1161304890 Q * shedi Quit: Leaving 1161309746 T * Bertl http://linux-vserver.org/ <- new and shiny | latest stable 2.02.1, exp 2.02.2-rc3, devel 2.1.0, exp 2.1.1-rc42, stable+grsec 2.0.2.1 | util-vserver-0.30.211 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 1161309753 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.18.1-vs2.0.2.2-rc3.diff 1161309757 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.18.1-vs2.1.1-rc42.diff 1161310177 M * Bertl ensc: please let me know if vs2.0.2.2-rc3 fixes the issues for you 1161310192 M * Bertl I'm off to bed now ... have a good one everyone! cya tomorrow! 1161310218 N * Bertl Bertl_zZ 1161310792 J * _dmax ~semaj@81.193.57.115 1161310822 Q * dmax Ping timeout: 480 seconds 1161310830 N * _dmax dmax 1161310920 Q * s0undt3ch Ping timeout: 480 seconds 1161311030 Q * samueltc Quit: BitchX-1.1-final -- just do it. 1161312905 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1161312915 J * ensc ~irc-ensc@p54B4F84F.dip.t-dialin.net 1161313072 Q * comfrey Ping timeout: 480 seconds 1161319784 J * comfrey ~comfrey@h-64-105-215-75.sttnwaho.covad.net 1161320808 Q * Aiken Quit: Leaving 1161323024 J * s0undt3ch ~s0undt3ch@81.193.57.115 1161325027 Q * comfrey Read error: Operation timed out 1161325437 Q * ensc|w Ping timeout: 480 seconds 1161327095 J * dna_ ~naucki@201-239-dsl.kielnet.net 1161332082 M * Hollow daniel_hozac: around? 1161332122 Q * _are_ Ping timeout: 480 seconds 1161332202 M * doener morning Hollow 1161332221 M * Hollow morning doener 1161332235 M * doener woohoo! ah new, official cscope release! 1161332244 M * Hollow heh 1161332262 M * Hollow i just switched from viewvc to a very minimal trac installation for our subversion browser 1161332267 M * Hollow really nice :) 1161332282 M * doener ah, released on sept 30th... anyway still 3 years and 26 days since the last one 1161332289 M * Hollow lol 1161332293 M * doener and more than 5 since the last non-bugfix release 1161332390 M * doener grr, udev manages to change my wlan interface name to "wlan0" only in about 1 of 20 module loadings... 1161332401 M * Hollow modules suck 1161332403 M * Hollow :) 1161332433 M * doener yeah, I'm looking into how to completely power off the thing without module unloading 1161332446 M * doener but until I grasped that, the module unloading will have to do it 1161332499 M * doener I'm not happy to waste a few watts for a wlan chip that does nothing 1161332544 M * Hollow heh 1161332575 M * Hollow ah.. tracs changeset view is so nice.. 1161332591 M * Hollow to bad you can't use the svn browser standalone 1161332612 M * Hollow but i have disabled all other components hehe 1161333796 J * meandtheshell ~markus@85-124-36-50.dynamic.xdsl-line.inode.at 1161334525 M * daniel_hozac Hollow: i am now :) 1161334837 M * Hollow daniel_hozac: trac is up an running :) 1161334848 M * daniel_hozac wee, where? :) 1161334860 M * Hollow http://svn.linux-vserver.org/projects 1161334871 M * Hollow most features disabled.. 1161334878 M * Hollow and a short message on each front page 1161334894 M * daniel_hozac ah, nice 1161334912 M * Hollow indeed :) 1161335865 M * Hollow gtg, already late for my lecture.. back later 1161335908 M * daniel_hozac hehe, have fun! 1161336691 J * _are_ ~are@62.112.159.81 1161336827 M * daniel_hozac Bertl_zZ: hmm, are futexes really isolated? grep futex patch-* doesn't give me anything relevant. 1161337118 J * Piet hiddenserv@tor.noreply.org 1161338600 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1161338720 Q * ntrs_ Ping timeout: 480 seconds 1161339778 M * daniel_hozac Bertl_zZ: i can't get 2.0.2.2-rc3 to crash (that way) :) 1161340499 J * chelli ~thomas@PC1.ideolabs.TGZ-Ilmenau.de 1161340566 Q * chelli 1161341396 J * dna___ ~naucki@67-235-dsl.kielnet.net 1161341670 Q * dna_ Ping timeout: 480 seconds 1161341750 J * lilalinux ~plasma@dslb-084-058-222-217.pools.arcor-ip.net 1161342695 Q * s0undt3ch Ping timeout: 480 seconds 1161343336 J * shedi ~siggi@inferno.lhi.is 1161345017 J * phreak`` ~phreak``@140.211.166.183 1161345050 Q * derjohn Quit: by(t)e 1161346259 M * harry pagesize = 4096 1161346269 M * harry legolas:/usr/local/etc/vservers/test# cat rlimits/rss.* 1161346270 M * harry 262144 1161346271 M * harry 131072 1161346283 M * harry so... i have a hard limit of 1GB, right? 1161346307 M * harry 1835 root 25 0 4094m 4.0g 236 R 99.7 25.5 0:07.69 bleh 1161346316 M * harry then how is that possible inside a vps? 1161346331 M * daniel_hozac what does /proc/virtual//limit say? 1161346365 M * harry RSS: 1050 1955 262144 0 1161346368 M * harry VM: 3724 1052851 -1 0 1161346381 M * daniel_hozac and ANON? 1161346387 M * harry ANON: 252 252 -1 0 1161346399 M * daniel_hozac so you're not even close to the limit. 1161346416 M * harry seems so... i think the 4GB limit is because it's 32 bit 1161346429 M * harry and it does only a malloc, but never initialises it 1161346455 M * daniel_hozac so it won't be in RSS. 1161346460 M * harry uhu 1161346465 M * harry just figured that out too :) 1161346479 M * harry kinky tough... 1161346491 M * harry 64 bit machine, with only (for now) 32 bit machines 1161346497 M * harry at least mm is better! :) 1161346617 M * harry Linux legolas 2.6.17.13-grsec2.1.9-vs2.0.2.1 #1 SMP Thu Oct 19 11:14:33 CEST 2006 x86_64 GNU/Linux 1161346620 M * harry waaaaaaaaaaa ;) 1161347235 J * SMuZZ ~smuzz@c213-158-248-202.static.sdsl.no 1161347337 M * SMuZZ anyone have any ideas why I get "quotacheck: error (22) while opening /dev/hdv1" when I try do complete the quotacheck in according to the howto on http://www.5dollarwhitebox.org/wiki/index.php/Howtos_Linux-Vserver_With_LVM_And_Quotas ? 1161347350 M * SMuZZ all steps are followed 1161347372 M * daniel_hozac have you run vrsetup since the last reboot, on /dev/hdv1? 1161347401 M * SMuZZ I use the command vsetup /dev/vroot0 /dev/vg0/vserver1 1161347413 M * daniel_hozac and /dev/vroot0 is the device you copied to the guest? 1161347416 M * SMuZZ and then copy the /dev/hdv1 dev file into the vserver dev dir. 1161347424 M * SMuZZ yep 1161347450 M * sid3windr is there an easy way to get a fedora core vserver on a debian host? 1161347468 M * daniel_hozac sid3windr: install yum. 1161347476 M * sid3windr from source? 1161347490 M * SMuZZ dont debian have yum in apt-get ? 1161347492 M * daniel_hozac should have a package in sid, IIRC. 1161347496 M * sid3windr in sid yup 1161347498 M * daniel_hozac probably etch too. 1161347498 M * sid3windr just saw it 1161347517 M * sid3windr backports have it \o/ 1161347538 M * SMuZZ the only thing tho that differ from the howto's setup and mine is that the lvm disk is on hdb instead of hda, I donno if this has anything to say. 1161347566 M * daniel_hozac SMuZZ: shouldn't matter at all. 1161347620 M * SMuZZ daniel_hozac: good, one more thing to not worry about. 1161347756 M * daniel_hozac SMuZZ: well, i've never used quotas, nor read that howto, but it looks sane to me. 1161347799 M * SMuZZ daniel_hozac: I agree it looks sane, but it dont help when it wont work here :) 1161347845 J * Rich_Estill ~restill@c-24-11-195-139.hsd1.mi.comcast.net 1161347848 M * daniel_hozac SMuZZ: could you post an strace of the quotacheck that fails? 1161347858 M * daniel_hozac SMuZZ: EINVAL on open seems rather peculiar. 1161347870 M * SMuZZ this is the last part of it 1161347876 M * sid3windr error: can't create transaction lock on /var/lib/rpm/__db.000 1161347876 M * sid3windr Warning, could not load sqlite, falling back to pickle 1161347877 M * sid3windr ;( 1161347881 M * SMuZZ this is from the guest where I run the quota check 1161347883 M * SMuZZ lstat64("/", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 1161347884 M * SMuZZ quotactl(Q_GETFMT|USRQUOTA, "/dev/hdv1", 0, 0xffff9dec) = -1 ESRCH (No such process) 1161347886 M * SMuZZ open("//aquota.user", O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory) 1161347889 M * SMuZZ quotactl(Q_GETFMT|GRPQUOTA, "/dev/hdv1", 0, 0xffff9dec) = -1 ESRCH (No such process) 1161347893 M * SMuZZ open("/dev/hdv1", O_RDONLY|O_LARGEFILE) = 3 1161347895 M * SMuZZ lseek(3, 1024, SEEK_SET) = -1 EINVAL (Invalid argument) 1161347898 M * SMuZZ close(3) = 0 1161347900 M * SMuZZ write(2, "quotacheck: ", 12quotacheck: ) = 12 1161347903 M * SMuZZ write(2, "error (22) while opening /dev/hd"..., 35error (22) while opening /dev/hdv1 1161347906 M * SMuZZ ) = 35 1161347908 M * SMuZZ exit_group(0) 1161347916 M * SMuZZ and the hdv1 device is there 1161347917 M * SMuZZ [root@dataguard ~]# ls -la /dev/hdv1 1161347917 M * SMuZZ brw-rw---- 1 root disk 4, 0 Oct 20 04:01 /dev/hdv1 1161347926 M * daniel_hozac ah, so it's trying to seek for whatever reason... 1161348155 M * daniel_hozac SMuZZ: you did mount the filesystem with usrquota,grpquota on the host, right? 1161348175 M * daniel_hozac or maybe that's not required... i never did understand how quota works. 1161348187 M * SMuZZ [root@dataguard /]# cat /proc/mounts 1161348187 M * SMuZZ rootfs / rootfs rw 0 0 1161348187 M * SMuZZ /dev/root / ext3 rw,data=ordered,usrquota,grpquota 0 0 1161348187 M * SMuZZ none /proc proc rw,nodiratime 0 0 1161348187 M * SMuZZ none /tmp tmpfs rw,nodev 0 0 1161348189 M * SMuZZ none /dev/pts devpts rw 0 0 1161348191 Q * renihs Read error: Operation timed out 1161348192 M * SMuZZ [root@dataguard /]# mount 1161348194 M * SMuZZ /dev/hdv1 on / type ext3 (rw,usrquota,grpquota) 1161348197 M * SMuZZ none on /proc type proc (rw) 1161348199 M * SMuZZ none on /sys type sysfs (rw) 1161348202 M * SMuZZ none on /dev/pts type devpts (rw,gid=5,mode=620) 1161348204 M * SMuZZ yep I did that. 1161348207 M * SMuZZ this is also from the guest host 1161348229 M * SMuZZ and this is from the host 1161348231 M * SMuZZ /dev/mapper/vg0-vserver1 on /vservers/centos4 type ext3 (rw,usrquota,grpquota) 1161348239 M * SMuZZ so I got quota all over :) 1161348246 M * daniel_hozac please use paste.linux-vserver.org for longer pastes. 1161348270 M * SMuZZ ofcourse, no problem 1161348988 M * daniel_hozac everything looks fine... can you run quotacheck on the host? 1161349084 M * daniel_hozac IMHO the only way to get ESRCH from quotactl is if the filesystem doesn't have quota enabled. 1161349184 M * daniel_hozac why it thinks that, i'm not sure. 1161349287 M * daniel_hozac what kernel are you using? 1161349384 M * SMuZZ 2.6.16-vs2.1.1-rc21-gentoo-r3 1161349435 M * daniel_hozac well, umm, you might want to upgrade that. 1161349530 M * SMuZZ oki, what version do you suggest? 1161349549 M * daniel_hozac well, 2.6.18.1-vs2.1.1-rc42 for instance. 1161349566 M * SMuZZ the quota works on the host. 1161349662 M * SMuZZ well I'll try and kernel upgrade here then. 1161350527 M * SMuZZ exit 1161350533 Q * SMuZZ Quit: leaving 1161352392 Q * Johnnie Remote host closed the connection 1161352453 J * SMuZZ ~smuzz@monster.dataguard.no 1161352468 M * SMuZZ got 2.6.18-vs2.0.2-gentoo-r8 up and running with the same result there. 1161352512 M * Hollow reminds me to commit rc42.. already have it in my working copy 1161352889 M * daniel_hozac no 2.0.2.2-rc3? :) 1161352962 M * Hollow niether rc42 nor rc3 atm 1161353026 M * daniel_hozac SMuZZ: that's very strange. you don't get anything in dmesg or so, right? 1161353186 M * Hollow apropos... i plead to poke Bertl for sane versioning *grml* 1161353188 N * Rich_Estill Rich_Estill_afk 1161353201 M * daniel_hozac Hollow: hmm? 1161353214 M * daniel_hozac what's unsane about the current versioning? :) 1161353238 M * Hollow why can't we just use 2.0.3? 1161353268 M * Hollow and why does he leave out the second dot every time it gets released 1161353270 M * Hollow :/ 1161353280 M * daniel_hozac the second dot thing i don't understand either :) 1161353297 M * daniel_hozac and i guess the changes aren't really worthy of being called 2.0.3. 1161353308 M * Hollow still bah :) 1161353309 M * daniel_hozac it's more a "two small fixes" thing than lots of new fixes. 1161353321 M * daniel_hozac s/new// 1161353371 M * doener SMuZZ: 32bit quota on 64bit system? 1161353417 M * SMuZZ doener : that wont work ? 1161353423 M * SMuZZ I have a 64 bit system yet 1161353429 M * SMuZZ yet =yes 1161353483 J * Johnnie ~jdlewis@jdlewis.org 1161353509 M * Hollow is 2.0.2.1-t8 the right one? 1161353516 M * doener SMuZZ: kernel is 64bit and userland is 32bit? 1161353523 M * daniel_hozac Hollow: "the right one"? 1161353530 M * Hollow the one mentioned in the topic 1161353533 M * doener SMuZZ: ... the guest's userland 1161353536 M * Hollow the one that should be 2.0.2.1 1161353547 M * daniel_hozac 2.0.2.1 is 2.6.17.13-vs2.0.2.1. 1161353552 M * Hollow ok.. 1161353557 M * SMuZZ doener : kernels is 64 bit yes, but I donno about the centos4 system im running here, my guess is that it's 32bit. 1161353562 M * daniel_hozac i don't believe there is one for 2.6.18. 1161353567 M * daniel_hozac (final, that is) 1161353568 M * doener SMuZZ: try file /bin/bash 1161353572 M * daniel_hozac it'll be 2.0.2.2 instead. 1161353575 M * Hollow in /Experimental there is, which is called *-t8 1161353576 M * doener that should tell you if it's 32bit or 64bit 1161353586 M * Hollow whatever t8 is 1161353589 M * daniel_hozac Hollow: -t8 is try8 ;) 1161353594 M * Hollow ah 1161353594 M * daniel_hozac or test8. 1161353609 M * SMuZZ doener : /bin/bash: ELF 32-bit LSB executable, so I guess im screwed then? 1161353612 M * doener SMuZZ: the OpenVZ guys discovered the bug with 32bit quota tools on 64bit kernels 1161353617 M * doener let me search the mail on lkml 1161353623 M * SMuZZ ahh cool 1161353633 M * doener http://lkml.org/lkml/2006/10/19/123 1161353646 M * daniel_hozac SMuZZ: try installing a 64-bit guest instead? or do you need it to be 64-bit for some reason? 1161353686 M * SMuZZ daniel_hozac : nope I dont need the 32 bit, problem was that I only found 32 bits centos4 images 1161353698 M * daniel_hozac SMuZZ: why don't you install it yourself? 1161353723 J * marcfiu ~mef@targe.CS.Princeton.EDU 1161353738 M * SMuZZ daniel_hozac : cause it's was simpler to download a pre-done image :) 1161353753 M * daniel_hozac is it? 1161353775 M * daniel_hozac i suppose if you don't have the necessary tools installed already, that might be true. 1161353799 M * SMuZZ well I just unzipped the img and started the vserver from the folder i unzipped it to. 1161353825 M * daniel_hozac ... while building it yourself is a single vserver ... build command. 1161353838 M * SMuZZ yep 1161353885 P * marcfiu 1161353886 M * SMuZZ but the 32 vs 64 bit error explains why, and I prolly need to get the 64 bit up and running instead. 1161353908 M * SMuZZ and I'll try that patch 1161354011 M * doener SMuZZ: that patch got some comments that say that it needs improvement 1161354028 M * doener SMuZZ: it might work for you, but it might as well fail horribly ;) (didn't really follow that thread) 1161354067 M * SMuZZ doener : well this isn't a prod system anyway so if it barfs at me I just have accept that :) 1161354099 M * doener as long as it doesn't kill your kitten, eh? ;) 1161354141 Q * fosco Quit: et hop, on change de box 1161354155 M * SMuZZ hehe 1161354269 M * SMuZZ well let's see what happens 1161354271 Q * SMuZZ Quit: brb 1161354621 J * fosco fosco@konoha.devnullteam.org 1161354635 Q * fosco 1161354667 J * SMuZZ ~smuzz@monster.dataguard.no 1161354676 M * SMuZZ that didn't help any 1161354689 J * fosco fosco@konoha.devnullteam.org 1161354692 M * SMuZZ I'll try todo a 64 bit system here instead. thanks for the help. 1161354866 N * Bertl_zZ Bertl 1161354871 M * Bertl morning folks! 1161354923 M * Bertl daniel_hozac: hmm, have to check, I remember doing futex patches at some point, maybe they never got in 1161355207 M * Hollow morning Bertl 1161355874 M * Hollow SMuZZ: FYI: 2.0.2.1, 2.0.2.2_rc3 and 2.1.1_rc42 are in the overlay now 1161357017 J * derjohn ~derjohn@80.69.37.19 1161357266 M * doener grr... stupid segfault... and according to gdb it even happens in a zombie process if I interpret the output right 1161357633 Q * lilalinux Remote host closed the connection 1161357949 M * daniel_hozac Bertl: btw, did you see my question yesterday about [vd]x_barrier and the lack of VX_WATCH? 1161357988 M * daniel_hozac i.e. is it intentional, and if so, what's the point? :) 1161358297 J * stefani ~stefani@tsipoor.banerian.org 1161358529 M * Bertl daniel_hozac: nope, missed it, but does a spectator process really do filesystem ops? otherwise I'd see it as 'bonus' barrier check :) 1161358540 M * Bertl welcome stefani! derjohn! 1161358565 M * Bertl daniel_hozac: means: we can change it if you have a reason for changing it ... 1161358574 M * daniel_hozac hmm, doesn't the spectator context get all mounts? 1161358643 M * Bertl 'all' as in 'host' or 'guest' or 'all guests'? 1161358695 M * derjohn Bertl, great foo ;) .... 2.6.18.1 rc42 runs on the frist maschine. debian packel upgraded. no issues so far. 1161358717 M * Bertl derjohn: *G* good to hear :) 1161358734 M * daniel_hozac Bertl: hmm, all guests? i'm not sure how to interpret the vx_check in show_vfsmnt. 1161358738 M * derjohn I hope daniel's issue (see above) will not cause an rc43 ... ? 1161358747 M * daniel_hozac nah, it's not an issue. 1161358763 M * Bertl daniel_hozac: no, spectator sees the 'full' current namespace 1161358784 M * daniel_hozac hmm, wouldn't the host already do that? 1161358785 M * Bertl so, as long as spectator is a cotnext, not a capability/flag 1161358799 M * Bertl it's identical with the host/admin view 1161358817 M * daniel_hozac ok. 1161358829 M * daniel_hozac how would that change if it was a capability/flag? 1161358866 M * Bertl with a process flag (that's what I plan) every process (even one inside a guest/context) could becomea spectator 1161358896 M * sid3windr anyone here tried building fc4 guest on debian host? can't seem to get it built 1161358915 M * daniel_hozac sid3windr: why would you want a FC4 guest? it's been moved to legacy already ;) 1161358918 M * Bertl this, for example, would allow seamless s/ptrace across context switches 1161358929 M * sid3windr daniel_hozac: because $software wants it to be fc4 1161358961 M * daniel_hozac sid3windr: i think you'll have to get the GPG keys and put them in /etc/vservers/.distributions/fc4/pubkeys or so. 1161359003 M * Bertl okay, have to leave now .. translocating, will be back later this evening ... 1161359009 M * daniel_hozac Bertl: ok, but why would you want a spectator inside a guest? how would it differ from a regular guest process? 1161359021 M * daniel_hozac Bertl: ok, cya! 1161359027 M * Bertl daniel_hozac: see my comment above (strace) 1161359038 M * daniel_hozac so just for migrated processes? 1161359040 M * derjohn Should I enable Inode Tag Propagation? 1161359040 N * Bertl Bertl_oO 1161359051 M * sid3windr hmm 1161359061 M * daniel_hozac derjohn: i don't know, should you? :) 1161359100 M * derjohn I dont currently know for what I should be good for.... tagging inodes instead of files? 1161359114 M * daniel_hozac derjohn: i think that means that if you do mount -o xid=120 /dev/... /vservers/...; mount /dev/...2 /vservers/.../dir, the latter mount will also get xid 120. 1161359120 M * daniel_hozac files are inodes ;) 1161359138 M * daniel_hozac or well, inode+dentry, i suppose. 1161359175 M * derjohn well, in my mind a file is some kind of chain-of-inodes. but maybe my mind am wrong :) 1161359184 M * daniel_hozac chain of blocks, perhaps. 1161359191 M * derjohn uh 1161359200 M * daniel_hozac the inode is the thing containing uid, gid, permissions, etc. 1161359207 M * daniel_hozac i.e. metadata. 1161359219 M * derjohn so one inode per file ? 1161359225 M * daniel_hozac yep. 1161359233 A * derjohn restructures his mind 1161359234 M * Hollow yummi, strace across context switches sounds nice 1161359240 M * daniel_hozac indeed. 1161359257 M * Hollow one thing i really miss 1161359269 M * daniel_hozac if only we could convince strace that vserver only takes 3 args :) 1161359274 M * Hollow heh 1161362175 J * Osgigi ~abigor@kurzweil.no-ip.org 1161362223 M * sid3windr daniel_hozac: thanks, that was indeed it... needed to puzzle somewhat, but I got the guest up and running now :) 1161362259 M * daniel_hozac :) 1161362402 Q * Osgiliath Ping timeout: 480 seconds 1161362457 J * Osgiliath ~abigor@kurzweil.no-ip.org 1161362687 Q * Osgigi Ping timeout: 480 seconds 1161362956 M * ensc Bertl_oO: vs2.0.2.2-rc3 seems to work fine; thx 1161363405 M * sid3windr yay for vserver, now using it at $company as well 1161363409 M * sid3windr "great solution" ;> 1161363417 M * sid3windr "yeah, been using it at home for a couple of years" :p 1161364188 Q * Piet Ping timeout: 480 seconds 1161364952 Q * derjohn2 Ping timeout: 480 seconds 1161364955 J * derjohn2 ~aj@dslb-084-059-028-003.pools.arcor-ip.net 1161365062 M * daniel_hozac do we protect the guest's init from strace? 1161365182 M * daniel_hozac we don't, do we? 1161365414 M * daniel_hozac ok, we do, my bad. 1161366773 M * daniel_hozac Bertl_oO: include/linux/config.h:6:2: warning: #warning Including config.h is deprecated. 1161367166 M * ensc ... and fails to build with 2.6.19 ... 1161367230 M * ensc Bertl_oO: I still get kernel warnings: http://paste.linux-vserver.org/530 1161367240 M * ensc happens sometime during boot 1161367640 M * daniel_hozac ensc: isn't there a line right above that? 1161367676 M * ensc EXT3-fs: mounted filesystem with ordered data mode. 1161367676 M * ensc BUG: warning at kernel/pid.c:270/pid_task() 1161367754 M * daniel_hozac ah, CONFIG_VSERVER_DEBUG disabled? 1161367764 M * ensc yes; should be a production system ;) 1161367764 M * daniel_hozac enabling that should tell you exactly why it's warning. 1161367794 M * ensc btw, happened just again... 1161368050 M * daniel_hozac do you have a test-case? 1161368293 M * ensc last warning might be related with a cfagent run and quickly spawned/died process 1161368433 M * daniel_hozac i think i can see how that would happen. 1161368755 M * daniel_hozac but enabling CONFIG_VSERVER_DEBUG should help... (maybe extending the warning message with the process name before rebuilding) 1161368849 M * daniel_hozac (if it is indeed a quickly spawned/died process, it won't stick around long enough for a pid<->name lookup post-logging, i guess) 1161368869 M * Osgiliath Hi everyone, is it possible to specify the debian's mirror when we build a guest ? 1161368912 M * daniel_hozac Osgiliath: yes, vserver ... build -m debootstrap --context ... -- -d -m 1161368927 M * Osgiliath ok, thanks :) 1161369080 M * sid3windr editing Vserver_users isn't all that straightforward 1161369128 A * sid3windr adds some comments 1161369312 M * sid3windr pity alt-e is no longer bound to edit on the wiki 1161369317 M * sid3windr got so used to it with mediawiki installs ;) 1161369338 M * daniel_hozac "no longer"? the old tavi had it? 1161369810 M * sid3windr no, as in opposed to a default mediawiki install 1161369827 J * coocoon ~coocoon@dslb-084-056-150-072.pools.arcor-ip.net 1161369836 M * coocoon hello 1161369841 M * daniel_hozac hi 1161369903 J * dna_ ~naucki@67-235-dsl.kielnet.net 1161369972 M * daniel_hozac sid3windr: are you sure? that seems like an odd feature to remove. 1161369999 M * sid3windr mine do, and I didn't do anything special to enable it :) 1161370018 M * sid3windr perhaps the person who installed it didn't like it messing with the browser's actual alt-e for edit menu :) 1161370355 Q * dna___ Ping timeout: 480 seconds 1161370852 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1161371547 Q * _are_ Ping timeout: 480 seconds 1161373125 J * _are_ ~are@62.112.159.81 1161373891 Q * kevinp Read error: Connection reset by peer 1161373968 Q * Blissex Remote host closed the connection 1161374028 J * kevinp ~kevinp@ny.webpipe.net 1161374142 M * Hollow sid3windr: this is a standard mediawiki installation 1161374254 M * sid3windr strange 1161374259 M * sid3windr perhaps it is in the skin? 1161374267 M * sid3windr come to think of it, it probably is :) 1161374282 M * sid3windr I installed 2 mediawiki's 6 months ago or so and they both have alt-e to edit 1161374416 M * sid3windr /index.php?title=-&action=raw&smaxage=0&gen=js seems to generate tooltips & access keys and is referred to inside 1161376159 J * bonbons ~bonbons@83.222.36.111 1161377217 Q * coocoon Ping timeout: 480 seconds 1161377849 J * coocoon ~coocoon@dslb-084-056-136-191.pools.arcor-ip.net 1161377899 Q * dna_ Quit: Verlassend 1161378070 M * doener http://lkml.org/lkml/2006/10/20/251 -- new generic containers patchset 1161378090 M * doener just discovered in my inbox, didn't read the mails yet ;) 1161378154 M * doener woah.. huge patches... 1161378809 M * bonbons doener: there are some other things in 2.6.19 that are isolation/virtualization related! 1161378885 M * doener I know, I already started a port but didn't get far yet 1161378897 M * doener IPC and UTS namespaces IIRC, right? 1161378914 M * bonbons yep, that's those I remember of-head 1161378946 M * doener they've also been in -mm for some time, but the announcements that akpm made for those -mm versions didn't exactly made me feel confident to try to do a port 1161378970 M * bonbons will be some more work to port using those features... 1161378982 M * daniel_hozac wasn't the pidspace stuff in as well? 1161379012 M * doener not sure about that 1161379875 Q * notasnark Ping timeout: 480 seconds 1161380286 Q * bonbons Quit: Leaving 1161380329 P * stefani I'm Parting (the water) 1161380737 Q * coocoon Ping timeout: 480 seconds 1161381663 J * coocoon ~coocoon@dslb-084-056-185-247.pools.arcor-ip.net 1161382234 Q * mire Quit: Leaving 1161383380 N * Bertl_oO Bertl 1161383386 M * Bertl evening folks! 1161383486 M * daniel_hozac evening Bertl! 1161383505 M * coocoon hello bertl 1161384323 M * daniel_hozac Bertl: hmm, seems the pid_task warning might be a bit too verbose. 1161384351 M * daniel_hozac i get it when running strace -p (i mistyped the pid) 1161384357 M * Bertl yeah, figured that 1161384390 M * Bertl but I'd like to identify specific cases we _want_ to have 1161384403 M * Bertl and silence the other ones if possible 1161384406 M * daniel_hozac are there any cases where we want to warn? 1161384458 M * Bertl yes, I think all access between contexts for example 1161384478 M * daniel_hozac why? 1161384483 M * Bertl i.e. from xid(a) > 1 to xid(b) > 1 1161384498 M * Bertl because IMHO those should not happen? 1161384518 M * daniel_hozac but we're already blocking the access? seems like a high cost for a typo :) 1161384591 M * Bertl hmm, well, the stack dump will go away 1161384605 M * Bertl the 'warning' itself should be fine, IMHO 1161384621 M * Bertl similar as the warning we throw on xid file access 1161384625 M * daniel_hozac yeah. 1161384672 M * Bertl I just wanted to get some ideas who are the callers (that's what the stack is for) 1161384709 M * daniel_hozac ah, ok. 1161384755 Q * Osgiliath Ping timeout: 480 seconds 1161384772 M * Bertl so, please report back those dumps if they are unusual :) 1161384851 Q * coocoon Quit: KVIrc 3.2.0 'Realia' 1161384873 M * daniel_hozac did you see ensc's dump? 1161384885 M * Bertl nope, url? 1161384889 M * daniel_hozac http://paste.linux-vserver.org/530 1161384906 M * daniel_hozac no CONFIG_VSERVER_DEBUG though. 1161384932 M * daniel_hozac (btw, shouldn't the warnings be printed even without that? i thought they were already) 1161384957 M * Bertl nope 1161384986 M * Bertl at least I don't think so 1161385010 M * daniel_hozac warnings don't really seem like debugging information, IMHO. 1161385034 M * Bertl yeah, we should fix that, or make it a separate option 1161385073 M * Bertl maybe even add some 'silencer' or treshhold 1161385105 M * daniel_hozac similar to the debugging? 1161385120 M * daniel_hozac (i.e. with a /proc/sys/vserver/warnings or similar) 1161385125 M * Bertl maybe, maybe even with rate limiting or so 1161385159 M * daniel_hozac rate limiting seems like a maybe a bit too much for the kernel. don't most syslogs do that anyway? 1161385179 M * Bertl right, was just an idea 1161385201 M * daniel_hozac i guess it would help with dmesg though. 1161385434 M * Bertl shouldn't be too hard to make it per context and similar to a bucket 1161385487 M * daniel_hozac heh, per context token bucket for warnings... 1161386275 M * Bertl doener: we should make a wiki page collecting all the virtualization links/patches/etc 1161386420 J * Osgiliath ~abigor@kurzweil.no-ip.org 1161386430 M * Bertl wb Osgiliath! 1161386440 M * Osgiliath hi :)