1268698552 M * neofutur jpic: using environment variables ? 1268698624 M * neofutur http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1268698634 M * neofutur /etc/vservers/.defaults/apps/init 1268698643 M * neofutur The environment to set when starting guests. Contains one VAR=VAL pair per line. 1268698648 M * neofutur could be simple ? 1268698780 M * jpic absolutely, thanks neofutur ! 1268702057 Q * imcsk8 Quit: This computer has gone to sleep 1268706252 M * jpic neofutur: how to use it exactly? apparently /etc/init.d/local doesn't inherit from this env ... 1268708123 J * balbir ~balbir@122.172.50.39 1268709091 M * neofutur http://linux-vserver.org/Frequently_Asked_Questions#What_is_the_initial_PATH.3F 1268709097 M * neofutur this _should_ work ! 1268709122 M * daniel_hozac what initstyle? 1268709188 M * neofutur its gentoo host and guests 1268709205 M * daniel_hozac then no 1268709207 M * daniel_hozac that won't work. 1268709219 M * neofutur how could we nake this work ? 1268709227 M * neofutur give environment variables to a host at restart ? 1268709228 M * daniel_hozac you want environmental variables? 1268709231 M * neofutur yes 1268709248 M * daniel_hozac you do realize they cannot be changed, right? 1268709251 M * neofutur different for each vserver, typically to have different backup times for all the vservers 1268709255 M * neofutur yes 1268709257 M * neofutur unchanged is ok 1268709463 M * neofutur could be something else than environment variables , but env seems percfect 1268709471 M * neofutur having in the vserver 1268709479 M * neofutur echo $BACKUP_TIME 1268709480 M * neofutur 13 1268709485 M * neofutur for example 1268709516 M * neofutur but : 1268709518 M * neofutur cat /etc/vservers/master_neofutur/apps/init/environment 1268709530 M * neofutur BACKUPHOUR=13 1268709543 M * neofutur does nothing in the vserver conext after restarting it 1268709580 M * neofutur hehe daniel_hozac your the author of the FAQ : http://linux-vserver.org/Frequently_Asked_Questions#What_is_the_initial_PATH.3F 1268709592 A * neofutur more optimistic now ;) 1268709646 M * daniel_hozac hmm, actually, i misremembered. 1268709661 M * daniel_hozac it's only for stop that the entire environment is cleaned for Gentoo guests. 1268709669 M * daniel_hozac so it should actually work. 1268709677 M * neofutur perhaps I dont understand the FAQ or the use of /etc/vservers/vservername/apps/init/environment ? 1268709745 M * daniel_hozac what does vserver --debug start show? 1268709754 M * neofutur i go test it 1268709868 M * neofutur thousands of lines 1268709873 M * daniel_hozac yeah 1268709878 M * neofutur you want a pastebin ? full or grep something ? 1268709879 M * daniel_hozac pastebin it 1268709932 M * neofutur ok i did 1268709934 M * neofutur vserver --debug master_neofutur restart 1>a 2>b 1268709944 M * neofutur i ll pastebin a and b separately 1268710010 M * neofutur pastebin cant accept 117 kb paste 1268710019 M * neofutur wait a minute 1268710052 M * neofutur http://dpaste.com/172392/ 1268710060 M * neofutur ( wgetpaste saved me again ;) 1268710078 M * neofutur oups wait 1268710113 M * neofutur yes b is too big even for wget paste, i ll give you an url 1268710207 M * neofutur http://quaoar.ww7.be/b 1268710210 M * neofutur here it is ! 1268710276 M * neofutur ( You have to run the vprocunhide command after every reboot in order to setup /proc permissions correctly for vserver use. An init script has been installed by util-vserver. To use it you should add it to a runlevel: <-- couñld this be related ? ) 1268710303 M * daniel_hozac no 1268710325 M * neofutur i can see /etc/vservers/master_neofutur /bin/env -i BACKUPHOUR=13 1268710331 M * neofutur in the output 1268710365 M * neofutur but sonthing in env after vserver restart 1268710371 M * neofutur nothing 1268710376 M * daniel_hozac how do you determine that? 1268710401 M * neofutur i tried echo $BA+tab 1268710406 M * neofutur end env | grep BA 1268710429 M * neofutur a better way to get it ? 1268710433 M * daniel_hozac where? 1268710437 M * daniel_hozac from an initscript? 1268710470 M * neofutur in a shell as user 1268710477 M * neofutur ok its true only in the initscripts ;) 1268710495 M * neofutur true ? 1268710506 M * neofutur ill ad an echo in the rc.local 1268710639 M * neofutur /etc/conf.d/local.start for gentoo :p 1268710680 M * daniel_hozac it's the init environment 1268710682 M * daniel_hozac nothing else 1268710851 M * neofutur jpic: here ? 1268710929 M * neofutur you understand what we need daniel_hozac ? do you see a better mean to have it work ? 1268710950 M * daniel_hozac why don't you just set it in the config file? 1268710951 M * neofutur some kinf of config from the host , different for each vserver 1268710969 M * neofutur i ll have a look at this 1268710969 M * daniel_hozac i don't see how /etc/vservers//apps/init/environment is any better than /etc/crontab in the guest. 1268711286 M * jpic because each vserver crontab needs ~30 variables, which we want to maintain in one place 1268711335 M * neofutur i cant see it either from /etc/conf.d/local.start 1268711346 M * neofutur ho to get this env variable in a initscript ? 1268711660 M * neofutur perhaps /etc/vservers/vserver-name/scripts could be useful too ? 1268711665 M * daniel_hozac no 1268711743 M * neofutur daniel_hozac: how to get this env variable in a initscript ? env dont see it neither does echo $BACKUPHOUR 1268711760 M * neofutur ( in /etc/conf.d/local.start ) 1268711763 M * daniel_hozac what util-vserver version are you using? 1268711769 M * daniel_hozac what does your initscript do? 1268711803 M * neofutur [ebuild U ] sys-cluster/util-vserver-0.30.216_pre2864 [0.30.216_pre2849] USE="bash-completion 1268711861 M * neofutur I did this : http://pastebin.com/N8uUpjpt 1268711877 M * neofutur and it prints env when a i restart the vserver 1268711894 M * neofutur anything i should do another way ? 1268711998 M * neofutur the results of the vserver restart : http://pastebin.com/eSu0vHw2 1268712014 M * neofutur it prints env but no BACKUPHOUR 1268712045 J * SauLus_ ~SauLus@c193110.adsl.hansenet.de 1268712377 Q * SauLus Read error: Connection reset by peer 1268712377 N * SauLus_ SauLus 1268712709 M * neofutur daniel_hozac: we are using the vserver+grsec patch 1268712713 M * neofutur could be related ? 1268712733 M * daniel_hozac unlikely 1268712742 M * daniel_hozac look into your guests boot scripts 1268714076 J * MeCooL mecool@94.128.87.117 1268714137 Q * balbir Ping timeout: 480 seconds 1268714775 J * balbir ~balbir@122.172.54.102 1268715515 M * neofutur they re only default gentoo initscripts 1268715526 M * neofutur nothing custom afaik 1268715986 M * neofutur in the host /etc/vservers/vserver-nam/cache is pointing to nothing 1268716028 M * neofutur /etc/vservers/.defaults/cachebase/ is not existing 1268716033 M * neofutur could this be a problem ? 1268716179 M * daniel_hozac no 1268716189 M * daniel_hozac well, it'll be a problem if you use disk limits 1268716195 M * daniel_hozac but not related to this issue 1268716268 M * neofutur ok 1268717475 Q * larsivi 1268717479 J * larsivi ~larsivi@188.113.74.106 1268717643 M * neofutur nothing special in our initscripts 1268717782 M * neofutur hum : 1268717785 M * neofutur env_whitelist:# /etc/conf.d/env_whitelist: Environment whitelist for rc-system 1268717785 M * neofutur env_whitelist:# Specify which variables are allowed to be passed from the environment to the 1268717788 M * neofutur env_whitelist:# rc-system. If it is not set by the environment, then the variable will be 1268717791 M * neofutur env_whitelist:# taken from /etc/profile.env - meaning, if you need to set LANG or such, 1268717794 M * neofutur env_whitelist:# do it in a /etc/env.d/99myownstuff file for example, and run env-update. 1268717797 M * neofutur could be this ;) 1268717909 A * neofutur tries echo BACKUPHOUR >> /etc/conf.d/env_whitelist 1268717978 J * sharkjaw ~gab@90.149.121.45 1268719005 M * neofutur daniel_hozac: echo if I put : 1268719006 M * neofutur sedna ~ # cat /etc/env.d/97mine 1268719006 M * neofutur BACKUPHOUR=$BACKUPHOUR 1268719021 M * neofutur i get a BACKUPHOUR env variable in the vserver after restart 1268719060 M * neofutur the the env var contains BACKUPHOUR not the value passed through /etc/vservers//apps/init/environment 1268719078 M * neofutur how to get this value in /etc/env.d/ script ? 1268719570 M * daniel_hozac ask #gentoo 1268719708 M * neofutur ok thanksĂ 1268720004 M * neofutur not easy to explain this . . . 1268720263 J * imcsk8 ~ichavero@evdomip-43-221.iusacell.net 1268720272 J * ghislain ~AQUEOS@adsl2.aqueos.com 1268720606 M * neofutur I tried to add it in /etc/conf.d/env_whitelist 1268720610 M * neofutur but no success 1268721403 J * ghislain1 ~AQUEOS@adsl2.aqueos.com 1268721516 Q * ghislain synthon.oftc.net charm.oftc.net 1268721516 Q * MeCooL synthon.oftc.net charm.oftc.net 1268721516 Q * FireEgl synthon.oftc.net charm.oftc.net 1268721516 Q * thalunil synthon.oftc.net charm.oftc.net 1268721580 J * ghislain ~AQUEOS@adsl2.aqueos.com 1268721580 J * MeCooL mecool@94.128.87.117 1268721580 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com 1268721582 Q * MeCooL Max SendQ exceeded 1268721622 Q * ghislain Ping timeout: 480 seconds 1268721632 Q * FireEgl Ping timeout: 480 seconds 1268721951 M * neofutur daniel_hozac: I got it ! 1268722043 M * neofutur /etc/rc.conf 1268722044 M * neofutur rc_env_allow="BACKUPHOUR" 1268722055 M * neofutur # By default we filter the environment for our running scripts. To allow other 1268722055 M * neofutur # variables through, add them here. Use a * to allow all variables through. 1268722055 M * neofutur rc_env_allow="BACKUPHOUR" 1268722232 M * neofutur now I got the value in /etc/conf.d/local 1268722251 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com 1268723264 Q * derjohn_mob Ping timeout: 480 seconds 1268724087 J * ntrs ~ntrs@77.29.0.126 1268724858 M * neofutur then echo in /etc/env.d/97myenv and then have it system wide 1268725641 Q * balbir Ping timeout: 480 seconds 1268725753 M * neofutur anyone here have an idea why linux-vserver chose to use openrc ? 1268725783 M * neofutur everything is so much harder with openrc / no docs / no answers on the internet . . . 1268725793 M * neofutur 07:06 < daniel_hozac> ask #gentoo 1268725806 M * neofutur #gentoo dont want to hear about openrc 1268726307 J * balbir ~balbir@122.172.51.69 1268728343 J * thalunil ~thalunil@walledcity.de 1268728365 Q * imcsk8 Quit: This computer has gone to sleep 1268728375 J * BenG ~bengreen@cpc2-aztw22-2-0-cust521.aztw.cable.virginmedia.com 1268728959 J * MeCooL mecool@94.128.87.117 1268729092 Q * balbir Ping timeout: 480 seconds 1268729238 N * Bertl_oO Bertl 1268729243 M * Bertl morning folks! 1268729775 J * balbir ~balbir@122.172.105.32 1268730397 M * hijacker morning 1268730447 J * m_o_d ~kane@host.ltv.pl 1268730925 Q * balbir Read error: Connection reset by peer 1268731890 J * balbir ~balbir@122.172.108.191 1268732025 M * ghislain1 morning 1268732195 M * ghislain1 bertl: did you play with 2.6.33 in production, does the perf show the same issue you saw on 2.6.32 ? 1268732214 M * ghislain1 i am playing with it on a non critical server, for now all seems fine 1268732382 M * ghislain1 woha 2.6.33.1 is allready there, lol not finished compiling everything that a new kernel arise 1268733776 Q * balbir Ping timeout: 480 seconds 1268734128 M * BenG Bertl, the 2.6.33 vs2.3.0.36.30.1 patch, it has the umask patches already applied yes? 1268734212 M * BenG I tried to run google-chrome with that kernel and patch and I still get the same result 1268734441 J * balbir ~balbir@122.172.63.226 1268734461 M * BenG "Failed to move to new PID namespace: Operation not permitted" 1268734470 Q * MeCooL Ping timeout: 480 seconds 1268734607 Q * BenG Quit: I Leave 1268734704 M * karasz moin 1268734746 M * karasz is work on vcd and vstatd stopped? or moved somewhere else than http://svn.linux-vserver.org/projects ? 1268735146 M * biz neofutur: what has linux-vserver to do with openrc? It's all up to your gentoo guest, you could also use baselayout-1... but openrc+baselayout-2 combined with vserver-init-style gentoo is way better 1268735495 Q * ntrs Ping timeout: 480 seconds 1268737835 Q * ghislain1 Quit: Leaving. 1268737844 J * ghislain ~AQUEOS@adsl2.aqueos.com 1268738198 M * neofutur hum ok, so its just the co-admin of this server who the openrc bad choice ;( 1268738203 M * neofutur jpic: grrrrrrr ! 1268738297 J * ghislain1 ~AQUEOS@adsl2.aqueos.com 1268738507 Q * ghislain Ping timeout: 480 seconds 1268740380 M * karasz Hollow: could you please shed some light on vcd and vstatd status? 1268740392 M * jpic neofutur: baselayout-1 is useable, but needs to be patched after emerge 1268740404 M * jpic what's the problem with baselayout 2 btw? 1268740620 J * MooingLe1ur ~troy@shells195.pinchaser.com 1268740762 Q * MooingLemur Ping timeout: 480 seconds 1268742271 J * kir ~kir@swsoft-msk-nat.sw.ru 1268742416 M * Bertl jpic: should work fine, IIRC it is the intended layout 1268743684 Q * jrdnyquist Quit: Leaving 1268744284 J * jrdnyquist ~jrdnyquis@slayer.caro.net 1268744795 P * kir Leaving. 1268745633 J * MeCooL mecool@94.129.166.19 1268746020 Q * infowolfe Read error: Operation timed out 1268746223 J * infowolfe ~infowolfe@c-71-236-152-35.hsd1.or.comcast.net 1268746228 Q * quasisane Remote host closed the connection 1268746546 J * quasisane ~sanep@c-76-24-80-97.hsd1.nh.comcast.net 1268746674 J * barismetin ~barismeti@zanzibar.inria.fr 1268747193 Q * barismetin Remote host closed the connection 1268747209 J * barismetin ~barismeti@zanzibar.inria.fr 1268747438 J * yarihm ~yarihm@80.219.150.83 1268748462 M * Bertl ghislain1: not yet, but I should get to test 2.6.33 end of the week 1268748477 M * Bertl (in production, that is :) 1268748710 Q * MeCooL Ping timeout: 480 seconds 1268749166 J * ntrs ~ntrs@77.28.57.216 1268749394 M * ghislain1 bertl: will then be very interested by your feelings about it :) , i am preparing to switch all my box to a new kernel with recent patche so my first target is 2.6.33 (because it has cgroup disk io balancing) 1268749415 M * ghislain1 bertl: for now i switched only one box, no issue so far 1268749669 J * jrklein ~jrklein@2001:0:53aa:64c:0:4395:b97d:df43 1268749690 Q * jrklein 1268749800 M * Bertl sounds good, did you test the cgroup IO stuff? 1268750091 J * dowdle ~dowdle@scott.coe.montana.edu 1268751766 J * dna ~dna@31-216-103-86.dynamic.dsl.tng.de 1268752382 J * swen ~quassel@217.72.66.253 1268752442 M * ghislain1 bertl: not again but i will try it i guess this week end as customers do not let me any time in the week :( 1268752673 M * swen hi all! 1268752683 M * swen i have a question. 1268752744 M * ghislain1 this is quite common, i have some too, any one you would like to voice out ? ;p 1268752762 M * swen Is it possible to rename interface to be seen under a different name in guest? Example: I want vlan99 to be named eth0 in guest 1268752775 M * swen :D 1268752835 M * swen I've checked some documents about IP aliases but as far as I can tell it obsolete 1268752858 M * swen it's obsolete 1268752898 M * swen I've also tryed "echo eth0 > interfaces/1/name", but without success 1268752903 M * Bertl nope 1268752923 M * Bertl the 'name' part in interfaces is used for defining an alias 1268752942 M * Bertl i.e. dev=eth0, name=karli will give you eth0:karli 1268752964 M * swen i see 1268752976 M * Bertl but it would be rather trivial to do such a mapping, just that there is/was no point in doing so (yet) 1268752981 M * swen so I'm stuck with "fuggly" names :D 1268753002 M * Bertl yep, but there is no need to put the IP in vlan99 :) 1268753020 M * Bertl i.e. you could as well put it on eth0, which will show up as eth0 :) 1268753148 M * swen I know. But I have various (isolated) VLANs defined and I wanted standard interface names for guests (so vserver user could not tell he is in VLAN) 1268753188 J * derjohn_mob ~aj@213.238.45.2 1268753242 M * swen Anyway, thanks for the answer. Now I need not to look any further 1268753245 M * ghislain1 swen: sorry i could not resist 1268753261 M * swen ghislain1: no problem 1268753303 N * swen swen_away 1268755200 Q * sharkjaw Remote host closed the connection 1268755333 J * MeCooL mecool@94.129.151.93 1268755619 Q * ncopa Quit: Ex-Chat 1268755760 Q * dna Quit: Verlassend 1268757248 J * pdilung ~pdilung@b7.ivosoft.sb.cust.gts.sk 1268757252 M * pdilung hello all 1268757257 M * pdilung a question 1268757278 M * pdilung how to virtualize the memory in guest when cgroups is used? 1268757303 M * Bertl http://linux-vserver.org/util-vserver:Cgroups 1268757316 J * dna ~dna@31-216-103-86.dynamic.dsl.tng.de 1268757317 M * Bertl with recent kernels, just enable VIRT_MEM with cgroups enabled 1268757360 M * pdilung kernel vserver-sources-2.3.0.36.28 i.e. 2.6.32 + 2.3.0.36.28 1268757372 M * Bertl probably not recent enough 1268757380 M * pdilung Bertl: i did already, do not work 1268757387 M * Bertl get 2.6.33+latest patch and you should be fine 1268757404 M * pdilung bertl: uff :) 1268757417 M * pdilung bertl: however is the memory capped? 1268757447 M * pdilung bertl: i still can write a simple malloc in loop :) 1268757457 M * pdilung bertl: sorry for a stupid question ;) 1268757468 M * Bertl you can configure the memory limits via the cgroup config 1268757481 M * Bertl (see the page I pasted, scroll down to memory) 1268757493 M * pdilung bertl: I did already :) 1268757522 M * pdilung bertl: but the virt_mem does not make my guest see the limited resources 1268757533 M * Bertl as I said, your kernel is too old 1268757537 M * pdilung bertl: i do not care though unless it works 1268757553 M * pdilung bertl: so I write a simple allocator program and see what happens 1268757555 M * Bertl the cgroup memory limit should work in 2.6.32 as well 1268757572 M * Bertl just the VIRT_MEM info is not there 1268757596 M * pdilung bertl: this is what I wanted to read, real values in free is AFAIK a cosmetic problem 1268757775 J * bliz42 ~chatzilla@64.190.170.210 1268757930 M * bliz42 any reason the latest stable release is 2.2.0.7, from two years ago? I use newer builds, but catching some flack from the production guys because the newer builds aren't "stable" 1268758042 M * Guy- bliz42: ask them why a "stable" label is important and how it makes the code better 1268758082 M * Bertl bliz42: not enough resources and time to do extensive testing required for 'stable' 1268758095 M * Bertl bliz42: but we are slowly getting there 1268758101 M * bliz42 the same reason they pay a big license fee to redhat *sigh* 1268758106 M * bliz42 they don't listen 1268758218 M * bliz42 my dev box which goes under extensive updates and crazy code has better uptime than their "stable" production servers.. so I guess it's all moot :) 1268758352 M * pdilung Bertl: ok, seem to work 1268758363 M * pdilung Bertl: nice job Herbert :) 1268758375 M * Bertl thanks! have fn! 1268758378 M * Bertl *fun 1268758551 M * pdilung Bertl: it can also be changed on the fly, super :)))) 1268758573 M * Bertl yep, basically all limits can be adjusted at runtime 1268758579 M * pdilung Bertl: so my recurrent 10M alocator eats as much as I allow :) 1268758596 M * pdilung Bertl: I always knew vserver rocks 1268758631 M * pdilung Bertl: so do I have to use 2.6.33 for VIRT_MEM? 1268758643 M * pdilung Bertl: or latest vserver patch will do fine? 1268758984 M * Bertl 2.6.32 will work in the future, but atm, 2.6.33 is the oĂly one with virt mem support (for cgroups) 1268758989 M * Bertl *only 1268758993 J * imcsk8 ~ichavero@nat.ti.uach.mx 1268759278 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1268759403 Q * dna Quit: Verlassend 1268759421 Q * tokkee Remote host closed the connection 1268759428 J * tokkee tokkee@osprey.tokkee.org 1268760966 M * Bertl off for now ... bbl 1268760976 N * Bertl Bertl_oO 1268761017 J * hijacker_ ~hijacker@87-126-142-51.btc-net.bg 1268761992 Q * pdilung Quit: leaving 1268762640 J * jrdnyquist2 ~jrdnyquis@slayer.caro.net 1268762647 Q * jrdnyquist2 1268764147 Q * barismetin Quit: Leaving... 1268765282 Q * derjohn_mob Ping timeout: 480 seconds 1268765640 J * manana ~mayday090@84.17.25.149 1268767394 Q * bliz42 Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158] 1268767837 J * BenG ~bengreen@cpc2-aztw22-2-0-cust521.aztw.cable.virginmedia.com 1268768005 Q * balbir Ping timeout: 480 seconds 1268768678 J * balbir ~balbir@122.172.57.20 1268769384 Q * BenG Quit: I Leave 1268773501 J * derjohn_mob ~aj@d045113.adsl.hansenet.de 1268773542 J * pdilung ~dilung@p54975561.dip.t-dialin.net 1268773568 M * pdilung Bertl_oO: thnx for the pointer regarding cgroups and virt_mem 1268773646 M * pdilung 3.6.33 + vs2.3.0.36.30.3 shows memory settings in free and top properly 1268774150 J * ktwilight_ ~keliew@91.178.145.134 1268774150 Q * ktwilight Read error: Connection reset by peer 1268775021 M * pdilung Hollow: Hey, I made a new vserver ebuild and patchset see 309861 @ bugs.gentoo.org 1268775111 Q * bonbons Quit: Leaving 1268775141 Q * yarihm Quit: Leaving 1268775657 J * larsivi_ ~larsivi@188.113.74.106 1268775657 Q * larsivi Read error: Connection reset by peer 1268776457 Q * infowolfe Quit: Leaving 1268777731 Q * hijacker_ Quit: Leaving 1268778098 J * infowolfe ~infowolfe@c-71-236-152-35.hsd1.or.comcast.net 1268780006 J * smash_ smash@newton.uk.to 1268780008 Q * smash Read error: Connection reset by peer 1268780177 Q * ntrs Ping timeout: 480 seconds 1268780437 J * Mr_Smoke_ smokey@layla.lecoyote.org 1268780473 Q * Mr_Smoke Remote host closed the connection 1268781393 Q * wibble Remote host closed the connection 1268781408 J * wibble ~wibble@vortex.ukshells.co.uk 1268782279 Q * tokkee Ping timeout: 480 seconds 1268783204 M * Bertl_oO pdilung: no problem .. glad to hear that it works as expected! 1268783214 M * Bertl_oO off to bed now .. have a good one everyone! 1268783220 N * Bertl_oO Bertl_zZ