1220573427 Q * dowdle Remote host closed the connection
1220573589 J * xdr ~xdr@57-173-96-87.cust.blixtvik.se
1220573594 Q * xdr 
1220573601 J * xdr ~xdr@57-173-96-87.cust.blixtvik.se
1220573615 Q * xdr 
1220575558 Q * ViRUS_ Quit: Leaving
1220579665 J * quinq ~quinq@quinq.eu.org
1220579670 M * quinq hello :)
1220580288 T * * http://linux-vserver.org/ |stable 2.2.0.7, devel 2.3.0.34, grsec 2.2.0.7|util-vserver-0.30.215|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 forget about the minute. 
1220580288 T * ChanServ -
1220580482 J * balbir_ ~balbir@122.167.116.172
1220581995 Q * quinq Quit: Quitte
1220583894 J * ghislainocfs2 ~Ghislain@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr
1220584184 Q * ghislainocfs21 Ping timeout: 480 seconds
1220585869 Q * derjohn_mob Ping timeout: 480 seconds
1220587972 Q * pisco_ Remote host closed the connection
1220588092 J * pisco_ ~pisco@tor.noreply.org
1220590752 J * xdr ~xdr@57-173-96-87.cust.blixtvik.se
1220591669 Q * xdr Ping timeout: 480 seconds
1220591869 N * Bertl_zZ Bertl
1220591879 M * Bertl morning folks!
1220592399 Q * Dinde Quit: leaving
1220592699 J * sharkjaw ~gab@216-241-9.231210.adsl.tele2.no
1220592826 J * xdr ~xdr@191-173-96-87.cust.blixtvik.se
1220593251 Q * mcp Read error: Connection reset by peer
1220593251 J * emcepe ~mcp@wolk-project.de
1220593267 N * emcepe mcp
1220593267 M * Bertl hey emcepe! LTNS!
1220593309 M * Bertl mcp: same goes for you :)
1220594606 Q * larsivi Quit: Konversation terminated!
1220594705 M * hijacker morning 
1220594853 M * Bertl hey, how's going?
1220595782 J * derjohn_mob ~aj@e180212153.adsl.alicedsl.de
1220596117 J * chigital ~chigital@tmo-127-1.customers.d1-online.com
1220596677 Q * chigital Ping timeout: 480 seconds
1220597358 M * hijacker not bad
1220597369 M * hijacker at work
1220597372 M * hijacker yourself?
1220597421 M * Bertl still busy, but otherwise fine, tx!
1220597445 M * hijacker how's the weather like in Austria?
1220597458 M * Bertl mostly sunny and warm
1220598171 P * ghislainocfs2 
1220598339 J * larsivi ~larsivi@85.221.53.194
1220598369 N * DoberMann[PullA] DoberMann
1220598542 M * MooingLemur Bertl: what do you think of what Google did to Europe in their Chrome comic? :)
1220598584 M * MooingLemur http://www.theregister.co.uk/2008/09/02/chrome_european_map/
1220598589 Q * derjohn_mob Ping timeout: 480 seconds
1220598676 M * Bertl hehe, well, not the first time that folks from oversea get europe wrong :)
1220599025 M * fb Bertl: but they have google maps, ain't they? ;)
1220599089 M * Bertl a lot of folks _have_ powerful tools, that doesn't mean they are able to use them :)
1220599414 M * ard well said :-)
1220599985 M * fb Bertl: they created i should say :)
1220600113 M * Bertl and you think the guys dawing those comics were part of that? I sure hope not :)
1220600632 J * emcepe ~mcp@wolk-project.de
1220601007 Q * mcp Ping timeout: 480 seconds
1220601007 N * emcepe mcp
1220601095 M * ktwilight_ mornin' Bertl :)
1220601134 M * ktwilight_ i reran the tests with in very verbose mode, and i tried to decipher it, but couldnt' quite get my head around it. if you have the time, would be great if you can take a look :)
1220601146 M * Bertl sure
1220601151 M * ktwilight_ http://irc.13thfloor.at/LOG/2008-09/LOG_2008-09-04.txt
1220601160 M * ktwilight_ way bottom
1220601171 M * ktwilight_ here's the paste again http://rafb.net/p/D4ueuY48.html and http://rafb.net/p/YPJ2o073.html
1220601270 M * Bertl so 045 and 047 fails, yes?
1220601307 M * ktwilight_ yup it fails. it fails with a permission denied
1220601331 M * ktwilight_ 45 fails with a "no such file or directory"
1220601336 M * ktwilight_ 47 fails with a permission denied
1220601363 M * Bertl yeah, let's try to do those tests manually
1220601395 M * Bertl let me check first where the file_1 originally comes from
1220601440 M * Bertl it looks to me as if the file still exists when 033 executes
1220601449 M * ktwilight_ hm, i've looked at other tests and the path to file_1 seems to be correct, so am unsure...
1220601532 M * Bertl actually it seems to be still there on 034
1220601578 M * Bertl hehe, and we have a similar failure in 024
1220601594 M * Bertl (just there it is kind-of undetected
1220601648 M * Bertl that looks like the do_tag_read() is failing and reporting the 'No such file or directory'
1220601678 M * Bertl i.e. it is not the argument which is missing, it is one of the binaries which fails (for a yet unknown reason)
1220601695 M * ktwilight_ ah i see
1220601713 M * Bertl so let's check for 'cat' first
1220601754 M * Bertl add a line like this in do_test() right before the do_sync
1220601760 M * ktwilight_ k
1220601810 M * Bertl echo "CAT: `which cat` : `cat /proc/uptime`"
1220601832 M * Bertl and rerun the script as you did before
1220601905 M * Bertl but I think it is actually related to something cat does differently .. maybe the files actually become inaccessible at some point
1220601933 M * ktwilight_ CAT: /bin/cat : 675836.76 674701.82
1220601936 M * ktwilight_ that's the CAT output
1220601940 M * Bertl okay, so cat works fine
1220601943 M * ktwilight_ :)
1220601955 M * Bertl let's mount the xfs filesystem and do some of the tag related tests
1220601965 M * Bertl mount it with the 'tag' option please
1220602028 M * ktwilight_ k, done
1220602041 M * ktwilight_ i did a  mount -t xfs -o rw,tag /dev/loop0 /home/mnt
1220602065 M * Bertl now let's touch a file there, like this: touch /home/mnt/test
1220602079 M * ktwilight_ k, done. np there
1220602083 M * Bertl then change the tag with vtag 
1220602119 M * ktwilight_ uh, example please :)
1220602141 M * Bertl unfortunately _all_ my devel systems are down or unreachable atm
1220602148 M * Bertl so you have to check with --help
1220602154 M * ktwilight_ yup am on it
1220602161 M * ktwilight_ so am creating or migrating?
1220602235 M * Bertl ah, use chtag :)
1220602253 M * Bertl chtag -c 100 /home/mnt/test
1220602283 M * Bertl check with lstag /home/mnt/test
1220602290 M * ktwilight_ chtag not found...
1220602291 M * ktwilight_ hmm..
1220602316 M * Bertl chxid for older tools
1220602318 M * ktwilight_ oh chxid
1220602323 M * ktwilight_ found itin the testfs
1220602346 M * ktwilight_  $ lsxid /home/mnt/test
1220602346 M * ktwilight_                  100  /home/mnt/test
1220602358 M * Bertl okay, good, now let's try:
1220602399 M * Bertl cat /home/mnt/test
1220602410 M * ktwilight_ blank
1220602420 M * Bertl okay, we are only interested in errors
1220602426 M * ktwilight_ :)
1220602432 M * Bertl chcontext --xid 1 -- cat /home/mnt/test
1220602447 M * ktwilight_ blank
1220602454 M * Bertl chcontext --xid 2 -- cat /home/mnt/test
1220602464 M * ktwilight_ aha!
1220602465 M * ktwilight_  $ chcontext --xid 2 -- cat /home/mnt/test
1220602465 M * ktwilight_ New security context is 2
1220602465 M * ktwilight_ cat: /home/mnt/test: No such file or directory
1220602471 M * Bertl okay, that's fine :)
1220602536 M * ktwilight_ may be a bit dumb, but i tried xid {3,4,5} and it all failed
1220602543 M * ktwilight_ when tried 1 again, it works.
1220602543 M * Bertl yes, that's okay
1220602570 M * ktwilight_ k
1220602588 M * Bertl okay, I think I need a better idea what actually fails and shouldn't fail ... let's rerun the -xvvv test with 'bash -x'
1220602620 M * Bertl (either add the -x to the hashbang line or invoke it with bash)
1220602641 M * Bertl will give a lot of output so store and upload it somewhere
1220602757 M * ktwilight_ k, gimme a sec :)
1220602918 M * ktwilight_ Bertl, http://rafb.net/p/vfjvst70.html
1220603087 M * Bertl well, that one failed quite early :)
1220603100 M * Bertl because you have the fs still mounted
1220603170 M * ktwilight_ doh
1220603320 M * ktwilight_ ah k, got it :)
1220603734 A * ktwilight_ scratches head
1220603744 M * ktwilight_ i can't get it to output to a file using > OUTPUT
1220603771 Q * Mojo1978 Read error: Connection reset by peer
1220603804 M * Bertl you probably want something like this:
1220603815 M * Bertl ... >file 2>&1 3>&1
1220603862 M * ktwilight_ nope, incomplete too. i tried it with 2>&1 before as well, same thing. :/
1220604008 A * ktwilight_ htinks he found the solution :)
1220604263 M * MooingLemur perhaps: command &> file
1220604285 M * MooingLemur redirect both stdout and stderr to the file
1220604306 M * Bertl is basically the same as > .. 2>&1
1220604314 M * MooingLemur not exactly
1220604342 M * MooingLemur I think that's stdout to file, and then stderr to stdout
1220604346 M * ktwilight_ sweet got it
1220604353 M * ktwilight_ ./testfs.sh -xvvv -D /dev/loop0 -F xfs -M /home/mnt 2>&1 3>&1 | tee OUTPUT
1220604354 M * MooingLemur in that order
1220604356 M * ktwilight_ that worked ^
1220604364 M * Bertl MooingLemur: see 'man bash' search for &>
1220604381 M * ktwilight_ gonna paste the file somewhere now
1220604408 M * MooingLemur Bertl: hmm.. I might be thinking another shell.. sorry :|
1220604411 M * Bertl ktwilight_: yep, that is correct if you use '|'
1220604419 M * ktwilight_ :)
1220604434 M * Bertl MooingLemur: I doubt that 'other shells' will handle that very differently
1220604483 M * ktwilight_ http://rafb.net/p/VElgw173.html
1220604498 M * Bertl tx
1220604525 M * MooingLemur Bertl: Then I don't know why I'm confused :)
1220604563 M * ktwilight_ am confused from the beginning :)
1220604581 M * ktwilight_ some googling helped though
1220604586 M * Bertl hehe, np :)
1220604590 M * MooingLemur was that behavior the same in original bourne shell?
1220604628 M * Bertl not sure it had the x>&y redirectors
1220604642 M * MooingLemur I'm remembering crontabs on an old solaris box at an old job
1220604650 M * MooingLemur I think it used subshells to get the desired redirections
1220604669 M * MooingLemur (command 2>&1) | sendmail blah
1220604805 M * Bertl ktwilight_: ah, I see now, it seems that the tagging is active even without 'tag' option
1220604819 M * ktwilight_ which line is that?
1220604845 M * Bertl 2173 we start to mount without tag
1220604879 M * Bertl 2241 the read with xid=2 fails
1220604909 M * ktwilight_ ah
1220604946 M * ktwilight_ though it still looks like gibberish to me :)
1220604951 M * ktwilight_ my bash skills must suck really bad :/
1220604952 M * Bertl as I presume you want to use tagging, I consider that a minor problem for you
1220604975 M * Bertl but I remember having fixed that (or something quite similar) some time ago
1220604976 M * ktwilight_ hm, IIRC glusterfs uses tagging, so it's quite important
1220605010 M * Bertl I doubt glusterfs knows _anything_ about taggin (Linux-VServer tag option that is)
1220605043 M * Bertl but let me check the 2.6.22.19 branch for xfs
1220605052 M * ktwilight_ oh, so it's not about ext attr
1220605063 A * ktwilight_ finds some docs about vserver tagging
1220605064 M * Bertl please upload a 'testme.sh' output somewhere
1220605150 M * ktwilight_ Bertl, http://rafb.net/p/5ntWj776.html
1220605160 J * FireEgl FireEgl@173.16.9.10
1220605203 M * ktwilight_ hm, strange, when i run ./testme.sh i get all succeeded. but with -vvvv 011 throws an "Operation not permitted"
1220605222 M * Bertl that is part of the test .)
1220605285 M * Bertl I think I need to comment the tests somewhere
1220605315 M * ktwilight_ oh...
1220605329 M * Bertl eexec 011   1 "chcontext --secure $SID $UXID mknod $tmpdir/node c 0 0"
1220605331 M * ktwilight_ ah i see...it's to test the security.
1220605337 M * Bertl that means, it tries to create a device node
1220605342 M * ktwilight_ so operation not permitted is correct.
1220605346 M * Bertl yup
1220605349 M * ktwilight_ :)
1220606787 M * Guy- is there an elegant way to share an nscd instance running on the host among all guests?
1220606943 M * Guy- it listens on a unix domain socket
1220606959 M * Guy- I supposed I could bind mount the directory with the socket in all guests, right?
1220606966 M * Bertl well, you should be able to copy that into the guest(S)
1220606985 M * Guy- copy?
1220606993 M * Bertl bind mount will do nicely
1220607020 M * Guy- what did you mean when you said copy, btw?
1220607126 M * Bertl cp -va
1220607178 M * Bertl but nscd needs to be aware of those sockets then
1220607189 M * Bertl so the --bind is the better option for you
1220607303 M * Guy- as far as I know, copying a unix socket makes no sense whatsoever :)
1220607324 M * Guy- you can't listen on an existing socket (you have to unlink it and recreate it)
1220607343 M * Guy- or do I have that wrong?
1220607443 M * Bertl nah, I was thinking of modifying nscd to listen to more than one socket
1220607489 M * Bertl (i.e. forget about the copy, was thinking pipe there :)
1220608125 M * MooingLemur Am I missing something simple here?  http://pastebin.ca/1194683
1220608138 M * MooingLemur the vserver gets the host's /usr/portage/local
1220608182 M * MooingLemur oh.. duh.. it didn't exist :P
1220608260 M * MooingLemur or.. hmm.. still doesn't work as I'd expect
1220608310 M * PowerKe Does the host have /usr/portage/local? (so the mountpoint exists)
1220608413 Q * esa Ping timeout: 480 seconds
1220608413 M * MooingLemur the host had it, but I want the vserver to get its own
1220608456 M * PowerKe You still need the local directory on the host, as you're mapping /usr/portage in the guest
1220608469 M * PowerKe If the host doesn't have the directory, the mountpoint inside the guest doesn't exist
1220608513 M * MooingLemur oh, the host had it
1220608519 M * MooingLemur and the vserver didn't at first
1220608548 M * MooingLemur and they both had it, and it still gave me a writeable /usr/portage/local from the host in the vserver..
1220608565 M * MooingLemur I just made /usr/portage/local a symlink.
1220608575 M * Bertl are you operating in the proper namespace?
1220608589 M * MooingLemur which will suit this installation
1220608594 M * PowerKe writeable? that's odd since the /usr/portage mount is ro
1220608624 M * Bertl I mean, when you test things like '/usr/portage/local' for the guest?
1220608647 M * MooingLemur vserver guest enter
1220608675 M * Bertl okay, and you don't see the mounts there or what?
1220608678 M * MooingLemur I'm sure.. and the df reaffirms where that mount is coming from
1220608809 M * MooingLemur I'm not sure where it's getting confused.. but I think something is when you're bind-mounting over the top of another bind mount..  I'll set up a test case later.. but in the words of Bertl, "nap attack". :)
1220608830 M * Bertl enjoy!
1220608842 M * MooingLemur thanks for looking :)
1220609046 J * esa bip@62.123.8.50
1220610827 Q * nkukard Quit: Leaving
1220611929 J * friendly ~friendly@ppp118-208-172-153.lns10.mel4.internode.on.net
1220611947 P * friendly 
1220614207 J * ghislainocfs2 ~Ghislain@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr
1220614216 M * ghislainocfs2 hello
1220614271 M * Bertl hey
1220614326 M * ghislainocfs2 how are you Bertl :)
1220614372 M * Bertl fine, thanks, and you?
1220614584 M * ghislainocfs2 fine fine, still struggling to get a vserver setup with a san that is a global FS :) i cannot beleive nobody use this kind of setup apart from using nfs servers eheh
1220614597 M * ghislainocfs2 how's your ocfs2 coding goes ?
1220614689 M * Bertl well, not much changed since yesterday .. except for the fact that one of my test systems is gone :/
1220614698 M * Bertl (unreachable atm)
1220614712 M * ghislainocfs2 erf
1220614721 M * Bertl but let's do some tests with the patch/setup you had yesterday
1220614740 M * Bertl so that I get an idea what could possibly be wrong
1220614794 M * Bertl please re-run the testfs.sh-0.20 with 'bash -x', or add the '-x' to the hashbang line of the script
1220614805 M * Bertl (for the OCFS2 setup, that is)
1220614837 M * Bertl redirect the output (> ... 2>&1) to some file, and upload that somewhere
1220614850 M * ghislainocfs2 in the paste bin ?
1220614895 M * Bertl yep, for example
1220614963 M * ktwilight_ ./testfs.sh -xvvv -D /dev/loop0 -F xfs -M /home/mnt 2>&1 3>&1 | tee OUTPUT
1220614964 M * ktwilight_ :)
1220614965 Q * balbir_ Read error: Operation timed out
1220615205 M * ghislainocfs2 bertl, will that fit : bash -x   ./testfs-.20.sh  -yvvv -D /dev/etherd/e10.1 -F ocfs2 -M /test
1220615236 M * Bertl yep
1220615636 J * loddafnir ~mike@chello080108111137.5.11.univie.teleweb.at
1220615870 M * ghislainocfs2 dam i have te restart it
1220615879 M * ghislainocfs2 i forgot the LANG=C
1220616770 J * quinq ~quinq@quinq.eu.org
1220616772 M * quinq hi :)
1220616824 M * Bertl hey
1220616897 J * biz biz@baze.de
1220616907 M * biz Hello :-)
1220616921 M * biz I've just bootstrapped an Ubuntu Hardy system and installed a vanilla 2.6.22.19 kernel with linux-vserver and grsec patches. The boot process stalls for about 2 minutes trying to start klogd. http://baze.de/~biz/pics/klogd_stalls.png - any ideas?
1220616960 M * daniel_hozac on the host?
1220616986 M * biz yes. Sorry I can't paste it, the png is a screenshot from a LARA remote console
1220617021 M * Bertl well, as you can see, grsec kills klogd 
1220617037 M * biz it's SIGSEGV
1220617052 M * biz Is this a missing CAP or something?
1220617061 M * Bertl yep, probably because it tries to access protected file or so
1220617079 M * Bertl but don't ask me for the details, I don't use grsec
1220617080 M * quinq that's a procps issue
1220617135 M * biz quinq: you got more info on it? There are some forum threads about it, but I was unable to fix it yet
1220617141 M * biz Eg. http://forums.grsecurity.net/viewtopic.php?f=3&t=1972
1220617177 M * Bertl disable grsec for a test?
1220617257 M * ghislainocfs2 http://paste.linux-vserver.org/12435
1220617263 M * ghislainocfs2 sorry for the delay
1220617267 M * Bertl np
1220617270 M * biz Ok, let me try it :-)
1220617338 M * quinq nop biz, I was just talking about googling ;)
1220617522 Q * pisco_ Ping timeout: 480 seconds
1220617551 M * biz Indeed, it works without grsec enabled, klogd starts up just fine
1220617586 M * biz This is the complete error fetched out of dmesg: (2 lines)
1220617599 M * biz klogd[1465:#0]: segfault at 0000000000000000 rip 00002b22b6cabc67 rsp 00007ffff4085b50 error 4
1220617602 M * biz grsec: signal 11 sent to /sbin/klogd[klogd:1465] uid/euid:103/103 gid/egid:104/104, parent /sbin/klogd[klogd:1464] uid/euid:103/103 gid/egid:104/104
1220617619 M * daniel_hozac run klogd under gdb, get a backtrace.
1220617774 J * pisco ~pisco@tor.noreply.org
1220617902 Q * Aiken Remote host closed the connection
1220617956 M * Bertl ghislainocfs2: can you try to figure what the 'T' flag means for OCFS2?
1220618109 M * ghislainocfs2 T flag on what ?
1220618132 M * Bertl files, chattr/lsattr
1220618166 M * Bertl ++ echo -e '\tverify /test/file_3896: -+(-)-i-+(-)?([AE])+(-) = ----i-----------T-'  
1220618169 M * ghislainocfs2 ok
1220618179 M * Bertl the 'i' is immutable, the 'T' normally only applies to dirs
1220618459 M * ghislainocfs2 yes
1220618765 Q * larsivi Quit: Konversation terminated!
1220618854 M * ghislainocfs2 i donot find anything
1220618898 M * Bertl okay, thanks!
1220618998 M * daniel_hozac you see files with T?
1220619016 M * Bertl yep
1220619033 M * Bertl maybe a flag collision somewhere ...
1220619114 M * Bertl I really hate when the build/test systems are not available
1220619115 M * daniel_hozac OCFS2_IXUNLINK_FL
1220619159 M * Bertl does that collide with whatever 'T' is?
1220619179 M * Bertl I thought that was unused, maybe actually is unused
1220619180 M * daniel_hozac yeah, FS_TOPDIR_FL is also 0x00020000.
1220619192 M * Bertl ah, okay, then we need to change that#
1220619199 M * biz daniel_hozac: I've just strace'd it and run it with 'klogd -dn' (debug, no detach) .. and I could solve the segfault using the klogd "-x" switch. So I'm on the next problem.. "klogd -x" just sits there, "stalling" again
1220619212 M * daniel_hozac i think we should just use FS_*_FL in fs/ocfs2/ocfs2_fs.h.
1220619227 M * daniel_hozac the comment above them seem to suggest that's what mainline should be doing too.
1220619250 M * Bertl the question is, why is mainline _not_ doing it :)
1220619285 M * Bertl but yeah, I'm fine with that
1220619296 M * daniel_hozac maybe OCFS2 doesn't get the love it should...
1220619307 M * biz it's the same behaviour klogd hat when run within a guest system.. no errors, klogd just waiting for something. I'll have a look at the last action in the strace
1220619314 M * daniel_hozac IIRC the FS_*_FL flags haven't been around for _that_ long.
1220619314 M * biz s/hat/has/
1220619335 M * Bertl biz: in a guest, hanging around is fine
1220619376 M * biz Bertl: out of interest, do you know what's causing this?
1220619407 M * Bertl in the guest, if you have the syslog virtualization on, klogd will not be able to read anything
1220619422 M * Bertl if it is off, klog will not be allowed to open the file
1220619437 M * Bertl IIRC, most klog implementations do not handle that case very well :)
1220619457 M * biz Ah, ok. I'm just trying to get an idea for my non-guest machine 
1220619478 M * Bertl did you try to gdb klog there and see why it segfaults?
1220619511 M * biz Yeah.. it was because of no usable System.map
1220619530 M * daniel_hozac that's, umm, "odd".
1220619539 M * Bertl ghislainocfs2: okay, I should have a working kernel build system in about 20 minutes
1220619542 M * biz The segfault is gone using "-x" (to not do EIP translations, so imho not use System.map)
1220619586 A * Bertl is not sure if he would want klogd messing with the kernel output :)
1220619612 M * biz Is it safe to disable klogd on the main system?
1220619634 M * Bertl yes, but you won't get any of those 'dmesg' infos stored on disk
1220619657 M * biz So klogd is just dmesg and syslogd is for the rest?
1220619665 M * Bertl yep
1220619688 M * Bertl guests, for example, do perfectly fine without klog
1220619698 M * Bertl (as there is nothing to log anyway)
1220619972 M * biz http://baze.de/~biz/files/klogd_strace_stalls_here.log
1220619986 M * ghislainocfs2 oky bertl
1220619989 M * biz this is where it stalls for about 3 minutes
1220620234 M * Bertl maybe it is doing a lot more than it is supposed to do, and some of that might involve name resolving or so?
1220620357 M * biz if you reload the file, you can see the appended rest that happens just about 5 minutes later
1220620412 M * biz Yeah.. that's possible. Though I have no DNS problems nor do I use netlogging in any form
1220620413 Q * sharkjaw Remote host closed the connection
1220621011 J * Mojo1978 ~Mojo1978@ip-88-152-62-183.hsi.ish.de
1220621206 M * biz I just got rid of it... thanks for your help, Bertl, daniel_hozac and quinq :-)
1220621217 M * Bertl np
1220621469 Q * xdr Remote host closed the connection
1220621887 Q * quinq Remote host closed the connection
1220621923 J * quinq ~quinq@quinq.eu.org
1220621995 M * Bertl hmm, kernel.org seems to be down/unreachable too ... strange
1220622089 M * fb Bertl: why strange?
1220622135 M * Bertl well, did you expect it to be down? do you know something I don't? :)
1220622176 M * fb Bertl: Bertl try www.eu.kernel.org ;)
1220622188 M * quinq you didn't know about the prophecy ?
1220622197 M * Bertl fb: yeah, no problem, I know a bunch of mirrors :)
1220622227 J * xdr ~xdr@191-173-96-87.cust.blixtvik.se
1220622257 M * fb Bertl: main host seems to be up, but isc's uplink lags strangely
1220622265 M * fb maybe some DoS
1220622307 M * Bertl DoK! a micro$oft(tm) plot?
1220622316 Q * Mojo1978 Remote host closed the connection
1220622317 M * Bertl (just kidding :)
1220622328 M * fb quinq: what prophecy?
1220622430 M * quinq about kernel.org going down on the fifth of september two thousand and nine
1220622463 M * Bertl well, that prophecy is off by one year then :)
1220622483 M * quinq hummm
1220622492 M * quinq meybe that wasn't our calendar at that time
1220622499 M * quinq or their bios battery was down
1220622685 M * fb quinq: why it's supposed to go down then?
1220622723 M * quinq I don't know, nobody was able to read the other part
1220622842 J * ntrs ~ntrs@77.29.74.58
1220623350 M * fb :>
1220623527 J * dowdle ~dowdle@scott.coe.montana.edu
1220623842 M * nox why does vhashify only work for running vservers?
1220623873 J * vargadani ~vargadani@catv-89-135-16-217.catv.broadband.hu
1220623892 M * vargadani hey there
1220623902 M * vargadani did any of you have problems with locales in a guest?
1220623912 M * vargadani i get locale warnings when using perl
1220623958 M * ghislainocfs2 bertl, do i missed your mesages i was not here for some time ?
1220624023 M * nox vargadani: debian?
1220624030 M * vargadani no... ubuntu
1220624035 M * vargadani so basically yes, debian :)
1220624046 M * quinq dpkg-reconfigure locales
1220624062 M * vargadani nox: what is that doesn't work?
1220624089 M * vargadani i tried that but the same warning seems to appear
1220624114 M * quinq which warning ?
1220624136 M * nox locales and console-data are normally the prob
1220624137 M * vargadani perl: warning: Setting locale failed.
1220624137 M * vargadani perl: warning: Please check that your locale settings:
1220624165 M * Bertl ghislainocfs2: nah, took a little longer than expected :(
1220624171 M * nox it is nothing vserver related
1220624175 M * quinq echo $LANG ?
1220624212 M * vargadani en_US.UTF-8
1220624265 M * vargadani locale -a only results in POSIX and C
1220624327 J * Rockj roysn@tihlde.org
1220624338 M * quinq dpkg-reconfigure -plow locales ? :)
1220624349 M * Bertl welcome Rockj!
1220624360 M * Rockj Ey. Is there any special patch for 2.6.26.3 kernel that I need to use? :)
1220624400 M * Bertl http://vserver.13thfloor.at/Experimental/
1220624437 M * vargadani alright... didn't work, I will try to do sth about it and let you know what i did...
1220624439 M * vargadani thanx though
1220624440 M * vargadani bye
1220624458 M * quinq good luck :)
1220624477 M * Bertl vargadani: 
1220624495 M * Bertl what kind of warnings do you see, and when?
1220624537 M * Bertl I presume, you get a bunch of them when you 'enter' a guest from your host, and do something with perl, right?
1220624551 M * Rockj Bertl: http://vserver.13thfloor.at/Experimental/patch-2.6.26.3-vs2.3.0.35.diff , I assume? :)
1220624573 M * Bertl yep, and you can add the cowbl-fix02, rpc-fix01 and ocfs2-fix02 too
1220624585 M * Bertl (then you should be up-to-date)
1220624591 M * daniel_hozac nox: it needs to get the list of config files from inside the guest.
1220624613 M * vargadani Bertl: yes... that is right... I entered the guest and did some things in perl
1220624619 M * Bertl daniel_hozac: do we still need that with CoW link breaking?
1220624625 M * vargadani the errors are like this: locale: Cannot set LC_CTYPE to default locale: No such file or directory
1220624638 M * vargadani locale -a returns only C and POSIX
1220624658 M * nox daniel_hozac: so only way is to start the server/hashify and stop them afterwards?
1220624660 M * Rockj Bertl: btw, is it bad to run experimental vserver? I normally use latest kernels on my server (which I use for mail and my websites) with no problems . I guess vserver wont add any difficulties :)
1220624667 M * Bertl vargadani: just do the following before you enter the guest: export LC_ALL=C LANG=C
1220624693 M * Bertl Rockj: I don't hope so, but if you encounter anything, please let us know
1220624713 M * vargadani Bertl: and what if I connect to the host through SSHd?
1220624730 M * vargadani do I just simple export these and set the env vars?
1220624733 M * Bertl vargadani: you mean, to the guest? then it should be fine
1220624740 M * vargadani alright
1220624747 M * vargadani i will make an attempt to accomplish this
1220624749 M * vargadani thanx
1220624766 M * Bertl the problem is that you bring your environment (and the locale) from the host into the guest (with the enter)
1220624784 M * daniel_hozac nox: well, you can do it manually if you want.
1220624790 M * Bertl vargadani: and the guest doesn't have the locale from your host installed yet
1220624831 M * daniel_hozac Bertl: vhashify doesn't really know about CoW yet.
1220624852 M * Bertl ah, okay, so future feature ...
1220624862 M * vargadani Bertl: to be honest it isn't that clear yet...
1220624887 M * vargadani do I need to install the locale of the host to the guest?
1220624899 M * Bertl vargadani: when you ssh to the host, and you do: 'locale', you'll see that you have a bunch of settings there
1220624907 Q * xdr Remote host closed the connection
1220624946 M * Bertl vargadani: when you 'enter' the guest (note: not via ssh, but via vserver guest enter), you basically bring your host locale into the guest
1220624973 M * Bertl vargadani: but that guest doesn't have the necessary files installed for your host's locale settings
1220625029 M * vargadani I c
1220625046 M * Bertl so your options are:
1220625057 M * Bertl - use ssh instead of enter (for the guest)
1220625071 M * Bertl - use the export line above to unset your locale (on the host)
1220625080 M * Bertl - install the host locales in the guest
1220625095 M * Bertl - well, and finally, ignore the warnings :)
1220625184 J * pmenier ~pmenier@ACaen-152-1-17-137.w83-115.abo.wanadoo.fr
1220625192 M * Rockj *compiling kernel* 
1220625254 M * vargadani Bertl: hehe... yes... the last one was the one that I was using, but I want to get it work well... how do I install the host's locales to the guest? isn't that enough if I let's say put the proper lines in the /etc/environment file in the guest?
1220625298 M * ghislainocfs2 is this debian ,
1220625307 M * quinq ubuntu
1220625314 M * ghislainocfs2 aptitude instal locales should do fine
1220625333 M * quinq or as said, reconfigure ^^
1220625349 M * vargadani well, I tried those :)
1220625358 M * daniel_hozac reconfigure doesn't really work if it's not installed.
1220625400 M * vargadani locales is installed
1220625405 M * quinq yeah but he said it is
1220625413 M * quinq so.
1220625413 M * vargadani i used dpkg-reconfigure locales
1220625418 M * Bertl vargadani: in the guest?
1220625421 M * vargadani yes
1220625426 M * quinq and choose your host's locales ?
1220625439 M * vargadani in fact, it didn't give me any output
1220625452 M * quinq with the -plow ?
1220625518 M * vargadani no
1220625523 M * vargadani it doesn't work
1220625562 M * quinq humm Ubuntu mayeb doesn't use locales packages as Debian, but another one
1220625565 M * quinq *maybe
1220625620 M * quinq you have to edit /var/lib/locales/supported.d/local
1220625624 M * quinq and then reconfigure locales
1220625652 M * vargadani i don't even have this file :)
1220625706 M * quinq damn Ubuntu
1220625708 M * ghislainocfs2 drop ubuntu or read the fine manual ;)
1220625709 J * [BiG^BrotheR] Dr-DreaM@87.101.133.222
1220625716 M * vargadani hehe :)
1220625725 M * vargadani i only wish to drop ubuntu
1220625755 M * vargadani quing: that one workes... I put en_US.UTF-8 UTF-8 into the non-existent file
1220625759 M * vargadani and reconfiged locales
1220625761 M * ghislainocfs2 less /usr/share/i18n/SUPPORTED
1220625763 M * vargadani now it seems to work
1220625776 M * ghislainocfs2 locale-gen <whatever locale>
1220625783 M * quinq yeah, or man locale
1220625806 M * vargadani :)
1220626130 Q * [BiG^BrotheR] Quit: ][DreaM-ScripT][
1220626837 Q * pisco Remote host closed the connection
1220626842 M * fb quinq: under ubuntu you need to install language-support files
1220626897 M * fb language-pack-{lang_code} 
1220626930 J * pisco ~pisco@tor.noreply.org
1220627056 M * quinq oh I see
1220627064 M * quinq thanks about the hint
1220627092 M * quinq and do you have to keep the whole lang_code locales or you can only keep the ones you want ?
1220627127 M * Bertl they shouldn't be that large, but you basically only need the ones you are using
1220627136 M * ghislainocfs2 by the way nobody use a cluster file system with vserver here ?
1220627141 M * ghislainocfs2 i feel lonely.. lol
1220627166 M * Bertl :)
1220627184 M * Bertl I should have my stuff ready shortly, still installing a few essential things
1220627216 M * Rockj oki, I think I got this stuff up and running actually. if host is running debian lenny, I could fine test other linux distro's yes? like ubuntu server? 
1220627226 M * Bertl yep
1220627285 Q * vargadani Remote host closed the connection
1220627357 M * quinq Bertl, what I was saying is, does language-pack-{lang_code} gives the possibility to actually choose which ones you want :)
1220627381 M * Bertl no idea, I don't use debian/ubuntu 
1220627422 M * Rockj I think I forgot something important when I compiled new kernel
1220627428 M * Rockj proc-security stuff
1220627477 M * Rockj which is enabled
1220627547 M * quinq they don't work the same, Bertl 
1220627584 M * daniel_hozac ghislainocfs2: not everyone has the hardware to set it up :-)
1220627849 M * Bertl daniel_hozac: do you remember what the reason for 'qemu not showing _anything_ when booting a kernel with nographic' was?
1220627876 M * Bertl I know, I figured that once, but I can't remember right now
1220627904 M * Rockj Bertl: what default stuff does guest need access to in /proc?  /proc/uptime and what more?
1220627927 M * Bertl do you get a message telling you that you don't have proc/uptime?
1220627938 M * Bertl if so, run vprocunhide on your host :)
1220627978 M * daniel_hozac Bertl: correct -append and serial console enabled? i don't know what else it would be.
1220628001 M * Bertl it just sits around (the kernel) in some delay loop
1220628001 J * xdr ~xdr@191-173-96-87.cust.blixtvik.se
1220628039 M * daniel_hozac strange. i don't recognize that though.
1220628070 M * Bertl I had that once, I still remember ... was some weird config option which needed to be turned off or on
1220628087 M * ghislainocfs2 daniel: yes of course, but i am a one man shop and can afford it (aoe blade is 3k$ still a lot but not so much for a san) so i would think that bigger players using vserver would have something setup like this
1220628104 M * Bertl daniel_hozac: do you have a qemu useable config for 2.6.26.3/vs somewhere?
1220628105 M * ghislainocfs2 daniel: perhaps they do not want to communicate also :)
1220628111 J * vargadani ~vargadani@catv-89-135-16-217.catv.broadband.hu
1220628139 M * Bertl ghislainocfs2: most Linux-VServer users do not want to afford anything :)
1220628147 M * ghislainocfs2 eheh
1220628175 M * ghislainocfs2 some big hosting company use it that's why i wondered
1220628185 N * DoberMann DoberMann[PullA]
1220628186 M * daniel_hozac Bertl: http://people.linux-vserver.org/~dhozac/t/linux-2.6.26.3-vs2.3.0.35.config
1220628204 M * Bertl ghislainocfs2: they mostly play safe, and cluster filesystems are not really well tested in linux
1220628206 M * daniel_hozac Bertl: tried explicitly setting -serial stdio and all that?
1220628213 M * vargadani by comping and then restoring the /etc/vservers directory, can I backup the guests and on a new install say use them?
1220628223 M * vargadani ohh. comping = coping :)
1220628247 M * daniel_hozac ghislainocfs2: i can second that, last time i tried OCFS2 on DRBD it crashed under any kind of load...
1220628297 M * ghislainocfs2 oh man all those cluster hype then are jsut marketing :(
1220628334 M * ghislainocfs2 so i will go for one partition for each server toooo bad :)
1220628335 M * daniel_hozac well, i didn't have time to investigate, so it could've been DRBD.
1220628353 M * daniel_hozac nonetheless, be sure to test whatever setup you do go with...
1220628365 M * ghislainocfs2 yes sur i'll do
1220628386 M * ghislainocfs2 ocfs2 do nto work with vserver so i must drop this, i will use a single partition for each node
1220628394 M * Bertl ghislainocfs2: actually we appreciate testing filesystem like OCFS2 and GFS(2) with linux-vserver
1220628419 M * Rockj 'vserver ... suexec' is supported for running vservers only; aborting...
1220628419 M * Rockj vc_migrate_context(): Function not implemented
1220628420 M * Bertl ghislainocfs2: also, I'm pretty sure we can make it work soon
1220628426 M * Rockj do I have wrong vtools or something? 
1220628439 M * daniel_hozac Rockj: lenny?
1220628442 M * Rockj fetched vserver-tools from lenny
1220628455 M * daniel_hozac get util-vserver-0.30.216~r2772-2 from sid.
1220628462 M * ghislainocfs2 yes but work on the 2.3 version that has no cpu scheduler etc so i will be unable to use it :)
1220628470 M * Rockj hum. 
1220628483 M * Rockj is it wise to mix packages from lenny and sid? o_O
1220628500 M * daniel_hozac the package should migrate to lenny in a couple of days anyway.
1220628512 M * quinq isn't lenny frozen ?
1220628518 M * daniel_hozac it is.
1220628521 M * Bertl ghislainocfs2: well, you can go for the stable branch too, you just need to fix OCFS2 for that?!
1220628552 M * ghislainocfs2 yes
1220628554 M * daniel_hozac actually, it should have already migrated...
1220628557 M * Rockj daniel_hozac: dont have a direct url for the deb? :)
1220628562 M * ghislainocfs2 the stable has not the nmap
1220628577 M * Bertl ghislainocfs2: which is not 'our' problem, no?
1220628602 M * ghislainocfs2 so no hashify :), ohh yes of course i speak in a general way not vserver only
1220628633 M * Bertl ghislainocfs2: for the missing hard cpu/token bucket sheduler .. feel free to join the 'Linux-VServer development fund' to improve on that 
1220628645 M * ghislainocfs2 if ocfs2 dev were quicker it would have been ok, i wonder if i can use the newer kernel ocfs code with the older kernel ;)
1220628690 M * ghislainocfs2 by quicker i mean as speedy as vservers one they missed you by a kernel release :)
1220628721 M * ghislainocfs2 yes i was thinking about it, if the big players do not help then the little must move i guess ;)
1220628724 M * Bertl you can try to backport the changes
1220628762 M * Bertl (or hire some kernel developer to do so)
1220628825 M * Bertl but as I said, I'm positive that we are able to fix the ocfs2 issues in 2.6.26/vs soon (as soon as I have my _last_ fallback devel system up and running)
1220628836 M * Rockj daniel_hozac: using apt-get install -t sid util-vserver didn't fetch any newer version then what's already in lenny
1220628863 M * daniel_hozac and you have that setup?
1220628878 M * ghislainocfs2 that would be rgreat
1220628879 M * ghislainocfs2 ok
1220628908 M * Bertl ghislainocfs2: it was an unfortunate series of events that I'm down to one system
1220628974 M * ghislainocfs2 bertl: murphy's law ;)
1220628996 M * Bertl yep, precisely
1220629045 M * Rockj daniel_hozac: fetched deb manually and used dpkg -i instead.
1220629464 M * Rockj can a "guest-os" , reboot its server? like - just rebooting his vserver instance?
1220629481 M * daniel_hozac sure.
1220629501 M * ghislainocfs2 bertl: iw ill try to put some money to help the latest release
1220629503 M * Rockj 3 lines paste ok? 
1220629509 M * ghislainocfs2 bertl: does it lack a lot ?
1220629536 M * Rockj if I inside a guest-os try to type "reboot", it says missing /dev/initctl  - so seems this is removed so it can't reboot the "master" hosting the vservers
1220629557 M * Rockj so I assume there is other commands for "rebooting" a guest-instance when your inside a guest-instance? :)
1220629577 M * Rockj (brb)
1220629622 M * daniel_hozac use reboot/halt -f
1220629653 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it
1220629706 J * nkukard ~nkukard@196.212.73.74
1220629767 M * Bertl ghislainocfs2: that would be great!
1220629910 M * ghislainocfs2 bertl: paypal donation to herbert@13thfloor.at ? the link on the webpage is in german so i cannot understand a thing lol
1220629954 M * ghislainocfs2 ahh found a way to be in french
1220629970 M * Bertl http://www.13thfloor.at/vserver/donate/ (should be english)
1220630138 M * Rockj Ill be back later tonight, vserver seems fun. 
1220630149 M * Bertl k, cya!
1220630167 Q * Rockj Quit: yai!
1220630310 Q * mrfree Quit: Leaving
1220630329 M * ghislainocfs2 bertl: i made the best effort i could for AQUEOS (my company)
1220630403 M * Bertl if you provide me with some company data, we'll add you to the Hall'o'Fame
1220630646 M * ghislainocfs2 oh thanks :) i will then !
1220631227 J * _gh_ ~gerrit@67.170.155.50
1220633243 M * Bertl ghislainocfs2: please remind me, what is required to get an ocfs2 filesystem (single node) active on a local partition?
1220633548 J * ntrs_ ~ntrs@77.29.68.145
1220633746 M * Bertl daniel_hozac: nope, no luck with your config either ... doesn't boot here :/
1220633967 Q * ntrs Ping timeout: 480 seconds
1220634276 M * Bertl now trying with an old 2.6.23 config I found here
1220635183 M * daniel_hozac Bertl: hmm, it's what i've been doing my testing with since 2.6.22 (well, tweaked ever so slightly for each new version, of course).
1220635232 M * Bertl yeah, problem here is an older qemu and gcc
1220635256 M * Bertl but if the next test fails too, I'll bite the bullet and reinstall the system
1220635398 M * Bertl ah, now I remember! it was the split/size
1220635440 M * Bertl no hardware support here, so I have to move the kernel out of kernel space
1220635818 Q * ntrs_ Read error: Operation timed out
1220636066 N * pmenier pmenier_off
1220636787 Q * vargadani Remote host closed the connection
1220637768 M * daniel_hozac ah, okay.
1220638613 M * Bertl ghislainocfs2: still around?
1220640226 Q * Guest3460 Quit: restarting session
1220640266 N * phedny Guest5477
1220640271 J * phedny ~mark@2a02:348:35:5a26::1
1220641698 M * quinq I've got a networking question
1220641706 M * Bertl shoot
1220641723 M * quinq my guest is bound to a dummy interface with a local network
1220641752 M * quinq when I access Internet from my guest,  connexion is directly made with my eth0 (Internet interface)
1220641763 M * quinq is there anyway to filter that way of traffic ?
1220641778 M * quinq I mean with iptables
1220641780 M * daniel_hozac iptables, as you normally would.
1220641782 M * Bertl sure, iptables and the OUTPUT chain will do nicely
1220641792 M * quinq but
1220641808 M * quinq I'm not able to get source IP
1220641817 M * quinq because that's only output
1220641830 M * Bertl hmm?
1220641852 M * quinq if I got two guests, rules can't be applied differently among guests
1220641871 M * daniel_hozac sure they can.
1220641891 M * quinq I tried :/
1220641898 M * daniel_hozac i do some serious filtering both for inter-guest traffic and outbound traffic...
1220641906 M * Bertl quinq: what rule?
1220641917 M * quinq exemple, my guest is 192.168.0.2
1220641930 M * quinq I can't filter OUTPUT on eth0 from 192.168.0.2 source
1220641942 M * quinq because output address is directly my eth0 address
1220641959 M * quinq and that' not a question, that's a fact :/
1220641963 M * quinq +s
1220641972 M * Bertl did you assign the 'public' ip to the guest?
1220641979 M * quinq no
1220641983 M * quinq I assigned 192.168.0.2
1220641991 M * quinq on a dummy interface on host
1220642001 M * quinq which have 192.168.0.254
1220642003 M * daniel_hozac OUTPUT is before NAT.
1220642024 M * daniel_hozac well, SNAT anyway.
1220642033 M * Bertl quinq: the dummy is of no relevance
1220642203 M * quinq for exemple if I do
1220642219 M * quinq iptables -A OUTPUT -o eth0 -s 192.168.0.2 -j DROP
1220642231 M * quinq i can still access Internet from my guest
1220642299 M * Bertl what kernel?
1220642322 M * daniel_hozac well, that doesn't really mean anything.
1220642339 M * daniel_hozac we'd need a complete dump of the OUTPUT chain to say anything.
1220642348 M * quinq errr
1220642350 M * quinq my bad
1220642353 M * quinq I'm really sorry
1220642357 M * quinq tried that yesterday
1220642361 M * quinq working now
1220642366 M * quinq I must have made a mistake
1220642372 M * quinq really sorry about that :)
1220642449 M * Bertl well, glad that it works as it is supposed to work :)
1220642570 M * quinq me too ;)
1220643044 J * blues blues@adf7.neoplus.adsl.tpnet.pl
1220643068 M * Bertl welcome blues!
1220643164 Q * blues_ Ping timeout: 480 seconds
1220643265 J * ntrs_ ~ntrs@77.29.68.145
1220643729 J * laptopnenolod ~nenolod@ip70-189-74-62.ok.ok.cox.net
1220645888 Q * esa Ping timeout: 480 seconds
1220646208 J * derjohn_mob ~aj@p5B23F1CB.dip.t-dialin.net
1220646576 Q * xdr Quit: leaving
1220646796 J * esa bip@62.123.8.50
1220648017 J * esa` bip@62.123.8.171
1220648028 Q * esa Ping timeout: 480 seconds
1220648032 Q * nenolod Read error: Connection reset by peer
1220648045 J * nenolod ~nenolod@ip70-189-74-62.ok.ok.cox.net
1220650099 J * Aiken ~Aiken@ppp118-208-124-125.lns4.bne4.internode.on.net
1220650749 J * ktwilight__ ~ktwilight@9.116-66-87.adsl-dyn.isp.belgacom.be
1220651042 Q * ktwilight_ Ping timeout: 480 seconds
1220651158 Q * ntrs_ Read error: Operation timed out
1220651583 M * blues re
1220651743 J * xdr ~xdr@48-173-96-87.cust.blixtvik.se
1220652004 J * doener ~doener@i577BB16D.versanet.de
1220652107 Q * doener_ Ping timeout: 480 seconds
1220654678 M * Bertl ghislainocfs2: I finally managed to get a working test environment, and fixed the ocfs2 issues AFAICT, please check the vs2.3.0.35.5 patch (usual place) with your setup and let us know how it goes (the 002 test will fail, but that is a 'bug' in testfs.sh :)
1220655325 N * DoberMann[PullA] DoberMann[ZZZzzz]
1220656718 Q * dowdle Remote host closed the connection
1220656887 Q * loddafnir Ping timeout: 480 seconds
1220658473 M * Bertl okay, off to bed now ... have a good one everyone! cya!
1220658478 N * Bertl Bertl_zZ
1220658767 Q * FireEgl Ping timeout: 480 seconds