1265242028 Q * bzed Read error: Connection reset by peer 1265242032 J * bzed ~bzed@devel.recluse.de 1265243464 Q * imcsk8 Quit: Leaving 1265245827 J * fosco fosco@91.208.40.1 1265246048 Q * dowdle Remote host closed the connection 1265247832 Q * gdm_ Server closed connection 1265247835 J * gdm ~gdm@pistol.redetoile.net 1265248743 N * Bertl_zZ Bertl 1265248747 M * Bertl back now ... 1265248785 M * fzylogic I'm about to leave for the day. It looks like cache info in general is hidden from the guests, though. 1265248800 M * fzylogic /proc/meminfo shows none in use 1265248825 M * fzylogic as a stop-gap, I patched our kernels to not include cache in the "used memory" calculation 1265249243 M * Bertl yes, we had caches and buffers set to 0, as they were not accounted 1265249265 M * Bertl what I'm curious is, you said you have per guest accounting info for that? 1265249326 M * fzylogic yeah, cgroups track it 1265249336 M * fzylogic looks to be correctly limited to the guest context 1265249353 M * Bertl well, then it shouldn't be too hard to show that info 1265249372 M * fzylogic that's what I figured. I just got a little lost figuring out where userspace got the info from :) 1265249372 M * Bertl have to dig into that, as all I saw was the rss and mlock limits 1265249388 M * fzylogic right 1265249400 M * Bertl you see the info 'on the host' in the cgroups dir, yes? 1265249407 M * fzylogic yes 1265249476 M * Bertl what does 'grep CGROUP .config' give for your kernel? (pastebin) 1265249523 M * fzylogic http://paste.linux-vserver.org/14228 1265249643 M * fzylogic need to run. I'll be back in the morning. 1265249646 Q * fzylogic Quit: fzylogic 1265249652 M * Bertl tx, will look into it and provide a patch to test with .. 1265250856 Q * micah Server closed connection 1265250858 J * micah ~micah@micah.riseup.net 1265250921 Q * derjohn_mob Ping timeout: 480 seconds 1265251465 J * derjohn_mob ~aj@c135091.adsl.hansenet.de 1265251528 Q * Guest2852 Server closed connection 1265251564 J * Genghis ~Genghis@ph34r.my.d-n-s.org.uk 1265251586 N * Genghis Guest698 1265253480 J * takeru ~takeru@nttkyo941188.tkyo.nt.ftth.ppp.infoweb.ne.jp 1265253964 Q * takeru Ping timeout: 480 seconds 1265255172 Q * ^Willie^ Read error: Connection reset by peer 1265261017 Q * balbir Read error: Connection reset by peer 1265261159 Q * infowolfe Server closed connection 1265261173 J * infowolfe ~infowolfe@c-71-236-152-35.hsd1.or.comcast.net 1265263050 J * petzsch ~markus@dslb-188-103-057-193.pools.arcor-ip.net 1265264080 Q * niki Quit: Leaving 1265264450 P * petzsch 1265264579 J * balbir ~balbir@122.248.161.59 1265265815 J * ghislain ~AQUEOS@adsl2.aqueos.com 1265266931 J * fzylogic ~fzylogic@wireless.fzylogic.net 1265267012 M * fzylogic Bertl: ugly, but I think it does what we need: http://karategerbil.com/kernel_debug/memdisplay.patch 1265267047 M * fzylogic there's probably a more direct way to get cgroup stats than mem_cgroup_get_local_stat though 1265267067 Q * fzylogic 1265267300 M * Bertl heh, yeah, there is :) 1265267339 J * sharkjaw ~gab@90.149.121.45 1265268424 Q * derjohn_mob Ping timeout: 480 seconds 1265270024 Q * SpComb Remote host closed the connection 1265270556 M * ghislain hello there 1265270584 M * ghislain bertl: i don't know if you tried to upgrade a host to squeeze 1265270624 M * ghislain bertl: but this fails from everywhere for me, insserv seems to do the requirement dance and fails on every hardware related services even if they are not enabled :( 1265270637 M * ghislain bertl: insserv: Service mountkernfs has to be enabled to start service networking 1265270654 M * ghislain and so on... dozen of messages like this one, this is just for information 1265270932 Q * ktwilight__ Remote host closed the connection 1265271002 J * ktwilight__ ~keliew@158.92-136-217.adsl-dyn.isp.belgacom.be 1265271036 M * Bertl ghislain: yeah, sounds like a debian issue to me 1265271090 M * Bertl read: please contact micah and/or the debian folks for possible solutions 1265271111 M * Bertl off for now ... bbl 1265271115 N * Bertl Bertl_oO 1265272850 J * ncopa ~ncopa@245.39.189.109.customer.cdi.no 1265275150 J * thierryp ~thierry@zankai.inria.fr 1265275465 Q * ncopa Quit: Ex-Chat 1265275640 Q * thierryp Quit: ciao folks 1265275654 J * thierryp ~thierry@zankai.inria.fr 1265275654 J * ncopa ~ncopa@245.39.189.109.customer.cdi.no 1265277202 J * kir ~kir@swsoft-msk-nat.sw.ru 1265277436 M * ghislain bertl: oh yes i know :) this was just for information as squeeze is comming (beware you mortals ! ) 1265279312 J * dna ~dna@170-198-103-86.dynamic.dsl.tng.de 1265279701 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1265280492 J * barismetin ~barismeti@zanzibar.inria.fr 1265280518 Q * barismetin 1265280530 J * barismetin ~barismeti@zanzibar.inria.fr 1265283384 Q * kir Ping timeout: 480 seconds 1265283783 J * kir ~kir@swsoft-msk-nat.sw.ru 1265284225 Q * dna Quit: Verlassend 1265285842 Q * fosco Quit: leaving 1265285947 J * fosco fosco@marx.wirefull.org 1265287231 Q * balbir Ping timeout: 480 seconds 1265287790 J * franklin_ ~quassel@95.108.8.74 1265287903 N * franklin_ franlkin1 1265288551 M * franlkin1 hi 1265288577 M * franlkin1 I've got a problem with barrier flag, can anuone help me with this? 1265288747 J * swen ~quassel@217.72.66.253 1265288755 M * franlkin1 from time to time this flag apperas inside vservers trees and generates log: Feb 3 10:46:38 xx kernel: vxW: [�php�,24801:#40000|40000|40000] did hit the barrier 1265288757 M * swen hi folks! 1265288782 M * swen franlkin1: do you use rlimits? 1265288796 M * franlkin1 swen: no I dont 1265288856 M * swen any hits here? cat /proc/virtual/{XID}/limit 1265288862 M * swen last column 1265288883 M * swen I have another question 1265288911 M * swen can I assign multiple IPs to same interface ? 1265288923 M * swen to a guest 1265288933 M * Bertl_oO franlkin1: kernel/util-vserver version? 1265288978 M * Mr_Smoke swen: yes :) 1265288984 M * Bertl_oO swen: you can assign multiple IPs to a single guest, and Linux (on the host) allows you to assign multiple IPs to the same interface 1265289026 M * Mr_Smoke Funny though, unless you use ifconfig-style aliases, some applications don't seem to be aware of additional IPs 1265289033 M * swen Bertl_oO: yes I know that. I guess I didn't ask right :\ 1265289035 M * Mr_Smoke e.g. djbdns setup script 1265289059 M * swen How do I permanently configure multtple IPs to same guest. 1265289071 M * swen Do I just copy inderface/0 entry? 1265289081 M * swen and change ip? 1265289084 M * Bertl_oO just make another sub-dir in /etc/vservers//interfaces 1265289093 M * swen oh 1265289102 M * Bertl_oO like .../interfaces/1/ 1265289112 M * swen it's just not that intuitive 1265289126 M * Bertl_oO they are processed in alphanumerical order, see flower page 1265289239 M * swen yes I understand that. It's just matter of semantics. You don't expose "interface" - just in IP. If you'd exposed interface you wouldn't need to expose same interface twice :D 1265289262 M * swen thanks for help! 1265289304 M * swen Support on this channel is extraortinary! I just wish I was in "the right" timezone :D 1265289438 M * Bertl_oO hmm, can't follow :) 1265289592 M * franlkin1 Bertl_oO: util-vserver 0.30.216~r2772-6/ kernel 2.6.26-2 - both from debian lenny 1265289633 M * Bertl_oO http://linux-vserver.org/Installation_on_Debian#Issues_with_the_current_2.6.26_Kernel 1265289648 M * swen Bertl_oO: It happens that all of Vserver gurus are offline 1265289649 M * Bertl_oO i.e. lieve with it, or get a working kernel 1265289670 M * Bertl_oO swen: is that so? when does that happen? 1265289701 M * swen when I come to work in the morning. Im on CET 1265289707 M * Bertl_oO swen: and I was more referring to the interface part 1265289730 M * swen oh 1265289775 M * swen I was just commenting on implementation :D 1265289851 M * Bertl_oO yeah, but what was it what you were trying to communicate? :) 1265289877 M * swen oh! 1265289910 M * swen that you name directory "interfaces", but in fact you dont expose interface - merely an IP 1265289945 M * Bertl_oO ah, got it, well, that name is kind of legacy 1265289973 M * Bertl_oO not sure there is a better name for that though 1265289999 M * Bertl_oO after all you specify ip,mask/prefix and device 1265290023 M * Bertl_oO (where device is what we usually refer to as interface) 1265290090 M * franlkin1 Bertl_oO: thanks for a link but I dont see solution for my problem. barrier attribute seems to apear randomly on working vserver during its operation 1265290120 M * franlkin1 Bertl_oO: it may wort 2 weeks or so without problems 1265290161 M * franlkin1 Bertl_oO: and after that barrier suddenly appears 1265290162 M * Bertl_oO the solution is to ditch the known-broken kernel and old util-vserver 1265290224 M * franlkin1 Bertl_oO: sp you think it's a problem with broken/old utils/kernel? 1265290348 M * Bertl_oO what else would cause barriers to appear inside your guest? 1265290413 M * Bertl_oO unless you have some cron script or control frontend which might be buggy (and be able to do that) I'd opt for the kernel 1265290832 M * franlkin1 Bertl_oO: have you ever heard of problems like that before? 1265290929 M * Bertl_oO you see the list of issues with that specific kernel? you see that the attributes are messed up? 1265291029 J * hparker ~hparker@linux.homershut.net 1265291112 M * swen Bertl_oO: one more quertion. Is Karmic kernel also broken? 1265291134 M * Bertl_oO no idea ... 1265291216 M * swen Bertl_oO: tnx 1265291277 M * franlkin1 Bertl_oO: yes but in different way 1265291546 M * swen franlkin1: can you tell more? 1265291674 J * bobnormal ~irc@87-194-32-179.bethere.co.uk 1265293029 M * franlkin1 swen: I've got a problem with secure barrier, it seems that it randomly appears in vservers tree 1265293109 M * swen franlkin1: I was refering to Karmic kernel being broken statement 1265293801 Q * sharkjaw Remote host closed the connection 1265294229 Q * franlkin1 Remote host closed the connection 1265294744 J * balbir ~balbir@122.172.58.41 1265296426 M * bobnormal whats the command to execute a command under a specific context again, and does it take vserver names instead of context ids? 1265296959 M * mnemoc chcontext 1265297045 M * bobnormal cheers 1265297835 Q * biz Quit: Reconnecting 1265297836 J * biz ~biz@node2.cluster1.pyrox.eu 1265298186 J * urbee ~urbee@93-103-199-233.dynamic.dsl.t-2.net 1265298188 M * urbee hoi 1265298208 M * urbee anyone got a clue, using openvcp, is a guest able to make a reinstall of his VPS ? 1265298408 M * urbee or, install some other distro 1265298589 N * DoberMann[PullA] DoberMann 1265299033 Q * opuk Ping timeout: 480 seconds 1265299254 J * opuk ~kupo@pipe.intertubez.net 1265299482 Q * Chlorek Quit: - 1265299555 J * Chlorek cokolwiek@c.sed.pl 1265299776 M * ghislain urbee: i don't know but i hope not this will be a "do an error and smash you in the face button" that no customer will misss and create payhem 1265299780 M * ghislain mayhem 1265299853 Q * hparker Quit: Quit 1265300713 Q * thierryp Quit: ciao folks 1265300732 J * dowdle ~dowdle@scott.coe.montana.edu 1265300891 J * niki ~niki@94.145.207.11 1265301329 J * petzsch ~markus@dslb-188-103-057-193.pools.arcor-ip.net 1265302046 Q * balbir Ping timeout: 480 seconds 1265302643 Q * ncopa Quit: Ex-Chat 1265302779 J * balbir ~balbir@122.172.108.117 1265303151 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1265303444 J * derjohn_mob ~aj@c189119.adsl.hansenet.de 1265305076 Q * barismetin Quit: Leaving... 1265305931 P * kir Leaving. 1265306153 J * fzylogic ~fzylogic@dsl081-243-128.sfo1.dsl.speakeasy.net 1265306355 Q * bobnormal Remote host closed the connection 1265307003 N * DoberMann DoberMann[PullA] 1265310186 Q * gnuk Quit: NoFeature 1265312092 Q * ghislain Quit: Leaving. 1265312683 Q * gdm Quit: leaving 1265314377 J * imcsk8 ~ichavero@148.229.1.11 1265314596 N * Bertl_oO Bertl 1265314603 M * Bertl back now ... 1265314609 M * petzsch wb 1265314620 J * ktwilight ~keliew@254.63-240-81.adsl-dyn.isp.belgacom.be 1265314844 Q * ktwilight__ Ping timeout: 480 seconds 1265315909 J * hijacker ~hijacker@87-126-142-51.btc-net.bg 1265316204 M * fzylogic Bertl: any chance you caught the PoC patch I posted last night? 1265316215 M * fzylogic seems to do the trick for reporting memory info 1265316264 M * Bertl yeah, I saw it, but it is somewhat problematic 1265316279 M * Bertl will provide one without the drawbacks 1265316288 M * fzylogic yeah, I figured as much :) 1265318198 J * mugwump ~samv@watts.utsl.gen.nz 1265319568 Q * bonbons Quit: Leaving 1265320354 J * yarihm ~yarihm@77-58-27-9.dclient.hispeed.ch 1265320465 Q * hijacker Quit: Leaving 1265320594 Q * derjohn_mob Ping timeout: 480 seconds 1265321163 J * derjohn_mob ~aj@c189076.adsl.hansenet.de 1265322642 Q * arekm Quit: leaving 1265322925 Q * imcsk8 Quit: Leaving 1265323042 J * arekm arekm@carme.pld-linux.org 1265323071 Q * arekm 1265323081 J * arekm arekm@carme.pld-linux.org 1265323087 Q * arekm 1265323113 J * arekm arekm@carme.pld-linux.org 1265323527 J * imcsk8 ~ichavero@148.229.1.11 1265323751 M * Bertl fzylogic: should have a patch to test with in a few minutes 1265323761 M * fzylogic alright, cool 1265323785 M * Bertl took a little longer as I had to sort out some stuff first 1265324210 M * Bertl http://vserver.13thfloor.at/ExperimentalT/delta-memcg-info03.diff ontop of info02 should do the trick 1265324239 M * Bertl hmm, info02 with s/10/PAGE_SHIFT/ 1265324402 M * Bertl (i.e. the last status we had + delta-memcg-info03) 1265324658 M * fzylogic ok 1265324773 M * petzsch good night 1265324783 M * Bertl cyz 1265324790 M * Bertl cya even 1265324798 Q * petzsch Quit: Leaving. 1265326686 Q * imcsk8 Quit: Leaving 1265327254 M * fzylogic Bertl: is that patch supposed to be showing cache usage in the guests? 1265327293 M * fzylogic hmm..looks like it is 1265327302 M * fzylogic is supposed to, that is 1265327305 M * fzylogic it isn't working 1265327344 M * Bertl hmm, what do you get instead? 1265327350 M * fzylogic 0 1265327354 M * fzylogic does virt_mem need to be disabled? 1265327365 M * fzylogic ahh, yes 1265327370 M * fzylogic it works with virt_mem disabled 1265327402 M * fzylogic except that's displaying the system-wide cache usage 1265327436 M * Bertl that is actually fine, the question is, what does the memory.stats show? 1265327447 M * Bertl i.e. does the cgroup account cached memory? 1265327491 M * fzylogic sure does 1265327495 M * fzylogic cache 815104 1265327527 M * Bertl okay, sec 1265327672 M * Bertl and you get 0 there, right? 1265327692 M * fzylogic yeah 1265327708 M * Bertl okay, I guess we have to add a debug printk to get more info 1265327711 M * fzylogic shared, cached, buffers all == 0 1265327938 M * Bertl so, how I see it, we have three possible causes for that 0 1265327952 M * Bertl a) vx_vsi_cached() returns 0 1265327968 M * Bertl b) vx_vsi_cached() returns < 0