--- Log opened śro kwi 28 00:00:11 2004 00:04:57>> shuri [~shushushu@69.28.239.24] has quit [Ping timeout: 480 seconds] 00:08:23>> micah [micah@micha.hampshire.edu] has joined #vserver 00:08:29< micah> hola compadres 00:09:05>> shuri [~shushushu@69.28.239.24] has joined #vserver 00:09:17< micah> looking for a 2.4.26 grsec2+vserver-1.27 patch 00:23:34>> Bertl_oO is now known as Bertl_zZ 01:00:20>> Pierre [~pierre@ivry-5-82-224-217-20.fbx.proxad.net] has joined #vserver 01:35:54>> _shuri [~shushushu@69.28.239.24] has joined #vserver 01:41:13>> shuri [~shushushu@69.28.239.24] has quit [Ping timeout: 480 seconds] 01:41:13>> _shuri is now known as shuri 02:02:42>> _shuri [~shushushu@69.28.239.24] has joined #vserver 02:07:58>> shuri [~shushushu@69.28.239.24] has quit [Ping timeout: 480 seconds] 02:08:01>> _shuri is now known as shuri 04:08:15>> talon_ [talon@host-63-149-223-100.irwinresearch.com] has quit [Ping timeout: 480 seconds] 04:37:53>> dvorak [~dvorak@ny-amherst-C4-1-bg4a-3-4.bflony.adelphia.net] has joined #vserver 04:40:04< dvorak> Hi all, what are the pros and cons of having complete separation of virtual machines over increased sharing of resources 04:40:34< dvorak> what I mean is instead of copying all binaries and libraries into separate virtual machines 04:40:45< dvorak> can you minimize this sharing 04:41:00< dvorak> subsequently, is it worth the effort? 04:41:19< dvorak> s/sharing/copying 04:43:04< click> you can mount --bind previous to starting the vserver 04:43:11< click> that would work 04:59:38< dvorak> true, but would it be worth the effort 05:15:36< click> depends... 05:15:56< dvorak> click: on? 05:16:18< click> what you want to have within this resource 05:16:30< click> usr-based programs, administrative tools etc 05:16:36< click> *user-based 05:16:51< click> for user-based programs, it might be a nice feature 05:18:27< dvorak> true. If I make separate vservers, then each can be independently administered, however, I need to copy everything over 06:04:13>> dvorak [~dvorak@ny-amherst-C4-1-bg4a-3-4.bflony.adelphia.net] has left #vserver [] 06:08:58>> tchan [~tchan@24.12.194.132] has quit [Remote host closed the connection] 06:09:03>> tchan [~tchan@c-24-12-194-132.client.comcast.net] has joined #vserver 06:13:57>> Pierre [~pierre@ivry-5-82-224-217-20.fbx.proxad.net] has left #vserver [a+] 06:22:35>> click [click@gonnamakeyou.com] has quit [Ping timeout: 480 seconds] 07:29:03>> tchan [~tchan@c-24-12-194-132.client.comcast.net] has quit [Quit: leaving] 07:58:01>> gst [~gst@eris.sysfrog.org] has joined #vserver 08:45:35>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has joined #vserver 09:27:40>> Bertl_zZ is now known as Bertl 09:27:54< Bertl> morning everyone! 09:32:49< eyck> morning 09:34:29< Bertl> okay, translocating now, back in an hour ... 09:34:40>> Bertl is now known as Bertl_oO 09:45:44< eyck> hmm, they are already making translocators? woow, we really do live in 21st century 09:45:47< eyck> or was it 22nd 10:04:53>> rs [rs@ice.aspic.com] has joined #vserver 10:07:22< rs> hello 10:38:06< eyck> hi rs! 11:00:29>> Bertl_oO is now known as Bertl 11:00:54< Bertl> okay, translocation successful ;) 11:01:53< Bertl> rs: did you find time to test the new kernel? 11:02:42< rs> yes 11:02:50< rs> I sent you a private message 11:03:27< rs> the memory enforcement seems to work well but the namespace cleanup is still broken 11:03:38< Bertl> ah okay, obvoiously missed that, sorry ... :( 11:04:51< rs> the trace is almost the same 11:05:12< rs> http://rs.rhapsodyk.net/vserver/namespace_cleanup_crash2.txt 11:05:30< Bertl> yep, found it, thanks ... 11:06:35< Bertl> did you compile the kernel with debug info enabled? 11:07:23< rs> hmm maybe, which one ? 11:09:16< Bertl> 2.6.5-vs1.9.0pre12.2 11:09:38< rs> which debug infos ? 11:09:55< Bertl> hmm, the kernel debug info (one option)? 11:10:25< Bertl> sec .. searching .config 11:10:52< rs> CONFIG_DEBUG_KERNEL=y 11:10:53< Bertl> CONFIG_DEBUG_INFO=y 11:10:56< Bertl> yep 11:10:59< Bertl> okay ... 11:11:00< rs> # CONFIG_DEBUG_INFO is not set 11:11:13< rs> so, no :) 11:11:21< Bertl> okay, for the 'older' trace? 11:11:39< Bertl> (as they are almost identical) 11:11:41< rs> was the same config 11:11:58< rs> do you need it ? 11:12:01< Bertl> bad ... okay, please enable this option from now on for every kernel ;) 11:12:14< rs> ok 11:12:27< Bertl> enable it for this one and recompile the kernel ... 11:12:42< Bertl> (should not change anything in the kernel, besides adding the symbols) 11:13:27< rs> btw, I tried to understand to network accounting ouput, but I can't understand the difference between count and total for each fields 11:13:54< Bertl> count = number of 'packets', total = amount of data 11:15:55< rs> why hadn't I thought of that ? :) 11:16:08< rs> and 3 pos ? 11:17:32< rs> I recompile a kernel with debug info to reproduce the trace 11:17:47< rs> the trace will have more info or I have to do something special ? 11:18:35< Bertl> ad 3 pos 11:18:47< Bertl> accounted is input/output/error(fault) 11:19:07< Bertl> ad info: no, not much more info on the trace, but you can use addr2line on that ... 11:19:27>> ensc [~ircensc@ultra.csn.tu-chemnitz.de] has joined #vserver 11:19:32< Bertl> morning enrico! 11:19:35< rs> hi enrico 11:20:30< rs> Bertl: after reboot I suppose ? :) 11:20:41< Bertl> no, not required ... 11:20:53< Bertl> just use the 'old' trace for the same kernel 11:21:13< Bertl> take the EIP value (c0174252) and do 11:21:22< Bertl> addr2line -e vmlinux c0174252 11:23:16>> gst [~gst@eris.sysfrog.org] has quit [Quit: Client exiting] 11:26:18< rs> you mean c0174409 11:26:32< rs> c0174252 was with 12.1 11:26:42< Bertl> yes, sorry ... 11:27:08< Bertl> it's early in the morning ... %-) 11:27:32< rs> hehe :) 11:33:36< rs> addr2line -e vmlinux c0174409 11:33:36< rs> include/asm/spinlock.h:123 11:33:57< Bertl> okay, now start decreasing the address by steps of 4-8 11:34:03< Bertl> so c0174401 11:34:15< Bertl> then c01743F9 11:34:18< rs> hmm maybe the EIP is wrong now 11:34:29< rs> I also enabled the spinlock debugin 11:34:30< Bertl> no that looks good ... 11:34:34< Bertl> arg! 11:34:51< Bertl> okay, new trace please! 8-) 11:34:53< rs> I will reinstall the kernel and redo the test to be sure 11:35:41< Bertl> .oO( why do people never do what I suggest ;) 11:36:44< rs> because I noticed that one of your old suggestion (enabling spinlock debug) wasn't done :) 11:50:53< rs> wow this time I got a segfault with a kernel trace 11:51:05< Bertl> ahh ... 11:51:13< rs> by spinlock debuging apparently 11:51:38< Bertl> so it made sense to enable it .. :) 11:51:50< rs> http://rs.rhapsodyk.net/vserver/namespace_cleanup_crash3.txt 11:51:51< rs> :) 11:52:27< Bertl> okay, now we use the addr2line ... 11:52:38< Bertl> starting with c0183c98 11:52:48< rs> include/linux/mount.h:49 11:52:59< rs> I have to say that it's a 2.6.5 :) 11:53:22< Bertl> didn't expect that (2.6.5-vs1.9.0pre12.2) ;) 11:54:42< Bertl> ah, okay, so we are taking a spinlock twice! 11:55:03< Bertl> no, actually we are unlocking a not locked spinlock ;) 11:55:11< rs> :) 11:55:22< rs> vserver issue ? 11:55:36< Bertl> probably, as cleanup is not in vanilla ;) 11:55:54< rs> but namespace is, isn't it ? 11:56:29< Bertl> yep, ... nevertheless, let's continue the decrementing ... 11:56:34< Bertl> c0183c94 11:56:36< Bertl> c0183c90 11:56:50< Bertl> c0183c8c 11:56:53< Bertl> c0183c88 11:57:23< rs> c0183c8c: include/asm/spinlock.h:120 11:57:52< rs> same for 88 11:57:59< Bertl> okay, keep going ... 11:58:26< rs> c0183c84 include/asm/spinlock.h:119 11:58:57< rs> c0183c70 fs/namespace.c:313 11:59:04< rs> got it :) 11:59:28< Bertl> what is in your code at this line? 11:59:32< Bertl> for (p = mnt; p; p = next_mnt(p, mnt)) { 11:59:34< Bertl> ? 11:59:34< rs> } 11:59:39< rs> :o) 11:59:48< rs> end of umount_unused 11:59:49< Bertl> okay, so the next line is 11:59:52< Bertl> __umount_tree(mnt, &kill); 11:59:59< rs> next is do_umount 12:00:04< rs> declaration 12:00:14< rs> __umount_tree(mnt, &kill); 12:00:30< rs> it's the line 312 12:02:17< Bertl> okay ... goto kernel/vserver/namespace.c, vc_cleanup_namespace 12:02:32< Bertl> and add the following to the code: 12:02:49< Bertl> + spin_lock(&vfsmount_lock); 12:02:49< Bertl> umount_unused(current->namespace->root, current->fs); 12:02:49< Bertl> + spin_unlock(&vfsmount_lock); 12:03:30< rs> I keep spin_lock(&dcache_lock); 12:03:32< rs> ? 12:03:45< Bertl> yep ... just add this around the umount_unused 12:07:15< eyck> oh goodie :) mcp just praised my bsd patchset on wolk ML ;) 12:08:12< Bertl> .oO( hmm ... is this actually a good or a bad sign ;) 12:08:26< mcp> eyck: paste the sentence and Bertl will know it ;) 12:08:46< eyck> right away: 12:08:46< eyck> Cooking up a -ck, an -lck, an 12:08:46< eyck> - -bsd or whatever you think of is, to be real, kindergarten, peanuts, child 12:08:46< eyck> birthday party or whatever may imply it's trivial :p 12:08:50< eyck> :))) 12:09:01< eyck> end quote is right before ":)))" ;) 12:09:31< mcp> heheehe 12:09:38< Bertl> mcp: good news, vserver for wolk is on my todo, scheduled for the next days ... 12:09:50< mcp> cooolio 12:10:00< Bertl> needed to stabilize it first ;) 12:12:19< Bertl> s/needed/had/ 12:13:09< Bertl> btw, qemu 0.5.4 is out ... looks good .. 12:18:34>> serving [~serving@213.186.189.95] has quit [Read error: Connection reset by peer] 12:23:35< Bertl> ensc: what changed in .211? 12:26:20< Bertl> okay, off for now, back in an hour or two ... 12:34:06< rs> Bertl: about the memory enforcement policy, isn't it possible to force swapping for a context that is over limit ? 12:34:37< rs> I know that is need the RSS enforcement before 12:34:40< rs> it 12:43:24< rs> enrico ? 12:51:09< rs> lunch time, brb 13:43:11>> Netsplit uranium.oftc.net <-> nucleon.oftc.net quits: nalfein, yarihm, cereal^away, lilo, mcp, UFOczek 13:44:16>> Netsplit over, joins: mcp, yarihm, UFOczek, lilo, cereal^away, nalfein 13:52:18>> Doener [~doener@p5082DD41.dip.t-dialin.net] has joined #vserver 13:53:05< Doener> Hi! 13:59:42< rs> hello Doener 14:12:25< Bertl> okay, I'm back ... 14:12:32< rs> wb Bertl 14:13:33< Bertl> I thougt about that, but it is not that simple ... 14:14:11< Bertl> the thing is, linux-vserver is mostly about _sharing_ resources ... and therfore data might (will) belong to different vservers ... 14:14:40>> serving [~serving@213.186.189.95] has joined #vserver 14:23:32< rs> Bertl: enforcing the virtual memory seems to break many stuff 14:36:24< rs> btw your patch seems to work but I can't fully testing it while the vserver command fail to start the vserver because it doesn't find the vdir 14:36:36< rs> I think that the namespace cleanup clean the vserver mountpoint before enter it 14:36:48< rs> or something like that 14:37:44>> no_maam_ [~erik@datenzone.de] has joined #vserver 14:37:52>> no_maam [~erik@datenzone.de] has quit [Read error: Connection reset by peer] 14:42:59< Bertl> wow .. our network is great today ... 8-) 14:43:36< rs> your network ? 14:44:01< Bertl> I'm at our General Hospital atm ... 14:44:18< Bertl> (and I meant the network here in this section ;) 14:44:19< rs> are you seek ? 14:44:40< Bertl> no, fortunately not, I'm doing some work here ;) 14:44:48< rs> hehe ok :) 14:46:10< Bertl> okay, I was talking about RSS and swapping, as the network went offline ... 14:46:36< Bertl> ad enforcing the virtual memory seems to break many stuff: 14:46:53< Bertl> yeah, because a linux machine (until recently) did not limit virtual memory at all! 14:47:24< rs> yep, for exemple, apache with php4 loaded takes 140M of virtual memory 14:47:28< rs> and does nothing 14:47:40< rs> I don't understand why 14:47:49< rs> and it takes only 5k of RSS :) 14:48:12< Bertl> try enabling strict no overcommit on a vanilla kernel, and you'll see how many apps break ;) 14:48:35< rs> do you know why ? 14:48:52< Bertl> because VM was never enforced ... that's why 14:49:14< Bertl> if you can alloc 2GB of memory, and use only a small fraction of that ... why should you care? 14:49:44< Bertl> but back to RSS and limits: 14:50:10< rs> yes limits potential number of realloc :) 14:50:29< Bertl> the idea on short term is to enforce the RSS hard limit as VM limit (or at the places where VM checks now happen, basically fork and friends) 14:51:12< rs> but ? 14:51:22< Bertl> and to use the already planned 'soft' limits to controls swapping behaviour and penalize the contexts 14:55:01< Bertl> will need some testing (any volunteers?) to find the proper spots to do this 'enforcement' but I think this should make such a limit useable for providers ... 14:56:45< rs> I will test it :) 14:56:57< Bertl> I thought so ... ;) 14:56:57< rs> with about 40 testers for moment :) 14:57:09< rs> maybe more soon 14:57:20< rs> good alpha test isn't it ? :) 14:57:24< Bertl> the thing is we have to do some memory accounting first ... 14:57:28< Bertl> not bad! 14:57:44< Bertl> or should I say: "Not too shabby!" ? 14:58:44< Bertl> so what we need would be a machine which is relative fast on processing logs (kernel logging) 14:59:03< Bertl> and can store some amount of log data ... 14:59:22< Bertl> the next step would be to enable the logging (memory VM/RSS accounting) 14:59:28< Bertl> without any enforcements 14:59:58< Bertl> then make a log over, hmm .. let's say 2-3 hours of work ... to collect some 'usage' patterns ... 15:00:32< rs> logs are sent thru network and a dedicated machine process it 15:00:49< rs> what kind of processing do you need ? 15:01:29< Bertl> let's see if that setup is fast enough not to impact normal use ... 15:01:40< Bertl> in your kernel sources, change the folowing: 15:02:23< rs> it have to be on an in use machine ? 15:02:38< rs> or a machine that does nothing can feet ? 15:02:45< Bertl> hmm, well no, but it would be good to have a typical setup, with some applications running ... 15:02:56< rs> ok 15:03:06< Bertl> because if there doesn't happen anything, we won't get usage patterns ;) 15:03:37< Bertl> include/linux/vinline.h: ~180 15:04:06< Bertl> -#define VX_DEBUG_ACC_RSS 0 15:04:06< Bertl> -#define VX_DEBUG_ACC_VM 0 15:04:06< Bertl> -#define VX_DEBUG_ACC_VML 0 15:04:06< Bertl> +#define VX_DEBUG_ACC_RSS 1 15:04:06< Bertl> +#define VX_DEBUG_ACC_VM 1 15:04:08< Bertl> +#define VX_DEBUG_ACC_VML 1 15:04:38< rs> that's all ? 15:04:40< Bertl> then recompile (with make) 15:04:58< Bertl> yep 15:05:25< rs> make not gcc ? *joke* :) 15:05:41< Bertl> *G* 15:05:59< Bertl> gnu make has improved a lot in the last years ... ;) 15:06:06< rs> hehe :) 15:06:32< rs> btw, do you can me to disable AS rlimits during the test ? 15:07:10< Bertl> yep, do not enable any limits ... 15:07:35< Bertl> but as I said, be prepared for _alotof_ loggin ... 15:08:04< rs> ok 15:08:51>> Apollo [~Apollo@caracal.norcomcable.ca] has joined #vserver 15:09:12< Bertl> hi Apollo! 15:09:19< Apollo> Hi! 15:22:27< Apollo> Bertl do you have a couple minutes to help me over a hurdel? 15:23:09< rs> Bertl: do you want me to disable CPU limits too ? 15:24:29< Bertl> Apollo: sure, but our network is funny today ... 15:24:41< Bertl> now I see you, a minute later I don't ... 15:25:01< Apollo> ouch, k. 15:26:12< Apollo> I'm trying to enable quota's on / and /var inside the vserver. I've created a /dev/hdv1 as / and /dev/hdv2 as /var. On the outside, they sit as /dev/hda6 and hda7. 15:26:45< Apollo> When I enter in the vserver repquota -aug 15:27:41< Apollo> dev/hdv1 responds normal 15:27:51< Apollo> I mean, dev/hdv2 responds normal 15:28:09< Apollo> dev/hdv1 does not, it comes back with, repquota: Device (/dev/hdv1) filesystem is mounted on isn't block or character device nor it's loopback or bind mount. Skipping. 15:30:08< Bertl> hmm .. you did setup a vroot device for hda6 and hda7 15:30:30< Bertl> rs: CPU limits do not matter ... 15:30:37< Apollo> umm... probably not. How does one do that. 15:32:11< Apollo> to create hdv2 I did a cp -av /dev/hda7 /vserver location/dev/hdv2 15:32:31< Apollo> I left, hdv1 as it was created by vserver bulid. 15:33:50< Bertl> okay, that explains it ;) 15:34:10< Bertl> for security reasons, you should not copy the block devices to the vserver ... 15:34:20>> monrad [~monrad@213083190231.sonofon.dk] has joined #vserver 15:34:29< Bertl> if I logon to such a vserver, I can take over your host without any toubles ;) 15:34:40< Apollo> oh... um, oops! 15:34:49< Bertl> that is, what the vroot device was designed for ... 15:35:01< Bertl> it proxies the quota ioctls to the 'real' block device ... 15:35:22< Bertl> and you need one for hdv1 and another for hdv2 ... 15:35:41< Bertl> they are in stable/devel, so you just have to enable them on kernel build (if not already done so) 15:36:00< Bertl> and configure them with the vrsetup utility ... 15:36:05< Apollo> ah ok, i'll look deeper into that. 15:36:14< Bertl> (which has a similar syntax as losetup) 15:36:53< Apollo> k, big thx... off on a mission now. 15:37:20< Bertl> a mission from glod? 15:37:53< Bertl> anyway, you are welcome ... 15:38:46>> Doener` [~doener@p5082BE21.dip.t-dialin.net] has joined #vserver 15:39:02< Bertl> hi Doener! hi monrad! 15:39:09< monrad> hi 15:43:28>> Doener [~doener@p5082DD41.dip.t-dialin.net] has quit [Ping timeout: 480 seconds] 15:51:56 * Doener` still wonders why his isp forces disconnection after 24 hours... 15:53:09< Bertl> hmm, amybe they use M$ routers? 15:59:14< eyck> because it's HIP, all the 'isp's' around here disconnect after few hours... 16:00:07< eyck> besides, as latest bugtraq hooplaa reminds us all, it's dangerous to rely on TCP connections being..well..reliable... 16:03:03< Bertl> rs: btw, could you scan your logs for messages like "!!! limit: ... on exit." ? 16:03:40< rs> I cant, network pb 16:04:03< Bertl> hehe, network problem day ;) 16:05:58< rs> yes 16:08:10< Bertl> I hope you didn't flood your network with the logging ... 16:19:55< eyck> hmm, I think you're really hoping that he DID flood his network, and you'll get some meaningfull logs from this.. 16:20:44< Bertl> shhht! 16:21:53< Bertl> btw, doesn't look good for BME ... 16:22:15< eyck> BME? 16:22:22< Bertl> bind mount extensions ... 16:23:21< eyck> pity, maybe in 2.8 16:32:33< Apollo> Bertl: Thanks... works now! 16:33:17< Bertl> perfect! 16:34:20< eyck> good work everyone, 16:34:21< eyck> ;) 16:49:28>> _id [~id@pD95E9BB8.dip.t-dialin.net] has joined #vserver 16:49:34< _id> hello =) 16:50:55< Bertl> hi _id! 16:52:06< _id> i am going to try your fine vserver - i am pretty amazed by its possibilities 16:52:46< Bertl> yeah, it's amazing what you can do with free software ;) 16:53:53< _id> i only use free/GNU software for at least 5 years for my personal pleasure# 16:54:45< _id> only my old company sued me to use Windoze and Solaris 16:55:45< _id> but this is history =) - i left ! 16:56:15< Bertl> hehe, solaris ... maybe even on intel? 16:57:05< _id> i tied the x86 branch - awful ... on PPC its OK 16:57:19< _id> but Debian is much more convienent 16:57:23< Bertl> SPARC ;) 16:57:24< _id> or gentoo 16:57:34< _id> arg yes 16:58:19< _id> i will be back in a few hours 16:58:21< _id> cya 16:58:26>> _id is now known as _id_qw 16:58:26< Bertl> cya! 17:06:33>> Apollo [~Apollo@caracal.norcomcable.ca] has quit [Quit: Trillian (http://www.ceruleanstudios.com)] 17:06:55>> Apollo [~throwaway@caracal.norcomcable.ca] has joined #vserver 17:12:34< rs> Bertl: you're right, it make a lot of log 17:12:51< rs> the ssh server can't start :o) 17:13:14< rs> vx_pages_avail[ -1, 9]: 1 > 0 + 1 in mm/mmap.c:1402 17:13:14< rs> vxd: vx_acc_pages [ -1, 9]: 0 += 1 in mm/mmap.c:1442 17:13:14< rs> vx_pages_avail[ -1, 9]: 1 > 0 + 1 in mm/mmap.c:1402 17:13:16< rs> ... :) 17:24:04< rs> I have to go, see you later 17:24:15>> rs [rs@ice.aspic.com] has quit [Quit: leaving] 17:26:19>> Bertl is now known as Bertl_oO 18:15:23>> _shuri [~shushushu@69.28.239.24] has joined #vserver 18:19:58>> shuri [~shushushu@69.28.239.24] has quit [Ping timeout: 480 seconds] 18:19:59>> _shuri is now known as shuri 18:47:03>> _id_qw is now known as _id 18:47:06< _id> re 19:20:58>> hadge [~me@h64-5-199-35.gtcust.grouptelecom.net] has joined #vserver 19:21:47< hadge> Hello all, I was wondering if there was a changelog somewhere that indicates what was changed, or what now works with util-vserver-0.29.211 19:23:53>> _id is now known as _id_23 20:05:29>> surriel [~riel@imladris.surriel.com] has joined #vserver 20:10:23>> dsanta [~santa@c68.190.156.105.roc.mn.charter.com] has quit [Ping timeout: 480 seconds] 20:16:29>> surriel is now known as riel 20:16:33>> riel [~riel@imladris.surriel.com] has left #vserver [Client exiting] 20:27:45>> dsanta [~santa@c68.190.156.105.roc.mn.charter.com] has joined #vserver 20:35:53>> dsanta [~santa@c68.190.156.105.roc.mn.charter.com] has quit [Ping timeout: 480 seconds] 20:56:43< ensc> hadge: http://savannah.nongnu.org/cgi-bin/viewcvs/util-vserver/util-vserver/doc/compatibility.xml?rev=1.2&content-type=text/vnd.viewcvs-markup lists changed commands (but is a little bit outdated) 20:57:02< ensc> hadge: configuration has been changed to 100% 20:59:06< ensc> a low level changelog is in the tarballs 21:01:42>> LeXo [~LeXo@lns-th2-4f-81-56-252-185.adsl.proxad.net] has joined #vserver 21:01:45>> LeXo [~LeXo@lns-th2-4f-81-56-252-185.adsl.proxad.net] has quit [Read error: Connection reset by peer] 21:01:55>> LeXo [~LeXo@lns-th2-4f-81-56-252-185.adsl.proxad.net] has joined #vserver 21:05:13>> LeXo_ [~LeXo@lns-th2-4f-81-56-252-185.adsl.proxad.net] has joined #vserver 21:06:58>> LeXo_ [~LeXo@lns-th2-4f-81-56-252-185.adsl.proxad.net] has quit [Quit: ] 21:07:15>> LeXo is now known as franck 21:11:04>> Apollo [~throwaway@caracal.norcomcable.ca] has quit [Quit: ] 21:32:05>> JonB [~NoSuchUse@129.142.112.33.ip.tele2adsl.dk] has joined #vserver 21:32:07>> hadge [~me@h64-5-199-35.gtcust.grouptelecom.net] has quit [Quit: ] 21:36:08>> Bertl_oO is now known as Bertl 21:36:25< Bertl> evening everyone! 21:36:26< JonB> hey Bertl 21:37:11< Bertl> hey Jon, how are you? 21:38:07< JonB> Bertl: busy, so i havent been working on my project :( 21:41:50< Bertl> hmm, busy like having fun, or busy like trying to stop the waterleakage in the basement? 21:41:51< JonB> Bertl: i installed subversion at my work, and did some cvs2svn conversion tests 21:42:10< JonB> Bertl: busy studying another project, and a little work 21:42:22< JonB> a good thing that i dont need to turn it in before the 1. aug 21:42:34< JonB> Bertl: what do you use ? 21:42:40< Bertl> for what? 21:42:52< Bertl> revision control? 21:43:23< JonB> Bertl: yes 21:44:07< Bertl> well, I have a bunch of my stuff in cvs, but I'm not very happy with that ... wanted to try out gnu arch, as it seemed very advanced .... 21:44:38< Bertl> but for the kernel stuff, I consider all this not really appropriate ... 21:45:18< JonB> Bertl: why not approciate? 21:45:33< JonB> Bertl: what was it with cvs you were unhappy with? 21:46:04< Bertl> well, for example, I have the patches for vserver 2.6 in a split form ... 21:46:19< JonB> split form ? 21:46:38< Bertl> so there are 18 (or 19) separate patches, which all together give what I upload as the vserver 1.9.x patch 21:47:04< JonB> why this many ? 21:47:34< Bertl> they are separated by topics, each dealing with a separate feature/area ... 21:48:03< Bertl> e.g. one dealing with the arch specific vserver syscalls ... 21:48:17< Bertl> another dealing with the immutable unlink stuff ... 21:48:50< Bertl> when the kernel changes (for example some syscalls are added) I just have to 'adapt' the patches affected (usually 1-2) 21:49:17< Bertl> which can be easily reviewed, as they only deal with one issue/topic at a time 21:50:05< Bertl> applying and maintaining them, is done by some shell scripts, so no additional work for me there ... 21:50:31< JonB> okay 21:50:53< Bertl> I know that this isn't suitable for userspace development ... 21:51:23< JonB> i suppose it is not suitable if you are more developers 21:51:48< Bertl> well, probably that would cause trouble too, yes ;) 21:53:06< Bertl> ensc: are you still around? 21:53:16< ensc> yep 21:53:29< Bertl> I'd like to discuss the cleanup issue, if you have some time? 21:54:05< ensc> I have to prepare the LinuxTag papers 21:54:57< Bertl> just let me know, when you take a break or so ... ;) 21:56:15< ensc> ok 21:56:54< JonB> howcome it isnt bertl who is going to linuxtag? 21:57:40< Bertl> is that so? ;) 21:58:45< JonB> Bertl: yeah, that is so 22:01:50< Bertl> well,what should I say? 22:02:48< JonB> Bertl: talk about vserver ofc. 22:03:02< Bertl> I will, I will ... 22:03:39< JonB> Bertl: good 22:08:56>> dsanta [~santa@c68.190.156.105.roc.mn.charter.com] has joined #vserver 22:09:07< Bertl> hi dsanta! 22:09:40>> JonB [~NoSuchUse@129.142.112.33.ip.tele2adsl.dk] has quit [Quit: Leaving] 22:20:07>> dsanta [~santa@c68.190.156.105.roc.mn.charter.com] has quit [Ping timeout: 480 seconds] 22:21:18>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has quit [Remote host closed the connection] 23:19:12< Bertl> okay, away for a little (back in30 or so .. ) 23:19:21>> Bertl is now known as Bertl_oO 23:37:47>> _shuri [~shushushu@69.28.239.24] has joined #vserver 23:41:46>> shuri [~shushushu@69.28.239.24] has quit [Ping timeout: 480 seconds] 23:41:49>> _shuri is now known as shuri 23:49:02>> Bertl_oO is now known as Bertl --- Log closed czw kwi 29 00:00:02 2004