hour 1138628080 Q * id23 Ping timeout: 480 seconds 1138628385 J * DoM_ ~DoM@151.56.232.170 1138628779 Q * DoM__ Ping timeout: 480 seconds 1138629078 M * SiD3WiNDR heh 1138629081 M * SiD3WiNDR funny people 1138629087 M * SiD3WiNDR join to say they'll be back, then quit 1138629271 J * Doener doener@i5387DDDC.versanet.de 1138629729 Q * balbir Quit: Leaving 1138630313 J * Viper0482 ~Viper0482@p54976CDA.dip.t-dialin.net 1138630915 Q * BartVB Ping timeout: 480 seconds 1138630963 M * Hollow bonbons: what do you mean by "enters the namespace, and still sees the host-equivalent root" 1138630985 M * Hollow i mean.. if you just enter the namespace you donÄt have chrooted actually 1138631052 M * bonbons the mount -rbind /path/to/guest / could have hidden original / (which it does not, or vnamespace remembers the hosts /) 1138631086 M * Hollow hm, try to do a secure chdir before and/or after namespace_enter 1138631091 M * Hollow like in vmount.c 1138631095 M * Doener bonbons: the namespace is saved at a point where you are still below the original root 1138631172 M * Doener when you enter the namespace, you use a saved copy of it, where your cwd is in the old tree 1138631216 M * bonbons the test was just to check wether bind-mounts from host to guest were still possible with corrected rbind 1138631366 J * ScoobyD00 ~foo@80-195-186-201.cable.ubr08.newm.blueyonder.co.uk 1138631391 M * ScoobyD00 hi 1138631410 M * ScoobyD00 i am still having problems with setting the xid 1138631427 M * ScoobyD00 ;o( 1138631477 M * Hollow -v 1138631505 M * ScoobyD00 i upgraded my tools so now they are on .209 1138631513 M * ScoobyD00 (0.30.209) 1138631519 M * ScoobyD00 i moved to ext3 1138631535 M * Hollow from? 1138631586 M * ScoobyD00 well yesterday i was using debians "stable" 0.30.204 1138631602 M * Hollow i mean which filesystem you had before 1138631607 M * ScoobyD00 oh, reiserfs 1138631616 M * Hollow ok, and which problems do you encounter now? 1138631630 M * ScoobyD00 same, everything says like this: /vservers/test1/usr/share/locale/zh_TW: Invalid argument 1138631637 M * ScoobyD00 "Invalid argument" 1138631659 M * ScoobyD00 i have mounted with tagxid 1138631663 M * ScoobyD00 dev/md3 on /var/lib/vservers type ext3 (rw,noatime,tagxid) 1138631681 M * Hollow hm, which vserver version? 1138631771 M * ScoobyD00 debians: kernel-patch-vserver 1.9.5.3 1138631781 M * ScoobyD00 on kernel 2.6.8 1138631839 M * Hollow oi 1138631866 M * Hollow this is so old i can't even remember *g* 1138631872 M * ScoobyD00 :o) 1138631885 M * ScoobyD00 bertl said that just the new tools should solve this problem for ext3 1138631940 M * Hollow hm, not sure.. guess you have to wait for him then.. i have no 1.9.5 available 1138631973 M * Hollow which command line did you use? 1138631986 M * ScoobyD00 chxid -c test1 -R /vservers/test1/ 1138632066 M * Hollow hm, is /vservers a symlink? 1138632076 M * ScoobyD00 yeah to /var/lib/vservers 1138632083 M * Hollow *shrug* 1138633315 Q * FireEgl Ping timeout: 480 seconds 1138633903 M * Hollow bonbons: what do you think about implementing the vserver command in C too? 1138633922 M * bonbons I like the idea 1138633949 M * bonbons especially with libvconfig I makes really sense 1138633950 M * Hollow yeah, i'm really unhappy with the bash side of things atm 1138634012 M * bonbons maybe we should re-reorganize the code then, build the whole thing in a more pugin-fashion (like portage-utils) 1138634038 M * Hollow hm, my idea is to create another lib(s) and put most code the tools in there, and use this lib for the vserver command but also adapt the tools so users can still ise them in bash scripts 1138634042 J * FireEgl ~FireEgl@Atlantica.Tcldrop.Com 1138634064 A * Hollow goes looking at portage-utils 1138634193 M * Hollow yeah, looks nice 1138634314 M * bonbons I would possibly see two binaries, one for operating on vservers, one for handling configuration operations (like config-editor/viewer) 1138634381 M * Hollow probably a third one, maybe a daemon, for state handling and statistics (over netlink or sysfs or whatever it will be) 1138634394 M * bonbons the config editor/viewer would not need root rights (except eventually to commit changes) => ${prefix}/bin, while the vserver command would require root privileges ${prefix}/sbin 1138634395 M * Hollow should also replace vshelper 1138634432 M * bonbons yep, the daemon is a good idea 1138634546 M * Hollow we once had the idea of a simple command interpreter.. e.g. you create a script: 1138634552 M * Hollow #!/bin/vshell 1138634562 M * Hollow start 1138634566 M * Hollow set option foo 1138634581 M * Hollow etc.. 1138634632 M * Hollow that could be combined with the rescue shell idea 1138634662 M * Hollow but i guess that's something for 1.5 or so *g* 1138634707 M * bonbons yep, that's for once the individual commands exist (just a matter of replacing main() by something better ;)) 1138634723 M * Hollow yup, guess that would be pretty straight then.. 1138634809 M * Hollow ok, so any suggestion for a good source organization? 1138634835 M * bonbons I don't think switching to shell-like interface would be complicated 1138634927 M * bonbons hmm... I would say we need a library part (config, ...), an applet part (with all subcommands) and then a tools part (which has code to reach the applets) 1138634980 J * liquid3649_ ~Viper0482@p549769E6.dip.t-dialin.net 1138634988 M * Hollow ok, so we create reusable (for tools, shell, etc) applets 1138635017 M * bonbons organization of the applets is another question, all in one dir, or sort by abstraction level (e.g. start/stop versus context/ncontext/namespace/...) 1138635041 M * bonbons sure, each applet is kind of a inline-tool 1138635155 M * Hollow i think we should just stick to a higher abstraction, if you want context/ncontext commands you can use libvserver directly 1138635224 M * Hollow or one of the libraries used by the applets 1138635240 M * bonbons that's true as well 1138635302 M * Hollow so, guess we can place them all in one dir 1138635378 M * bonbons I would think we should have the low-level commands available as "expert" applets, but not visible by default 1138635413 M * bonbons this is especially useful once we add the shell feature 1138635414 Q * Viper0482 Ping timeout: 480 seconds 1138635462 M * Hollow we could distinguish between commands and applets 1138635490 M * Hollow and probably let applets use commands, but my wrapper alarm light seems to flash 1138635621 M * bonbons hard to say beforehand how much high-level commands can call lower-level ones 1138635657 M * Hollow guess start/stop need nearly all, the rest nearly none 1138635663 M * Hollow ;) 1138635711 M * bonbons if it's more or less just a single call to libvserver function, then calling a command for that is overkill, if there's more homework, it might be useful 1138635814 M * bonbons start/stop should mainly be a hardcoded (for now) script 1138635817 M * Hollow hm, what about putting repetitive high-level functions to the internal (or whatever) lib, and just them together with libvserver functions in the appelet 1138635866 M * bonbons which ones do you think about? 1138635966 M * Hollow hm, *shrug* and because of that it seems like overkill.. i think we should just stick to applets of a higher abstraction 1138635988 M * Hollow *reminder* do not overcomplicate things 1138636130 M * bonbons that's sure, keep it simple. 1138636203 M * bonbons we should just keep the lowlevel ones available in background (e.g. for debugging) 1138636251 M * Hollow apropos debugging... we need a elegant solution here too i'm afraid 1138636278 M * daniel_hozac hmm, why go for a shell if you expose the API to various higher-level languages? 1138636292 J * Lars ~adminuser@you.are.the.lamer.la 1138636306 M * Lars Got a problem a vserver has stoned with a proftpd proccess i tryd kill -9 the pid but it wount stop 1138636308 M * Lars what do i do 1138636331 M * daniel_hozac Lars: i guess it's in state D? 1138636346 M * bonbons some printf like DEBUG(...) functions which would be fine for this (theses would be generally #defined to printf or void) 1138636348 M * Lars state D? 1138636359 M * daniel_hozac yes, i.e. in ps output. 1138636365 M * Lars oh yes. 1138636390 M * daniel_hozac that means it's uninterruptible, i.e. there's nothing you can do. 1138636400 M * Lars 11336 49171 hansa ? Ds 0:00 proftpd: (accepting connections) 1138636400 M * Lars ---------------------------------------------------------------------- 1138636400 M * Lars Vserver '/etc/vservers/hansa' still running unexpectedly; please investigate it manually... 1138636411 M * Lars tryd kill -9 11336 1138636430 M * daniel_hozac (sidenote: you should stop using dynamic contexts ;)) 1138636437 M * daniel_hozac yes, kill -9 won't do anything. 1138636455 M * Hollow bonbons: yup.. we'll talk later, off for a while 1138636469 M * Lars daniel_hozac: so i cant do anything? 1138636478 M * daniel_hozac Lars: not really, no. 1138636479 M * bonbons Hollow: no prob, let's ideas crystalize till then 1138636505 M * Lars daniel_hozac: great 1138637568 M * Lars hmm 1138637571 M * Lars reboot then 1138638474 J * lares ~lares@S0106000f664610bc.cg.shawcable.net 1138638479 M * lares Heya all :) 1138639032 Q * liquid3649_ Remote host closed the connection 1138639555 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il 1138639701 Q * ScoobyD00 Quit: 1138640325 M * Hollow bonbons: back now.. so, what does your crystal ball tell us? 1138640327 M * Hollow :) 1138640375 M * bonbons it got an interrupt ;) will be back in a few minutes 1138641065 M * bonbons back, request satisfied 1138641150 M * bonbons Hollow: I would suggest setting up a list of high-level commands the we want to have (be it current feature or whish) 1138641252 M * Hollow ok 1138641400 M * bonbons I think of: start, stop, restart/reboot, kill, enter, status for guestspecific operations 1138641447 M * Hollow config get/set 1138641484 M * bonbons in addition I have statistics, ps+kill, unify for general handling 1138641511 J * dearaujo ~dan@pixpat.austin.ibm.com 1138641512 M * bonbons configuration (get/set) and guest creation should probably be a separate tool 1138641524 A * Hollow nods 1138641871 Q * prae Quit: Execute Order 69 ! 1138642245 M * bonbons I'm writing a page on the wiki with list of commands and what they should do 1138642418 J * Viper0482 ~Viper0482@p549769E6.dip.t-dialin.net 1138642513 M * derjohn bonbons, consider the feature "management user" or "admin user" per guest (i.e. user X on host may stop/start it own guest) 1138642624 M * bonbons derjohn, so you mean extend permission on some commands to non-root users? 1138642689 M * derjohn bonbons, yes, for /me as hoster it would be nice to give vserver customers self-service rights ;) 1138642714 M * derjohn bonbons, at the time there would be only sudo with is a brute as ... judo or so ;) 1138642809 M * bonbons derjohn, jup, that could done by sudo or equivalent, or even a suid-wrapper. The problem is that handling the guests requires root privileges, so vserver-utils would need to be suid 1138642872 M * bonbons a specialized such wrapper would be better I think, even if such group-info could be added to configuration 1138642919 Q * Lars Quit: 1138642952 M * bonbons Hollow: look here for my applet-draft: http://dev.croup.de/proj/vserver-utils/wiki/Applets 1138643008 M * derjohn bonbons, are a .de based? Wound you like join the linux-vserver community @ linuxtag 2006 ? 1138643069 M * bonbons no, I'm .lu based 1138643089 M * derjohn bonbons, ok ;( i just saw "croup.de" 1138643140 M * bonbons that's where you can find vserver-utils' homepage 1138643147 N * Bertl_oO Bertl 1138643153 M * Bertl back now ... 1138643218 M * Hollow bonbons: very nice :) 1138643242 M * Hollow will be fully available later again 1138643278 M * bonbons "fully available later", who/what? you? 1138643357 M * Bertl guess so :) 1138644208 Q * Viper0482 Quit: bin raus, 1138644475 M * Hollow yeah, me.. *g* 1138644484 M * Hollow i'm just planning a promo mix 1138644498 M * Hollow i.e. music 1138644807 Q * sladen Read error: Connection reset by peer 1138644810 J * sladen_ paul@starsky.19inch.net 1138645366 J * stefani ~stefani@superquan.apl.washington.edu 1138645490 M * Bertl welcome stefani! 1138646025 J * cehteh foobar@cehteh.homeunix.org 1138646036 M * cehteh re 1138646045 M * cehteh (back form vacation) 1138646049 M * Bertl welcome cehteh! 1138646298 J * ScoobyD00 ~foo@80-195-186-201.cable.ubr08.newm.blueyonder.co.uk 1138646321 M * ScoobyD00 hi 1138646321 M * Bertl welcome ScoobyD00! 1138646332 M * ScoobyD00 hey bertl 1138646349 M * ScoobyD00 well, i managed to upgrade my util-vserver package to .209 1138646370 M * daniel_hozac you know .210 is current, right? 1138646372 M * ScoobyD00 and it didnt help the xid problem on either ext3 nor reiserfs :o( 1138646390 M * Bertl yeah, your kernel version was? 1138646391 M * ScoobyD00 yeah - just 209 was in debian testing 1138646443 M * ScoobyD00 vserver patch 1.9.5.3 on 2.6.8 1138646463 M * Bertl okay, so I'd say you should really try with a newer kernel 1138646481 M * ScoobyD00 anyways, i figured as i would have to repartition anyways, i may aswell use LVM, which in my test VMWARE image works really well 1138646493 M * ScoobyD00 so, file limits sorted :o) 1138646511 M * Bertl okay, you can do that too, just be aware that there might be other issues 1138646523 M * ScoobyD00 what kind? 1138646531 M * Bertl did you check with the testme and testfs scripts yet? 1138646547 M * ScoobyD00 forgot the urls - can you paste and i will give them a whirl 1138646575 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/ 1138646591 M * Bertl the testfs.sh requires a loopback (or lvm) partition to work on 1138646606 M * cehteh Bertl: what are the reasons that (k)nfs is broken ibn vserver? 1138646643 M * Bertl AFAIK knfs is not broken for vserver guests 1138646685 M * cehteh ah ok i only tried from root 1138646711 M * Bertl hmm? 1138646765 Q * cryo Ping timeout: 480 seconds 1138646815 M * cehteh just tried it again .. and it didnt worked which was my experience long time ago .. search the ML and there is no note that it is fixed, only complaints that it doesnt work .. and didnt investigated the cause any further .. 1138646840 M * Bertl no idea what you are trying .. maybe you can elaborate? 1138646884 M * cehteh actually not .. but i come back on that when i try again, i tried before my vacation some weeks ago 1138646901 M * cehteh just remembered that i wanted to ask you if it is something trivial 1138646918 M * ScoobyD00 http://pastebin.com/530774 1138646923 M * ScoobyD00 looks good? 1138646957 M * Bertl yep, looks good 1138646968 M * ScoobyD00 :o) 1138647042 M * ScoobyD00 one issue i am having right now tho... 1138647054 M * ScoobyD00 is with stopping my test vserver 1138647086 M * Bertl yes? 1138647089 M * ScoobyD00 it takes ages, and then spits out a message beginning with "A timeout occured while waiting for the vserver to finish" 1138647115 M * ScoobyD00 and then i think it is supposed to list some offending processes, but the list appears to be empty 1138647116 M * Bertl well, your guest doesn't stop as expected 1138647136 M * Bertl could be caused by the very old kernel version 1138647147 M * Bertl but should not be a problem 1138647176 M * daniel_hozac i wonder, why would you choose to run an experimental version of the patch on an ancient kernel? 1138647194 Q * _are_ Ping timeout: 480 seconds 1138647218 M * blizz short OT question, anyboy knows how to do linebreaks in LaTeX? (openoffice formula editor here, i think it was the same) 1138647223 M * Bertl daniel_hozac: because provides it? 1138647226 M * cehteh \\ 1138647228 M * ScoobyD00 it used to stop ok before i moved to .209 1138647240 M * cehteh \\ was meant for blizz 1138647248 M * ScoobyD00 and there is nothing in the list it provides 1138647260 M * ScoobyD00 i wondered if there was a bug in that timeout loop 1138647262 M * ScoobyD00 in .209 1138647276 J * undefined ~undefined@adsl-68-93-109-94.dsl.rcsntx.swbell.net 1138647297 M * blizz cehteh, thanks :) 1138647298 M * Bertl ScoobyD00: well, let me put it this way, 209 and 210 stop many guests perfectly fine here ... 1138647311 M * Bertl welcome undefined! 1138647319 M * undefined howdy Bertl 1138647324 M * blizz seems like it doesnt work in OO :/ 1138647337 M * cehteh that was tex 1138647338 M * Bertl ScoobyD00: so if that is so, it is a very special loop, but your system seems to be very special anyway :) 1138647358 M * ScoobyD00 :o) 1138647361 M * blizz cehteh, yeah, but sometimes its similar 1138647396 M * Bertl ScoobyD00: you might set a smaller timeout, or look through the shutdown scripts 1138647415 M * undefined is "sth" an abbrevation for "something"? (seen it used a few times here and today in a post to the mailing list) 1138647426 M * ScoobyD00 yeah, it doesnt seem to do any harm, just a delay 1138647455 M * Bertl undefined: depends, could be ... 1138647490 M * ScoobyD00 do you know where the script that has that loop in it is? 1138647509 M * Bertl the vwait tool should do that 1138647523 M * Bertl it's a C program IIRC 1138647541 M * ScoobyD00 hmm yeah 1138647552 M * ScoobyD00 sure i hit CTRL-C earlier and it dumped the loop to the screen 1138647609 M * Bertl http://linux-vserver.org/ChangeLog26 1138647643 M * Bertl # wakeup_vx_info added (vx_ctxwait) 1138647667 M * blizz cehteh, its "newline" *lol* 1138647723 M * Hollow Bertl: any news wrt state handling/accounting (netlink, sysfs, whatever) 1138647751 M * Bertl Hollow: parse error? 1138647767 M * Hollow *g* 1138647786 M * Hollow the communication layer we talked about 1138647798 M * Bertl aah, no, not yet 1138647837 M * Hollow ok, did you follow the discussion about vserver-utils above? 1138647864 M * Bertl only partially 1138647891 M * Hollow to sum up.. we'll implement the vserver command in C too 1138647907 M * Bertl aha, okay ... 1138647912 J * shed| ~siggi@inferno.lhi.is 1138647933 M * Bertl Hollow: with script helpers or completely in C? 1138647954 M * ScoobyD00 at the point whilst it is waiting in this loop, there is no processes showing in "vps aux" for the vserver 1138647976 M * Bertl ScoobyD00: did you read the changelog line above? 1138647989 M * bonbons Bertl: about completely in C 1138647997 J * _are_ ~are@62.112.159.81 1138648002 M * Hollow well, we will introduce "applets" which can be used by the vserver command (e.g. start, stop, etc) and a set of command line tools which can be used in scripts 1138648024 M * Bertl bonbons: hmm, okay, maybe something like the fsck or mount mechanisms would be nice 1138648044 M * ScoobyD00 bertl: couldnt see my version so didnt think it was for me 1138648051 J * phycho ~phych0@209.8.233.12 1138648054 M * Bertl bonbons: i.e. do some stuff by default, but use helper programs/scripts if available ... 1138648060 M * Bertl welcome phycho! 1138648079 M * Bertl ScoobyD00: simple, those were all after your release :) 1138648105 J * DoM__ ~DoM@151.56.238.63 1138648109 M * bonbons that depends on the operation, but for those that are distro-dependent certainly (like building guest filesystem) 1138648139 M * Bertl bonbons: I'd even call out for _unknown_ commands (if the command scheme is structured) to allow for extensions ... 1138648144 M * phycho hey there. 1138648164 M * phycho just reading up on this whole vserver thing =) 1138648167 M * phycho im used to using UML. 1138648176 M * ScoobyD00 ;o) 1138648188 M * phycho which, is garbage 1138648190 M * phycho to put it nicely 1138648191 M * phycho :-) 1138648195 M * bonbons Bertl: took note :) for the start will just be "unknown" but in future will look at extensions-location 1138648202 M * phycho going to try vserver instead 1138648236 M * Bertl phycho: well, it will be a little different, but defintiely faster and resource friendlier 1138648247 M * phycho hehe sounds good 1138648265 M * phycho reading howto atm 1138648293 M * Bertl well, first, forget all what you know about UML networking and filesystems :) 1138648309 Q * lares Read error: Connection reset by peer 1138648310 Q * shedi Ping timeout: 480 seconds 1138648345 M * phycho hehe 1138648351 M * Hollow bonbons: whatabout keeping the current tools (probably remove the v prefix) and put the to sth like /usr/lib/vserver-utils/commands, so script users can use them, but they shouldn't be in $PATH 1138648376 M * phycho damn.. need to have a dedicated partition for it according to this tutorial 1138648377 M * phycho =( 1138648387 M * phycho looks like ill need to move /usr. 1138648396 M * Bertl it's advised, but not 100% required 1138648412 M * Bertl it is very strongly advised if you use tagging though :) 1138648413 M * phycho well, if its worth doing, its worth doing right ;) 1138648433 M * phycho well, i need quota support at any rate. 1138648441 M * phycho dont want one user hogging the entire disk ;) 1138648447 M * Bertl quota or disk limits ? 1138648461 M * bonbons we can either keep them as is, or just make them 'expert' applets 1138648461 M * Bertl (quota is per user not per guest) 1138648487 M * phycho welll, i need per guest. 1138648487 M * Hollow hm 1138648503 M * Bertl phycho: which we call (per guest) disk limits 1138648514 Q * DoM_ Ping timeout: 480 seconds 1138648518 M * phycho bertl - do i need a seperate filesystem for that? 1138648530 M * Bertl yes 1138648541 M * phycho http://www.5dollarwhitebox.org/wiki/index.php/Howtos_Linux-Vserver_With_LVM_And_Quotas 1138648542 M * phycho im reading this 1138648543 M * phycho k. 1138648556 M * bonbons Hollow: being applets they can be used by scripts as well. This would keep down the count of binaries and make #!vshell things easier 1138648562 M * phycho split-2.6.14.3-vs2.01.tar 1138648565 M * phycho whats that for? 1138648570 M * Bertl well, that uses LVM IIRC, but you can utilize a _shared_ partition too 1138648570 A * Hollow nods 1138648571 M * phycho i know what patch-2.6.14.x is for 1138648575 M * phycho so whats split for? 1138648601 M * Bertl the split is mostly targeted at developers and interested folks, it's a broken out version of the patch 1138648604 M * phycho http://www.5dollarwhitebox.org/wiki/index.php/Howtos_Linux-Vserver_Debian_Sarge 1138648610 M * phycho im going to be following that 1138648613 M * phycho ah k, i dont need that then 1138648626 M * Bertl not really 1138648638 M * Hollow so, do you plan to make the applets own executables, each with mean, or library-like and linked into the whole vserver binary in the end? 1138648645 M * Bertl phycho: if you are going to use sarge, in general be careful! 1138648658 M * Bertl phycho: sarge has _many_ outdated and broken packages ... 1138648658 M * Hollow s/mean/main/ 1138648663 P * dearaujo 1138648664 M * phycho i know 1138648670 M * phycho debian is pretty junk 1138648677 M * phycho im going to be using sarge as the hostOS 1138648683 M * phycho but probably not as the guest. 1138648700 M * Bertl it works fine if you get the right kernel and tools 1138648710 M * Bertl it has one major drawback though 1138648723 M * phycho whats that? 1138648728 M * bonbons I though of portage-utils style, everything is compiled into a single binary, but each applet has it's applet_main and applet_* functions (this keeps forking to a minimum, same for global space usage) 1138648739 M * Bertl you cannot install rpm based guests easily with a debian host, as the rpm tools (rpm, yum, apt-rpm) do not work ... 1138648760 M * Bertl phycho: you can run them though ... 1138648761 M * Hollow and scripts use the applets via the vserver command..!? 1138648788 M * phycho bertl - what host os would you recommend? 1138648806 M * bonbons exactly, or "directly" if $0 == applet-name 1138648826 M * Bertl phycho: that is your preference, but any rpm based host distro will be able to install debian guests too 1138648832 M * Hollow ah yeah.. bash alias come to my mind 1138648837 M * phycho k 1138648902 M * Hollow so, we could create symlinks to the vserver command for each applet and script users just set the path with the symlinks in it, and voila 1138648914 M * bonbons exactly 1138648956 M * Hollow ok, sounds good :) 1138649017 M * phycho is there totally no way to run seperate kernels on linux vserver? 1138649028 M * phycho i know on UML you could have seperate kernel executables 1138649035 M * mnemoc phycho: xen 1138649039 M * phycho i know 1138649040 M * Bertl no, that's the idea behind linux-vserver, to avoid a kernel 1138649045 M * phycho but xen is no use to me 1138649049 M * Hollow so, how do we distinguish between expert and normal applets? 1138649050 M * phycho xen cant oversubscribe resources 1138649051 M * phycho ;) 1138649062 M * phycho which i need to do considering server has 512mb ram :D 1138649069 M * Bertl phycho: but the cases where you need a different kernel are pretty seldom 1138649081 M * phycho bertl - i know 1138649081 M * phycho it was more for testin things 1138649084 M * phycho *testing 1138649091 M * phycho i might use both UML and vserver on one machine 1138649095 M * phycho =) 1138649099 M * phycho uml for testing seperate kernels 1138649105 M * phycho and vserver for running services 1138649110 M * phycho since vserver=lot less overheads 1138649118 M * phycho so should be quicker than UML. 1138649119 M * Bertl well, you can run linux-vserver in xen domU and start an UML guest inside the linux-vserver guest :) 1138649124 M * phycho UML = *REALLY* slow 1138649133 M * phycho hehe cool 1138649161 M * phycho ill stick to learning vserver first =) 1138649169 A * phycho hopes vserver patches compile cleanly with grsecurity 1138649178 J * Smutje_ ~Smutje@xdsl-84-44-247-106.netcologne.de 1138649184 A * Hollow giggles 1138649206 M * Bertl phycho: with the mainline grsec yes :) 1138649216 M * phycho yep 1138649216 M * phycho k sweet 1138649234 M * Bertl I mean the grsec patches _in_ mainline :) 1138649241 M * Hollow :D 1138649243 M * phycho does mount work on vserver? 1138649250 M * phycho can i mount/unmount from within in? 1138649252 M * phycho *it 1138649259 M * Bertl if you allow it, yes 1138649263 M * phycho k sweet 1138649273 M * phycho just some of my users want the unmount command ;) 1138649279 M * mnemoc grsec got into mainline?? 1138649284 M * Hollow LOL! 1138649304 M * Bertl it's great how mnemoc pops up :) 1138649310 M * mnemoc p 1138649311 M * mnemoc :p 1138649315 M * phycho bertl - i assume you dont mean the grsec patches from www.grsecurity.net? 1138649329 Q * Smutje Ping timeout: 480 seconds 1138649337 M * Bertl phycho: thing is, folks always try to mix/match vserver with grsec 1138649350 M * phycho k 1138649355 M * phycho and it works yes? 1138649357 M * Bertl phycho: and most folks do it because they 'think' just adding the patches helps 1138649366 M * phycho heh 1138649377 M * Roey hi 1138649385 M * Bertl phycho: fact is, there are some things which work/help, and others which don't 1138649385 M * Roey hey Bertl, psycho, memnoc the devil 1138649392 M * phycho :) 1138649396 M * phycho hey 1138649398 M * mnemoc and most folks think grsec is good 1138649401 M * phycho lol 1138649402 M * Bertl phycho: and, depending on the versions, there are even patches available 1138649420 M * phycho mnemoc - security by obscurity doesnt owrk 1138649423 M * phycho =) 1138649434 M * phycho aka hiding list of processes 1138649435 M * phycho which grsec can do 1138649451 M * phycho bertl - thanks 1138649464 M * Bertl np 1138649477 M * mnemoc hi Roey :) 1138649563 M * phycho so.. a user in a vserver cant break out of it at all.. right? 1138649569 M * phycho (even if they exploit the kernel?) 1138649598 M * Bertl well, if they _manage_ to exploit the kernel properly, they can get out too ... 1138649609 M * phycho heh. damn 1138649612 M * Bertl just remember the VMware exploit which let you escape 1138649621 M * phycho yeah 1138649642 M * phycho well, the plan is to give users free public linux boxes 1138649642 M * phycho with root 1138649651 M * phycho so the more secure it is, the better 1138649652 M * phycho =) 1138649655 M * Bertl that's what linux-vserver is for 1138649660 M * phycho its going to be a public service 1138649667 M * phycho aka, anyone can register for one 1138649681 M * phycho so im going to get a lot of people trying to exploit it :P 1138649717 M * Bertl which will keep us in good contact, and might even benefit the project, no? 1138649730 M * phycho yeah probably :) 1138649762 M * phycho i provide freebsd jail based ones already 1138649764 M * phycho which is pretty tight 1138649764 M * phycho just wanting to move to linux as well 1138649765 M * phycho have two servers to offer them on =) 1138649779 M * bonbons Hollow: distinction can be done in multiple ways, be it a flag on the applet's description structure, or wether it's listed by default with other applets (vserver -h could should major applets, vserver --applets would also show expert ones) 1138649786 M * Roey mnemoc: hey hey :) 1138649803 J * cryo ~say@212.86.233.146 1138649820 M * Hollow ok, so we put them still all in one location.. 1138649821 M * Bertl phycho: sounds good, do you advertise them somehow? 1138649859 M * phycho i will do once they are running properly, yes 1138649863 M * phycho at the moment its only friends 1138649869 M * bonbons yep all in one location, just visibility to user is different (symlinks can be put in different locations in either case) 1138650047 M * bonbons for the sourcecode we can organize them in groups as we like (e.g. expert, ...) 1138650100 M * Hollow i guess it's ok for now to put them all together 1138650280 M * bonbons it's ok, but we could also split into two groups, expert/lowlevel and higherlevel 1138650325 M * Hollow then i'd prefer to do applets and commands 1138650357 M * Hollow commands for the low-level ones, and applets for the abstracted 1138650400 M * Hollow maybe we can even find another name for applets... it always reminds me of java :) 1138650442 M * bonbons we can also call it 'plugin' 1138650536 Q * meandtheshell Quit: a brave heart left ... 1138650629 M * Hollow well, it doesn't really matter 1138650639 M * Hollow plugin may be a bit ambiguous 1138650654 M * Hollow sounds like dynamic loading 1138650685 M * Hollow guess we'll stick with applets 1138650731 M * Bertl call it apples :) 1138650734 M * bonbons we basically have a few "synonyms": applet (generic), command, operation 1138650757 M * Hollow Bertl: you can't compare apples with oranges.. 1138650758 M * Hollow ;) 1138650786 M * Hollow Bertl: any better suggestion than plugin or applet? 1138650816 M * Bertl gizmo? thingy? 1138650828 M * Bertl wossnames? 1138650843 M * Hollow wrzlbrmpft 1138650894 M * bonbons what about `head -c 10 /dev/random` 1138652061 M * bonbons instead of debating on names I'm putting together some skeleton code, and will port a few commands to "vapplet-style" 1138652080 M * Bertl vapplet good name :) 1138652436 M * Hollow good idea :) 1138652449 M * Hollow i'm too tired already... 1138652668 J * Aiken ~james@tooax6-003.dialup.optusnet.com.au 1138652792 M * Bertl welcome Aiken! 1138652797 M * Bertl Vudumen: ping! 1138652818 J * tudenbart ~willi@xdsl-213-196-243-18.netcologne.de 1138652901 M * Aiken hi Bertl 1138653058 M * Vudumen Bertl ?? 1138653086 M * Bertl Vudumen: hey, seems moon stumbled somehow ... 1138653251 Q * dothebart Ping timeout: 480 seconds 1138653454 M * Vudumen Bertl: is it okay if i check it tomorrow? 1138653456 M * Vudumen currently i'm at home 1138653471 M * Bertl np 1138653476 M * Vudumen fine :) 1138653629 J * dearauj1 ~dan@pixpat.austin.ibm.com 1138653658 M * Bertl welcome dearauj1! 1138653685 M * dearauj1 hi Bertl 1138653689 M * dearauj1 I have a question 1138653727 M * dearauj1 Is it true that I cannot configure a vserver to the same ip address as the root server? 1138653733 M * dearauj1 i assume no 1138653739 M * Bertl no, that's not true 1138653815 M * dearauj1 so how does that exactly work then? 1138653830 M * Bertl well, the host has a number of ip addresses 1138653841 M * dearauj1 host meaning root? 1138653844 M * Bertl the guest is limited to a subset of them 1138653888 M * Bertl we call the host host and the guest guest :) 1138653888 M * Bertl i.e. the phyiscal machine is the host 1138653899 M * dearauj1 :) ok 1138653949 M * dearauj1 is this where the ip/number comes into play 1138653966 M * dearauj1 eg, 192.168.0.3/32 1138653978 M * Bertl well, that's an ipv4 ip with a prefix 1138654033 Q * phycho Read error: Connection reset by peer 1138654060 M * dearauj1 when you say the host has a number of ip addresses, how is that so? my apologies for my misunderstanding 1138654078 M * Bertl well, you know how a linux machine works regarding networking? 1138654085 M * dearauj1 vaguely :) 1138654106 M * dearauj1 bit a newbie with networking 1138654122 M * Bertl okay, you have network interfaces and they carry so called ip addresses 1138654127 M * dearauj1 yes 1138654135 M * Bertl that's what you see when you do 'ip addr ls' 1138654141 M * dearauj1 yep 1138654146 M * dearauj1 that I know :) 1138654149 M * dearauj1 heh 1138654151 M * Bertl now that is noting linux-vserver specific 1138654157 M * Bertl *nothing 1138654173 M * Bertl linux-vserver does not change that in any way ... 1138654188 M * Bertl it only 'restricts' the guests to a subset of the available ips 1138654210 M * dearauj1 available within the ost 1138654213 M * dearauj1 *host 1138654222 M * Bertl (similar to restricting the filesystem to a subset of files) 1138654257 M * Bertl all IPs whether used by the guests or not are present on the host 1138654322 M * Bertl does that make sense? 1138654360 M * dearauj1 that part i am confused about is the available IPs within the host.. 1138654379 M * Bertl all IPs are available in the host (physical machine) 1138654380 M * dearauj1 how is it that you have many IPs in one host when only one IP is assigned 1138654406 M * Bertl try to use the right nomenclature and it will match 1138654414 M * Bertl host = physical machine 1138654464 M * dearauj1 but the host (physical machine) is assigned 1 IP 1138654465 M * Bertl guest = vps = virtual server 1138654496 M * Bertl the host will _always_ carry _all_ ip addresses used 1138654564 Q * shed| Quit: Leaving 1138654600 M * dearauj1 for instance 1138654622 M * dearauj1 if the host if configured as having an IP of 198.162.0.1 1138654645 M * dearauj1 and the vserver is assigned 198.162.0.2 1138654662 M * dearauj1 the host carries both 198.162.0.1 and 192.168.0.2 1138654668 M * Bertl well, probably you mean 192.168.x.x 1138654691 M * Bertl but yes, the host will have both IPs and the guest will be limited to 192.168.0.2 1138654728 M * dearauj1 but that would assume the dhcp server has .0.2 available yes? 1138654761 M * Bertl which DHCP server? 1138654788 M * dearauj1 dhcp server on the network - not on the host 1138654800 M * ScoobyD00 are you running a "DHCP client" on the vserver guest? 1138654802 M * Bertl well, not related what your dhcp server does 1138654813 M * Bertl no, the guest does not use dhcp 1138654832 M * dearauj1 and now my confusion begins again 1138654834 M * dearauj1 :) 1138654838 M * dearauj1 ScoobyD00: no 1138654839 M * Bertl why? 1138654853 M * ScoobyD00 the host must get all its IP's directly (either from DHCP, or static) 1138654854 M * Bertl dearauj1: look, assume you have a linux server 1138654864 M * dearauj1 on my host machine - I have a dhcp client 1138654871 M * dearauj1 ScoobyD00: yes - i understand that 1138654881 M * Bertl dearauj1: further assume you have three addresses assigned to your host (192.168.0.1, 192.168.0.2 and 192.168.0.3) 1138654899 M * Bertl derjohn: I do not care where you get them from, dhcp, static config, whatever 1138654906 M * Bertl derjohn: *sorry* 1138654910 M * ScoobyD00 ok, so you need to tell linux to use interface "aliases", and set each to get a DHCP address 1138654915 M * Bertl dearauj1: I do not care where you get them from, dhcp, static config, whatever 1138654921 M * dearauj1 ok 1138654935 M * Bertl now you decide to put linux-vserver kernel on that machine 1138654950 M * Bertl dearauj1: nothing changes, still three ips assigned (as above) 1138654982 M * Bertl but, you now can start a guest and assign' one (or more) of the ip addresses to the guest 1138654990 M * dearauj1 ok - i understand 1138655003 M * Bertl the geust will only see the addresses assigned to it 1138655015 M * Bertl the host will still see everything 1138655028 M * dearauj1 but I can also assign the host IP to my vserver IP 1138655044 M * dearauj1 ie, i am only assigned 1 IP address (not 3 like in your example) 1138655049 M * Bertl as all of the IPs I mentioned _are_ host ips ... yes 1138655067 M * dearauj1 yes ok 1138655095 M * Bertl of course, if you use the same IP for host services and for guest services, you might get collisions 1138655108 M * Bertl i.e. an apache on the host might clash with one in the guest 1138655128 M * dearauj1 yes 1138655132 M * dearauj1 that makes sense 1138655166 M * dearauj1 so since im on a laptop and somewhat mobile - is there a way to configure dhcp on each vserver? 1138655178 M * dearauj1 *dhcp client 1138655187 M * Bertl there is, but it doesn't make much sense 1138655195 M * dearauj1 no? 1138655203 M * Bertl because the guest propably uses the same interface card 1138655214 M * Bertl (unless you have more than one network card) 1138655223 M * dearauj1 unfortunately no 1138655225 M * Bertl the dhcp will always request the same ip 1138655238 M * Bertl as the dhcp protocol is based on the MAC address 1138655266 M * dearauj1 ok 1138655269 M * Bertl but, nevertheless you can use a dhcp client request with a special client identifier to get _another_ ip 1138655283 M * Bertl and you can then assign that IP to the guest 1138655308 M * Bertl but let me say it is _much_ simpler to use a private ip for the guest and do SNAT 1138655314 M * Hollow off to bed now, cu tomorrow 1138655324 M * Bertl Hollow: good night! 1138655367 M * dearauj1 Bertl: that makes sense 1138655378 M * dearauj1 Bertl: thank you very much for your help 1138655383 M * Bertl you're welcome! 1138655481 M * bonbons Hollow: good night 1138655502 M * dearauj1 Bertl: one more thing (sorry) is there documentation that shows how to request another ip and assign that to the guest? 1138655532 N * sladen_ sladen 1138655560 M * Bertl dearauj1: I'd look into the dhclient manpage, I don't know of a HowTo or so especially for linux-vserver 1138655576 M * dearauj1 ok will do - thank you 1138655627 M * daniel_hozac well, jacques' utils support it, don't they? 1138655649 M * Bertl supposedly, yes 1138657175 A * sladen ponders about reparenting processes 1138657227 M * Bertl hey sladen! everything fine? 1138657255 M * sladen just been told a load of PA space is vanishing with a few hours notice 1138657285 M * sladen Bertl: ah, while we're at it, can you repoint archive.linux-vserver.org -> 193.28.45.41 1138657295 M * Bertl sure 1138657400 M * Bertl sladen: really archive? 1138657424 M * sladen Bertl: or whatever the current entry is 1138657428 M * sladen Bertl: archives? 1138657433 M * Bertl okay 1138657496 M * sladen the scary thing is that the first time I deployed vserver in anger was the last time my boxes had to move IPs 1138657602 M * Bertl hmm, interesting, my nameserver refuses to accept the new domain :) 1138657610 M * Bertl s/domain/entry/ 1138657628 M * Bertl ah, my fault ... 1138657782 M * Bertl sladen: okay, modulo propagation, it should be done 1138657811 M * Bertl sladen: what issues do you see with proces reparenting (or lack thereof)? 1138657816 M * sladen 4 years, 9 months ago I think 1138657854 M * sladen let existing processes inherit new priviliges (access to additional IPs) 1138657898 M * sladen in a limited way you can do a vserver enter and restart certain daemons but existing processes basically need the whole vserver restarting 1138657911 M * Bertl you want to hot-fix/patch the ip isolation? 1138657984 M * sladen would seem to be the case. The above is the use-case 1138658008 M * Bertl well, this has various issues 1138658027 M * Bertl first, it largely depends on the previous setup 1138658053 M * Bertl for example, you are better off with a two ip setup than with a single ip setup 1138658076 M * Bertl then it depends on the service, and how it was bound 1138658906 J * phycho ~phych0@209.8.233.12 1138658913 M * Bertl wb phycho! 1138658924 M * Roey http://zanyvgquotes.com/finalfantasy2/ff4spoony.png 1138658930 M * phycho ty 1138658940 M * phycho http://bbs.adslguide.org.uk/showthreaded.php?Cat=&Board=metronet&Number=2251544&page=0&view=expanded&sb=5&o= 1138658942 M * phycho read that thread 1138658948 M * phycho some sysadmins (aka so called unix engineer) 1138658950 M * phycho really dont know shit 1138658953 M * phycho lol 1138658956 M * phycho its ironic, im a student and i know more than him lol 1138659013 M * phycho :D 1138659014 M * Bertl phycho: well, you know Albert Einstein? 1138659019 M * phycho lol yeah 1138659041 M * phycho quite funny coz he tries to shoot down my arguements to show off his knowledge to thers 1138659044 M * phycho *others 1138659047 M * phycho and i shoot down everything he says lol 1138659068 M * phycho its shocking to think people like him are taking our jobs :x 1138659119 M * Bertl "Zwei Dinge sind unendlich: Das Universum und die menschliche Dummheit. Aber bei dem Universum bin ich mir noch nicht ganz sicher." -- Albert Einstein 1138659204 M * Bertl basically means: "two things are infinite: the universe and the stupidity of mankind. but I'm not sure about the universe :)" 1138659233 M * phycho hehehe 1138659934 Q * bonbons Quit: good night 1138659970 M * Doener phycho: regarding noexec (just read the thread), that doesn't help if you have access to /lib/ld-x.x.x.so, at least on linux (FJYI) 1138659999 M * Doener err, s/FJYI/JFYI/ 1138660041 M * daniel_hozac hmm, i seem to recall reading that that was "fixed"? 1138660065 M * daniel_hozac that may have been in relation to SELinux or similar though. 1138660168 M * Doener probably 1138660289 M * daniel_hozac hmm, doesn't seem to work here... may still be a Fedora thing. 1138660308 M * phycho doener - this is freebsd dude. 1138660318 M * phycho =) 1138660320 M * daniel_hozac (/mnt/true: error while loading shared libraries: /mnt/true: failed to map segment from shared object: Operation not permitted) 1138660326 M * phycho heh 1138660358 A * phycho is waiting to see if he replies 1138660360 M * phycho i bet he doesnt lol 1138660371 M * Doener daniel_hozac: i get that with ld-linux.so.2 but works with ld-2.3.5.so 1138660378 M * phycho heh 1138660393 M * daniel_hozac Doener: that was with ld-2.3.5.so. 1138660423 M * Doener phycho: thus I added 'at least on Linux', no offense intended, nice to know that i works 'the right way' on FBSD 1138660440 M * Doener daniel_hozac: hm, interesting... 1138660496 M * daniel_hozac (mount -t tmpfs -o rw,size=4m,noexec none /mnt; cp /bin/true /mnt; /lib/ld-2.3.5.so /mnt/true) 1138660508 M * phycho doener - no offense taken dude =) 1138660528 M * phycho personally i dont like to use linux that much 1138660535 M * phycho im a freebsd man myself 1138660545 Q * Roey Quit: Leaving 1138660658 M * michal_ ld hack was fixed in 2.6 1138660662 M * michal_ in vanilla 2.6 1138660680 M * michal_ not releated to any MAC system, just on a plain kernel 1138660692 M * matti Hueh. 1138660700 M * michal_ matti: ? ;p 1138660704 M * daniel_hozac SELinux is mainline ;) 1138660709 M * michal_ who cares ;] 1138660710 M * phycho yeah 1138660713 A * phycho enabled it 1138660724 M * phycho havent configged it yet tho 1138660730 M * michal_ just yet another bad decision of linus, purely politic releated 1138660747 M * michal_ eot here 'cause i do not want a flame 1138660767 M * phycho heh 1138660858 M * michal_ 5[albeiro@atlantis ~]% /lib/ld-linux.so.1 23:40 1138660862 M * michal_ zsh: 18340 segmentation fault (core dumped) /lib/ld-linux.so.1 1138660864 M * michal_ [albeiro@atlantis ~]% /lib/ld-linux.so.1 /bin/ls 23:40 1138660867 M * michal_ zsh: 18341 segmentation fault (core dumped) /lib/ld-linux.so.1 /bin/ls 1138660870 M * michal_ [albeiro@atlantis ~]% uname -a 23:40 1138660873 M * michal_ Linux atlantis.knm.org.pl 2.6.14-gentoo-r5 #2 SMP Mon Dec 26 20:12:52 CET 2005 i686 i686 i386 GNU/Linux 1138660876 M * michal_ try that on 2.4 ;p 1138660884 M * michal_ matti: how many linux variations are there on atlantis ? 1138660907 M * matti michal_: I don't know. I don't care about atlantis. 1138660920 M * michal_ well, nobody does ;) 1138660923 M * matti michal_: QNeX is crazy, poor admin, and he didn't care much also. 1138660942 M * michal_ i see 1138660942 M * michal_ eot ;] 1138660943 M * matti michal_: Remeber last discovered hole? 1138660947 M * michal_ yep 1138660950 M * michal_ have send him sms 1138660955 M * matti michal_: That was funny ;] 1138660959 M * matti michal_: Yeah, I remeber :) 1138660960 M * michal_ to make him fix it 1138660967 M * michal_ and he did :) 1138660967 M * matti michal_: He was scared like shit. 1138660970 M * michal_ :P 1138660988 M * michal_ no suprise ;] 1138660996 M * matti michal_: But, he still allows other user to access his home directory, so... 1138660997 M * michal_ probably his first root owned machine 1138661009 M * matti michal_: If you want his personal and private files... Well, enjoy. 1138661010 M * matti ;] 1138661027 M * michal_ thx you, got plenty of mine own 1138661036 M * matti No dobut :) 1138661037 M * matti Hehehe. 1138661055 M * phycho lol 1138661076 M * matti phycho: No, no :) 1138661083 M * matti phycho: This is so called "reality" ;] 1138661089 M * michal_ http://pastebin.de/4083 1138661091 M * michal_ worth lookikng ;p 1138661099 M * matti phycho: Users and poor admins don't care about security... 1138661114 M * matti phycho: Everybody knows, that security is for hackers and american movies ;p 1138661114 M * matti ;p 1138661147 M * michal_ civic is having an account there ? 1138661149 M * michal_ uh 1138661149 M * michal_ lol ! 1138661155 M * michal_ if it is the same civic i know 1138661167 M * michal_ he is ehm... %%%^$%$%$#$# up ;p 1138661221 M * michal_ and eryk 1138661226 M * Bertl michal_: RPM reminds me of espdiff ... 1138661226 M * michal_ your favorite 1138661228 M * Doener daniel_hozac: haha, i had the line like this "mount -t tmpfs -o rw noexec /mnt", that way I could've even done /mnt/true... doesn't work with a real noexec here either... I'll go hiding now 1138661246 M * daniel_hozac Doener: haha. 1138661272 M * michal_ Bertl: lol 1138661285 M * phycho matti - only morons dont care about security 1138661288 M * phycho put it this way 1138661293 M * matti phycho: Indeed. 1138661298 M * phycho my wifi has ipsec over the top, with WPA2-enterprise underneath 1138661302 M * phycho hidden ssid + mac filtering 1138661306 M * phycho + RADIUS auth 1138661307 M * phycho lol 1138661314 M * matti phycho: Well... 1138661316 M * phycho like to see someone break into that haha 1138661321 M * matti phycho: Paranoia is not good either. 1138661321 M * matti ;] 1138661321 M * phycho major overkill! 1138661325 M * phycho lol. 1138661326 M * Bertl phycho: what ssid and keys do you use :) 1138661335 M * phycho bertl - even if i told you 1138661338 A * michal_ will easily break in 1138661342 M * phycho you still wouldnt get in 1138661344 M * phycho lol 1138661345 M * michal_ where do you live ? 1138661348 M * phycho my neighbours only has wep 1138661349 M * phycho ;) 1138661355 M * phycho i was bored one day so cracked it 1138661355 M * michal_ i'l take a nice gun, walk to ya 1138661356 M * daniel_hozac matti: being paranoid doesn't mean no one is following you/breaking in ;) 1138661357 M * phycho took <30mins 1138661360 M * michal_ put the gun to your head 1138661364 M * michal_ and ask for access 1138661366 M * michal_ kindly :] 1138661367 M * phycho lol 1138661391 M * phycho michal_ - put it this way dude, you are more likely to manage to break into my house and steal the server 1138661392 M * phycho than break in via the wireless 1138661393 M * phycho lol 1138661399 M * michal_ ;p 1138661408 M * michal_ no problem for Those Who Neeed It 1138661413 M * phycho quite funny, i love the madwifi virtual AP's feature 1138661413 M * michal_ (steal the server) 1138661416 M * matti phycho: Don't be so sure :) 1138661419 M * phycho i setup one called 'default' 1138661421 M * michal_ (with you if necesary ;p) 1138661424 M * matti phycho: Watch any american movie ;p 1138661427 M * phycho running honeyd with a shitload of nt4 servers 1138661437 M * phycho and ive had so many people have a pop at it <3 1138661438 M * matti phycho: Y'll see, how to breat top secret goverment network ;p 1138661441 M * matti phycho: Instantly ;] 1138661441 M * phycho who think its a real network 1138661442 M * phycho lol 1138661446 M * phycho hehe :P 1138661449 M * matti phycho: By using Microsoft Office ;] 1138661457 M * matti phycho: No shit. 1138661457 M * matti ;]; 1138661510 M * michal_ indeed 1138661530 M * matti We also have our own regional hacker movie. 1138661551 M * matti And in this movie, some hacker penetrate out goverment system with Emacs and Sendmail. 1138661554 M * matti So. 1138661556 M * matti ;p 1138661565 M * phycho lol 1138661567 M * michal_ emacs throught sendmail way 1138661571 M * michal_ you know 1138661575 M * michal_ it is quite logical 1138661582 M * matti michal_: Yeeeh, that the trick itself. 1138661584 M * matti ;] 1138661596 M * michal_ you are sending emacs throught sendmail 1138661597 M * phycho id be more likely to believe it if they wrote their own exploit in vi 1138661598 M * matti michal_: But! This is TOP SECRET! I need to kill ya now. 1138661599 M * michal_ source of it 1138661607 M * michal_ and there ya go 1138661611 M * phycho :P 1138661611 M * michal_ BO ;p 1138661620 M * michal_ matti: again ? 1138661629 A * michal_ enters zombie mode 1138661630 M * michal_ try now !;p 1138661642 A * phycho goes to look for more stuff to watch 1138661650 M * michal_ anyway 1138661652 M * michal_ http://pastebin.de/4080 1138661657 M * michal_ lokk there 1138661658 M * phycho Lethal.Weapon.4.1998.DVDRip.AC3.XViD.iNT-MASTER 1138661660 M * phycho old school 1138661661 M * phycho <3 1138661665 A * phycho watches 1138661675 M * mnemoc dune rules 1138661691 M * phycho i got that somewher 1138661693 M * phycho *somewhere 1138661714 M * Bertl phycho: you know this channel is logged, yes? 1138661720 M * matti phycho: Oh you bloody pirat! 1138661721 M * phycho yeah 1138661724 A * matti is calling RIAA now! 1138661725 M * matti ;] 1138661727 M * phycho lol 1138661728 M * matti ;p 1138661739 M * phycho who said i dont own the origional? :) 1138661747 A * michal_ 1138661764 M * matti phycho: Yeah, and make such warez scene directory name for funny ;] 1138661771 M * phycho lol 1138661771 M * michal_ your neighbours have steal it from ya 1138661778 M * phycho haha 1138661778 M * matti phycho: I can believe in that, really ;p 1138661781 M * michal_ along with all prof that you have ever buy it 1138661784 M * michal_ remember 1138661787 M * phycho hehe 1138661788 M * michal_ we're living in a matrix 1138661797 M * michal_ and i can make you vanish like you never existed 1138661802 M * michal_ anytime 1138661806 M * phycho matti - i could have released it for all you know ;) 1138661808 M * phycho lol 1138661813 M * michal_ uhm that was some other movie actualy ;p 1138661820 M * phycho hehe 1138661828 M * Bertl okay, let's cut the crap .. plz move #offtopic ... 1138661829 M * phycho oooh.. /me thinks hes got replies 1138661839 M * phycho http://bbs.adslguide.org.uk/postlist.php?Cat=&Board=metronet 1138661839 M * phycho wooo 1138661840 A * phycho watches 1138661873 A * michal_ quiets 1138661875 M * michal_ matti too 1138661899 M * ebiederm Bertl: Are you around? 1138661905 M * Bertl ebiederm: yep 1138661910 M * matti phycho: Yeah... And most of warez guys should be called something like Santa Claus I know ;] 1138661923 M * matti phycho: Steal the rich, copy and give the poor one. 1138661925 M * ebiederm I just discovered that the vserver list is terrible for posting patches to. 1138661938 M * matti phycho: So melodramatic. 1138661940 M * matti phycho: :) 1138661944 M * ebiederm It kicked out my patches because I Cc'd too many people. 1138661951 M * ebiederm Is that deliberate? 1138661959 M * matti Bertl: Sorry :( 1138661992 M * Bertl ebiederm: well, you have to talk to martin about that, he maintains the list, but I'm reading your patches/postings anyway 1138662040 M * Bertl ebiederm: but I haven't seen any posting on the ML yet :) 1138662050 M * ebiederm Bertl: What is the best way to ask that question? 1138662071 M * Bertl ebiederm: are you subscribed to the list? 1138662075 M * ebiederm Bertl: Yes. 1138662083 M * phycho lol 1138662093 M * ebiederm If I just post to the list it should get through. 1138662095 M * Bertl ebiederm: well, just mail him and cc the ML 1138662097 M * matti phycho: Sorry for flood. 1138662103 M * matti phycho: :) I'll cool now. 1138662103 J * Skram ~mark@shells.vpses.net 1138662104 M * ebiederm Bertl: Ok sounds good. 1138662105 M * Skram Hi All. 1138662109 M * matti Hi Skram ;] 1138662116 M * Bertl welcome Skram! 1138662118 M * Skram Anyone have troubles restarting a VPS from within. 1138662124 M * michal_ hi Skram :) 1138662127 M * Bertl Skram: you maybe? 1138662129 M * Skram Ive spoken with you all a bit ago, but I have been curious about this.. 1138662131 M * Skram Bertl: Well, yes. 1138662146 M * Skram Like when a user issues "reboot" it shuts down their vserver, and it doesnt come back up :( 1138662151 M * ebiederm Bertl: I did have fun today. One of my patches help reveal a long standing bug in 2.6. 1138662179 M * ebiederm init was getting pgrp == session == 0 instead of 1. 1138662201 M * Bertl Skram: so seeking for fellow admins with the same issues, or more for a solution? 1138662213 M * Skram Solution :) 1138662240 M * Bertl well, then let's first heck the installtion, with the testme.sh (and upload the output to pastebin.com or so) 1138662247 M * Bertl s/heck/check/ 1138662250 M * Skram erm.. 1138662254 M * Skram I didnt install it on this server 1138662256 M * Skram but okay. 1138662264 M * Skram I will do this on the semi-dev server l) 1138662269 M * michal_ http://piotr.milczarek.com/rotfl/unixoid_hell.gif 1138662277 M * Bertl Skram: http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh-0.15 1138662284 M * Bertl Skram: it's harmess ... 1138662290 M * Skram *harmless? 1138662296 M * Bertl yep 1138662316 M * Skram everything "succeeded." 1138662327 M * Bertl I'm interested in the other info, so plz upload 1138662340 M * Skram Linux 2.6.14-vs2.0.1-gentoo #1 SMP Mon Jan 16 23:43:02 UTC 2006 i686 1138662340 M * Skram Ea 0.30.209 273/glibc (DSa) 1138662340 M * Skram VCI: 0002:0001 273 03000034 (TbnP) 1138662344 M * Skram :) 1138662398 M * Bertl okay, guess the guests are init based, yes? 1138662400 P * dearauj1 1138662407 M * Skram Austin?! Im in austin! 1138662417 M * Skram Sorry 1138662418 M * Skram Erm... 1138662432 M * Skram We always start and stop them with "vserver 1138662433 M * Skram " 1138662448 M * Skram there is a 1138662452 M * Bertl okay, yes, fine, but the config? 1138662456 M * Skram /etc/init.d/vservers, but just for all the servers. 1138662490 M * Bertl do you have a directory in /etc/vservers for each guest? 1138662498 M * Skram in /etc/vservers//init/stypem it says "plain" 1138662504 M * Skram Yes, /etc/vservers/ 1138662509 M * Bertl okay 1138662518 M * Bertl and the vshelper is configured properly? 1138662535 M * Skram Never used that :) 1138662537 M * Skram * :( 1138662553 M * matti http://piotr.milczarek.com/rotfl/unixoid_hell.gif 1138662584 M * Bertl Skram: cat /proc/sys/kernel/vshelper 1138662587 M * Skram matti: ha ha. 1138662590 M * Skram Bertl: okay one second 1138662597 M * Skram /usr/lib/util-vserver/vshelper 1138662601 M * Skram is the output of that "cat" 1138662638 M * Bertl ls -la /usr/lib/util-vserver/vshelper 1138662645 M * Skram its there. 1138662652 M * Skram -rwxr-xr-x 1 root root 5921 1138662654 M * Bertl okay, that's at least something :) 1138662658 M * Skram Right :) 1138662675 M * Bertl how do you 'restart' the guest from inside 1138662686 M * Skram "reboot" 1138662712 M * Bertl could you try 'reboot -f' just for a test? 1138662713 J * shedi ~siggi@inferno.lhi.is 1138662717 M * Skram sure 1138662761 P * stefani I'm Parting (the water) 1138662779 M * Skram it hasnt shown back up in "vserver-stat" :( 1138662799 M * Bertl so it really stops and isn't started ... interesting 1138662842 M * Bertl are the tools gentoo tools? 1138662847 M * Skram ? 1138662862 M * Bertl are they mainline or isntalled via gentoo 1138662866 M * Bertl (util-vserver) 1138662868 M * Skram i believe gentoo 1138662939 M * Bertl well, the problem is, atm I cannot test it myself ... 1138662943 M * Skram * sys-cluster/util-vserver 1138662943 M * Skram Latest version available: 0.30.210-r1 1138662944 M * Skram Right 1138662955 M * Skram It was installed via portage, so gentoo version 1138662961 M * Skram well, is there something i can run for you? 1138662966 M * Bertl it might be a gentoo related issue ... 1138662974 M * Skram I have some customer data on here, so I cant give you root to the main server. 1138662978 M * Skram well, vpshost that is 1138662996 M * Bertl well, if you like you can install a vanilla kernel plus vanilla tools 1138663006 Q * ScoobyD00 Quit: 1138663007 M * Bertl and then try again 1138663020 M * daniel_hozac i guess enabling vshelper logging might help too. 1138663022 M * Bertl but I'd suggest to wait for Hollow and check this with him first 1138663035 M * Skram Ill just have users issue open a reboot ticket :) 1138663047 M * Skram Hmm, ill talk wit Hollow.. 1138663047 M * Bertl daniel_hozac: yes, but I'm not sure this is enabled in the gentoo kernel 1138663064 M * daniel_hozac (i.e. mkdir -p /etc/vservers/.defaults/apps/vshelper; touch /etc/vservers/.defaults/apps/vshelper/logfile) 1138663078 M * Skram # CONFIG_VSERVER_DEBUG is not set 1138663081 M * Bertl daniel_hozac: ah, you mean userspace ... 1138663090 M * Bertl Skram: yeah, thanks, I saw that 1138663100 M * Skram what do you mean? 1138663102 M * Skram okay.. 1138663130 M * Bertl let's try what daniel_hozac suggests ... 1138663152 M * Skram done 1138663160 M * Skram okay.. 1138663162 M * Skram rebooting 1138663219 M * Skram vpshost linux # cat /etc/vservers/.defaults/apps/vshelper/logfile 1138663219 M * Skram Mon Jan 30 23:18:27 UTC 2006: vshelper restart 6184 1138663219 M * Skram Restarting vserver '/etc/vservers/e' 1138663219 M * Skram vserver 'e' already running 1138663220 M * Skram weird 1138663235 M * Skram i accidently typed something wron 1138663236 M * Skram g 1138663238 M * Skram let me re-do 1138663312 M * Skram waiting for it to reboot... 1138663320 M * Skram (or turn off at all ) 1138663322 M * Skram okay its down 1138663331 M * Skram vpshost linux # cat /etc/vservers/.defaults/apps/vshelper/logfile 1138663331 M * Skram Mon Jan 30 23:20:25 UTC 2006: vshelper restart 6184 1138663331 M * Skram Restarting vserver '/etc/vservers/e' 1138663331 M * Skram vserver 'e' already running 1138663333 M * Skram thats what it says 1138663335 M * Skram weird! 1138663359 M * phycho omg 1138663361 M * phycho guys 1138663365 M * phycho remember i posted that link 1138663367 M * phycho i got a pm from that guy 1138663374 M * phycho When you've graduated you should look us up for an interview, you're clearly very talented and we're always after good minds to join. We currently have a development and support job going, so if you know anyone whose skills are on a par with yours then please let them know. 1138663375 M * phycho lol 1138663445 M * Skram ..? 1138663571 M * daniel_hozac Skram: that is odd. there are no leftover processes, right? 1138663610 M * daniel_hozac Skram: does vserver e restart work as expected? 1138663614 M * Skram right 1138663615 M * Skram yes. 1138663652 M * daniel_hozac does vserver --defaulttty e restart work as well? 1138663672 M * Skram lemme start it 1138663793 Q * zobel Ping timeout: 480 seconds 1138664139 J * zobel zobel@zobel.irc.ftbfs.de 1138664468 J * miRC ~Q@ip70-178-107-144.ma.dl.cox.net 1138664514 M * Bertl welcome miRC! 1138664535 M * miRC hi 1138664887 A * Skram has to recompile php :( 1138665200 Q * sladen Remote host closed the connection 1138665249 M * marl_ just a quicky :) does anyone know an easy way to get ONLY the ip addy from 'ip' without any of the other stuff it normally displays? (so i can use it withing config files etc. to allow for ip addys changing 1138665293 M * Bertl why not use 'ip addr ls'? 1138665404 M * Bertl ip addr ls eth0 | gawk '/inet/ { print $2 }' 1138665411 M * marl_ im trying to grab just the ip addy out of the command, not all the other stuff like mac addys etc. can do it with a wee program, but wanted to ask if there was a shorter way of doin git :) 1138665454 J * sladen ~paul@193.28.45.41 1138665457 M * marl_ thanks Bertl 1138665466 M * marl_ knew there was a shorter way :) 1138665504 M * daniel_hozac guess you want /inet / if you have IPv6 support in the kernel. 1138665561 M * Bertl and maybe: 1138665563 M * Bertl ip addr ls eth0 | gawk 'BEGIN {FS="[ /]*"} /inet / { print $3 }'