1120781727 J * Moo ~troy@phx200.pinchaser.com 1120781734 M * Bertl welcome Moo! 1120781742 M * Moo moo :) 1120781805 N * Moo MooingLemur 1120781813 M * MooingLemur hmm.. someone else has Moo here.. 1120781850 N * jonsmel jonsmel_zZ 1120781870 M * Bertl well, moo is short ... 1120781990 M * MooingLemur so's my patience right now with the small number of packages that infer the compiler from the kernel version, i.e., assuming a 64 bit compiler in a vserver 32 bit guest due to the 64 bit kernel. 1120782027 M * MooingLemur :P 1120782045 M * Bertl well, why not give them a 32bit kernel then? 1120782050 M * MooingLemur openssl and mozilla-* to be precise. 1120782058 M * MooingLemur because there are 64 bit guests running, too 1120782060 M * Bertl kernel version, to be precise :) 1120782545 Q * alexx Ping timeout: 480 seconds 1120782928 J * alexx ~alexx@82.225.136.176 1120784435 J * explasm__ explasm@p549F7D9B.dip.t-dialin.net 1120784875 Q * eXplasm2 Ping timeout: 480 seconds 1120785401 M * nokoya hello 1120785411 M * nokoya morning :D 1120785444 M * nokoya I got this problem when i create vs use debian-newvserver.sh 1120785552 M * nokoya I saw something wrong with emacs20 1120785559 M * nokoya is that important ? 1120785610 M * Bertl might be, but debian-newvserver.sh (as the name suggests) is not part of the mainline tools, so you have to contact the debian developers (or file a bugreport) 1120787508 J * tanjix ~info@office.star-hosting.de 1120787512 M * tanjix hello together 1120787517 M * Bertl hey tanjix! LTNS! 1120787572 M * tanjix can i setup new vserver host image with debian sources from apt and the 2.6 kernel image ? 1120787623 M * Bertl probably ... no idea what the debian sources provide ... 1120787637 M * tanjix kernel-patch-vserver - context switching virtual private servers - kernel patch 1120787637 M * tanjix util-vserver - tools for Virtual private servers and context switching 1120787637 M * tanjix vserver-debiantools - Tools to manage debian virtual servers 1120787637 M * tanjix vserver - Virtual private servers and context switching 1120787647 M * tanjix this is what i find when searching apt fpr "vserver" 1120787668 M * Bertl yep, I know most of the 'sarge' tools are broken (vserver wise) 1120787690 M * Bertl I was told that 'recent' versions (in unstable & testing) do work 1120787726 M * tanjix ok, but can i patch the 2.6kernel image for vserver-usage ? 1120787736 M * tanjix or must i compile my own 2.6 kernel ? 1120787766 M * Bertl there are different versions in debian, the 2.6.8 is working, but it lacks some fixes 1120787782 M * Bertl I've also heard that there is a 2.6.11 and 2.6.12 vserver image 1120787853 M * tanjix i have the debian kernel image "kernel-image-2.6-686-smp" running 1120787931 M * tanjix is there a howto somewhere hos to successfully patch it ? 1120787994 M * Bertl http://linux-vserver.org/Step-by-Step+Guide+2.6 1120788132 M * tanjix hm but that kernel must be compiled... no way to use that kernel already installed? it's because that kernel i now have has all drivers inside that we need for all of our possible hardware configuration 1120788201 M * tanjix or is there a way to read out the actual kernel config and "import" that to the vanilla kernel ? 1120788225 M * Bertl well, there is a patch for the 2.6.8-16 debian kernel 1120788250 M * Bertl aside from that, I can not tell you how to 'modify' the existing debian kernel 1120788264 M * tanjix apt-get source kernel ... 1120788281 M * tanjix would give me the sources for that kernel 1120788293 M * tanjix i think i'll try the 2.6.11 version 1120788498 M * tanjix is there a way to read out the actual kernel config and "import" that to the vanilla kernel ? 1120788520 M * Bertl my kernels all store the kernel .config in /proc 1120788532 M * Bertl but I doubt the default debian kernel will do so ... 1120788558 M * Bertl nevertheless, I guess they include some .config file for the kernel in their sources ... otherwise it would not build 1120788603 M * tanjix hm i'm failing on running menuconfig.. missing some files 1120788693 M * tanjix now :) 1120788879 M * Bertl very likely ncurses/dev 1120788885 M * tanjix yep :) 1120788894 M * tanjix in /proc there is no config file 1120788914 M * Bertl thought so ... would have been to easy ... 1120788960 M * tanjix the sources contain a pre-compiled kernel as i see that 1120788992 M * Bertl maybe with a .config file in place? 1120789006 M * tanjix nope :( 1120789058 M * tanjix but not that matter i'll jest select everythinh again in the kernel config.. only takes some time.. thought there was an easier way 1120789074 M * daniel_hozac /boot/config-*? 1120789085 M * Bertl tanjix: again? so you already selected it? 1120789093 M * daniel_hozac i don't know Debian, but that's where FC puts them. 1120789093 M * tanjix master:/boot# ls config-2.6.8-2-686-smp 1120789094 M * tanjix config-2.6.8-2-686-smp 1120789113 M * Bertl well, that could do the job ... 1120789125 M * tanjix i'll try to load that config 1120789128 M * Bertl just copy it into your 'new' kernel tree as .config 1120789135 M * Bertl and then do 'make oldconfig' 1120789142 Q * [maple] Read error: Connection reset by peer 1120789151 M * Bertl it should just ask you stuff which changed 1120789162 M * Bertl (i.e. new options, etc) 1120789171 M * tanjix before or after patching ? 1120789179 M * Bertl you can do both 1120789189 M * Bertl before to see changes in mainline 1120789201 M * Bertl after to check vserver related stuff 1120789435 J * [maple] ~maple@strace-nat0.ua-online.net 1120789443 M * Bertl wb [maple]! 1120789517 M * tanjix compiling ... 1120789573 M * tanjix the howto says only to run "make".. is there more to do with the 2.6 kernel ? 1120789593 M * Bertl once the kernel is built, use make modules_install 1120789619 M * Bertl to install the modules, and either make install or copy the resulting kernel into place 1120789653 M * Bertl great! we now have _working_ xid support for xfs,jfs,reiserfs,ext2 and ext3 1120789893 M * tanjix hmm make stopped with errors 1120789930 M * Bertl where did you get the kernel source tree again? 1120789950 M * tanjix the 2.6.11 kernel ? 1120789972 M * Bertl whatever you did compile right now ... 1120790001 Q * nokoya Quit: 1120790011 J * nokoya young@hi-230-82.tm.net.org.my 1120790021 M * tanjix its the kernel source from the howto you gave me 1120790073 M * Bertl so from kernel.org? 1120790096 M * tanjix yep 1120790104 M * Bertl should be fine then ... 1120790117 M * Bertl could you upload the errors somewhere (e.g. pastebin.com) 1120790202 M * tanjix http://pastebin.com/309263 1120790269 M * Bertl do you use that drm driver? 1120790285 M * tanjix i dont know what that is :) ? 1120790301 M * Bertl okay, in this case I'd suggest to remove it from the .config 1120790310 M * Bertl (it seems it is broken in that version) 1120790333 M * tanjix deleteing every line that contains drm ? 1120790333 M * Bertl you might also consider using a recent kernel instead of the one used in the Getting Started example 1120790355 M * Bertl (but it probably works when you disable the drm stuff) 1120790423 M * tanjix i did so and ran make oldconfig again 1120790670 M * tanjix hmm 2 gigs ram are too little :) 1120790716 M * Aiken I have a test case for my syslogd in an alpha guest problem 1120790737 M * Aiken the function flock is not working in an alpha guest 1120790762 M * Bertl hmm, we had that, IIRC it was a dietlibc issue, no? 1120790763 M * Aiken the test case works on the host and on x86 host & guest 1120790782 M * Aiken both syslogd and my test case are compiled against glibc 1120790790 M * Bertl ah, okay .. 1120790808 M * Bertl source for your test case? 1120790826 M * Aiken via pastbin? 1120790842 M * Bertl whatever you prefer ... 1120790926 M * Aiken http://pastebin.com/309270 1120790947 M * Aiken it is a combination of the flock failing and the way syslogd exists 1120790958 M * Aiken exits even 1120791053 M * Bertl and it works quite fine on the host? 1120791185 M * Aiken yes 1120791343 M * Aiken flock fails followed by the child exiting and going zombie 1120791364 M * Aiken the parent never receives it's SIGTERM so has to wait for that big sleep to finish 1120791375 M * Aiken hence the apparent lock up 1120791437 M * Bertl and I guess it only happens on alpha, right? 1120791541 M * Aiken only on the alpha and only in the guest 1120791565 M * Bertl okay, my test machine is currently not available ... (alpha) so it will have to wait a little 1120791604 M * Aiken the problem is not annoying me as much now I have a test case 1120791626 M * Bertl yeah, that will help to track it down quite easily I guess 1120791775 M * Aiken and I did not have a problem with 2.6.11.11 2.0-rc4 1120791810 M * Bertl hum, so that test case works with 2.6.11.11-vs2.0-rc4 ? 1120791859 M * Aiken syslogd worked, just rebooting to make sure the test case is ok 1120792323 M * tanjix the new kernel does not boot, stopping with kernel panic 1120792339 M * Bertl do you have a copy of the panic? 1120792372 M * tanjix there were errors that modules.dep could not be found 1120792373 M * Bertl (it's probably related to the initrd, debian has some changes there IIRC) 1120792753 M * Aiken the test case works with 2.6.11.11 1120792791 M * Aiken just had a worrying moment when I lost the machine, had forgotten about the alpha specific changes the other day :( 1120792820 M * Bertl so it's either a change in the vserver code rc4->rc5, or, which is more likely, a change in the kernel code 2.6.11->2.6.12 ... 1120792892 M * Bertl Aiken: how do you feel about another kernel test? 1120792916 M * Aiken ok 1120792919 M * Bertl http://www.13thfloor.at/~doener/vserver/patches/patch-2.6.12-vs2.0-rc4.diff 1120792933 M * Bertl this was the forward port Doener did for rc4 ... 1120792940 M * Aiken so a std kernel with that patch? 1120792958 M * Bertl so the linux-vserver code is supposed to be equiv to the 2.6.11.11-vs2.0-rc4 1120792985 M * Bertl yes, it should give us a good hint who is to blame :) 1120793103 M * nokoya Bertl: what kind of script u know is good to make vs ? 1120793124 M * nokoya except debian-newvserver.sh 1120793137 M * Bertl the tools support a bunch of build methods, see the following url for details: 1120793163 M * Bertl http://linux-vserver.org/alpha+util-vserver 1120793174 M * Bertl (section Vserver creation) 1120793207 M * nokoya aha 1120793401 M * Doener morning 1120793410 M * Bertl morning Doener! 1120793424 M * Bertl Doener: rise and shine? 1120793482 M * Doener no, didn't have coffee yet ;) 1120794800 M * Aiken 2.6.12-vs2.0-rc4 fails as well 1120794812 M * Bertl ah, excellent ... 1120794835 M * Bertl Doener: somehow the 2.6.12 port did break something ... 1120794857 M * Doener hum... 1120794860 M * Bertl Aiken has created a test case ... 1120794876 M * Doener saw it 1120794876 M * Bertl a pretty nice one actually ... 1120794940 A * Doener would like a perror in there ;) 1120794951 M * Bertl http://vserver.13thfloor.at/Stuff/aiken_flock_test.c 1120794990 M * Bertl change it to your preference, Aiken will probably compile/run it for you :) 1120795042 M * Bertl I don't understand why this bites just alpha/guests ... 1120795066 M * Bertl at the first look I'd blame the signal handler (session/pgroup issues) 1120795153 M * Aiken just dropping it back to the the file stuff only it fails 1120795312 M * Doener Aiken: can you make the printf("Can't lock...") a perror, please? 1120795377 M * Aiken ok 1120795460 M * Aiken the file code I copied out of syslogd 1120795577 M * DaPhreak Greek0: morning Doener, Bertl Aiken, nokoya ... 1120795579 M * DaPhreak *g* 1120795585 M * DaPhreak gonna get to work ;P 1120795596 M * Bertl morning! 1120795623 M * Doener morning DaPhreak 1120795638 M * Aiken I am starting to think it is a nfs issue 1120795747 M * Aiken I want to try something with a std 2.6.12.2 kernel 1120795786 M * Bertl hmm hmm ... how is nfs involved? 1120795824 M * Aiken /vserver is a nfs mount point 1120795905 M * Bertl hmm, and when you test it on the host, you did not use the /vserver dir? 1120795938 M * Aiken no 1120795942 M * Bertl i.c. 1120795966 M * Aiken did not think to, have not had any problem with nfs before 1120796150 M * Bertl what are the nfs mount options? and did you enable nfs xid tagging? 1120796269 M * Aiken the tagging under the vserver options? no, that looks like it is server related, the machine is a nfs client 1120796285 M * Bertl is it enabled or disabled? 1120796292 M * Aiken disabled 1120796301 M * Aiken default mount options rw,v3,rsize=8192,wsize=8192,hard,udp,lock,addr=barney 1120796315 M * Bertl okay .. on the server it is disabled too, right? 1120796341 M * Bertl could you try with tcp instead of udp? 1120796365 M * Aiken the server is 2.4.30-vs-1.2.10 1120796400 M * Doener "A. flock()/BSD locks act only locally on Linux NFS clients prior to 2.6.12."... 1120796403 M * Doener http://nfs.sourceforge.net/ 1120796404 M * Bertl hmm, so probably no tcp nfs support 1120796432 M * Bertl Aiken: is nfslockd running on the server? 1120796531 M * Aiken no change using tc and nfslockd is not running 1120796637 M * Bertl yeah, seems to be a 'simple' change in nfs behaviour with 2.6.12, thanks Doener! 1120796651 M * Aiken nfs fails with this machine as well now 1120796651 M * Doener yw 1120796670 M * Bertl D10. I'm trying to use flock()/BSD locks to lock files used on multiple clients, but the files become corrupted. How come? 1120796676 M * Aiken this is only since my last kernel upgrade :( 1120796688 M * Bertl yep, read the url Doener pasted 1120796765 M * Doener http://www.linux-nfs.org/Linux-2.6.x/2.6.11/linux-2.6.11-18-flock.dif 1120796788 M * Doener that looks like being guilty ;) 1120796978 M * Bertl looks like NFS is quite in flux -- we have to watch it carefully ... 1120796984 M * Aiken I'll admitt a change in nfs behavour is not what I expected 1120796997 M * Bertl well, me neither :) 1120797046 M * Doener Aiken: does it fail with -ENOLCK? 1120797081 M * Doener erhm... nvm 1120797354 M * Aiken yes 1120797389 M * Aiken error 77 No locks available 1120798551 M * Aiken the delta-alpha patch also works against 2.6.11.11 1120798580 M * Bertl good to know ... 1120798861 M * Doener the -ENOLCK makes absolutely no sense to me... 1120798862 M * DaPhreak ... 1120798876 M * Doener the only location for it i find is in nfs_flock() 1120798884 M * Doener if (!(fl->fl_flags & FL_FLOCK)) 1120798884 M * Doener return -ENOLCK; 1120798898 M * Bertl NFS has a separate error handler 1120798910 M * Bertl i.e. it uses different defines 1120798929 M * Bertl search for the explicit error code ... 1120800616 M * Bertl Doener: if you get around, I added a bunch of patches to FOR-2.0 (they are supposed to fix all known xid/barrier/etc issues, and add the legacy mode) 1120800655 M * Hollow morning 1120800729 M * Doener ok 1120800729 M * Bertl a new testfs.sh is also available ... 1120800729 M * Bertl except for jfs (which doesn't support xattrs) everything works fine (so far) 1120800944 M * eyck great! 1120800944 M * eyck is there any chance for vs1.2.11? or 1.2.10.11? 1120800945 M * Bertl eyck: hmm, have we still issues pending there? 1120800945 M * Aiken definate nfs issue, a std 2.6.12.2 does the same :( 1120800945 M * Bertl morning Hollow! eyck! btw ... 1120800968 M * Hollow gtg to school anyway.. :/ so cu later 1120801053 M * eyck Bertl: I know only about this barrier thingy, 1120801053 M * eyck morning, morning 1120801053 M * Bertl eyck: hmm, which fs? 1120801053 M * Bertl xfs? 1120801053 M * eyck yes, that too, 1120801054 M * eyck and ext2 and reiserfs IIRC 1120801073 M * Bertl hmm, guess I have to search the logs ... 1120801154 M * eyck so that's your secret, you've got large cache, but no permanent storage, so as long as everything stays in your cache, you're brilliant;) 1120801341 M * Bertl yeah, probably ... :) 1120801358 M * Bertl but, I have a second and third level cache too :) 1120801369 M * eyck good for you 1120801401 M * Bertl yeah, and folks like you can work as 4th level cache *G* 1120801438 M * eyck off-chip 1120801439 M * Bertl but seriously, I remember that we 'lowered' the barrier criterion ... 1120801450 M * Bertl to allow chmod 000 inside a guest, right? 1120801465 M * eyck you lost me, 1120801478 M * Bertl so that isn't what you are referring to? 1120801483 M * eyck you changed something and barrier stopped working, 1120801496 M * eyck it might be it, I didn't know the reason for this change 1120801517 M * Bertl so you are saying 1.2.x has a broken barrier? 1120801568 M * eyck yes, we already talked about it, and then you fixed it 1120801570 M * eyck + if (di_flags & XFS_DIFLAG_IUNLINK) 1120801570 M * eyck + flags |= XFS_XFLAG_IUNLINK; 1120801588 M * Bertl that's just for xfs ... 1120801660 M * eyck hmm, I'm looking at the patch, and there are only XFS changes, correct. 1120801750 M * Bertl yep, I have that one still in my queue 1120803417 Q * _are_ Quit: bbl 1120803430 M * Aiken I have done something, later on I hope I will know exactly what the something was 1120803444 M * Aiken upgrading nfs utils and flock now seems to be working 1120803505 M * Bertl sounds good, please let us know, when you found out ... 1120803586 M * Bertl okay, enough fs testing for me for today ... I'm off to bed : 1120803597 M * Bertl cya later this evening I guess ... 1120803599 M * DaPhreak night Bertl :) 1120803608 N * Bertl Bertl_zZ 1120803635 Q * tanjix Ping timeout: 480 seconds 1120803637 M * Aiken bye 1120807105 J * Doener` ~doener@p54874F49.dip.t-dialin.net 1120807534 Q * Doener Ping timeout: 480 seconds 1120808598 J * prae ~prae@ezoffice.mandriva.com 1120811906 Q * Aiken Quit: Leaving 1120812926 Q * Hollow Quit: /dev/zero 1120813295 J * Hollow ~Hollow@home.xnull.de 1120815750 M * id hi 1120816978 N * dlp ddlp 1120817164 J * pflanze ~chris@62.12.148.89 1120817268 M * pflanze Hello. Is there any announcement text somewhere on 2.0-rc5? 1120817295 M * pflanze Or: did anything important change since rc4? 1120819434 J * martinh ~martinh@isa.voelcker.com 1120819471 M * martinh hello! how do i make pty devices available to my vserver guests? 1120820011 M * martinh i need to run things like xterm 1120821825 Q * martinh Remote host closed the connection 1120822989 Q * dsoul Ping timeout: 480 seconds 1120824774 J * dsoul darksoul@pingu.ii.uj.edu.pl 1120824957 Q * xbing Ping timeout: 480 seconds 1120828860 Q * mcp Read error: Connection reset by peer 1120828870 J * mcp ~hightower@wolk-project.de 1120829880 J * Val ~val@v41.ath.cx 1120829926 M * Val Hi :) 1120831107 Q * Hollow Quit: /dev/zero 1120831265 J * Hollow ~Hollow@home.xnull.de 1120831834 N * jonsmel_zZ jonsmel 1120832042 Q * explasm__ Ping timeout: 480 seconds 1120832181 J * explasm__ explasm@p549F6F48.dip.t-dialin.net 1120832607 Q * AprilDL Quit: Chatzilla 0.9.67 [Mozilla rv:1.8b2/20050226] 1120833280 Q * complexho Ping timeout: 480 seconds 1120833936 J * Pazzo ~Pazzo@host130-250.pool8172.interbusiness.it 1120834281 Q * BWare Quit: using sirc version 2.211+KSIRC/1.3.12 1120834431 Q * Hollow Quit: /dev/zero 1120834472 Q * albeiro Ping timeout: 480 seconds 1120834722 J * albeiro ~albeiro@graffias.estrefa.pl 1120838299 J * renihs ~renihs___@193.170.52.70 1120838326 Q * renihs Quit: 1120838380 J * renihs ~renihs___@193.170.52.70 1120838492 M * renihs hmm i am running samba within a util-vserver 0.30.207 (kernel 2.6.11.12-vs2.0), but it complains about not beeing able to open the requested 10000 files, i did try to set with vlimit (nofile) but that doesnt affect 1120838516 M * renihs how can i set the open-file max? for a vserver individually? 1120839194 Q * renihs Read error: Connection reset by peer 1120839240 J * renihs ~renihs___@193.170.52.70 1120839268 M * renihs hmm channel seems quite idle :) 1120839307 M * renihs another question, does it pose a security thread in any way to have images within /vservers wich are loopmounted to /vservers/vs1 etc 1120839554 N * Bertl_zZ Bertl 1120839610 M * Bertl renihs: for each guest, check the rlimits in /proc/virtual//limits 1120839632 M * Bertl renihs: but I suspect you are hitting either an ulimit or a global max (host) 1120839676 M * Bertl renihs: (ad loop mounting) if done properly, no .. most security issues arise from bad/broken configs though ... 1120839687 J * Rushmoom ~me@80-219-253-246.dclient.hispeed.ch 1120839699 M * Bertl welcome Rushmoom! 1120839714 M * Rushmoom Hello! :) 1120839765 M * Greek0 Bertl: should I repost my vs_memory.h includes patch to the vserver ml, or is it ok already? 1120839777 M * baggins Bertl: where does ewrite come from? it's used in testfs.sh 1120839793 M * Greek0 as you can probably read in the backlog I "fixed" my posting problem yesterday 1120839804 M * Bertl Greek0: well, I know you aptch location now, so it should be fine ... jsut do not change any patches in place, if possible 1120839825 M * Bertl baggins: it's in the same dir (a trivial script) 1120839847 M * Bertl I didn't want to bother with escaping all that stuff (so that it works inline) 1120839859 Q * prae Quit: Execute Order 69 ! 1120839872 M * Greek0 Bertl: I'll append revision numbers if I put new versions online (leaving the old ones in place) 1120839917 M * Bertl Greek0: excellent! 1120840015 M * renihs Bertl, re 1120840062 M * Rushmoom I was running VServer (1.9.5) on Gentoo, init-style = plain. Starting up the Vserver worked (was pingable), but did not give *any* output at all. Entering the VServer seemed to work at first, but I always got "kicked" after 1 sec (Killed). 1120840077 M * Rushmoom Anyone has an idea what could cause such a behaviour? 1120840095 M * Bertl broken guest setup ... 1120840122 M * renihs hmmm /proc/virtual//limit shows 10000 (the setting i tried to set with vlimit -c xid nofile=10000 1120840135 M * renihs but ulimit -n (within the vservers) shows 1024 1120840139 M * renihs grumbl 1120840145 M * Bertl Rushmoom: your guest has to start some services (it probably failed to do so, so the context shuts down after some while 1120840155 M * Rushmoom Bertl: I set up the guest system (Gentoo) after a standarized procedure, which worked fine on my local machine, but not on a remote one :( 1120840176 M * Bertl renihs: you know that there are soft and hard ulimits? 1120840206 M * Rushmoom Bertl: Ah, ok. But shouldn't there be anything in the logs? (logfiles were empty) 1120840224 M * renihs Bertl, hmm ya i "know" that hmm hmm *thinking* 1120840229 M * Bertl Rushmoom: well, maybe you just forgot to clean up the /var/run/* whatever entries, which are written by the runlevel scripts 1120840256 M * Rushmoom Bertl: Nope, I did that, as stated somewhere in the docs... 1120840274 M * renihs btw, i guess i found a little script "bug", if i have a samba share in fstab.remote and i start the server i get the error: .: invalid option -- n (and it asks the password for the samba share which is defined in fstab.remote password=bla) 1120840302 M * Rushmoom Bertl: But "Killed" means, just the context has been shut down, not the server? Knowing this already would help me :) 1120840306 M * Bertl renihs: please file a bug report on savannah then ... 1120840348 M * Bertl Rushmoom: well, I don't know 'what' kills you, at least a dozent of different options come to my mind ... 1120840396 M * Rushmoom Bertl: :-( I don't see any way how to narrow it down, because on the host-system, VServer does not seem to log anything 1120840435 M * renihs /etc/vservers//rlimits/nproc.hard <--- this does not seem to get used hmm 1120840449 M * Greek0 Rushmoom: in the guest /var/log there isn't anything too? 1120840466 M * Rushmoom Greek0: No :( 1120840481 M * Rushmoom Greek0: Maybe even syslogd fails on startup... 1120840521 M * renihs i also *always* get a Vserver "bla" still running unexpectedly; please investigate it manually... <-- which isnt true, any idea what i could have done wrong? 1120840533 M * renihs it starts and stops flawlessly 1120840590 M * Bertl that usually points to a script which hasn't stopped/exited (or broken userspace tools) 1120840618 M * Bertl you should update to latest kernel aptch and tools and see if it persists ... 1120840634 M * Rushmoom Also the VServer shows another *very* weird behaviour: after starting it, vserver status sometimes gives "started", sometimes "stopped" 1120840677 M * Bertl renihs: ad ad *.hard you probably already set that with *, because * sets both .. (*=nproc here) 1120840681 M * Rushmoom And vserver-stat shows completely wrong uptimes, sometimes "N/A", just a few seconds later... 1120840720 M * Bertl Rushmoom: okay, time for a) testme.sh b) probably kernel/tool update 1120840743 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh 1120840769 M * Rushmoom Bertl: Ah, did not know about that, thanks! 1120840835 M * SiD3WiNDR some stuff should be linked more prominently I guess 1120840839 M * SiD3WiNDR like a troubleshooting guide 1120840841 M * Rushmoom Bertl: But to b): I only could do a downgrade, as I already am running devel kernel sources, and devel utils. (This was advised in the Gentoo Howto) 1120840843 M * SiD3WiNDR first step, run testme 1120840845 M * SiD3WiNDR download it at .. 1120840846 M * SiD3WiNDR :) 1120840861 M * SiD3WiNDR or "before making a patch, check if Bertl or Doener` already did it at http://..." ;) 1120841007 M * renihs ha, my first bug report :) 1120841029 M * Rushmoom Bertl: But anyway, thanks for your help, I think I've now got some things to try and to check, I got some new motivation for getting it running :) 1120841055 A * Rushmoom is very pleased to have found this channel :) 1120841074 M * renihs Bertl, your last message wasnt entirely clear to me :), nofile= max files? (softlimit?) 1120841096 M * renihs argl, i am not making myself very clear neither i guess...grumbl 1120841096 M * Bertl Rushmoom: could you upload the output of the testme.sh? 1120841138 M * Rushmoom Bertl: Not yet, sorry, I don't have access to the server right now :( 1120841159 M * Bertl renihs: there are different kinds of limits, ulimt vs rlimit and hard vs soft (or even min) 1120841239 M * Bertl Rushmoom: okay, np, just let us know when you have ... 1120841240 M * renihs Bertl, but which do i need to set to affect the output of ulimit -n within the vserver? 1120841250 M * renihs i mean ... 1120841252 M * Bertl several ... 1120841254 M * renihs yeah :) 1120841273 M * renihs i did set rlimit, vlimit (nofile=10000), /proc/sys/fs/file-max on host machine hmm 1120841280 M * Bertl first, you have to make sure that the hosts ulimit is at least that high (for soft and hard) 1120841288 M * renihs yap that i am sure about 1120841305 M * renihs ahnoo 1120841306 M * renihs argl 1120841339 M * renihs Bertl, never mind, me is just too confused, works now 1120841395 M * Bertl excellent! 1120841411 M * renihs Bertl, a bindmount cant be configured in fstab? it does work with vnamespace though ... 1120841421 M * renihs i mean within a vserver fstab 1120841444 M * Bertl it should work ... just depends what and where you bind :) 1120841450 M * renihs hmm 1120841465 M * renihs i tried to bind /tmp to /usr/portage/distfiles 1120841466 M * renihs :) 1120841481 M * renihs since my /usr/portage is samba mounted (ro) 1120841491 M * renihs (gentoo vserver :) 1120841506 M * renihs lemme try again :) 1120841615 M * renihs hmm one more thing, i just noticed that after i stopped and started a vserver which has a samba mount within, this mount is shown multiple times on the host machine??? 1120841619 M * renihs like mount shows 3x 1120841620 M * renihs /192.168.5.185/portage on /vservers/forums/usr/portage type smbfs (0 1120841636 M * renihs on the host box 1120841696 M * renihs (they remain after i shut down the vserver) 1120841734 M * renihs hmm never mind, if i umount it -> i get error umount: //192.168.5.185/portage: not found -> and they are dissapeared... 1120841735 M * renihs strange 1120841778 M * Bertl don't trust mount, check /proc/mounts for the truth :) 1120841785 M * renihs yah :) 1120841813 M * renihs it isnt shown there :) 1120841831 M * renihs grumbl, but i guess the output of mount is ugly after a few months running :) 1120841852 M * Bertl well, mount just shows /etc/mtab ... 1120841879 M * Bertl so if you clean that up and/or find the guilty 'mount' command and make it -n (noentry) 1120841903 M * renihs Bertl, i guess the -n error i posted above is exactly the reason for that maybe 1120841916 M * Bertl very likely ... 1120841926 M * renihs i ll just kick the mtab from time to time 1120841935 M * Bertl that's why you definitely should file a bugreport :) 1120841938 M * renihs i did already 1120841939 M * renihs :) 1120841943 M * renihs my first one :) 1120841944 M * Bertl excellent! 1120841973 M * renihs i can post strace output from my startup if that makes sense 1120841986 M * renihs (didnt find where the -n comes to place) 1120842019 M * baggins ahh, I just ran testfs.sh on x86_64, on XFS in all XID/UID/GID combinations 1120842021 M * Bertl could you show me your mount entry? 1120842036 M * Bertl baggins: hey, great work! thanks! 1120842045 M * renihs Bertl, //192.168.5.185/portage /usr/portage smbfs defaults,password=bla 0 0 (fstab.remote) 1120842087 M * Bertl is 'bla' the password or does it contain 'special' chars? 1120842090 M * renihs for some reason nfs doesnt work for me (from a vserver) 1120842091 M * renihs bla is the pass 1120842111 M * renihs no security yet 1120842122 M * baggins Bertl: everytime xattr tests 104, 106 and 114 failed 1120842137 M * Bertl baggins: that's not good ... 1120842165 M * baggins Bertl: with XID-Runtime and XID-Disabled xid tests 14,15,16,17 failed 1120842174 M * Bertl baggins: could you upload the output please? 1120842182 M * Bertl and did you also run the xid tests? 1120842212 M * baggins Bertl: yes, see my previous line 1120842249 M * Bertl ah, yep, right! 1120842267 M * baggins Bertl: I didn't have delta-xfs-fix01.diff and delta-xid-clean01.diff applied 1120842268 M * Bertl yes, that is fine, the test script has no notion of the xid tagging method yet 1120842294 Q * pflanze Quit: [x]chat 1120842299 M * Bertl baggins: okay, that explains the issues 1120842335 M * Bertl baggins: http://vserver.13thfloor.at/Experimental/FOR-2.0/?C=M&O=A 1120842345 J * xbing ~nb@dsl081-044-121.lax1.dsl.speakeasy.net 1120842351 M * Bertl everything from the 8th of july should fix them ... 1120842356 M * Bertl welcome xbing! 1120842371 M * xbing w00t :) 1120842392 M * baggins I'll try them, then 1120842424 M * Bertl if you compile the other fs in (kernel side) and use some loopback device (or lvm partition or whatever) 1120842443 M * Bertl you could do me a favor and run it on all filesystems (on x86_64) 1120842484 M * baggins no problem, but I won't make it till monday 1120842487 M * baggins Bertl: BTW, enabling CONFIG_VSERVER_HISTORY creates a semi-non-working kernel 1120842511 M * Bertl on x86_64? 1120842525 M * baggins yep, from dmesg: 1120842527 M * baggins kernel: No per-cpu room for modules. 1120842542 M * Bertl cool! 1120842543 M * baggins and then I couldn't load ipv6.ko and some netfilter modules 1120842561 M * Bertl yeah, probably we exhausted the kernel space *G* 1120842584 M * Bertl will look into it, as soon as I have an x86_64 available again ... 1120842593 M * Bertl Vudumen: any news from the sunny side? 1120842693 M * baggins hmm, is this bit from delta-xid-clean01.diff ok? 1120842695 M * baggins - ip->i_d.di_xid = current_fsxid(cr); 1120842695 M * baggins + ip->i_d.di_xid = current_fsxid(cr, vp); 1120842736 M * baggins it doesn't make sense to me, as current_fsxid is declare to have one parameter 1120842990 M * Bertl baggins: take the rc5 kernel, then apply 1120843009 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-alpha-fix01.diff 1120843016 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-debug-feat01.diff 1120843020 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-legacyv-feat01.diff 1120843025 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-xid-clean01.diff 1120843030 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-jfs-feat01.diff 1120843035 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-reiser-fix01.diff 1120843039 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-xfs-fix01.diff 1120843055 M * Bertl in this sequence ... and you should have everything in place 1120843066 M * Bertl (you are missing the xid-clean01 part here) 1120843093 M * baggins yeah, I see it 1120843194 M * baggins and also delta-attr-fix01.diff and delta-xfs-feat01.diff for a real XFS test ;) 1120843313 J * complexho ~mark@funk.gotadsl.co.uk 1120843471 M * Bertl baggins: ah, yeah, forgot about those ... 1120843485 M * Bertl welcome complexho! 1120843832 J * prae ~prae@sherpadown.net 1120844039 M * Bertl wb prae! 1120844106 M * prae Hi bertl 1120844697 Q * xbing uranium.oftc.net kinetic.oftc.net 1120844697 Q * albeiro uranium.oftc.net kinetic.oftc.net 1120844697 Q * Pazzo uranium.oftc.net kinetic.oftc.net 1120844697 Q * Doener` uranium.oftc.net kinetic.oftc.net 1120844697 Q * baggins uranium.oftc.net kinetic.oftc.net 1120844697 Q * Loki|muh uranium.oftc.net kinetic.oftc.net 1120844697 Q * zimbo uranium.oftc.net kinetic.oftc.net 1120844697 Q * id uranium.oftc.net kinetic.oftc.net 1120844697 Q * FaUl uranium.oftc.net kinetic.oftc.net 1120844697 Q * Greek0 uranium.oftc.net kinetic.oftc.net 1120844697 Q * neofutur uranium.oftc.net kinetic.oftc.net 1120844706 Q * nox uranium.oftc.net jupiter.oftc.net 1120844706 Q * Rushmoom uranium.oftc.net jupiter.oftc.net 1120844706 Q * renihs uranium.oftc.net jupiter.oftc.net 1120844706 Q * explasm__ uranium.oftc.net jupiter.oftc.net 1120844706 Q * Val uranium.oftc.net jupiter.oftc.net 1120844706 Q * dsoul uranium.oftc.net jupiter.oftc.net 1120844706 Q * MooingLemur uranium.oftc.net jupiter.oftc.net 1120844706 Q * jonsmel uranium.oftc.net jupiter.oftc.net 1120844706 Q * tchan uranium.oftc.net jupiter.oftc.net 1120844706 Q * flock uranium.oftc.net jupiter.oftc.net 1120844706 Q * sladen uranium.oftc.net jupiter.oftc.net 1120844706 Q * ddlp uranium.oftc.net jupiter.oftc.net 1120844706 Q * eyck uranium.oftc.net jupiter.oftc.net 1120844706 Q * case uranium.oftc.net jupiter.oftc.net 1120844706 Q * micah uranium.oftc.net jupiter.oftc.net 1120844706 Q * mugwump uranium.oftc.net jupiter.oftc.net 1120844706 Q * stupidawy uranium.oftc.net jupiter.oftc.net 1120844706 Q * janra uranium.oftc.net jupiter.oftc.net 1120844706 Q * jkl uranium.oftc.net jupiter.oftc.net 1120844706 Q * Zoiah uranium.oftc.net jupiter.oftc.net 1120844706 Q * cereal uranium.oftc.net jupiter.oftc.net 1120844706 Q * skceb uranium.oftc.net jupiter.oftc.net 1120844706 Q * Vudumen uranium.oftc.net jupiter.oftc.net 1120844706 Q * SNy uranium.oftc.net jupiter.oftc.net 1120844706 Q * locksy uranium.oftc.net jupiter.oftc.net 1120844706 Q * lilo uranium.oftc.net jupiter.oftc.net 1120844706 Q * Bertl uranium.oftc.net jupiter.oftc.net 1120844706 Q * pusling uranium.oftc.net jupiter.oftc.net 1120844706 Q * kestrel uranium.oftc.net jupiter.oftc.net 1120844706 Q * meebey uranium.oftc.net jupiter.oftc.net 1120844706 Q * gregster uranium.oftc.net jupiter.oftc.net 1120844706 Q * Medivh uranium.oftc.net jupiter.oftc.net 1120844706 Q * SiD3WiNDR uranium.oftc.net jupiter.oftc.net 1120844706 Q * DaPhreak uranium.oftc.net jupiter.oftc.net 1120844706 Q * DaCa-fi uranium.oftc.net jupiter.oftc.net 1120844706 Q * Hunger uranium.oftc.net jupiter.oftc.net 1120844706 Q * aba uranium.oftc.net jupiter.oftc.net 1120844711 J * Rushmoom ~me@80-219-253-246.dclient.hispeed.ch 1120844711 J * renihs ~renihs___@193.170.52.70 1120844711 J * explasm__ explasm@p549F6F48.dip.t-dialin.net 1120844711 J * Val ~val@v41.ath.cx 1120844711 J * dsoul darksoul@pingu.ii.uj.edu.pl 1120844711 J * MooingLemur ~troy@phx200.pinchaser.com 1120844711 J * jonsmel ~jscottorn@209.33.206.3 1120844711 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net 1120844711 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il 1120844711 J * sladen paul@starsky.19inch.net 1120844711 J * ddlp ~daniel@205.234.152.200 1120844711 J * Vudumen vudumen@perverz.hu 1120844711 J * Bertl ~herbert@212.16.62.52 1120844711 J * pusling ~pusling@195.215.29.124 1120844711 J * SNy ~mfr@bmx-chemnitz.de 1120844711 J * skceb skceb@bang-bang.feuer-frei.com 1120844711 J * cereal ~cereal@217.20.124.153 1120844711 J * Zoiah Zoiah@matryoshka.zoiah.net 1120844711 J * jkl ~eric@ool-182cef46.dyn.optonline.net 1120844711 J * janra janra@paradox.homeip.net 1120844711 J * stupidawy foo@you.wish.you.were.pimp.olicio.us 1120844711 J * mugwump ~samv@210-54-92-184.ipnets.xtra.co.nz 1120844711 J * micah micah@micha.hampshire.edu 1120844711 J * lilo ~lilo@lilo.usercloak.oftc.net 1120844711 J * locksy ~locksy@mrtg.sisgroup.com.au 1120844711 J * case ~case@donpanic.faveve.uni-stuttgart.de 1120844711 J * eyck eyck@81.219.64.71 1120844712 J * aba ~aba@eos.turmzimmer.net 1120844714 J * DaPhreak ~phreak@styx.xnull.de 1120844714 J * SiD3WiNDR luser@bastard-operator.from-hell.be 1120844714 J * Medivh ck@paradise.by.the.dashboardlight.de 1120844714 J * gregster ~gregor@greart.de 1120844714 J * meebey meebey@booster.qnetp.net 1120844714 J * kestrel ~athomas@vsrouter.swapoff.org 1120844714 J * DaCa-fi ~danny@mail.limehouse.org 1120844722 J * nox ~nox@noxlux.de 1120844744 J * Hunger Hunger.hu@levnor.hu 1120844754 J * xbing ~nb@dsl081-044-121.lax1.dsl.speakeasy.net 1120844754 J * albeiro ~albeiro@albeiro.usercloak.oftc.net 1120844754 J * Pazzo ~Pazzo@host130-250.pool8172.interbusiness.it 1120844754 J * Doener` ~doener@p54874F49.dip.t-dialin.net 1120844754 J * neofutur ~neofutur@neofutur.net 1120844754 J * baggins baggins@kenny.mimuw.edu.pl 1120844754 J * Loki|muh loki@satanix.de 1120844754 J * Greek0 ~greek0@81.189.246.175 1120844754 J * zimbo ~zimbo@callisto.dom.bonis.de 1120844754 J * id ~id@relax-media.softwarezentrum.de 1120844754 J * FaUl ~immo@ip88.164.1211G-CUD12K-01.ish.de 1120844902 M * Bertl Greek0: I'm not too happy with the vs_memory in sched.h ... 1120845165 M * jonsmel Hi Bertl, how are things going? 1120845199 M * Bertl fine, and for you? 1120845216 M * jonsmel oh, they're coming 1120845231 M * jonsmel ran into gfs corruption problems 1120845241 M * Bertl yeah? 1120845242 M * jonsmel It was working great but then the FS crashed 1120845248 M * jonsmel YEAH!! :) 1120845384 M * jonsmel Hey, I've got a quick question. 1120845395 M * Bertl let's hear! 1120845411 M * jonsmel I have a script that I have put into the pre-start dir but it does not get run when I start the vserver 1120845416 M * jonsmel what am I missing there 1120845488 M * jonsmel basically my script needs to do a mount before the vserver is started 1120845510 M * jonsmel any thoughts 1120845562 M * Bertl o gosh, sorry I have to run ... back later ... 1120845570 N * Bertl Bertl_oO 1120845657 M * daniel_hozac jonsmel: so it's in the pre-start.d dir and it's executable? 1120845670 M * daniel_hozac jonsmel: any reason you can't do the mount from fstab/fstab.remote? 1120845706 M * jonsmel well, i have under the vserver dir a dir called scripts and then one called pre-start 1120845712 M * jonsmel it is in there and is executable 1120845729 M * jonsmel well, the mount was just one example 1120845739 M * daniel_hozac pre-start.d 1120845741 M * daniel_hozac not pre-start 1120845741 M * jonsmel there are other scripts for other vservers that I will need to setup 1120845746 M * daniel_hozac pre-start is a script, IIRC. 1120845754 M * jonsmel pre-start.d is a dir correct than? 1120845773 M * daniel_hozac yeah. 1120845820 M * jonsmel ok, i will try that 1120847659 J * Hollow ~Hollow@home.xnull.de 1120848106 M * Rushmoom Ah, Hollow, nice to see you :) 1120848279 M * Rushmoom Oh, nevermind, my Pizza's ready, I'll bother you some other time.. Sorry :) 1120849307 N * jonsmel jonsmel_zZ 1120850614 Q * renihs Ping timeout: 480 seconds 1120852907 J * renihs ~renihs___@193.170.52.70 1120853354 Q * Vudumen Ping timeout: 480 seconds 1120853567 J * Aiken ~james@tooax8-173.dialup.optusnet.com.au 1120853586 J * Vudumen vudumen@perverz.hu 1120858786 Q * renihs Remote host closed the connection 1120861324 Q * prae Quit: Pwet 1120861625 Q * Doener` Quit: Leaving 1120861671 J * Doener ~doener@p54874F49.dip.t-dialin.net 1120862700 J * AprilDL ~chatzilla@ip68-9-200-247.ri.ri.cox.net 1120864311 Q * Val Quit: zZz 1120864501 Q * Rushmoom Quit: 1120865517 J * maharaja maharaja@ipax.at 1120866015 Q * sukria Ping timeout: 480 seconds 1120866481 N * Bertl_oO Bertl 1120866495 M * Bertl back now ... 1120866532 M * maharaja Bertl: hi 1120866561 M * maharaja is there any news about kernel oopses/crashes/etc. (maybe in combination with xfs?) 1120866594 M * Bertl no, do you bring some? everything working fine so far ... 1120866614 M * maharaja Bertl: unfortunatly not :-/ 1120866619 M * maharaja minicom is not working as expected 1120866640 M * maharaja it's not logging everything and i'm not able to use the magic req key 1120866640 M * Bertl minicom is working fine here in at least 5 cases ;) 1120866673 M * maharaja the server crashes every few houers 1120866688 J * sukria ~sukria@www.sukria.net 1120866689 M * maharaja 99% of the times it crashes while accessing the hd 1120866705 M * maharaja rith now, i flushed my postfix queue with 350 emails (mailserver disk has been full) 1120866707 M * Bertl really looks like an hardware issue to me ... 1120866707 M * maharaja -> crash 1120866728 M * maharaja Bertl: i switche the whole hardware (except of the hd which are running in an md-raid1 1120866831 M * maharaja i never occured any crashes like that which have been related to the harddisks 1120866876 M * Bertl well, what can I say .. several folks are using simialr setups without any issues 1120866889 M * maharaja may i get your minicom rc file? 1120866890 M * Bertl even xfs got some 'stress' testing recently ... 1120866928 M * Bertl sure, which one would you like? 1120866940 M * maharaja a working one for monitoring/debugging a server :) 1120867016 M * Bertl http://vserver.13thfloor.at/Stuff/minirc.TEST 1120867038 M * Bertl works here with my local testsetup, including debugging 1120867163 M * maharaja thank you very much! 1120867191 M * Bertl yp 1120867197 M * Bertl *yw