1186015713 Q * bzed Quit: Leaving
1186015858 J * DoberMann_ ~james@AToulouse-156-1-46-191.w90-16.abo.wanadoo.fr
1186015965 Q * DoberMann[ZZZzzz] Ping timeout: 480 seconds
1186018987 J * friendly12345 ~friendly@ppp121-44-229-164.lns2.mel4.internode.on.net
1186020639 J * Freax ~Linus@bl7-129-104.dsl.telepac.pt
1186021068 Q * Linus Ping timeout: 480 seconds
1186025008 Q * Aiken Quit: Leaving
1186025105 M * Ashsong Evening folks.
1186025141 M * Ashsong I have a question about the COW immutable links.
1186025149 M * Ashsong Basically, they don't seem to be working for me.
1186025182 M * daniel_hozac is that on an XO?
1186025188 M * Ashsong Yeah.
1186025191 M * daniel_hozac i.e. with JFFS2?
1186025196 M * Ashsong I'm probably just missing something basic.
1186025198 M * Ashsong Yes.
1186025242 M * daniel_hozac well, the code is there...
1186025246 M * daniel_hozac so how doesn't it work?
1186025326 M * Ashsong I've been playing with the example on page 69 of 'main-en.pdf'
1186025346 M * daniel_hozac hmm? what PDF is that?
1186025353 M * Ashsong touch a; ln a b; setattr --iunlink a; echo 'hi' > a
1186025356 M * Ashsong basically.
1186025364 M * Ashsong With the echo done inside a container.
1186025367 M * daniel_hozac yeah, that should do it.
1186025373 M * Ashsong What's supposed to happen?
1186025408 M * daniel_hozac a should be a new file containing hi, b should have the iunlink attributes set still and be the old inode.
1186025428 M * daniel_hozac i take it that's not what's happening here?
1186025473 M * Ashsong After calling setattr on a, I ran vcontext --create --xid 2 -- echo 'hi' > a
1186025490 M * daniel_hozac not really necessary, but okay.
1186025494 M * cehteh eh
1186025509 M * cehteh > redirection is done in the calling shell .. 
1186025519 M * Ashsong Hmm.
1186025521 M * daniel_hozac heh, was just about to point that out.
1186025525 M * daniel_hozac but still, that doesn't matter.
1186025535 M * cehteh well .. just note that ;)
1186025536 M * daniel_hozac COW doesn't require you to be in a context.
1186025556 M * Ashsong Anyway, bash reports "a: Too many links"
1186025556 M * Ashsong and a file called "a?" is created.
1186025590 M * daniel_hozac is that with the latest kernel?
1186025592 M * Ashsong also, 'cat a' shows that the write was not performed.
1186025630 M * daniel_hozac it sounds as if there are some vserver fixes missing from your kernel.
1186025644 M * daniel_hozac not quite sure why it's failing in this case though.
1186025727 M * Ashsong it's with the stable branch of the olpc-2.6 repository: http://dev.laptop.org/git?p=olpc-2.6;a=tree;h=stable;hb=stable
1186025784 M * daniel_hozac yeah, that's missing two patches.
1186025793 M * Ashsong Ah.
1186025796 M * daniel_hozac http://vserver.13thfloor.at/Stuff/OLPC/delta-OLPC.0.4.1-0.4.2.diff and http://vserver.13thfloor.at/Stuff/OLPC/delta-OLPC.0.4.2-0.4.3.diff
1186025814 M * Ashsong Hmm.
1186025821 M * Ashsong Perhaps these are in the vserver branch but not yet merged into stable?
1186025840 M * daniel_hozac may be, i'm not sure how it works.
1186025859 M * Ashsong http://dev.laptop.org/git?p=olpc-2.6;a=shortlog;h=vserver
1186025896 M * Ashsong In any case, thanks for the tip.
1186025901 M * Ashsong I'll poke our kernel folks tomorrow.
1186026086 Q * eyck Ping timeout: 480 seconds
1186026108 M * daniel_hozac actually, those two patches don't seem complete either...
1186026182 M * daniel_hozac the relevant deltas are http://people.linux-vserver.org/~dhozac/p/k/delta-cow-fix10.diff and http://people.linux-vserver.org/~dhozac/p/k/delta-cow-fix11.diff, though i guess Bertl_zZ will update the OLPC branch.
1186026282 M * Ashsong Okay. I'll pass that along, and talk to Bertl tomorrow.
1186026418 M * daniel_hozac does your kernel have vserver debugging enabled?
1186026426 M * Ashsong yes.
1186026447 M * Ashsong i.e. sysctl vserver.debug_switch=255 dumps lots of output
1186026459 M * daniel_hozac okay, for this we want debug_misc=255
1186026467 M * daniel_hozac and then redo the test.
1186026474 M * Ashsong okay.
1186026481 M * daniel_hozac it might shed some light on why it's not working...
1186026512 M * Ashsong vserver.debug_misc is an unknown key.
1186026526 M * daniel_hozac hmm?
1186026526 M * Ashsong ah, wait.
1186026529 M * Ashsong probably my fault.
1186026543 M * Ashsong bingo.
1186026565 M * Ashsong (I'm in the process of building a new /sbin/init that does some filesystem twiddling before execing the real sysv-init
1186026566 M * Ashsong )
1186026577 M * Ashsong Consequently, /proc wasn't mounted. :)
1186026583 M * daniel_hozac heh.
1186026789 M * Ashsong well, it spewed some output.
1186026800 M * Ashsong I'll let you know as soon as I manage to get it off the system. :)
1186027356 M * Ashsong http://paste.lisp.org/display/45488
1186027471 M * daniel_hozac EINVAL... that's weird.
1186027589 M * Ashsong Ack, it appears that the pastebin misinterpreted some of the characters in that.
1186027609 M * Ashsong The real dump is at http://dev.laptop.org/~mstone/out
1186027610 M * daniel_hozac they're not important.
1186027613 M * Ashsong Look at time 153.37
1186027617 M * Ashsong Okay then.
1186027656 M * Ashsong The EINVAL is the vfs_sendfile return code?
1186027669 M * daniel_hozac yeah.
1186027779 M * Ashsong daniel_hozac: so it would seem.
1186027792 M * Ashsong Ack.
1186027817 M * Ashsong What's the status of ipv6 support these days?
1186027837 M * Ashsong That's another thing that appears to be absent from the olpc/vserver code I've seen so far.
1186027858 M * daniel_hozac oh really? i thought the OLPC branch was supposed to have that already...
1186027859 M * Ashsong vc_net_add() in particular seems to have cases only for ipv4
1186027896 M * daniel_hozac 2.3 should have IPv6 RSN, and the old patches seem to work alright.
1186027946 M * Ashsong e.g. http://dev.laptop.org/git?p=olpc-2.6;a=blob;f=kernel/vserver/network.c;h=67c311e64e068468d6aaccf36d14e887e1bf083d;hb=be30006dad097b85047dccc1a0b22b9ebc34da74#l503
1186027997 M * daniel_hozac indeed, that's completely lacking IPv6.
1186028056 M * Ashsong So I thought.
1186028078 M * Ashsong Well. It appears that we're going to have a very exciting time in the next two weeks... :)
1186028082 M * daniel_hozac another thing to poke your kernel folks about, i guess :)
1186028153 M * Ashsong Are there olpc/vserver patches I haven't seen yet that include IPv6 support?
1186028186 M * Ashsong Or is this also something that needs to be excerpted from the master VServer patches?
1186028224 M * daniel_hozac only patch i can find is timestamped march 2007, so that's not very up-to-date...
1186028253 M * daniel_hozac i guess it's something that got lost in the forward porting.
1186028276 M * Ashsong Sigh. :)
1186028295 Q * bored2sleep Ping timeout: 480 seconds
1186028296 M * Ashsong Well, I'm glad I found out today and not next week.
1186028307 M * neuralis hm, yeah, not sure what happened there
1186028498 M * Ashsong neuralis: we're going to have to think carefully in the next day or two (in concert with cscott, dilinger, and Bertl) about how to continue forward towards Trial-3.
1186028728 M * neuralis yeah. let's get the scoop from bertl first.
1186028734 M * Ashsong indeed.
1186028942 J * FireEgl FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa
1186033310 Q * Ashsong Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.5/2007071317]
1186035076 J * bored2sleep ~bored2sle@66.111.53.150
1186035141 N * DoberMann_ DoberMann
1186035273 J * Aiken ~james@ppp121-45-205-95.lns1.bne1.internode.on.net
1186035480 Q * friendly12345 Ping timeout: 480 seconds
1186036382 N * DoberMann DoberMann[PullA]
1186039171 N * BobR_zZ BobR
1186039314 J * dna ~naucki@116-204-dsl.kielnet.net
1186039590 J * friendly12345 ~friendly@ppp59-167-70-90.lns1.mel4.internode.on.net
1186039807 N * DoberMann[PullA] DoberMann
1186040020 J * bzed ~bzed@dslb-084-059-127-168.pools.arcor-ip.net
1186041348 N * Bertl_zZ Bertl
1186041386 M * Bertl morning folks!
1186041652 M * Supaplex ahoy :)
1186041695 M * Supaplex I thought about harassing some linux channel, and asking if 7 instances of cron, and 7 instances of syslog were a bad thing. ;)
1186041702 M * Supaplex but vpstree would be cheating eh? :P
1186041727 M * Bertl hmm?
1186041754 M * Supaplex one instance per host+guest
1186041777 M * Bertl ah
1186041810 M * Supaplex are you familar with how fakeroot operates?
1186041814 J * DavidS ~david@p57A49E04.dip0.t-ipconnect.de
1186041833 M * Bertl not really
1186041836 N * DavidS DavidS|Juelich
1186041856 M * DavidS|Juelich moin bertl :)
1186041857 M * Supaplex it uses a ld_preload library to intercept system calls
1186041909 M * Bertl ah, so it works only for glibc programs?
1186041913 M * Supaplex I think you could do the same thing instead of maintaining patches for all the system tools like top, pstree, etc.
1186041953 M * Bertl we don't have patches for top, pstree, etc :)
1186041987 M * Supaplex util-vserver: /usr/sbin/vpstree
1186041995 M * Bertl a simple wrapper
1186042006 M * Supaplex ahh I see
1186042014 A * Supaplex investigates further
1186042033 M * Supaplex so I could do the same for htop. sweet.
1186042044 M * Bertl probably :)
1186042056 M * Bertl in any case, there is no real point in a preload library
1186042085 M * Bertl because you cannot _add_ data easily there (and that is what you would need for the spectator context)
1186042106 M * Bertl and the isolation is already taken care of
1186042126 M * Supaplex groovy.
1186042139 M * Supaplex it's great to be wrong sometimes. :P
1186042243 M * Supaplex it works for lsof and htop so far. :)
1186042252 M * Bertl excellent!
1186042356 M * Supaplex want a patch?
1186042379 M * Bertl you mean for wrappers? sure! please send to daniel_hozac ...
1186042380 M * Supaplex http://riddle.daxal.com/~supaplex/vhtop.patch
1186042415 M * Supaplex it works on $0 so you can use it like vhtop vlsof etc.
1186042458 M * Bertl you want to consider the entwork context too (you probably want the spectator context there)
1186042495 M * Supaplex more than --ctx 1 ?
1186042623 M * Supaplex well you can view the patch now. I apparently had apache stopped.
1186042667 M * Bertl --xid 1 and --nid 1 :)
1186042675 M * Supaplex ok
1186042699 M * Supaplex exec $__SBINDIR/chcontext --silent --xid 1 and --nid 1 --ctx 1 $(basename $0 | sed s/^v//) "$@"
1186042753 M * Supaplex haha oops.
1186042763 M * Supaplex literal paste of and. :P
1186042766 M * Bertl nah, you need vcontext and ncontext for that
1186042821 A * Freax hello sir's :)
1186043181 M * Bertl hello Freax!
1186043335 M * Supaplex not a bunch of Freax here. only one huh? ;)
1186045412 Q * fatgoose synthon.oftc.net charm.oftc.net
1186045412 Q * balbir synthon.oftc.net charm.oftc.net
1186045446 J * fatgoose ~samuel@76-10-147-136.dsl.teksavvy.com
1186045446 J * balbir ~balbir@122.167.0.71
1186045853 N * BobR BobR_afk
1186047026 Q * bzed Quit: Leaving
1186048846 Q * FireEgl Ping timeout: 480 seconds
1186049829 J * yarihm ~yarihm@whitehead2.nine.ch
1186050222 J * lilalinux ~plasma@dslb-084-058-192-157.pools.arcor-ip.net
1186050269 J * Baby ~miry@195.37.62.208
1186051880 M * Bertl welcome Baby!
1186054236 J * ktwilight_ ~ktwilight@241.80-66-87.adsl-dyn.isp.belgacom.be
1186054459 M * daniel_hozac Bertl: did you see the discussion with the OLPC folks?
1186054480 M * Bertl yep, skimmed over it
1186054526 M * Bertl daniel_hozac: do we have tool side support for VCMD_*_match_ipv*?
1186054549 M * daniel_hozac not yet...
1186054561 M * Bertl okay, we will need that pretty soon
1186054579 M * Bertl i.e. 2.3.x will use/require that to get new functionality
1186054590 M * daniel_hozac right.
1186054648 Q * ktwilight Ping timeout: 480 seconds
1186054681 M * Bertl or should we do a new version of vc_net_add() for now?
1186054745 M * daniel_hozac well, it doesn't really matter i guess, the new functionality will still require tool support, no?
1186054769 M * Bertl yes, the question is just what kind of interface we use for now
1186054798 M * Bertl the first implementation will work with a simple sequential list with addr, net and range
1186054838 M * daniel_hozac i suppose for the utils a new version of vc_net_add would be most logical (as that's likely how i'll implement it for backwards compatibility)
1186054842 M * Bertl we can map that to the parent concept of the *match* functions or we can extend the net_add to handle that
1186054859 M * Bertl okay, then we add that for now
1186056054 M * Bertl daniel_hozac: does that look fine for you? http://vserver.13thfloor.at/Experimental/network_cmd.h
1186056088 M * Bertl mask is used as mask or second ip for ranges
1186056105 M * daniel_hozac so we'll need a mask in IPv6 too, no?
1186056125 M * Bertl depends on what we want to do for ipv6 addresses :)
1186056167 M * Bertl but we can add it, just to make sure we can use that too
1186056221 M * daniel_hozac looks good to me.
1186056222 M * Bertl reload
1186056290 M * daniel_hozac i'll try to add support in the utils later today...
1186056324 M * Bertl excellent, TIA!
1186056352 J * eyck_ ~eyck@nat.nowanet.pl
1186057013 M * Bertl daniel_hozac: just to clarify, I'll implement the following cases:
1186057066 M * Bertl NXA_TYPE_ANY, NXA_TYPE_ADDR, NXA_TYPE_MASK, NXA_TYPE_RANGE, NXA_TYPE_ADDR|NXA_MOD_BCAST and NXA_TYPE_ADDR|NXA_MOD_LBACK for ipv4
1186057092 M * Bertl and NXA_TYPE_ANY, NXA_TYPE_ADDR and NXA_TYPE_MASK (for now) for ipv6
1186057126 M * daniel_hozac does NXA_TYPE_MASK for IPv6 use the prefix or the mask?
1186057155 M * Bertl I think we should use the mask for that (and have the prefix too) no?
1186057203 M * daniel_hozac so there'll be an NXA_TYPE_NET which uses the prefix?
1186057234 M * Bertl yes, maybe we should add that now too
1186057269 M * Bertl probably the best is to fill in both, prefix and mask
1186057307 M * Bertl actually NXA_TYPE_ADDR allready includes the mask for v4 and the prefix for v6
1186057328 M * Bertl (single ip in a network)
1186057403 M * daniel_hozac okay
1186057434 M * Bertl we will have to do a bunch of tests to get a feeling for what matches are possible and/or efficient
1186057461 M * Bertl e.g. it would make sense to get the broadcast address from the single ip/mask/prefix
1186057720 Q * Supaplex Ping timeout: 480 seconds
1186057808 M * Bertl daniel_hozac: I would like to change NXA_TYPE_ANY to 0 and NXA_TYPE_ALL to ~0
1186057830 M * Bertl then NXA_TYPE_ALL could be used to remove all IPs
1186057865 M * Bertl actually we could even make that a single type
1186057880 M * Bertl and define that vc_remove* removes all _matching_ IPs?
1186057893 M * Bertl s/IPs/nx_addrs/
1186057917 M * Bertl does that make sense to you?
1186058011 J * FireEgl FireEgl@FireEgl.CJB.Net
1186058388 Q * the-me Ping timeout: 480 seconds
1186058519 M * daniel_hozac yeah, sure
1186058544 Q * cehteh Ping timeout: 480 seconds
1186058763 M * Bertl daniel_hozac: what about the endianess?
1186058775 M * daniel_hozac ?
1186058780 M * Bertl daniel_hozac: ah, never mind that is covered by the in_addr change :)
1186058802 M * daniel_hozac yeah, they should all be network byte order.
1186058948 M * Bertl the return code is not really meaningfull in the v1, i.e. I will switch to reporting errors only there
1186058957 M * Bertl net_add*v1 that is
1186059298 M * matti ;]
1186059458 M * Bertl daniel_hozac: I assume, you current implementation will only use NXA_TYPE_ANY for remove, for now?
1186059922 M * Bertl daniel_hozac: okay, here an update to the types: 
1186059924 M * Bertl #define NXA_TYPE_NONE           0x0000
1186059924 M * Bertl #define NXA_TYPE_ANY            0x00FF
1186059937 M * Bertl NXA_TYPE_NONE is not really relevant for userspace
1186060389 M * daniel_hozac okay.
1186060520 J * ema ~ema@rtfm.galliera.it
1186060746 Q * FireEgl Read error: Connection reset by peer
1186061131 J * Pazzo ~ugelt@195.254.225.136
1186061202 M * Pazzo hi @ll! 2.6.20.15, 2.6.21.6 or 2.6.22.1? I would opt for 2.6.20.15...
1186061324 Q * Aiken Remote host closed the connection
1186061565 J * FireEgl FireEgl@Sebastian.Atlantica.US.TO
1186061741 J * Piet ~piet@tor.noreply.org
1186062813 J * lilalinux_ ~plasma@dslb-084-058-247-189.pools.arcor-ip.net
1186063152 J * lilalinux__ ~plasma@dslb-084-059-030-082.pools.arcor-ip.net
1186063193 Q * lilalinux Ping timeout: 480 seconds
1186063373 Q * lilalinux_ Ping timeout: 480 seconds
1186063514 Q * Piet Remote host closed the connection
1186063975 Q * ||Cobra|| Remote host closed the connection
1186064222 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1186065190 P * friendly12345 
1186065728 J * Piet ~piet@tor.noreply.org
1186068952 Q * Freax Quit: I'll Be Back !
1186069833 M * Bertl okay, off for dinner ... back shortly ...
1186069838 N * Bertl Bertl_oO
1186069917 Q * gerrit Ping timeout: 480 seconds
1186070052 Q * Greek0_ Ping timeout: 480 seconds
1186070526 Q * yarihm Quit: Leaving
1186070609 N * ensc Guest678
1186070619 J * ensc ~irc-ensc@p54B4CA35.dip.t-dialin.net
1186070729 Q * Guest678 Ping timeout: 480 seconds
1186070817 J * derjohn_mobil ~aj@213.55.131.23
1186071167 Q * ||Cobra|| Remote host closed the connection
1186071228 Q * Pazzo Quit: ...
1186071444 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1186071884 J * gerrit ~gerrit@bi01p1.co.us.ibm.com
1186073019 N * Bertl_oO Bertl
1186073594 J * bonbons ~bonbons@2001:5c0:85e2:0:20b:5dff:fec7:6b33
1186074615 J * linus ~linus@bl7-129-104.dsl.telepac.pt
1186074619 Q * Piet Remote host closed the connection
1186074670 J * cehteh ~ct@pipapo.org
1186074746 J * Piet ~piet@tor.noreply.org
1186074759 Q * linus 
1186074902 J * linus ~linus@bl7-129-104.dsl.telepac.pt
1186075645 M * Bertl wb linus!
1186075743 M * matti B.
1186075744 M * matti :)
1186075787 M * daniel_hozac happy birthday matti!
1186075901 M * matti daniel_hozac: Thank you :)
1186076133 J * fatgoose_ ~samuel@76-10-147-136.dsl.teksavvy.com
1186076133 Q * fatgoose Read error: Connection reset by peer
1186076172 M * eyck_ matti? as in 'Top Gear Polar Expedition' matti?
1186076249 M * matti eyck_: Sorry, I don't follow.
1186076325 M * eyck_ no matter, my hash tables got updated with new associations, happy birthday.
1186076421 A * linus o.Ô
1186076546 M * eyck_ the more linuses the merrier?
1186076626 Q * ag- Read error: Connection reset by peer
1186076643 M * matti eyck_: Thanks mate.
1186076765 J * ag- ~ag@fedaykin.roxor.cx
1186076962 Q * ensc|w Remote host closed the connection
1186077071 Q * ema Quit: leaving
1186078994 M * baldy Bertl: `?
1186079008 M * baldy 20:23:11 up  9:15,  2 users,  load average: 455.04, 386.60, 230.69
1186079017 M * baldy second time.. only this day
1186079034 M * daniel_hozac so... figure out what's wrong. what are they all waiting for?
1186079045 M * daniel_hozac have you been able to get ps output?
1186079055 M * baldy daniel_hozac: nope ;(
1186079063 M * baldy vps faux wont work
1186079063 M * Bertl baldy: vps, vpstree 
1186079075 M * baldy vhost04:~# vpstree
1186079075 M * baldy vcontext: execvp("pstree"): No such file or directory
1186079093 M * Bertl maybe you should install pstree? :)
1186079103 M * daniel_hozac does ps faux work?
1186079113 M * baldy vhost04:~# vps aux
1186079117 M * baldy nothing.....
1186079136 M * daniel_hozac no v.
1186079147 M * baldy ahh oke
1186079150 M * baldy yep works
1186079207 M * daniel_hozac does vcontext --xid 1 --migrate ps faux work?
1186079224 M * baldy nope
1186079260 M * daniel_hozac does vcontext --xid 42 --create true work?
1186079364 M * baldy vhost04:~# vcontext --xid 42 --create true
1186079364 M * baldy New security context is 42
1186079385 M * daniel_hozac and it returns?
1186079399 M * baldy New security context is 42 <-- that
1186079414 M * daniel_hozac no, i mean, does the command return?
1186079421 M * daniel_hozac i.e. are you now back at your prompt?
1186079423 M * baldy ah yeo
1186079424 M * baldy p
1186079446 M * daniel_hozac okay, so replace true with ps faux.
1186079477 M * baldy vhost04:~# vcontext --xid 42 --create ps faux
1186079477 M * baldy New security context is 42
1186079477 M * baldy USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
1186079477 M * baldy root     22673  0.0  0.0   2316   716 pts/3    R+   20:30   0:00 ps faux
1186079477 M * baldy vhost04:~#
1186079581 M * daniel_hozac okay, so how about vcontext --xid <some guest> --migrate ps faux?
1186079608 M * baldy guest id or name?
1186079619 M * daniel_hozac it doesn't matter.
1186079644 M * baldy nothing
1186079666 M * daniel_hozac does ls -l /proc work?
1186079681 M * daniel_hozac (i.e. replace ps with that)
1186079692 M * baldy yep works
1186079713 M * daniel_hozac hmm.
1186079723 M * daniel_hozac ls -l /proc/1?
1186079736 M * baldy works too
1186079772 M * baldy lsof works too
1186079787 M * daniel_hozac vcontext --xid <some guest> --migrate bash -c 'ls -l /proc/*'
1186079841 M * daniel_hozac i guess cat /proc/*/status might be more useful.
1186079841 M * baldy after
1186079844 M * baldy -r--r--r-- 1 root root 0 2007-08-02 20:35 vinfo
1186079844 M * baldy -r--r--r-- 1 root root 0 2007-08-02 20:35 wchan
1186079851 M * baldy hangs it
1186079865 M * daniel_hozac hmm, that's interesting.
1186079870 M * daniel_hozac which process?
1186079890 M * daniel_hozac i.e. upload the ls -l /proc and all of the output from the current command.
1186079904 M * baldy http://paste.linux-vserver.org/4682
1186079908 M * baldy thats what it see
1186079913 M * baldy then it hangs
1186079950 M * daniel_hozac so... try ls -l /proc/<some pid inside>
1186079978 M * baldy i dont see any pid?
1186079993 M * daniel_hozac you got the ls -l /proc output, no?
1186080042 M * baldy ohh server hangs completely now
1186080061 M * baldy let me reset
1186080072 M * daniel_hozac what kernel is this again?
1186080109 M * baldy 2.6.22.1-vs2.2.0.3-rc1
1186080164 M * Bertl any kernel traces?
1186080188 M * baldy iptotal uses obsolete (PF_INET,SOCK_PACKET)
1186080188 M * baldy UDP: bad checksum. From 61.184.191.149:1541 to 89.106.73.5:1434 ulen 384
1186080190 M * baldy nothing
1186080250 M * daniel_hozac Bertl: sounds like some task_unlock missing, doesn't it?
1186080268 M * daniel_hozac not necessarily in vserver code...
1186080297 M * baldy funny is... IO load 100%
1186080320 M * Bertl daniel_hozac: could be, but could be anything from bad memory to broken driver code
1186080334 M * daniel_hozac well, sure.
1186080375 M * Bertl without a trace or some debug info via (serial) console
1186080388 M * baldy Bertl: ssh still works
1186080395 M * baldy but not all commands
1186080410 M * Bertl that sounds like I/O error on the root device
1186080419 M * Bertl does dmesg work?
1186080427 M * baldy yep
1186080438 M * Bertl save that and upload it somewhere
1186080483 Q * Piet Remote host closed the connection
1186080657 J * Piet ~piet@tor.noreply.org
1186080809 M * baldy http://paste.linux-vserver.org/4683
1186080904 M * Bertl hmm, nothing unusual
1186080925 M * Bertl IPVS, interesting ...
1186082108 M * baldy strange
1186082116 M * baldy cd /vservers/
1186082119 M * baldy then it ends
1186082132 M * daniel_hozac "ends"?
1186082142 M * baldy no returns and ssh freenz
1186082145 M * baldy -n
1186082164 M * daniel_hozac so everything gets back to normal?
1186082165 M * Bertl filesystem?
1186082365 Q * DavidS|Juelich Quit: Leaving.
1186082389 M * baldy seems so... fs broken...
1186082397 M * baldy maybe thats the problem with the IO
1186082420 M * Bertl it is just strange that you do not get any kernel messages
1186082430 M * daniel_hozac indeed...
1186082469 M * baldy when i wanne access /vservers/
1186082475 M * baldy then is ssh freez
1186082476 M * baldy strange
1186082480 M * baldy realy strange
1186082492 M * baldy VM: killing process couriertcpd(4072:#7019)
1186082493 M * daniel_hozac it's a separate filesystem, no?
1186082494 M * baldy first message
1186082500 M * baldy nope same :-/
1186082503 M * baldy same hdd
1186082518 M * daniel_hozac you have /vservers on the root filesystem?
1186082531 M * baldy ahh nope
1186082540 M * baldy sda2 is / and sda3 is /vserver
1186082549 M * daniel_hozac so... separate filesystem.
1186082879 Q * linus Quit: Democracy is also a form of worship.  It is the worship of Jackals by Jackasses.
1186083280 J * linus ~linus@bl7-129-104.dsl.telepac.pt
1186083376 Q * linus 
1186083849 J * Piet_ ~piet@tor.noreply.org
1186084221 Q * Piet Ping timeout: 480 seconds
1186084327 M * baldy any way to umount a fs which is busy?
1186084333 M * baldy --force or so?
1186084356 M * ard for nfs you can use -f -l
1186084362 M * ard -f for force and -l for lazy
1186084403 M * ard it means umount the fs, make it inaccessible for every process except the already open files/directories
1186084424 M * baldy yep
1186084425 M * baldy damn
1186084431 M * baldy fsck wont wirk
1186084433 M * baldy work
1186084463 M * ard for real block-devices you can do an mount -o remount,ro
1186084468 M * ard to force it readonly
1186084497 M * ard so any open files will suddenly become readonly
1186084540 M * ard that's really handy if you are planning a forced reboot, because you still have everything
1186084564 M * baldy the problem is load average: 987.22, ;)
1186084590 M * ard ah
1186084602 M * ard you have some blockdevice which is hanging?
1186084612 M * baldy dont know...
1186084636 M * ard it can rocket your load skyhigh, but make the system usable
1186084678 M * baldy e2fsck 1.40-WIP (14-Nov-2006)
1186084678 M * baldy /dev/sda3: clean, 990092/17104896 files, 6136954/34188328 blocks
1186084681 M * baldy funny is
1186084686 M * baldy when i wanne access /vsevers
1186084688 M * baldy ssh freez
1186084694 M * ard you will probably have a lot of processes in state D 
1186084698 M * baldy but normal conmmands works
1186084709 M * baldy what means D?
1186084735 M * ard uninterruptable sleep, waiting for something I/O which cannot be interrupted
1186084746 M * ard like reading a block from disk
1186084748 M * baldy yep IO load is 100%
1186084768 M * ard and iostat 1 shows no disk i/o I presume :-)
1186084792 M * baldy iostat?
1186084797 M * baldy where can i see it?
1186084802 M * ard just type it in:
1186084803 M * ard iostat 1
1186084809 M * ard shows i/o stats every second
1186084839 M * baldy on a debian host?
1186084844 Q * Baby Remote host closed the connection
1186084846 M * baldy command not found
1186084850 J * linus ~linus@bl7-129-104.dsl.telepac.pt
1186084858 M * ard ah
1186084862 M * ard aptitude install sysstat
1186084868 M * ard it's invaluable
1186084902 M * baldy ahhh
1186084949 M * ard there are probably other tools, but you must remember iostat, since it works on any real unix system
1186084956 M * ard aix is not a real unix system
1186084964 M * baldy hehe
1186084972 M * baldy i think he hhd is broken or so
1186084975 M * baldy hdd
1186085009 M * ard that's what I meant
1186085025 M * ard state D, which you can verify with top
1186085035 M * ard no cpu used
1186085057 M * ard usually means something bad in the I/O path, and usually because something is fried
1186085063 M * ard (or a kernel bug)
1186085110 Q * derjohn_mobil Quit: Verlassend
1186085273 J * fatgoose ~samuel@76-10-147-136.dsl.teksavvy.com
1186085684 Q * fatgoose_ Ping timeout: 480 seconds
1186085724 Q * linus Quit: Democracy is also a form of worship.  It is the worship of Jackals by Jackasses.
1186085816 J * linus ~linus@bl7-129-104.dsl.telepac.pt
1186085827 Q * fatgoose Read error: No route to host
1186087569 Q * nox Ping timeout: 480 seconds
1186087958 J * nox ~nox@static.88-198-17-175.clients.your-server.de
1186088057 Q * bonbons Quit: Leaving
1186088389 J * Aiken ~james@ppp121-45-205-95.lns1.bne1.internode.on.net
1186088777 M * Bertl okay, off to bed now ... have a good one everyone!
1186088783 M * daniel_hozac good night!
1186088788 N * Bertl Bertl_zZ
1186088791 M * Bertl_zZ daniel_hozac: you too
1186091088 J * c7a7uk ~c7a7uk@adsl-75-33-225-68.dsl.chcgil.sbcglobal.net
1186091439 Q * michal` Ping timeout: 480 seconds
1186091441 P * c7a7uk 
1186091930 Q * FireEgl Read error: Connection reset by peer
1186092648 J * michal` ~michal@www.rsbac.org
1186092904 J * FireEgl FireEgl@Sebastian.Atlantica.US.TO
1186092938 J * derjohn_mobil ~aj@213.55.131.23
1186093028 J * duisburg6681m ~duisburg6@dslb-084-062-140-149.pools.arcor-ip.net
1186093073 M * duisburg6681m hello ???
1186093081 M * daniel_hozac hi.
1186093262 Q * dna Quit: Verlassend
1186093557 N * DoberMann DoberMann[ZZZzzz]
1186093853 J * Baby ~miry@195.37.62.208
1186094423 Q * linus Quit: leaving
1186094909 Q * derjohn_mobil Ping timeout: 480 seconds
1186095083 Q * duisburg6681m Quit: Verlassend
1186095629 J * lilalinux_ ~plasma@dslb-084-058-217-006.pools.arcor-ip.net
1186095733 M * coderanger Bertl_zZ: Ping
1186095744 M * daniel_hozac _zZ ;)
1186095751 M * daniel_hozac doubt he's up already.
1186095755 A * coderanger hopes it is just out of date
1186095757 M * coderanger heh
1186095768 M * daniel_hozac well, he set it two hours ago.
1186095786 M * coderanger hrmm, okay, thanky
1186095804 M * daniel_hozac about COW/IPv6, or something else?
1186095808 M * coderanger v6
1186095826 M * daniel_hozac well, i think he's aware.
1186095844 M * daniel_hozac i pointed him to my conversation with Ashsong last night, at least.
1186095856 M * coderanger We just need to know if the lack of patch was an oversight or if it will take time to correct, and we need that ASAP
1186095977 M * daniel_hozac i think he's working on it. what you want is basically the network code from 2.3, which i think is nearing completion.
1186096004 J * derjohn_mobil ~aj@213.55.131.23
1186096038 Q * derjohn_mobil 
1186096059 Q * lilalinux__ Ping timeout: 480 seconds
1186096260 M * coderanger daniel_hozac: The issue is if we will have it in enough time that we should count on it for our networking security for first-ship
1186096311 M * daniel_hozac well, worst case scenario, you could just grab the IPv6 patch.
1186096314 M * coderanger we would need it in the next week or two at the latest to use it for our first run
1186096336 M * coderanger Okay, so it is already far enough along that we can count on _something_ being ready
1186096354 M * daniel_hozac (though that won't be API-compatible with the final version, i guess)
1186096566 Q * Piet_ Ping timeout: 480 seconds
1186097501 J * Piet_ ~piet@tor.noreply.org
1186097765 J * the-me Patrick@Linux-Dev.org