1146441635 M * baggins Bertl: ok, can you leave a patch or some dircetions what to include in the message? then i test it on my own if I can't find you
1146441669 M * doener_ no need for yet... for most of my work i just connect to my box at home via ssh anyway. i still need to adjust my workflow to having a notebook actually ;)
1146441674 M * Bertl baggins: okay, look for a dir BAGGINS on http://vserver.13thfloor.at/Experimental/
1146441688 M * Bertl baggins: it will contain the appropriate mini patch
1146441728 M * baggins Bertl: thanks, I'll start testing after some sleep :)
1146441881 M * daniel_hozac Bertl: i can reproduce that.
1146441895 M * doener_ Bertl: mbcap-fix01 is included in mbcap-feat01 isn't it?
1146441915 M * doener_ erhm... nvm
1146441921 M * Bertl daniel_hozac: the gdb?
1146441921 M * doener_ didn't parse context lines...
1146441925 M * daniel_hozac Bertl: yeah.
1146441937 M * Bertl okay, let me upload the patch now :)
1146442058 M * Bertl http://vserver.13thfloor.at/Experimental/BAGGINS/delta-reaper-fix01.diff
1146442289 J * ntrs_ ~ntrs@dsl-203-81-rt2-bras.wnvl.centurytel.net
1146442290 Q * ntrs Read error: Connection reset by peer
1146442366 M * doener_ Bertl: ok, so we use cap_bset to mask set eff/inh/perm caps at migration time, and apply the bcaps mask at the time that actual permission checks are done, right?
1146442379 M * Bertl yep
1146442442 M * doener_ why would I want to turn off bit X in cap_bset instead of bcaps?
1146442451 Q * Dr4g Quit: Open Source Development :: http://dynamichell.org
1146442477 M * Bertl doener_: the bits in cap_bset will be visible for the apps checking that inside the guest, e.g. bind9#
1146442498 M * Bertl doener_: if you apply limits there, they are not changed on the fly as with vx_bcaps
1146442499 M * doener_ ah, virtualization issues :)
1146442532 M * doener_ to fool bind9 into thinking "hey, great I can do all the stuff I ever wanted to"
1146442543 M * Bertl yep, precisely
1146442552 M * Bertl works fine
1146442562 J * ntrs__ ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com
1146442600 M * doener_ stupid thing... may I do that? No -> Bail out. Yes -> Great. ... Does it actually work? Well, I don't care...
1146442650 M * doener_ uhm... everything after "stupid thing.." was meant to be said by bind9...
1146442663 A * doener_ .oO( should try to express such stuff on irc)
1146442663 M * Bertl that's the problem when folks misunderstand security features
1146442669 M * doener_ s/should/shouldn't/
1146442716 M * daniel_hozac i guess it bails out if it doesn't work too.
1146442767 M * daniel_hozac it's just that most configurations don't use those particular features :)
1146442818 M * doener_ ok, the patch itself looks fine... anything you want me to look out for? like places that should also make use of that masking?
1146442846 J * ntrs ~ntrs@dsl-203-81-rt2-bras.wnvl.centurytel.net
1146442853 Q * ntrs_ Read error: Connection reset by peer
1146442903 M * Bertl doener_: yes, for example we missed the netlink (as the fix sows)
1146442906 M * Bertl *shows
1146442926 M * Bertl doener_: so it would be great if you have a look at the nfs and quota places too
1146442983 M * doener_ nfs will be 'fun' I guess :)
1146443034 J * ntrs_ ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com
1146443074 Q * ntrs__ Read error: Connection reset by peer
1146443174 M * nebuchadnezzar arf, my new kernel do not want to boot
1146443209 M * nebuchadnezzar snif
1146443210 M * Bertl what error/panic? do you get?
1146443224 M * nebuchadnezzar juste after booting linux:
1146443226 M * nebuchadnezzar IRQ: INO 0220 (13:0000040004701504:0000040004701104) --> (13:0:0000040004601504:0000040004601104), halting...
1146443226 M * nebuchadnezzar ProgFast Data Access MMU Miss
1146443226 M * nebuchadnezzar  
1146443226 Q * kir_home Quit: Ухожу я от вас
1146443252 M * nebuchadnezzar I'll try a 2.6.16.11 without all yours patch to see
1146443258 M * Bertl hmm, yeah, please
1146443269 M * Bertl seems like an IRQ issue
1146443290 M * daniel_hozac mm/oom_kill.c:106,116?
1146443302 M * nebuchadnezzar 2.6.16-vs2.1.1-rc15 works great :-/
1146443328 M * daniel_hozac Bertl: vxD: reaper mismatch: d54e9aa0[1650,#8000] <> c1379aa0[1,#0]
1146443350 M * Bertl so where comes the real init from?
1146443368 M * Bertl daniel_hozac: is that an init less client?
1146443373 Q * ntrs Ping timeout: 480 seconds
1146443374 M * _coocoon_ bertl: hello, i need a little bit help, having problems since some days testme.sh output http://paste.linux-vserver.org/13
1146443375 M * daniel_hozac yes.
1146443383 M * Bertl daniel_hozac: ah, hmm, good point then
1146443396 M * Bertl daniel_hozac: that's definitely a case we do not handle properly
1146443425 M * daniel_hozac _coocoon_: didn't i tell you what to do a few days ago?
1146443442 M * daniel_hozac _coocoon_: either you disable the legacy version in the kernel, or rebuild the utils with legacy support.
1146443462 M * Bertl _coocoon_: best is to compile the tools with:
1146443477 M * Bertl --with-apis=NOLEGACY (or --enable-apis?)
1146443484 M * nebuchadnezzar arf, I prefer disable legacy in my kernel :-)
1146443506 M * Bertl _coocoon_: could you try that?
1146443513 M * _coocoon_ daniel_hozac: yes right u have told me but it doesn't work
1146443518 M * _coocoon_ bertl: i will
1146443522 M * _coocoon_ thanx
1146443545 M * daniel_hozac --enable-apis.
1146443546 A * nebuchadnezzar compile a new kernel :-|
1146443559 M * Bertl daniel_hozac: ad mm/oom_kill.c ... that seems still valid to me
1146443579 M * Bertl daniel_hozac: maybe we should add a comment there?
1146443580 M * daniel_hozac _coocoon_: err, how do you know it doesn't work if you haven't tried it? (as that's what Bertl just suggested...)
1146443642 M * daniel_hozac Bertl: so we want to be nice to guest root too?
1146443672 M * Bertl at least a little more nice than to other processes
1146443804 M * _coocoon_ daniel_hozac: i have tried it believe me have enabled it in kernel, configure without any switch, also have used the switch and disabled in the kernel
1146443868 M * nebuchadnezzar If I understand, if it's enabled in the kernel you should use the switch --enable-apis, not the inverse ? :-)
1146443890 M * nebuchadnezzar hope I'm right
1146443928 M * Bertl yes, the option is a little confusing
1146443939 M * Bertl as NOLEGACY actually enables the legacy stuff
1146443960 M * _coocoon_ sorry but it doesn't work in any way, maybe i have made mistakes
1146443978 M * nebuchadnezzar Bertl: ok, so who's right ? you've lost me
1146444036 M * _coocoon_ so i have enabled it in the kernel set --with-apis=NOLEGACY and i got the same error
1146444042 M * _coocoon_ chbind failed
1146444087 M * Bertl _coocoon_: can you update the testme.sh for us?
1146444102 M * daniel_hozac _coocoon_: you need --enable-apis=NOLEGACY, as i said.
1146444123 M * _coocoon_ i used the latest i think so but now i think it looks better
1146444133 M * _coocoon_ i hope so
1146444233 M * _coocoon_ ok now it works fine sorry 
1146444255 M * _coocoon_ i am a little bit confused because of never has need this function
1146444275 M * _coocoon_ before util-vserver-0.30.210
1146444295 M * daniel_hozac a) the legacy version ID thing is new. b) all the legacy APIs are disabled by default since 0.30.210.
1146444301 M * _coocoon_ daniel_hozac: thanx u r right i do not know why it failed
1146444327 M * _coocoon_ ok now i know this
1146444514 M * _coocoon_ but i got this message  chattr: Unpassender IOCTL (I/O-Control) fr das Ger� beim Lesens der Flags in /etc/vservers/.defaults/vdirbase/apache further
1146444547 M * nebuchadnezzar arf, France 1, German 1 :-)
1146444553 M * Bertl _coocoon_: okay, could you retry after doing: export LANG=C L_ALL=C
1146444554 M * nebuchadnezzar LANG=C is your friend
1146444578 M * Bertl and upload the output (maybe with --debug)
1146444579 M * _coocoon_ i should set it to english right
1146444603 M * _coocoon_ it is german output
1146444625 M * _coocoon_ i got this mesaage building a vserver
1146444713 M * _coocoon_ chattr: Inappropriate ioctl for device while reading flags on /etc/vservers/.defaults/vdirbase/apache
1146444714 M * daniel_hozac _coocoon_: http://daniel.hozac.com/vserver/util-vserver/util-vserver-0.30.210-chattr.patch
1146444851 M * _coocoon_ daniel_hozac: i do not know where to patch it before configuring util-vserver-0.30.210 or in the util-vserver dir
1146444926 J * romantik20m ~gOmDrFsO@85.108.52.211
1146444951 M * _coocoon_ bertl: http://paste.linux-vserver.org/14
1146444969 Q * romantik20m Quit: 
1146444990 J * shedi ~siggi@inferno.lhi.is
1146445017 M * nebuchadnezzar what the hell is that kernel
1146445029 M * nebuchadnezzar Bertl: it does not works without your patches too
1146445036 M * nebuchadnezzar so I have a problem with the kernel
1146445047 M * Bertl nebuchadnezzar: thought so, maybe some missing config option?
1146445073 M * Bertl _coocoon_: looks good?
1146445074 M * nebuchadnezzar well, the same as 2.6.11
1146445082 M * nebuchadnezzar Time To Sleep now
1146445091 M * nebuchadnezzar be back when the washing machine will wake me up
1146445108 M * Bertl okay, cya!
1146445113 M * nebuchadnezzar good night
1146445167 M * _coocoon_ daniel_hozac: sorry but how to patch it
1146445212 M * _coocoon_ in the util-vserver-0.30.210 dir or in /usr/lib/util-vserver dir
1146445225 M * _coocoon_ i mean before installing or after 
1146445501 M * _coocoon_ daniel_hozac: thanx now it works fine thanx a lot
1146445542 M * _coocoon_ daniel_hozac: i am nerving a little bit right
1146445545 M * _coocoon_ ;-)
1146445713 M * Bertl _coocoon_: np
1146445726 M * _coocoon_ but now all works fine
1146445731 M * _coocoon_ zhanx a lot
1146445848 M * Bertl daniel_hozac: so, makin an execption for host init should be enough, yes?
1146445877 M * daniel_hozac Bertl: i think so.
1146446197 M * _coocoon_ ok thanx and good night
1146446201 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1146447925 J * MrX ~urk@219.95.5.156
1146447944 M * Bertl welcome MrX!
1146448497 M * Bertl yes!
1146448936 M * Bertl http://vserver.13thfloor.at/Experimental/delta-lo0.05.diff (first time that it passes my tests :)
1146448968 M * daniel_hozac cool!
1146449120 M * Bertl well, probably just means that my tests are incomplete :)
1146449131 M * daniel_hozac lol
1146449321 M * Bertl and right I am, did break raw packets :)
1146449352 M * Bertl but it's just a simple negated check
1146449479 Q * cemil Server closed connection
1146449488 J * cemil ~cemil@defiant.wavecon.de
1146449619 M * s0undt3ch :( I've just made my remote newly vserver un-available
1146449629 M * Bertl s0undt3ch: how that?
1146449646 A * doener_ bets on missing nodev
1146449658 M * Bertl nah, nobody would do that ...
1146449666 M * daniel_hozac except for eyck :)
1146449673 M * Bertl *shht*
1146449681 M * s0undt3ch well, rebooting was going fine, I added default to /vserver/new_vs/apps/init 
1146449713 M * s0undt3ch then I was trying to update my vs's portage tree and it was failing and I say, well, I just setup the vs so let's reboot
1146449726 M * s0undt3ch From mail.ufsoft.org (10.1.0.1) icmp_seq=2107 Destination Host Unreachable
1146449764 M * s0undt3ch and if it's powered down, I could power it up, I tried, and still no luck
1146449778 M * Bertl hmm, you rebooted the host?
1146449782 M * s0undt3ch so my guess is that it forzen somewere on poweroff or poweron
1146449789 M * s0undt3ch yup
1146449798 M * Bertl did you change the one-time default back?
1146449806 M * Bertl (grub)
1146449819 M * s0undt3ch I was still using the grub-set....
1146449821 M * Bertl otherwise it is now booting with the old-non-vserver kernel
1146449823 M * s0undt3ch and I did set it
1146449845 M * s0undt3ch if that was the case I would be able to ssh in
1146449870 M * s0undt3ch no I'll have to wait 5 hours to see what went wrong
1146449882 M * Bertl well, could be a 100 things, that's why I prefer remote consoles
1146449904 M * s0undt3ch remote serial consoles right?
1146449906 M * Bertl btw, if you have another server there you could have your remote console :)
1146449919 M * s0undt3ch how so?
1146449934 M * Bertl jsut connect a null modem cable (actually two) between them
1146449973 M * s0undt3ch well, can't do that now ;) but that's a thought
1146449975 M * Bertl http://vserver.13thfloor.at/Null-Modem/
1146449985 M * Bertl that's what I do :)
1146450070 M * s0undt3ch still when all thing are set up correctly internet won't be accessible on both comps only one which runs as a router so serial console like that won't be possible
1146450116 M * s0undt3ch since major f*ckup usually happen on the server(router) ;)
1146450140 M * s0undt3ch is there?
1146450141 M * Bertl ah, well, I was talking about two (almost) independant machines
1146450195 M * s0undt3ch but if I end up running an old P2 just a router/firewall, that might be possible(some day)
1146450197 M * Bertl grub works quite fine with serial, you can even have both, serial and normal console
1146450260 M * Bertl daniel_hozac, doener_: updated to delta-lo0.05.1.diff
1146450272 M * Bertl would appreciate a check and later test with guests
1146450286 M * Bertl (have to cleanup the debug stuff though)
1146450529 J * doener__ ~doener@i5387C2AF.versanet.de
1146450542 M * s0undt3ch by the way how do I get into the guests? chroot?
1146450549 M * daniel_hozac vserver ... enter
1146450552 M * daniel_hozac or ssh.
1146450576 M * s0undt3ch k, thanks, won't be able to do it right now but great ;)
1146450600 M * Bertl let us know what it was (tomorrow or so)
1146450630 M * s0undt3ch k, I think I'll here from now on since I''ll be running a totaly new server from now on...
1146450665 M * Bertl well, everybody is welcome here ... feel free to idle around
1146450801 M * s0undt3ch and, your probably gonna flame me, but why vserver over xen or openvz?
1146450816 M * Bertl ah, no, we are not flaming ...
1146450830 M * Bertl first, xen works quite fine _with_ Linux-VServer
1146450841 M * Bertl we consider it complementary
1146450862 M * s0undt3ch isnt it suposed to be the same or equal to vserver?
1146450877 M * Bertl i.e. if you need to run different kernels or operating systems, xen is your choice
1146450893 M * s0undt3ch do note that I've come in touch with any kind o virtualization very recently
1146450898 M * Bertl if you want to be resource efficient, on the same kernel, Linux-VServer is your choice
1146450914 Q * doener Ping timeout: 480 seconds
1146450918 M * s0undt3ch well, the I guess I chose the right one
1146450919 M * Bertl if you need both, just put a Linux-VServer kernel inside your xen domU
1146450945 M * s0undt3ch my server won't be a "vserver hoster"
1146451021 M * s0undt3ch I'm simply trying to change from hardened to a better one, cuz recent problem on the mail part of my system is making the whole system frozen, and I'd be happier if it just froze part of it, better if none of course ;) 
1146451037 M * s0undt3ch s/better one/better approach/
1146451070 M * Bertl i.c., well depends on how/why it did freeze your system
1146451175 M * s0undt3ch well, right now my gues is clamav, It's been failing hard, but could not figure out why because accidently( :] ) I ended up deleting my /etc/init.d/ so 12H fixing that part without needing to rebuild the whole system(since gentoo builds everything from source)
1146451251 M * s0undt3ch after that it stoped failing so don't know why, but I had to staty away for 2 days to come back to a sys frozen again, so, for now it's disabled to see if that the source of the problem
1146451270 M * s0undt3ch and since I was already thinking of using vserver on my server, this might be the time
1146451273 M * Bertl with frozen you mean it is not responding?
1146451283 M * Bertl or do you mean a kernel panic or trace?
1146451285 M * s0undt3ch yup, it completely died
1146451308 M * s0undt3ch powered on, but dead, inactive
1146451314 M * s0undt3ch un-resposive
1146451341 M * Bertl hmm, without console that is hard to classify ...
1146451364 M * s0undt3ch the last time, there wast anything on screen, it was just black
1146451381 M * s0undt3ch and since I was in a run, I just powered it up
1146451394 M * s0undt3ch ir, re-booted it
1146451404 M * s0undt3ch s/ir/ie/
1146451457 M * Bertl yeah, probably the console blanked ... although I remember that they added an unblank some time ago
1146451497 M * s0undt3ch so, right now I just want to migrate the need services temporarely to my desktop comp(with vserver already if possible ;) )
1146451518 M * Bertl should work ...
1146451557 M * s0undt3ch then I'll try to figure out what's happening on the server, and then completely re-install a gentoo stage1 with LVM2 which I'm not using now
1146451590 M * s0undt3ch basicly I'll wan't 4 guests web/mail/shells/my_own_dev_guest
1146451633 M * Bertl yes, service separate is one of the main uses aside from vps ...
1146451639 M * Bertl *separation
1146451650 M * s0undt3ch vps?
1146451657 M * Bertl virtual private servers
1146451661 M * s0undt3ch sorry I'm a noob :)
1146451665 M * s0undt3ch yeah
1146451681 M * Bertl I doubt that, and nothing to be sorry about :)
1146451727 M * s0undt3ch well, I not a noob I was 1 year ago of course, but still am on alot of stuff, but then again who isn't :)
1146451786 M * s0undt3ch I started only for the fun of not having to use Win, now I don't wan't anything else :)
1146451878 M * Bertl yeah, MS* can be very annoing
1146452081 M * s0undt3ch extremely anoying
1146452178 M * s0undt3ch well, since I can't setup my temp vserver gonna code a bit of python, see you soon, probably when I get home
1146452184 M * s0undt3ch thanks for all so far
1146452195 M * Bertl you're welcome! cya!
1146452690 M * doener_ i'm off to bed, good night folks!
1146452700 M * daniel_hozac good night!
1146452703 M * Bertl night!
1146452717 Q * doener_ Quit: leaving
1146453066 P * hillct Kopete 0.11.1 : http://kopete.kde.org
1146453217 M * daniel_hozac http://daniel.hozac.com/vserver/delta-dccp-fix01.diff
1146453266 M * Bertl ah, cool!
1146453276 M * Bertl that was what nebuchadnezzar mentioned
1146453283 M * daniel_hozac yeah, i had it configured too.
1146453326 M * Bertl pardon my ignorance, but what it is used for?
1146453333 M * daniel_hozac i have no idea.
1146453340 M * daniel_hozac i was just about to ask that myself.
1146453474 M * Bertl Datagram Congestion Control Protocol
1146453483 M * Bertl http://www.icir.org/kohler/dcp/draft-ietf-dccp-spec-11.txt
1146453596 M * daniel_hozac sounds like a merger of UDP and TCP.
1146453638 M * Bertl yeah, well, more like somebody implementing a variant of tcp over udp, if you ask me :)
1146453731 M * Bertl but probably it has good use in streaming and the game world :)
1146453749 M * daniel_hozac that's what it says :)
1146453832 J * hillct ~H@client200-5.dsl.intrex.net
1146453855 M * hillct Bertl: ping
1146453867 M * Bertl hey hillct!
1146453923 M * hillct my reading of the testme.sh output and vci config header was that I had not compiled in vsched support
1146453927 M * hillct however
1146453948 M * daniel_hozac HARDCPU != vsched support.
1146453959 M * hillct it appears that I have a /proc/virtual/<xid>/sched file
1146453964 M * hillct ah
1146453987 M * hillct daniel_hozac: I knew they weren't the same. thought they went hand in hand
1146453996 M * hillct perhaps I misunderstood on that point
1146454003 M * daniel_hozac HARDCPU puts the context on hold when you go over the limit.
1146454010 M * hillct ah
1146454015 M * daniel_hozac (i.e. you have too few tokens)
1146454022 M * Bertl you can as well have prio adjustments based on the bucket
1146454046 M * hillct which I believe is what I need
1146454161 M * hillct so, is the format found in /proc/virtual/<xid>/sched the format needed for /etc/vservers/<server>/schedule ?
1146454184 M * hillct the wiki describes the command line format, but not the file format
1146454194 M * Bertl yes, it is somewhat identical
1146454207 M * hillct somewhat
1146454217 M * Bertl the kernel format evolved
1146454226 M * hillct minus the CPU lines ?
1146454227 M * daniel_hozac same order, first 5 lines, no description.
1146454229 M * Bertl the userspace format is still the one descibed
1146454270 M * Bertl btw, we probably want that changed soon, especially to reflect the per cpu settings
1146454276 Q * Vudumen Server closed connection
1146454279 M * daniel_hozac s/5/6/
1146454281 J * Vudumen ~vudumen@perverz.hu
1146454329 M * daniel_hozac i'm not really sure how to handle that. Enrico seems to only import the kernel headers from the stable release.
1146454330 M * hillct [root@vmaster ~]# cat /proc/virtual/40/sched
1146454330 M * hillct Token:                62
1146454330 M * hillct FillRate:              1
1146454330 M * hillct Interval:              4
1146454330 M * hillct TokensMin:            15
1146454332 M * hillct TokensMax:           125
1146454332 M * hillct PrioBias:              0
1146454334 M * hillct VaVaVoom:              0
1146454334 M * hillct cpu 0: 4282 361 0
1146454336 M * hillct cpu 1: 2342 502 0
1146454359 M * hillct daniel_hozac: up to and including PrioBias
1146454364 M * hillct is that correct?
1146454382 M * daniel_hozac and the logic for deciding which interface to use would get a lot more complicated as interfaces move between devel and stable...
1146454408 M * hillct oh fun. 
1146454418 M * hillct I'm using 2.01-pre2
1146454425 M * daniel_hozac hillct: yes.
1146454469 M * hillct so using this format would make for a less than seemless upgrade path going forward?
1146454506 M * hillct I guess it would be a trivial enough manual change...
1146454571 M * hillct daniel_hozac: I didn't see PrioBias described in the vsched wiki page
1146454576 A * hillct checks again
1146454616 M * daniel_hozac hillct: it's not required, but it's supported.
1146454677 M * hillct the name gives some idea of it's intent, but could you elaborate (so I'm not guessing)?
1146454837 M * Bertl it basically does a hard-renicing of a context
1146454867 M * Bertl while nice values will still be handled inside, the bias will change the resulting priority
1146454938 M * hillct ah
1146454954 M * hillct I gather this http://linux-vserver.org/Scheduler+Parameters is out of date
1146454978 M * hillct in that 6th line isn't a dummy line
1146455011 M * Bertl I'm not sure the tools do handle it (yet) but recent patches support the prio_bias
1146455023 M * hillct but correct in that the file should contain values only - as daniel_hozac indicated
1146455073 M * mugwump hmm, get niced too far and you have less priority than the idle task
1146455086 M * mugwump or 'realtime' priority, which has its own caveats
1146455125 M * Bertl the limits are clamped to the normal limits
1146455166 M * daniel_hozac the priobias should be supported by the utils.
1146455197 M * hillct 209?
1146455254 M * hillct hmm
1146455256 M * daniel_hozac IIRC.
1146455273 M * hillct why the hell do I have 208??? (question to self)
1146455288 M * hillct hmm
1146455319 M * Bertl hillct: probably because 210 is more than 3 months old now?
1146455324 M * daniel_hozac should be supported since 0.30.199.
1146455333 M * Bertl hillct: and you are using debian sarge?
1146455341 M * hillct FC4
1146455351 M * Bertl wow, I'm impressed :)
1146455354 M * daniel_hozac isn't sarge at 0.30.204 without bpo?
1146455359 M * hillct I last updated maybe 6 months back
1146455364 M * Bertl daniel_hozac: ah, yeah, right
1146455380 M * hillct Bertl: about what?
1146455393 M * Bertl 0.30.204 for sare
1146455395 M * Bertl *sarge
1146455401 M * hillct ah
1146455459 M * hillct so, in /proc/virtual/40/sched, what's the first line Token about?
1146455475 M * hillct doesn't seem to correspond to a defined parameter
1146455483 M * hillct the next 5 of course do
1146455510 M * hillct is that the current token ?
1146455519 M * Bertl the current token value
1146455527 M * hillct makes sense
1146455533 M * Bertl (the number of tokens in the bucket)
1146455544 M * hillct K
1146455859 M * daniel_hozac hmm, doesn't the watch context normally see all of the host too?
1146455880 M * Bertl yes, usually it does, why?
1146455899 M * daniel_hozac seems to be broken with lo0.05.1.
1146455919 M * Bertl ah, you need the nid watch context for that to work
1146455925 M * daniel_hozac ah, ok.
1146455932 M * Bertl (and I'm not sure that works yet)
1146455946 M * daniel_hozac EINVAL from vc_net_create(), so i'd say no :)
1146455959 M * Bertl try to enter it (migrate)
1146455987 M * Bertl (you can probably use vcmd for that)
1146456048 M * daniel_hozac i'll have to, the chbind logic only uses migrate if vc_net_create returns EEXIST.
1146456229 M * daniel_hozac hmm, does it support net_migrate?
1146456243 M * Bertl I hope so .. let me check
1146456299 M * daniel_hozac ok, so maybe i just don't understand how to use it then :)
1146456346 M * Bertl /tmp/vcmd -dnC net_migrate
1146456366 M * Bertl will tell you the command and that it doesn't take any arguments
1146456384 M * Bertl vcmd -i <id> -C net_migrate -- true
1146456408 M * daniel_hozac ESRCH
1146456437 M * Bertl okay, that means the kernel still blocks that
1146456447 M * daniel_hozac yep.
1146456499 M * Bertl was hoping we already added that, but I guess the dynamic id changes are missing for networking too
1146456835 M * Bertl we could simply (for now) sync the nid on xid=1 enter
1146456911 M * Bertl ah, now I see why I didn't do that for nid yet
1146456915 M * Bertl         /* dirty hack until Spectator becomes a cap */
1146456921 M * Bertl (from context.c :)
1146457001 M * Bertl okay, I'm off to bed now ... have a good one everyone!
1146457012 N * Bertl Bertl_zZ
1146457773 P * hillct Kopete 0.11.1 : http://kopete.kde.org
1146459711 Q * samuel_ Server closed connection
1146459727 J * samuel_ ~samuel@levinux.UQAR.UQUEBEC.CA
1146459747 A * samuel_ is away: (Auto-Away after 10 mins) [BX-MsgLog On]
1146469855 M * s0undt3ch Bertl_zZ: Fail Error on powerdown: "The tar command failed with error: .udev/queue/class@vc@vcs1: cannot stat: no such file/dir" and hanged until I Hit CRTL-D to keep going or enter root's pass for maintenence
1146471291 Q * Skram Server closed connection
1146471297 J * Skram ~mark@admins.sentiensystems.net
1146473934 J * bonbons ~bonbons@83.222.39.134
1146477554 Q * h01ger Quit: h01ger
1146477591 J * h01ger ~holger@socket.layer-acht.org
1146478830 J * Viper0482 ~Viper0482@p54977205.dip.t-dialin.net
1146479351 Q * _jake- Remote host closed the connection
1146479370 J * _jake- qscand@murlocs.org
1146480055 J * _coocoon_ ~coocoon@p54A06FB2.dip.t-dialin.net
1146480065 M * _coocoon_ morning
1146481088 Q * pusling Ping timeout: 480 seconds
1146481263 Q * softi42 Ping timeout: 480 seconds
1146481382 J * pusling pusling@195.215.29.124
1146482394 J * softi42 ~softi@p549D52BF.dip.t-dialin.net
1146485170 J * Nights Nights@62.150.149.163
1146485631 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1146487524 M * Nights hi
1146487645 Q * Nights Quit: If i die, don't cry, just look at sky, and say bye bye.
1146493456 N * Bertl_zZ Bertl
1146493460 M * Bertl morning folks!
1146493471 Q * doener__ Remote host closed the connection
1146493506 M * Bertl s0undt3ch: interesting, never heard of tar running on shutdown?
1146493594 J * doener ~doener@i5387C2AF.versanet.de
1146493704 M * Bertl morning doener!
1146493716 M * doener hey Bertl :)
1146493743 M * doener I'm up for a few hours now... that reconnect was just forced by a crash of that goddamn nvidia driver
1146493807 M * doener the infamous Xid bug...
1146494083 M * Bertl hmm, nvidia xid bug?
1146494195 M * doener ah, sorry... didn't think of the xid term in Linux-VServer... yeah, the syslog entry for the crash always says "Xid blah blah"
1146494228 M * doener totally unrelated to 'our' xids
1146494304 M * Bertl ah, good ...
1146494326 M * Bertl most of my systems have ati or matrox cards ...
1146494386 J * sim0n ~simon@212.76.255.43
1146494401 M * Bertl welcome sim0n!
1146494407 M * sim0n hi bertl
1146494431 M * doener got an ati x300 in my thinkpad... not happy with that either, s3/s4 sleep breaks if it lasts longer than about 5 minutes...
1146494463 M * doener ie. screen contains garbage and sometimes the system locks up hard
1146494482 M * sim0n i got an offtopic question...maybe somebody can help me
1146494510 M * Bertl doener: with 2.6 kernels?
1146494511 M * sim0n i just upgraded glibc, and accidently droppe root priviledges
1146494524 M * sim0n when i try to su i get this: su: U???}?]?u?7?????? 8: Error 18446744072499232852
1146494546 M * sim0n the new glibc should work though.... any idea how i can manage to become root again ? 
1146494569 M * doener Bertl: 2.6.16.something (gentoo kernel), using the x.org 7 ati drivers
1146494635 M * doener tried a lot of stuff... vbetool, dummy x server, s3_bios, etc. doesn't help
1146494663 M * Bertl sim0n: execute your suid sh?
1146494708 M * sim0n uhm... neither bash nor any other shell are suid masked :-(
1146494728 M * Bertl use sudo?
1146494740 M * sim0n not on this system
1146494748 M * doener maybe you forgot to turn off root ssh logins?
1146494750 M * sim0n i tried to preload or something...didn work
1146494755 M * daniel_hozac runuser?
1146494773 M * sim0n i'm on the server... ssh is broken as well...can't login anymore
1146494789 M * sim0n runuser isn't installed 
1146494803 M * Bertl sim0n: reinstall the system properly?
1146494850 M * Bertl sim0n: maybe a reboot helps (if you managed to update glibc successfully
1146494851 M * sim0n strange thing is...when i upgraded glibc via the pkg-manager, some programs were broken. when i manually extracted glibc, everything worked..... 
1146494855 M * daniel_hozac runuser is part of coreutils...
1146494890 M * sim0n can't find runuser (gentoo system)
1146494926 M * sim0n (i tried again to upgrade via the pkg manager and somehow dropped root.. and that's where i am now)
1146494965 M * sim0n could ld_preload work? if so, what should i preload ?
1146495209 J * hillct ~H@client200-5.dsl.intrex.net
1146495226 M * hillct good morning all
1146495227 M * Bertl welcome hillct!
1146495235 M * Bertl sim0n: what do you want to preload?
1146495321 M * hillct still playing with sched_prio. Am I correct that VaVaVoom in /proc/virtual/<xid>/sched is simply a numerical representation of the relative performance of the given vserver resulting from the token values set for it?
1146495357 M * hillct or is it something I can set as a modifier to the resulting performance?
1146495367 M * sim0n Bertl: tried libc ... i hoped that maybe su tries to use the wrong libc (nptl nonnptl) and i could preload the right one. hmmm
1146495377 M * Bertl hillct: more the power-bonus it gets because it is well behaving
1146495394 M * hillct I gathered so it's something I can set?
1146495398 M * sim0n Bertl: guess i have to get onsite and fix things there arghf
1146495423 M * Bertl sim0n: remote console?
1146495431 M * Bertl sim0n: rescue system boot?
1146495449 M * sim0n Bertl: can't login...neither serial nor ssh
1146495473 M * Bertl (that's why I said rescue boot)
1146495489 M * sim0n Bertl: rescue system boot not for this system unfortunately 
1146495491 M * hillct sim0n: I know I came to this late, what happened (briefly)
1146495518 M * sim0n hillct: upgraded glibc.... accidently dropped root, now many things like "su" are broken
1146495525 M * sim0n ~ $ su
1146495526 M * sim0n sukria: U???}?]?u?7?????? 8: Error 18446744072499232852
1146495543 M * hillct oh, fun
1146495550 M * sim0n hillct: not really ;-)
1146495556 M * hillct yeah, I've been there
1146495570 M * sim0n :-)
1146495579 M * Bertl what about a static shell btw?
1146495591 M * Bertl sim0n: let me guess, you have none?
1146495601 M * sim0n i do... but still unable to become root
1146495616 M * hillct what distro?
1146495619 M * sim0n gentoo
1146495651 M * sim0n it actually looks like everything using pam is broken
1146495652 M * daniel_hozac sim0n: runuser is in /sbin. you'll probably have to supply the path.
1146495710 M * sim0n daniel_hozac: checked on three systems... there is no runuser
1146495716 M * daniel_hozac i guess it might not work for users though.
1146495733 M * doener daniel_hozac: i don't have runuser either
1146495758 M * hillct sim0n: you said no rescue CD. Gentoo distro is a de facto rescue CD
1146495795 M * sim0n hillct: no no... i said i can't reboot this server to a rescue system
1146495802 M * hillct ah
1146495805 M * hillct I see
1146495814 M * sim0n hillct: i sure will soon get there and fix it with a livecd :-)
1146495843 M * hillct come on though, you missed the server. You wanted to visit it anyway right?
1146495847 M * hillct ;)
1146495858 M * daniel_hozac i guess runuser must be a fairly recent feature... maybe just present in 5.9x?
1146495867 M * sim0n haha.... 
1146495889 M * sim0n running stable here, so dunno
1146495927 M * sim0n anyway..that's kinda strange... i emerge glibc = many things broken.   i manually extract glibc (same one!), everythings works as usual
1146495946 M * daniel_hozac oh, nevermind. it's a Fedora patch. sorry.
1146495957 M * sim0n :-)
1146495983 M * sim0n are there any tools which don't depend on pam, to become root ?
1146496039 A * Bertl .o( hmm, I know why I removed pam :)
1146496069 M * sim0n :-D ... 
1146496108 M * sim0n if i build sudo without pam support...will it work on a pam-system ?
1146496111 M * hillct busybox maybe
1146496139 M * sim0n yeah but busybox is non-suid....i guess would be a *bad* idea to suid busybox ...
1146496140 J * restill ~restill@c-71-197-23-172.hsd1.mi.comcast.net
1146496155 M * Bertl welcome restill!
1146496161 M * restill hello all
1146496215 M * restill sorry I haven't been around in the past week.  I have been working on my distributed computing project.
1146496307 M * Bertl np
1146496378 M * Bertl daniel_hozac: I think I will make the xid=1 set nid=1 too for now, your opinion?
1146496415 M * daniel_hozac sure, that will make using it a lot easier too.
1146496505 M * daniel_hozac (at least with util-vserver)
1146496632 J * mkhl ~mkhl@200-148-41-58.dsl.telesp.net.br
1146496653 M * Bertl welcome mkhl!
1146496779 M * hillct good luck sim0n
1146496783 M * hillct later all...
1146496786 P * hillct Kopete 0.11.1 : http://kopete.kde.org
1146496794 M * sim0n thnx hillct
1146497024 M * orzel i'm sure this is a faq but can't find it on the website
1146497037 M * orzel how am i supposed to deal with daemon expecting to use syslog ?
1146497048 M * orzel it seems they try to open /dev/something and fails
1146497056 M * Bertl which daemon?
1146497058 M * daniel_hozac umm... start syslog?
1146497194 M * orzel lol, that could be that easy ?
1146497401 M * orzel Awh-ui- /proc/kcore..  and syslog-ng doesn't seem to be able to read it in the vserver
1146497430 M * Bertl syslog-ng is fine, you either want to enable syslog virtualization
1146497440 M * Bertl or disable the klog part in the config
1146497453 M * Bertl (the latter one makes more sense IMHO)
1146497505 M * orzel i was disabling klog in the config
1146497520 M * orzel now syslog-ng starts, but my daemon still doesn't
1146497533 M * orzel getservbyname failed for tcp at /opt/sympa/bin/Log.pm line 75        <--- on line 75 it tries to do things with syslog
1146497562 M * Bertl getservbyname looks like service resolving
1146497582 M * Bertl try logger to verify that syslog works
1146497584 M * orzel yes, i've looked but evrything seems ok /etc/services is readable by all and the same as on my other comptuers
1146497674 M * orzel logger works :) 
1146497678 M * orzel fun, i didn't know this one command
1146497696 M * orzel damn, and now it works
1146498205 M * Bertl okay, off for dinner now ... back later
1146498211 N * Bertl Bertl_oO
1146499349 Q * mire Quit: Leaving
1146499505 J * mire ~mire@1-167-222-85.COOL.ADSL.VLine.verat.net
1146499542 Q * bubulak Remote host closed the connection
1146500225 J * bubulak ~bubulak@cicka.wnet.sk
1146500241 N * Bertl_oO Bertl
1146500243 M * Bertl back
1146500300 M * ray6 front
1146500328 M * Bertl :)
1146500510 M * ray6 bertl: are you going to see that krenn/virtuozzo stuff on linuxtag? I'd be interested in what exactly they're doing, I've got an invitation to their "Hausmesse" presenting their "HA out of the box" solution which seems to be based on virtuozzo?
1146500538 M * ray6 (at least SWSoft is a prominet partner on the flyer :)
1146500571 M * Bertl ray6: definitely
1146500612 Q * mkhl Quit: 
1146500727 M * ray6 I really would have liked to see the friday and saturday presentations of openvz, vserver and xen, but I really won't find the time :/
1146500761 M * Bertl I will report ...
1146500827 M * ray6 thanks :)
1146500966 M * ray6 Bertl: BTW I've now got my first virtual server with over 1000 days real uptime (an UML guest)
1146500991 M * Bertl cool
1146501106 M * ray6 had to reboot my two highest xen hosts (401 and 430 days) last week, but xen has less potential for such high uptimes, as well as vserver I guess
1146501153 M * ray6 the advantage of UML is that you don't have to reboot the host for an UML version update. So a guest which doesn't need the update is not affected by an update of other guests
1146501153 M * Bertl well, if it wasn't for a power outage, I'd had a vserver with roughly 600 days
1146501188 M * Bertl a machine which is not really reachable from outside, just monitoring a few other machines, so no need to update
1146501366 M * Bertl oops, actually that would be ~850 days ...
1146501368 M * ray6 sure if that's true for all guests on the host. The fun part about UML is that the guests can run on different UML versions
1146501516 M * ray6 but I'm moving away from UML, so there are good chances that the uptime of this server is the highest uptime I'll ever have on a virtual guest. But who knows, possibly it runs for another few years :)
1146501606 M * ray6 the fun part is that the uptime seems to have rolled over at 2x497 days. Explanation for this? Was the jiffies counter changed from signed to unsigned sometime in the past?
1146501615 M * ray6 the host rolled over at 497 already
1146501664 M * ray6 hm, my highest vserver has 8 days :) But if I really move my faxgateway into one this could change :)
1146501676 M * Bertl well, I guess so, or the HZ value change
1146501687 M * Bertl *changed
1146501719 M * ray6 Bertl: but by a factor of two, so to 50? Hmm. It's a long time ago I compiled that kernel :)
1146501806 M * Bertl we had 1000, 250, 100 and for a short time 200
1146501854 Q * Viper0482 Ping timeout: 480 seconds
1146501993 M * Bertl ray6: the theory with signed vs unsigned is interesting, but why would that change the max?
1146502027 M * Bertl ray6: IIRC jiffies are just incremented, the uptime is calculated differently
1146502147 M * Bertl but the time function was also changed, and IIRC they moved the initial jiffie value, so maybe that caused the difference
1146502401 J * Viper0482 ~Viper0482@p549777D0.dip.t-dialin.net
1146502742 Q * nokoya Quit: changing servers
1146502919 J * ben_ ~Lukas@88.134.54.56
1146502920 M * ben_ hello
1146502925 M * Bertl hey ben_!
1146502963 M * ben_ Bertl, vserver and xfs going?
1146502992 Q * sim0n Quit: Client exiting
1146503004 J * nokoya young@hi-230-82.tm.net.org.my
1146503305 M * Bertl ben_: going, yes :)
1146503314 M * Bertl off for now .. back later ...
1146503317 M * ben_ cu
1146503321 N * Bertl Bertl_oO
1146503834 Q * lilalinux Ping timeout: 480 seconds
1146504340 J * lilalinux ~plasma@dslb-084-058-252-074.pools.arcor-ip.net
1146505122 J * mkhl ~mkhl@200-153-181-136.dsl.telesp.net.br
1146505533 Q * shedi Quit: Leaving
1146505783 J * Dr4g ~Dr4g@82-40-40-135.cable.ubr06.uddi.blueyonder.co.uk
1146506090 Q * orzel Remote host closed the connection
1146506237 J * shedi ~siggi@inferno.lhi.is
1146506252 J * orzel ~orzel@berlioz.ethernet.freehackers.org
1146506432 J * _coocoon_ ~coocoon@p54A07343.dip.t-dialin.net
1146506442 M * _coocoon_ evening
1146508287 Q * s0undt3ch Ping timeout: 480 seconds
1146509312 J * Aiken ~james@tooax6-088.dialup.optusnet.com.au
1146510725 T * * http://linux-vserver.org/ | latest stable 2.01, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.{0.2,1.1}-rc18 | util-vserver-0.30.210 | 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 ;)
1146510725 T * Bertl -
1146511981 J * s0undt3ch ~s0undt3ch@bl7-252-179.dsl.telepac.pt
1146512120 J * BASH_GUARD ~BASH_GUAR@ACB6A590.ipt.aol.com
1146512136 P * BASH_GUARD 
1146512711 Q * mkhl Quit: 
1146514496 Q * Viper0482 Remote host closed the connection
1146516574 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1146516604 J * _coocoon_ ~coocoon@p54A07343.dip.t-dialin.net
1146517373 Q * s0undt3ch Quit: leaving
1146518532 Q * bonbons Quit: Leaving
1146518813 Q * mire Quit: Leaving
1146519703 J * mire ~mire@73-167-222-85.COOL.ADSL.VLine.Verat.NET
1146521498 Q * Skram Ping timeout: 480 seconds
1146522637 Q * Dr4g Read error: Connection reset by peer
1146522901 Q * ben_ Quit: 
1146523040 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1146523974 J * s0undt3ch ~s0undt3ch@bl7-248-177.dsl.telepac.pt
1146524582 Q * shedi Quit: Leaving
1146525037 N * Bertl_oO Bertl
1146525041 M * Bertl evening folks!
1146525303 M * doener wb Bertl 
1146525429 M * Bertl hey doener!
1146525446 M * Bertl doener: did you find some time to check/test the lo0.05.1?
1146525471 M * doener not yet...struggling with php gotchas atm
1146525489 M * Bertl ah, np, was just curios
1146525633 M * doener a website needs some real-time background processing, decided to go with a daemon listening on a datagram unix socket... client demands it to be written in php... took quite a while till i figured out various things about signal handling in php
1146525679 M * Bertl ah, sounds tricky
1146525785 M * doener well, the functions are there, but signals are only handled between php instructions, so blocking reads also block signals. had to use select for my single socket, because that one is interruptible... (IIRC reads on blocking sockets were interruptible in C, right?)
1146525818 M * Bertl hehe, yeah
1146526224 M * derjohn Bertl, the lt0x-boxen will shutdown tomorrow 1100. if you want to upload rc18+ patches, do it now pls ;) [I think we will habe inet connection at linuxtag, but ....]
1146526286 J * _coocoon_ ~coocoon@p54A072E9.dip.t-dialin.net
1146526292 M * _coocoon_ hello
1146526416 M * daniel_hozac hi
1146526508 M * s0undt3ch hello ppl, I know this ain't the place but do you know of a ethernet capable box for hd's?
1146526698 M * _coocoon_ sorry but i am looking for network aliasing in the 2.6.16 kernel can't find ti
1146526700 M * _coocoon_ it
1146526763 J * Aiken_ ~james@tooax7-022.dialup.optusnet.com.au
1146526861 J * are|lunch are@dslb-084-057-130-113.pools.arcor-ip.net
1146527078 Q * Aiken Ping timeout: 480 seconds
1146527128 J * shedi ~siggi@inferno.lhi.is
1146527197 M * Bertl _coocoon_: hmm?
1146527215 M * Bertl s0undt3ch: like network attached storage?
1146527245 M * _coocoon_ ip aliasing i can't find it
1146527277 Q * _are_ Ping timeout: 480 seconds
1146527751 M * doener _coocoon_: you mean like support for eth0:foo?
1146527759 M * _coocoon_ yes
1146527762 M * _coocoon_ right
1146527800 M * doener I doubt that can be turned off at all
1146527826 M * _coocoon_ really hm?????
1146527830 M * Bertl _coocoon_: do you want to disable or enable it?
1146527838 M * doener the support that is, as aliasing isn't mandatory for multiple ip address on a single interface anymore since 2.4 IIRC
1146527839 M * _coocoon_ enable
1146527864 M * Bertl _coocoon_: aha, *waves magic wand* *pling* enabled :)
1146527883 M * Bertl try ifconfig eth0:test 10.0.0.1 now
1146527887 A * doener wonders what Bertl would have done for disabling it
1146527901 M * Bertl hehe
1146527912 M * _coocoon_ ok so i must set it manually with the latest kernel
1146527923 M * Bertl no, you are talking about vservers, right?
1146527927 M * doener you always had to do that manually...
1146527943 M * doener for vservers, you need to specify a name in the config if you want aliases
1146527947 M * Bertl _coocoon_: just add an entry 'name' in your tree based config
1146527968 M * Bertl or when you use --interface, then make it name=eth0:...
1146527984 M * doener in the same directory as the 'dev' file. the name should be without the leading 'eth0:' IIRC