1242691821 Q * Medivh Ping timeout: 480 seconds 1242692793 Q * balbir_ Ping timeout: 480 seconds 1242693053 J * kiorky ~kiorky@cryptelium.net 1242694141 J * imcsk8 ~ichavero@189.155.79.208 1242696186 Q * doener_ Ping timeout: 480 seconds 1242696902 Q * imcsk8 Quit: This computer has gone to sleep 1242697393 M * Bertl off to bed now ... enjoy! 1242697398 N * Bertl Bertl_zZ 1242698458 J * balbir_ ~balbir@202.81.30.16 1242702143 J * imcsk8 ~ichavero@189.155.79.208 1242702194 J * geb ~geb@earth.gebura.eu.org 1242703920 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1242709144 Q * _nono_ Ping timeout: 480 seconds 1242709307 J * jfsworld ~jf@202.157.130.21 1242709321 M * jfsworld hey folks 1242709376 M * jfsworld i'm trying to tune scheduling parameters, but i'm running into a problem starting contexts 1242709384 M * jfsworld 'vsched: interval2: is not an integer' 1242709473 M * jfsworld anybody know anything about this? I have /usr/local/etc/vservers//sched/{fill-rate,fill-rate2,interval,interval2} - all set with integers, so i dont get why i get this error message.. 1242709644 J * _nono_ ~gomes@libation.ircam.fr 1242710301 Q * balbir_ Remote host closed the connection 1242711126 Q * _nono_ Ping timeout: 480 seconds 1242711398 Q * bzed Ping timeout: 480 seconds 1242711595 M * jfsworld quick qn on /etc/vservers/sched - must i put in all files? ie. fill-rate, fill-rate2, blah blah down the list 1242711609 M * jfsworld (i'm assuming /etc/vservers/vserver-name/sched/cpu-id is optional) 1242711707 J * _nono_ ~gomes@libation.ircam.fr 1242711986 J * rofe ~rofe@83.221.146.177 1242712046 Q * rofe 1242712058 J * rofe ~rofe@83.221.146.177 1242712265 J * bzed ~bzed@devel.recluse.de 1242712608 J * sharkjaw ~gab@149-242-57.oke2-bras6.adsl.tele2.no 1242713156 Q * sharkjaw Quit: Leaving 1242713170 J * sharkjaw ~gab@149-242-57.oke2-bras6.adsl.tele2.no 1242716005 J * dna ~dna@186-204-103-86.dynamic.dsl.tng.de 1242716677 J * davidkarban ~david@193.85.217.71 1242716752 P * jfsworld 1242717049 J * thierryp ~thierry@zanzibar.inria.fr 1242717969 Q * imcsk8 Quit: This computer has gone to sleep 1242718246 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1242718534 Q * davidkarban Read error: Connection reset by peer 1242718548 J * davidkarban ~david@193.85.217.71 1242718890 Q * sharkjaw Remote host closed the connection 1242718951 J * sharkjaw ~gab@149-242-57.oke2-bras6.adsl.tele2.no 1242719116 J * scientes ~scientes@174-21-91-35.tukw.qwest.net 1242719239 J * cga ~weechat@194.244.1.35 1242720561 Q * geb Ping timeout: 480 seconds 1242720717 J * geb ~geb@earth.gebura.eu.org 1242720980 Q * dna Quit: Verlassend 1242722091 M * ghislainocfs2 hello there 1242722121 M * ghislainocfs2 jfsworld: give us a pastebin of tail 1242722128 M * ghislainocfs2 tail /usr/local/etc/vservers//sched/ 1242722156 M * ghislainocfs2 ping: daniel_hozac 1242722198 M * ghislainocfs2 daniel if you are here, do you know a way to see which dir is not recognized when you get: vdlimit: vc_get_dlimit(): No such file or directory 1242722229 M * ghislainocfs2 i got a bunch of bind mount so i wondered if we can see which one is the problem or if i must comment them one by one to find the issue 1242723082 Q * harobed Quit: Ex-Chat 1242723112 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1242723218 Q * harobed 1242723548 N * Bertl_zZ Bertl 1242723554 M * Bertl morning folks! 1242723632 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1242723940 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1242723988 Q * BenG 1242724696 J * jfsworld ~jf@202.157.130.21 1242724709 M * jfsworld hello 1242724852 M * Bertl jfsworld: hey, to answer your question: you only need to fill in those you want to change from the default 1242724871 M * jfsworld yeah got it thx 1242724887 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it 1242724898 M * jfsworld do u know if there's a way to restrict memory usage for contexts? I looked at the flower page - but only mention of it was for "cpuset" 1242724944 M * Bertl http://linux-vserver.org/Documentation (see Resource and Memory limits a little down) 1242724979 M * jfsworld ah! ok 1242724980 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1242725235 J * Keeper longhorn40@94.50.101.236 1242725353 J * kir ~kir@nat-ip.sw.ru 1242725591 M * Keeper Hi 1242725855 Q * grobie Quit: Coyote finally caught me 1242726005 J * dna ~dna@186-204-103-86.dynamic.dsl.tng.de 1242726184 M * nkukard Bertl, daniel_hozac i just upgraded to 2.6.29 with latest development vserver patch 1242726214 M * nkukard still getting an "operation not permitted" when trying to mkdir /dev/cgroup/test-server but when I do it myself it works fine from commandline 1242726257 M * nkukard in vserver.functions, line 1478 +- 1242726266 M * nkukard that mkdir "$CGROUP_MNT/$name" 1242726420 M * Bertl which util-vserver is that? 1242726471 A * nkukard checks 1242726548 M * nkukard onone of the prereleases berth 1242726553 M * nkukard *bertl 1242726611 M * nkukard busy trying to see which one 1242726682 M * nkukard util-vserver-0.30.216-pre2827.tar.bz2 1242726759 M * Bertl try to output $CGROUP_MNT and $name 1242726783 M * Bertl also make sure not to run selinux or similar 1242726791 M * nkukard /dev/cgroup and test-server 1242726796 M * nkukard not got selinux or anything 1242726804 M * nkukard i can mkdir fine myself from the prompt 1242726809 M * nkukard and can assign tasks to it 1242726816 M * nkukard but, I get operation not permitted inside the script 1242726835 M * Bertl then please put an strace -fF before that mkdir 1242727001 J * doener ~doener@i59F5439D.versanet.de 1242727005 M * nkukard http://paste.linuxassist.net/214874 1242727009 M * nkukard I'm root :) 1242727017 M * nkukard running udev aswell if that makes any diff 1242727077 M * nkukard http://paste.linuxassist.net/214875 <= that is me doing it manually just after the above error 1242727261 M * Bertl any 'unusual' flags on your filesystem, like e.g. a barrier on / or so? 1242727298 M * nkukard nope .... i'm busy checking the umask 1242727305 M * Bertl also, please extend the script by ls -la "$CGROUP_MNT" 1242727306 M * nkukard mkdir("/dev/cgroup/debian-build", 0777) = 0 1242727309 M * nkukard 999 looks odd 1242727315 M * nkukard grr, 777 1242727558 M * nkukard http://paste.linuxassist.net/214876 1242727696 M * nkukard berth, don't i need any perms? like CAP_SYS_ADMIN? 1242727704 A * nkukard just reading some google results 1242727732 M * Bertl you are starting a guest, so you are doing that as root, yes? 1242727747 M * Bertl root on the host, by default, has all capabilities 1242727770 M * nkukard yes 1242727846 M * Bertl what's the output of the last paste (at the beginning)? 1242727929 M * nkukard the TEST? that is an echo TEST: $CGROUP_MNT 1242727951 M * Bertl so $CGROUP_MNT is already /dev/cgroup/wug-server ? 1242727968 M * Bertl where does the wug-server come from? 1242727975 M * nkukard CGROUP_MNT is /dev/cgroup 1242727987 M * nkukard wug-server is /etc/vservers/wug-server/cgroup/name 1242727995 M * Bertl so then why does 'echo TEST: $CGROUP_MNT' output that? 1242728012 M * nkukard echo TEST: "$CGROUP_MNT/$name" 1242728014 M * nkukard sorry, my bad 1242728028 M * Bertl aha, okay, and the ls below is from what command (exactly)? 1242728046 M * nkukard also, please extend the script by ls -la "$CGROUP_MNT" 1242728048 M * nkukard from that :) 1242728074 M * jfsworld Bertl: a qn about memory limits 1242728098 M * Bertl nkukard: please try with the latest snapshot, and let me know if the issue remains 1242728101 M * jfsworld since writing to oom_adj is not permitted in contexts, then what happens when sshd is killed inside a context? 1242728122 M * nkukard http://paste.linuxassist.net/214877 <= that is what i have so far, Bertl i will build and try now :) 1242728140 M * Bertl jfsworld: the same as when it is killed on the host, it dies and makes room for other processes 1242728174 M * jfsworld yeah, but on the host, oom_adj prevents that ? 1242728196 M * jfsworld not a major point, i guess. Cos u can always login to host and restart sshd but... just a thought 1242728254 M * jfsworld in my guest, i notice that all processes, except for init, have oom_adj -17 1242728258 M * Bertl jfsworld: the point is, if you allow a 'guest' to set the oom_adj to disable the OOM killer, what will keep john doe from making all his processes resilient (in the guest)? 1242728294 M * Bertl so, if you really want to protect sshd inside a guest, then please do so from the host 1242728304 M * jfsworld right. So the thing is, how? 1242728347 M * Bertl with xid=1, you should be able to see all processes, and thus make the proper adjustments 1242728371 M * Bertl if that fails (for whatever reason) please let me know 1242728589 M * jfsworld so i have to create a separate dedicated guest for admin? 1242728775 M * Bertl no, just use 'chcontext --xid 1 ...' or 'vcontext --migrate --xid 1 ...' 1242728888 M * jfsworld ok thx 1242728904 M * jfsworld one more qn: thing is like i said, all processes in my guest seem to be oom_adj -17 1242728906 M * jfsworld why is that? 1242728946 M * Bertl most likely because the process starting them had this 1242729038 M * Bertl on my system, they are unbiased (i.e. 0) 1242729049 M * jfsworld for real? 1242729060 M * jfsworld they were started just like any other context - 'vserver start ' - that's it! 1242729061 M * Bertl # chcontext --xid 1 -- cat /proc/24460/oom_adj 1242729061 M * Bertl 0 1242729084 M * Bertl what oom_adj has your logon? 1242729103 M * Bertl 'cat /proc/self/oom_adj' 1242729113 M * jfsworld hm, -17 1242729122 M * Bertl see, there you go :) 1242729161 M * jfsworld didnt know that would affect things.. blah. Hm. Learn somethiing new there then thankx 1242729224 M * nkukard Bertl, waiting for my staff to get off the vservers before i try upgrading :), may take another hour or so 1242729268 M * Bertl jfsworld, nkukard: np 1242729274 J * ben00001 ~ben@host90-152-0-26.ipv4.regusnet.com 1242729317 M * ben00001 Hello troops 1242729328 J * Piet ~piet@tor-irc.dnsbl.oftc.net 1242729333 P * jfsworld 1242729404 M * ben00001 anyone want to offer help (and be bored) about a series of performance questions 1242729483 M * Bertl hey, what's up? 1242729617 M * ben00001 I'm troubleshooting some performance now, I want my Apache RPS to go through the roof 1242729662 M * Bertl well, then I would contact the apache folks, no? 1242729667 M * ben00001 yes and no 1242729673 M * ben00001 Remember the server I let you on the other day, for an ecommerce install I put on - I tweaked the settings to get 20 rps 1242729689 M * ben00001 since sticking apache in a vserver, its more around 16-17 rps 1242729708 M * ben00001 before the cpu was loaded to 70% - now, its only loaded to 50% 1242729724 M * ben00001 is there some kind of cpu limitation implied with a vserver? 1242729743 M * Bertl not if you haven't enabled it 1242729747 M * ben00001 hmm 1242729765 M * ben00001 no matter how many concurrent users i throw at it, its still the same result 1242729799 M * ben00001 guess that idea goes out the window then 1242729801 M * ben00001 lol 1242729856 M * Bertl what kernel version do you use? 1242729888 M * Bertl (and how do you measure the RPS, where 20 seems awfully slow btw :) 1242729978 M * ben00001 I am using ... 2.6.28.7-grsec2.1.13-vs2.3.0.36.7 1242729997 M * ben00001 the RPS is through ab (apache bench), 35 concurrent users 1242730006 M * Bertl did you try disabling grsec? 1242730022 M * ben00001 ... not yet 1242730031 M * ben00001 its 20 rps on a Magento install 1242730037 M * ben00001 that's considered good :| 1242730075 M * Bertl well, IIRC, 2000 requests for static pages is not a big deal for apache, but of course, on dynamic page systems, it's probably different 1242730104 M * Bertl (althoug it should use caches for that too) 1242730526 Q * PowerKe Server closed connection 1242730538 J * PowerKe ~tom@d5153A5EC.access.telenet.be 1242730868 M * ben00001 to help the people with the issue i was having the other day 1242730897 M * ben00001 it was more my ignorance and stupidity that it wasn't working (isn't it always!) 1242730923 M * ben00001 where the ip's were secondaries of the main network interface, i had set them up wrong 1242730948 M * ben00001 in the /etc/vservers/X/interfaces/0/ -- there is three files 1242730957 M * ben00001 ip dev prefix 1242730992 M * ben00001 after removing the prefix file, and appending it to the end of the IP (like in earlier versions of util-vserver), it appeared to solve my issue 1242731035 M * ben00001 also, i had entered different prefixes for each ip by mistake, so i had a 32 and a few 24s 1242731063 M * Bertl yeah, that pretty much explains it :) 1242731073 M * ben00001 so i fixed my issue (i hope) by changing the ip to, xxx.xxx.xxx.xxx/32 (in the /etc/vservers/X/interfaces/0/ip) file 1242731089 M * ben00001 yup - i'm *special* 1242731100 M * ben00001 for all ips 1242731102 M * mnemoc ben00001: /32 doesn't go there 1242731105 M * Bertl well /32 isn't the best choice 1242731121 M * Bertl and you want to put it into the prefix as '32' if at all 1242731259 M * ben00001 i'll bear it in mind (if i get any hiccups) 1242731266 M * ben00001 what would you pick over 32? 1242731277 M * ben00001 oh, you mean, not the number, but the placement 1242731278 M * Bertl the actual netmask 1242731279 M * ben00001 gotcha ;) 1242731300 M * ben00001 i *think* its 29 on my network 1242731302 M * Bertl if you got a complete class C network, you put 24 there, if you got a subnet, adjust it 1242731303 M * ben00001 i'll check 1242731544 M * ben00001 28 then for me 1242731584 M * ben00001 so getting my RPS up - i've tweaked every tweak for apache and php and mysql - any info on where else i could hunt for general debian performance increases 1242731767 M * Bertl probably on a debian channel, like #debian? 1242731770 M * ben00001 lol 1242731772 M * ben00001 gotcha ;) 1242731834 M * Bertl in general, look for system optimizations first, then optimize the applications 1242731846 M * ben00001 i will have a look 1242731854 M * Bertl often simple stuff like apache logs can hog down a system pretty well 1242731862 M * ben00001 apache logs are on "emerg" 1242731872 M * ben00001 i never found much for system optimizations actually 1242731915 M * Bertl 64bit kernel, proper memory access, remove multilevel stuff, remove test & debug stuff 1242731946 M * Bertl fine tune the I/O subsystem (slowest component usually), optimize caching/preload/readahead 1242731982 M * Bertl reduce network stack to the minimum required, optimize rules and routes 1242732967 M * Mr_Smoke Hi there 1242732976 M * Bertl hey 1242732981 M * Mr_Smoke Hi Bertl :) 1242732985 M * Mr_Smoke Got a question for you 1242732992 M * Mr_Smoke More a confirmation of an assumption actually 1242733011 M * Mr_Smoke You told me the other day that networking was done at the IP level (at least in 2.2) 1242733058 M * Mr_Smoke Does that mean that if two guests (or 1 guest and the host for that matter) are to communicate via IPs assigned to eth0, this will actually happen via the host's lo ? 1242733086 M * Bertl yes, that is true for all Linux-VServer versions 1242733109 M * Mr_Smoke Say the host has lo (obviously) and eth0 1.2.3.4, and the guest has eth0 1.2.3.5 ... will 1.2.3.4 <-> 1.2.3.5 happen at lo then ? 1242733112 M * Mr_Smoke Ok 1242733119 M * Mr_Smoke That feels right too 1242733170 M * Mr_Smoke I'm thinking of logs centralizatoin here, using syslog over the "private" guest network 1242733182 M * Bertl unless you use network namespaces on a very recent kernel, all host-guest and guest-guest traffic will use lo 1242733189 M * Mr_Smoke I *could* also use private IP ranges for that, by the way? 1242733202 M * Mr_Smoke s/\?/\./ 1242733211 M * Mr_Smoke Good. 1242733351 M * Bertl have to leave now .. bbl 1242733355 N * Bertl Bertl_oO 1242733377 M * Mr_Smoke Seeya :) 1242733579 Q * DelTree Server closed connection 1242733591 J * DelTree ~deplagne@goldorak3.eric.deplagne.name 1242736291 Q * rofe Ping timeout: 480 seconds 1242737738 M * nkukard Bertl_oO, daniel_hozac using latest util-vserver and latest development I have still got the same problem with cgroup giving operation not permitted, but creating cgroups from the cmdline is working fine 1242737777 M * daniel_hozac what's your configuration? 1242737875 M * nkukard mkdir /dev/cgroup ; mount -t cgroup none /dev/cgroup <= making groups and all works nicely 1242737880 M * nkukard then in /etc/vservers 1242737910 M * nkukard wug-server/cgroup/name cpus mems , name is wug-server , cpus is 0 , mems is 0 1242737932 M * nkukard i can create that myself, but if util-vserver tries to create it, I'm getting operation not permitted, I did a bit of debugging earlier with Bertl_oO 1242738051 J * xdr ~xdr@gote2.244.cust.blixtvik.net 1242738385 Q * Keeper Remote host closed the connection 1242738468 M * nkukard daniel_hozac, its rather odd .... if i do the mkdir /dev/cgroup/wug-server from commandline it works 1242738492 M * nkukard if i do that EXACT same command in vserver.functions I get operation not permitted 1242738523 M * nkukard I added my own mkdir without the redirect to devnull between line 1477 and line 1478 1242738747 M * daniel_hozac and you're sure it's the mkdir that's failing? 1242738760 M * daniel_hozac also, you don't have /cpuset, do you? 1242738781 M * nkukard one sec, I'll add back bertl's debug info 1242738811 M * nkukard http://paste.linuxassist.net/214877 <= there is the debug I added 1242738855 M * nkukard http://paste.linuxassist.net/214876 <= there is the output and strace of mkdir 1242738864 M * nkukard # 1242738865 M * nkukard mkdir("/dev/cgroup/wug-server", 0777) = -1 EPERM (Operation not permitted) 1242738907 M * nkukard i have an empty /etc/vservers/wug-server/cpuset/ directory, and I have cpus, mems in /etc/vservers/wug-server/cgroup/ 1242738956 M * nkukard using util-vserver-0.30.216-pre2833.tar.bz2 1242739052 M * nkukard but ..... mkdir /dev/cgroup/wug-server from commandline is fine 1242739065 M * nkukard no error, and the cgroup is created 1242739074 M * daniel_hozac where is the --debug output? 1242739129 M * daniel_hozac you need cpuset.cpus and cpuset.mems, btw. 1242739144 M * daniel_hozac though you don't actually need mems if you don't want to restrict the NUMA zones. 1242739225 M * daniel_hozac also, your kernel doesn't have CONFIG_CGROUP_NS, does it? 1242739236 M * nkukard so I must have cgroup/ and cpuset/ with the same files in both? 1242739273 M * daniel_hozac no. 1242739276 M * daniel_hozac remove cpuset entirely. 1242739283 M * daniel_hozac it is an either-or thing. 1242739355 M * nkukard i have cgroup/cpus and cpuset/mems now 1242739362 M * nkukard busy checking CGROUP_NS 1242739540 M * nkukard daniel_hozac, i do have CONFIG_CGROUP_NS 1242739552 M * nkukard and cgroups are working perfectly fine on the host system 1242739563 M * nkukard vserver cgroups isn't :) 1242739744 M * nkukard daniel_hozac, the --debug shows a failure to echo $$ to the tasks file, it will because /dev/cgroup/wug-server/ doesn't exist as the mkdir fails 1242739750 M * nkukard if i unmount /dev/cgroup 1242739761 M * nkukard and start up the vserver, it creates /dev/cgroup/wug-server/tasks 1242739767 M * nkukard and contains a pid 1242739795 M * nkukard but if I mount /dev/cgroup mount -t cgroup none /dev/cgroup I can do it from the command prompt fine, but vserver.functions gets -EPERM 1242739808 M * nkukard no selinux running, no grsecurity or anything funny 1242741256 Q * mrfree Quit: Leaving 1242742152 M * nkukard daniel_hozac, very odd, huh? 1242742279 M * daniel_hozac no, entirely expected. 1242742284 M * daniel_hozac disable CONFIG_CGROUP_NS. 1242742312 N * Bertl_oO Bertl 1242742316 M * Bertl back now ... 1242742333 M * Bertl daniel_hozac: we probably should block that for now (in the kernel config) no? 1242742369 M * daniel_hozac yeah, i think so. 1242742378 M * nkukard what does CONFIG_CGROUP_NS do? 1242742412 M * daniel_hozac creates a cgroup every time you clone/unshare any namespace. 1242742473 M * nkukard so I must disable it and it hopefully solves my problems? :) 1242742483 M * daniel_hozac yes. 1242742629 M * nkukard thanks daniel_hozac Bertl :) 1242742634 M * nkukard i'll try it out without that shortly 1242745034 J * dowdle ~dowdle@scott.coe.montana.edu 1242745088 M * Bertl daniel_hozac: trying to build util-vserver (0.30.216-0.pre2833), I get this (on a newly installed mandriva 2009.1 system): http://paste.linux-vserver.org/12928 1242745103 M * Bertl any clues what I'm missing or is mandriva broken somehow? 1242745200 Q * sharkjaw Remote host closed the connection 1242745356 M * nkukard Bertl, libtool maybe? 1242745373 M * nkukard when I get errors like that I libtoolize, aclocal, autoconf, automake 1242745403 M * nkukard actually had that about 2hr ago with graphviz :) 1242745543 M * Bertl hmm, yes, libtool is new (2.2.6) maybe something in the utils need to be adjusted? 1242745564 M * nkukard i'd libtoolize, aclocal, autoconf, automake and try again :) 1242745642 M * Bertl automake bails out horribly 1242745757 M * nkukard :( 1242745789 M * bzed .oO(migrate to cmake)Oo. 1242745807 M * nkukard (assuming it uses automake ... sorry) 1242746630 Q * Kamping_Kaiser Read error: Operation timed out 1242746706 Q * kir Quit: Leaving. 1242747228 M * mnemoc Bertl: I just found that in 2.6.27+2.3 the ppid of process 1 is 1, but in older (2.6.17 for example) it's 0 1242747240 M * mnemoc Bertl: is this a kernel change of vserver related? 1242747276 M * mnemoc just tested on unpatched .27, it's vserver 1242747302 M * mnemoc bug or feature? :) 1242747394 J * Kamping_Kaiser ~kgoetz@ppp121-45-15-69.lns10.adl2.internode.on.net 1242747471 M * Bertl what 2.6.27 version? 1242747617 M * mnemoc 2.6.27.21-vs2.3.0.36.4-dist 1242747624 M * mnemoc s/-dist$// 1242747642 M * mnemoc known? 1242747716 M * _nono_ hi 1242747730 M * mnemoc hi _nono_ 1242747768 M * _nono_ is linux-vserver 2.6.22.19-vs2.3.0.34.1 (the latest in dhozac's yum repo) supposed to be stable? 1242747793 M * _nono_ I know it is a development version, but can I expect it to work in practice? 1242747796 M * mnemoc Bertl: in the host it's 1/0 as it should, but in the guests 1/1 1242747832 M * mnemoc Bertl: I assume it's because the guest's pid 1 is actually a child of host's pid 1 1242747869 M * Bertl known issue, should be fixed in 2.6.28+, will see if I can add it to 2.6.27 (I thought I already did :) 1242747898 M * mnemoc ^_^ 1242747924 M * mnemoc I should upgrade to the lastest .27.z anyway 1242747946 M * Bertl yep, checking with that would be helpful 1242747994 M * Bertl _nono_: most devel versions are pretty stable (not to say rock solid :), but if you worry about that, get the stable vs2.2 branch 1242748034 M * _nono_ well, I guess pretty stable is enough if it brings me IPv6 ;) 1242748392 M * daniel_hozac i've been running it in production since it was built, over a year ago. 1242748414 J * imcsk8 ~ichavero@148.229.1.11 1242748431 M * pmjdebruijn Bertl: I've been doing some very preliminary tests with 2.6.27.23-vs2.3.0.36.{4|5} and things look quite encouraging thus far 1242748579 Q * dowdle Remote host closed the connection 1242748611 M * Bertl pmjdebruijn: good to hear! 1242748645 J * dowdle ~dowdle@scott.coe.montana.edu 1242748673 M * micah erp, I got a weird situations with two guests 1242748693 M * micah i tried to start one with the same context as the other, and it was refused, so I changed the context 1242748700 M * micah and started it, but now both of them are running with no name 1242748716 M * pmjdebruijn Bertl: I'll be keeping you up to date 1242748731 M * pmjdebruijn though what problems could the pid1 issue implicate? 1242748735 M * Bertl micah: most likely one removed the back link for the other 1242748741 M * pmjdebruijn thus far I've only tested with a single vserver... 1242748760 M * Bertl pmjdebruijn: do you see it on your kernels? 1242748776 M * micah Bertl: which backlink is that? 1242748799 M * pmjdebruijn Bertl: I'll have to check tomorrow 1242748805 A * pmjdebruijn is home now 1242748807 M * micah yeah, removed the /var/run/ 1242748857 M * Bertl thos ein /var/run/vservers* 1242748883 M * micah serves me right :) 1242748956 M * micah (for doing something silly) 1242749177 J * hparker ~hparker@2001:470:1f0f:32c:290:96ff:fe50:40fa 1242749210 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1242749589 M * ben00001 thanks for the assistance again today guys 1242749594 M * ben00001 home time ... ;) 1242749602 M * Bertl you're welcome! 1242749618 M * ben00001 if you think of any quick and easy performance fixes email me :) 1242750496 Q * harobed Ping timeout: 480 seconds 1242750812 M * ghislainocfs2 ben: trhow 1000k€ at the hardware, quick and efficient 1242750949 M * ghislainocfs2 ben: just kidding;) 1242752506 Q * geos_one Ping timeout: 480 seconds 1242753088 Q * nou Ping timeout: 480 seconds 1242753117 J * nou Chaton@2001:6f8:328:bbc:6666:6667:: 1242753288 J * geos_one ~chatzilla@chello084115149052.4.graz.surfer.at 1242753710 Q * Kamping_Kaiser Ping timeout: 480 seconds 1242753766 J * ktwilight_ ~ktwilight@45.255-64-87.adsl-dyn.isp.belgacom.be 1242753996 Q * ktwilight Ping timeout: 480 seconds 1242754231 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1242754356 J * Kamping_Kaiser ~kgoetz@ppp121-45-95-170.lns10.adl6.internode.on.net 1242755511 Q * thierryp Ping timeout: 480 seconds 1242755759 J * doener_ ~doener@i59F55158.versanet.de 1242755861 Q * doener Ping timeout: 480 seconds 1242756274 Q * gnuk Quit: NoFeature 1242758621 M * mnemoc Bertl: 2.6.27.23-vs2.3.0.36.5 and still 1/1 :( 1242758800 J * Medivh ck@dolphin.serverbox.de 1242760209 Q * uva Quit: Leaving 1242761946 Q * geb Remote host closed the connection 1242762279 J * Chlorek ~cokolwiek@c.sed.pl 1242763612 M * Chlorek yes 1242763615 M * Chlorek hi there 1242763621 M * Chlorek i was found some bug 1242763631 M * Chlorek VIRT_LOAD not work 1242763913 M * nkukard working, thanks daniel_hozac , Bertl .... i definitely think that the vserver config should disalbe CONFIG_CGROUP_NS 1242763979 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net 1242764301 J * floopsy ~baihu@65.48.182.150 1242764518 M * Bertl Chlorek: please elabporate? 1242764562 M * Chlorek elab..what? 1242764630 M * Bertl kernel version/ patch version, how you tested, what ydoesn't work? 1242764662 Q * davidkarban Quit: Ex-Chat 1242764696 Q * floopsie Ping timeout: 480 seconds 1242764700 M * Chlorek yes 1242764718 M * Chlorek kernel ver 2.6.29, patch 1242764719 M * Chlorek wait 1242764842 M * Chlorek 2.6.29.2-vs2.3.0.36.12 1242764887 M * Chlorek i was started vserver and compilation on host 1242765609 M * Bertl ah, indeed, seems I didn't port that part to 2.6.27+ 1242765619 J * cga ~weechat@82.84.130.131 1242765672 Q * imcsk8 Quit: This computer has gone to sleep 1242765924 J * uva bno@118-168-234-171.dynamic.hinet.net 1242765936 M * Bertl daniel_hozac: okay, building a recent util-vserver on a recent mandriva 2009.1 with your spec file fails horribly .. note that they updated the build system and a bunch of tools including libtool which is now 2.2.6 1242766574 J * imcsk8 ~ichavero@nat.ti.uach.mx 1242766989 Q * bonbons Quit: Leaving 1242767091 J * grobie ~grobie@tyr.schnuckelig.eu 1242768720 Q * dna Quit: Verlassend 1242768956 Q * geos_one Quit: ChatZilla 0.9.84 [Firefox 3.0.10/2009050120] 1242769835 J * geos_one ~chatzilla@chello084115149052.4.graz.surfer.at 1242769881 Q * geos_one 1242770648 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1242771389 J * thierryp ~thierry@home.parmentelat.net 1242772051 Q * harobed Ping timeout: 480 seconds 1242772421 Q * scientes Ping timeout: 480 seconds 1242772446 Q * ard Ping timeout: 480 seconds 1242773367 J * cga ~weechat@194.244.1.164 1242775012 Q * Chlorek Quit: Changing server 1242775012 J * Chlorek ~cokolwiek@c.sed.pl 1242775058 Q * imcsk8 Quit: This computer has gone to sleep 1242775199 Q * thierryp Remote host closed the connection 1242775668 Q * doener_ Ping timeout: 480 seconds 1242775708 Q * opuk Server closed connection 1242775714 J * opuk ~kupo@pipe.intertubez.net 1242776491 Q * Piet Remote host closed the connection 1242776551 J * Piet ~piet@tor-irc.dnsbl.oftc.net 1242776938 J * saulus_ ~saulus@c207192.adsl.hansenet.de 1242777350 Q * saulus Ping timeout: 480 seconds 1242777352 N * saulus_ SauLus