--- Log opened czw maj 13 00:00:53 2004
00:05:01>> eyck [~eyck@62.233.189.138] has quit [Remote host closed the connection]
00:08:32>> eyck [~eyck@62.233.189.138] has joined #vserver
00:13:42>> _id [~id@pD9519056.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
00:24:39>> _id [~id@pD9519250.dip.t-dialin.net] has joined #vserver
00:24:53< Bertl> hello _id!
00:24:57>> taxcollector [~taxcollec@192.16.167.161] has joined #vserver
00:25:06< Bertl> greetings ryan!
00:25:12< taxcollector> Howdy
00:30:32< taxcollector> I am not too familiar with the non-alpha util-vserver, but from the chatter on the mailing list it has a "vserver-from-disk-image" feature? Is this a feature that needs to be brought to the alpha tools, or was it a feature that was dropped on purpose?
00:35:33< Bertl> hmm, vserver-from-disk-image, not sure what you are referring to ...
00:37:20< taxcollector> Like the ones mentioned on this page: http://www.jvds.com/vserver/
00:39:12< taxcollector> Perhaps I misunderstand how people use these images and it is ad-hoc
00:40:03< Bertl> most of them are just unpacked into a directory ... IIRC
00:40:23< Bertl> beats me why they are called 'images' ;)
00:41:10< taxcollector> OK; Thanks
00:41:57< Bertl> old tools (especially the newvserver script) support 'cloning' existing vservers or the root installation
00:42:19< taxcollector> Was that just an rsync?
00:42:46< Bertl> newer tools (alpha) support creation from vservers out of thin air ;) and the cloning process but not for the installed host system afaik.
00:43:03< Bertl> actually just a cp ...
00:43:59< monrad> thin air cool
00:44:25< Bertl> amazing isn't it? ;)
00:44:33< taxcollector> I am familiar with the way alpha creates vsevers with apt-get, etc. but by "cloning" do you mean unification and reference vsevers?
00:44:47< Bertl> yep
00:44:54< monrad> it would be nice if we could create vserver from the waste h2o that hydrogen cars produces
00:45:34< Bertl> probably will be done when they are more wide spread ... *G* 
01:05:01< Bertl> anybody interested in testing a highly experimental patch for 1.9.0 8-) ?
01:06:55< monrad> well my server is in the "if it aint broke dont fix it" state for the moment :)
01:07:52< taxcollector> I don't have any robust testing methodology at the moment, but I could try basic stuff with it
01:08:16< Bertl> sounds good, UP or SMP?
01:08:24< taxcollector> SMP
01:08:30< Bertl> that would be great!
01:09:39< Bertl> okay, let me upload a patch, I'll explain the changes and how to test when the kernel is compiling ...
01:09:46< Bertl> okay?
01:09:58< taxcollector> OK
01:14:59< Bertl> http://vserver.13thfloor.at/Experimental/patch-vs1.9.0-rcu0.1.diff
01:15:12< Bertl> (this is ontop of vs1.9.0 for 2.6.6)
01:15:36< taxcollector> OK; give me a couple of minutes to get the compiling going
01:15:54< Bertl> sure, let me know when you are ready ...
01:16:08>> gilbert [~gilbert@208-186-222-203.nrp4.brv.mn.frontiernet.net] has joined #vserver
01:16:18< Bertl> hello gilbert!
01:16:22< gilbert> hey
01:16:25< gilbert> got a question
01:16:30< gilbert> about mysql
01:16:40< Bertl> hey this is linux-vserver ;)
01:16:54< gilbert> well they colide
01:16:57< Bertl> at least try to make it look vserver related *G*
01:17:12< Bertl> okay, what's the issue?
01:17:13< gilbert> okie dokie dude :)
01:17:30< gilbert> how do i get mysql to run on each seperate server
01:17:41< gilbert> each seperate vserver
01:18:04< Bertl> hmm /etc/rc.d/init.d/mysql start ? (probably not what you want to hear, right?)
01:18:36< gilbert> thats half the batte :)
01:18:42< gilbert> but where im stuck
01:18:47< gilbert> is
01:18:56< gilbert> on the network configuration
01:19:17< Bertl> okay, you use different 'real' IPs for the servers?
01:19:44< gilbert> ok go here: http://www.linux-vserver.org/index.php?page=linux-vserver_administrators_guide
01:19:58< gilbert> down tooo 2. Daemons on the host server
01:20:04< Bertl> on my way ...
01:20:06< gilbert> thats where i am looking
01:20:08< gilbert> thanks man
01:20:12< gilbert> you rock 
01:20:37< gilbert> can i use 'v_xinetd.sysv' in some way
01:20:39< Bertl> okay, 2. Daemons on the host server that is?
01:21:39< Bertl> most likely you should take one step back, and rethink what is addressed here ...
01:21:58< Bertl> Host is the machine 'hosting' the Guests ...
01:22:10< Bertl> (the linux-vservers)
01:22:14< gilbert> oh ic
01:22:31< Bertl> the v_* wrapper are there to 'limit' the Host services to a subset of the available IPs
01:22:59< Bertl> so that for example the sshd (on the Host) doesn't interfere with the sshd on your vservers ...
01:23:22< gilbert> mmmmmk
01:23:31< Bertl> _inside_ a vserver, you do not need _any_ of those scripts ... on the contrary they will cause you trouble ...
01:24:08< gilbert> so if i move the main servers services into a vserver
01:24:18< gilbert> and then make another
01:24:26< gilbert> they wont conflict either
01:24:28< Bertl> the configuration of each vserver _limits_ the vserver to a specific set of addresses, and services binding from inside a vserver, automatically get those restrictions applied ...
01:24:55< Bertl> so if your vps1 has 192.168.0.1 and you have another one with 192.168.0.2
01:25:11< Bertl> then starting an sshd in both will work as expected ...
01:25:24< gilbert> interesting
01:25:28< Bertl> the first one will be bound to 192.168.0.1, the second one to 192.168.0.2
01:26:25< Bertl> but! if you have an unrestricted (no v_* wrapper) sshd on the host already running, both sshds inside the vservers will fail, with something like address/port already in use ...
01:26:46< gilbert> ok thanks bert!
01:26:54< Bertl> you're welcome!
01:27:34< gilbert> when i get back in here again ill have to give you an update
01:28:14< Bertl> I'd appreciate that ... (actually I appreciate any feedback)
01:29:58< taxcollector> Bertl: Going take a few more minutes; haven't configured 2.6 on this box before
01:30:08< Bertl> np, take your time ...
01:33:31< Bertl> taxcollector: just a side note: don't use an existing 2.4 config as base for 2.6 kernels ... but you probably know that already ...
01:33:46< taxcollector> Yup. Thanks though.
01:34:34< Bertl> ah, and another one: make sure to activate spin lock debugging in the kernel hacking section ... 
01:34:54< Bertl> (as well as magic sysreq key)
01:35:09< Bertl> hmm, and debug info would be nice too 8-)
01:35:33< taxcollector> OK
01:40:15< taxcollector> Does enabling legacy kernel API mess up not-legacy tools?
01:40:51< Bertl> in which section is that? linux-vserver?
01:40:59< taxcollector> Yeah
01:41:25< Bertl> no on the contrary, legacy is still required for recent alpha tools ...
01:41:49< Bertl> you can disable the proc security if you do not want to mess with that ...
01:42:22< Bertl> (although recent tools should handle it fine)
01:44:06< Bertl> btw, you know your surename is actually a german word?
01:44:19< taxcollector> Yes; "Court tailor" :)
01:44:31< taxcollector> My great-great grandfather was a German sailor
01:44:56< taxcollector> babelfish translates it as "yard cutter" which reminds me of a lawnmower
01:45:05< Bertl> *LOL*
01:45:11< taxcollector>  CC kernel/vserver/context.o
01:45:11< taxcollector> kernel/vserver/context.c:682: `free_vx_info' undeclared here (not in a function)
01:45:21< taxcollector> Have I misapplied something?
01:45:23< Bertl> compiling as modules?
01:45:35< Bertl> something, anything?
01:45:52< taxcollector> yes to first question
01:45:59< taxcollector> either or to second
01:46:11< Bertl> okay could you upload your .config somewhere?
01:46:27< taxcollector> OK
01:46:51< Bertl> hmm, is this a warning or an error, btw?
01:47:11< Bertl> I mean does it continue with the compilation after the message?
01:47:30< taxcollector> Continues...
01:47:32< taxcollector> kernel/vserver/context.c:682: initializer element is not constant
01:47:32< taxcollector> kernel/vserver/context.c:682: (near initialization for `__ksymtab_free_vx_info.value')
01:47:33< taxcollector> kernel/vserver/context.c:683: `vxlist_lock' undeclared here (not in a function)
01:47:33< taxcollector> kernel/vserver/context.c:683: initializer element is not constant
01:47:33< taxcollector> kernel/vserver/context.c:683: (near initialization for `__ksymtab_vxlist_lock.value')
01:47:33< taxcollector> make[2]: *** [kernel/vserver/context.o] Error 1
01:47:33< taxcollector> make[1]: *** [kernel/vserver] Error 2
01:47:35< taxcollector> make: *** [kernel] Error 2
01:47:47< Bertl> okay, that's an error though ...
01:48:50< Bertl> ah, I see, it's because of the modules, didn't test this yet ... give me a second to fix it up ...
01:49:13< taxcollector> OK
01:49:38< taxcollector> Was forced to change my password on my websever today and I'm having trouble remembering it now anyways
01:50:00< Bertl> hmm, actually it might be sufficient to just comment out the two lines at kernel/vserver/context.c 682/683 ...
01:50:30< Bertl> (as long as you didn't decide to make everything modular)
01:54:49< taxcollector> Still churning, forgot to use "-j 4"
01:54:57>> gilbert [~gilbert@208-186-222-203.nrp4.brv.mn.frontiernet.net] has quit [Quit: ]
01:55:32< Bertl> again some notes: you can stop the current build system at any time, you do not need to do a 'make clean' if you changed something ...
01:55:58< taxcollector> Yeah; I killed it and restarted it
01:56:00< Bertl> (this has improved a lot over the 2.4 build system)
01:56:09< taxcollector> building the modules now
01:56:26< taxcollector> Something I've never been sure of, maybe you know...
01:56:55< taxcollector> If you've chosen to use modules...
01:57:01< Bertl> yep?
01:57:02< taxcollector> you always have to rebuild them too, right
01:57:20< Bertl> you mean if you change/rebuild the kernel?
01:57:24< taxcollector> correct
01:58:07< Bertl> normally with 2.6 you just do a 'make' (no target specified) and all dependancies are resolved, modules are rebuilt where required
01:58:15< taxcollector> OK
01:58:24< taxcollector> Do you recall if this is different for 2.4?
01:58:50< Bertl> you still have to do a separate 'make modules_install' or 'make install', and it's a little different for 2.4 ...
01:59:43< Bertl> in 2.4 the modules might get strange versioning, depending on the build and the config options, so they might require a complete rebuild of the kernel/modules ...
02:01:29< Bertl> okay, I wanted to explain the purpose of the experimental patch to you ...
02:01:54< Bertl> (if you are interested what it actually does ;)
02:02:10< taxcollector> Now is a good time...
02:02:10< taxcollector> and I am interested :)
02:02:51< Bertl> there is a lot of passing structure references between the different 'objects' in the vs1.9.0 kernel patch
02:03:21< Bertl> for example the context is referenced from each task starting within or migrating to the context ...
02:04:16< Bertl> this involves a big deal of locking/unlocking in various kernel locations ... and I was thinking for a while now, that there must be a better way of doing this (with less overhead)
02:04:31< taxcollector> OK
02:05:29< Bertl> actually the current overhead isn't measurable except with micro benchmarking ... but there is a limit to the scalability, because locks tend to degrade performance when it comes to heavy multiprocessing ... 
02:06:00< Bertl> (probably not measurable up to 4 or 8 cpus)
02:06:27< Bertl> nevertheless, I think I found a good method to get rid of most of the locks used for now ...
02:07:20< Bertl> what I do not know yet is if there are cases where this 'new' approach will fail horrably, so we have to do some testing ...
02:08:39< Bertl> this 'hack' now replaces the entire context allocation and reference counting by something called 'read copy update' (RCU) and a hash table to store the actual context structures ...
02:08:51< taxcollector> OK
02:09:33< Bertl> If it works as expected, I'll explain why ;)
02:10:09< taxcollector> Everything has compiled and been installed; I'm just double checking I have a current version of mod-init-tools
02:10:39< Bertl> can you reboot this machine in case of a panic?
02:10:53< taxcollector> Yeah
02:11:04< Bertl> okay, do you have a direct or serial console?
02:11:05< taxcollector> but it is about a 15 minute round trip
02:11:36< taxcollector> There is a console
02:11:37< Bertl> do you use grub or lilo (or maybe loadlin?)
02:11:45< taxcollector> grub
02:12:13< taxcollector> 15 minute walking
02:12:30< taxcollector> just across campus so it isn't a big deal
02:12:36< Bertl> okay, make sure there is an additional boot argument 'panic=10' there .. so the machine will reboot 10 secs after a panic ...
02:12:59< taxcollector> Didn't know about that one; thanks.
02:13:37< Bertl> you
02:13:42< Bertl> 'r welcome ...
02:13:46< Zoiah> Bertl: me neither, thanks!
02:15:02< Bertl> there is an interesting docu about all those options, and guess what, it's included in the kernel source ;)
02:15:26< Bertl> Documentation/kernel-parameters.txt
02:15:26< taxcollector> :)
02:16:14< Zoiah> panic=          [KNL] kernel behaviour on panic.
02:16:15< Zoiah> Hmm...
02:16:30< taxcollector> Not coming back up...
02:16:46< taxcollector> I'm off to kick it...
02:17:06< taxcollector> Will you still be around Bertl?
02:17:12< Bertl> hmm, probably something unrelated ... did you setup grub to only boot once?
02:17:25< taxcollector> I don't recall
02:17:34< Bertl> sure I'll be around ... now that it's interesting ...
02:18:23< Bertl> okay, be prepared to either add an initrd (to provide the filesystem module) or to reconfigure the partition name ...
02:26:03< Bertl> Zoiah: yeah, the descriptions are somewhat sparse ...
02:28:41>> monrad [~monrad@213.83.190.226] has quit [Quit: Leaving]
02:42:11>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has quit [Quit: Leaving]
02:48:13>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has joined #vserver
02:51:04< SHu_here> (Bertl): okay, make sure there is an additional boot argument 'panic=10' there .. so the machine will reboot 10 secs after a panic ...
02:51:24< SHu_here> but if the kernel crash first it will alway crash even with rebbot
02:51:51< Bertl> well, that is why there was ...
02:51:54< Bertl> 02:17 < Bertl> hmm, probably something unrelated ... did you setup grub to only                boot once?
02:52:18< taxcollector> OK; I'm back...
02:52:21< SHu_here> hehe
02:52:22< SHu_here> ok
02:52:38< taxcollector> I probably wasn't the best guinea pig for this test I guess, but I am not going to give up...
02:52:51< Bertl> that's the spirit!
02:53:15< Bertl> now tell us, what was the reason for the 'panic?'
02:53:27< taxcollector> I updated mod-init-tools, enabled ext3 as built-in
02:53:28< taxcollector> ...
02:53:53< taxcollector> It is complaining that it can't read sda2
02:54:00< taxcollector> I still having quite figured out why not
02:54:10< taxcollector> There are other errors above it that I think are the culprit...
02:54:38< Bertl> hmm, sda2, scsi compiled as modules?
02:54:46< taxcollector> I was hoping it was just because I compiled ext3 as a module
02:54:53< taxcollector> I'll check those now
02:55:29< Bertl> hmm, maybe you should 'just' use an initial ramdisk ...
02:55:54< Bertl> mkinitrd should be able to build a nice one ...
03:01:14< Bertl> taxcollector: and don't worry, you are doing well for the first 2.6 run on that machine ;)
03:01:24< taxcollector> :)
03:05:25< Bertl> telling from the kernel compile time, the machine seems quite fast, what is it, dual PIII?
03:06:18< taxcollector> duel xeon
03:06:25< taxcollector> s/duel/dual
03:06:48< Bertl> hmm, cool, so we actually get 4 cpus with HT ... nice ;)
03:06:57< taxcollector> Yes :)
03:08:01< taxcollector> I hate to say this Bertl, but I'm getting compile errors now that I've included the appropriate SCSI drivers into the kernel, and not as modules. Not Vserver errors of course...
03:08:11< taxcollector> I'm going to have to step out for few hours...
--- Log opened czw maj 13 13:52:54 2004
13:52:54>> albeiro [albeiro@linux.gentoo.pl] has joined #vserver
13:52:54>> Irssi: #vserver: Total of 35 nicks [0 ops, 0 halfops, 0 voices, 35 normal]
13:52:54>> Irssi: Join to #vserver was synced in 1 secs
14:08:50>> Napalm [~napalm@host81-7-20-232.adsl.v21.co.uk] has joined #vserver
14:08:55< Napalm> hello everyone
14:09:08< Napalm> aww berts still asleep
14:09:09< Napalm> :)
14:15:46>> Napalm [~napalm@host81-7-20-232.adsl.v21.co.uk] has quit [Quit: ]
14:33:37>> serving [~serving@213.186.191.31] has joined #vserver
14:40:03>> franck [~franck@62.240.242.34] has quit [Ping timeout: 480 seconds]
14:47:17>> Bertl_zZ is now known as Bertl
14:47:22< Bertl> morning everyone!
14:48:13>> Brucey [cinacio@snape.aspic.com] has joined #vserver
14:48:46< Brucey> hi all :)
14:48:48< Bertl> eyck: wossname is a Terry Pratchett term, meaning 'whateverit is called' or 'what was it's name again?'
14:48:55< Bertl> hi Brucey!
14:55:43>> franck [~franck@62.240.242.34] has joined #vserver
14:58:09>> pflanze [~chris@195.226.6.75] has quit [Ping timeout: 480 seconds]
14:58:12< Bertl> hello franck!
15:01:30< kestrel> hey herbert
15:01:58< Bertl> greeting Alec!
15:08:18>> Napalm [~napalm@81.7.20.232] has joined #vserver
15:08:22< Napalm> hello bert
15:08:28< Napalm> i see your up now
15:08:33< Napalm> :D
15:09:42< eyck> Bertl: thnx.
15:09:46< Bertl> yep, I'm awake now ;)
15:09:56< kestrel> ah, finally! :P
15:10:01< Napalm> has anyone seen error messages like "SIOCSIFADDR: File exists"
15:10:02< kestrel> err, sorry, wrong window :)
15:10:24< Napalm> ive also got "SIOCSIFNETMASK: Cannot assign requested address"
15:11:11< Napalm> when trying to run util-vservers-0.29.4
15:11:19< Bertl> yep, seen that several times ...
15:12:22< Bertl> try to run the testme.sh on the host, and let me know what it says ...
15:15:13< Napalm> newvserver does link or copy any files whatsoever
15:15:26< Napalm> dev and proc only
15:16:40< Napalm> brb
15:16:43>> Napalm [~napalm@81.7.20.232] has quit [Quit: ]
15:25:40>> hiaslboy_oOOo [matthias@e-16.vc-graz.ac.at] has quit [Ping timeout: 480 seconds]
15:27:42>> Napalm [~napalm@host81-7-20-232.adsl.v21.co.uk] has joined #vserver
15:27:43< Napalm> back
15:27:59< Napalm> sorry about that bert, router troubles
15:28:16< Bertl> np
15:30:16< Napalm> Bert: i think i asked you this question before, but here goes again because i've forgotten the answer. I should be able to run debian Vservers on a RH9 Host server yea?
15:30:38< Bertl> yes
15:30:47< Napalm> interesting
15:30:48< Napalm> :)
15:32:35< Bertl> yeah, it's a fascinating technology ;)
15:42:20>> Apollo [~Apollo@caracal.norcomcable.ca] has joined #vserver
15:42:43< Bertl> hi Apollo!
15:42:57< Apollo> hi Bertl
16:07:13>> monrad [~monrad@213083190226.sonofon.dk] has joined #vserver
16:07:26< Bertl> greetings monrad!
16:15:34>> franck [~franck@62.240.242.34] has quit [Quit: User pushed the X - because it's Xtra, baby]
16:15:44>> eyck [~eyck@62.233.189.138] has quit [Ping timeout: 480 seconds]
16:17:14< monrad> hi
16:22:38< Filther> hey Bertl
16:22:51< Bertl> hi Filther!
16:22:54< Filther> can you see my issue? :)
16:23:35< Filther> it occurs when I stop a vps.. all others seem to lose network-awareness
16:23:48< Bertl> hmm, I saw that you had a lengthy discussion, but I didn't read it completely ...
16:24:19< Bertl> thing is probably that your vps for whatever reason doesn't create secondaries, but a primary address
16:24:31< Bertl> this on shutdown will take down the entire interface ...
16:24:36< Filther> hmm.. how can I create secondarys?
16:24:38< Filther> :)
16:25:09< Bertl> recent tools should do that by default, but a simple workaround would be to create 1-2 dummy addresses in the same network on the same interface
16:25:15< Bertl> (from the host side)
16:25:41< Bertl> if there _is_ a primary, then every further alias (in the same network) will become a secondary
16:26:07< Bertl> another method would be to create /32 addresses, because they are always singles ...
16:26:42< Filther> so, you say that if I create a new vps with the latest tools, everything should be fine?
16:26:52< Filther> (and with the latest kernel patch)
16:27:24< Bertl> the alpha tools should handle this fine afaik, but no guarantees ...
16:30:24>> kestrel [athomas@38.6.233.220.exetel.com.au] has quit [Quit: bye]
16:36:27< Filther> ok
16:36:29< Filther> I'll try
16:36:30< Filther> thx
16:36:33< Filther> byebye!
16:36:36>> Filther [~root@212.52.166.22] has quit [Quit: Filther has no reason]
16:39:05< Doener`> re
16:47:14>> _id [~id@pD951914A.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
16:47:39< Bertl> hi Doener`!
16:47:51>> eyck [~eyck@62.233.189.138] has joined #vserver
16:48:23< Doener`> hmm.. must have missed that the tools also bring up the interface on 'exec'...
16:49:37< Bertl> hmm, I'd say, no big deal ...
16:55:25< Doener`> Bertl: do you want me to test anything? i'm going to upgrade my kernel anyways, so i can stand some crashing today :)
16:56:01< Bertl> sure, UP or SMP system?
16:56:08< Doener`> UP
16:56:47< Bertl> I'll have a patch in about 30min that could use special testing ... (with PREEMPT in this case)
16:56:55< Doener`> ok
17:04:54>> pflanze [~chris@gate.wyona.com] has joined #vserver
17:05:41< Bertl> hello chris!
17:06:26< Doener`> hmm.. i feel exhausted... i'll take a nap... be back in about an hour ...
17:06:52< Bertl> cya
17:14:34< SHu_here> hi Bertl
17:14:50< Bertl> morning Shuri!
17:15:51< SHu_here> i am boring of rain!
17:16:04< SHu_here> and cold weather
17:16:25< SHu_here> where is the summer!:P
17:16:42< Bertl> probably hiding behind the next month ...
17:17:33< Bertl> but I agree, mesa don't like cold weather either ...
17:17:34>> pflanze [~chris@gate.wyona.com] has quit [Ping timeout: 480 seconds]
17:19:37< Apollo> we got a foot of snow couple days ago... still diggin out :(
17:20:39< Bertl> hmm, looks like vs1.9.0 has a proc race issue :(
17:21:59< SHu_here> wht that mean
17:23:06< Bertl> that I can crash vs1.9.0 with hitting the proc heavily ...
17:23:30< Bertl> well, not the kernel, but the userspace process ...
17:23:46< SHu_here> k
17:25:48< Bertl> hmm, interesting, it is in __copy_from_user() so probably not even vserver related ...
17:33:36< SHu_here> Bertl
17:33:47< Bertl> yep?
17:33:52< SHu_here> do you know how it work
17:33:53< SHu_here> A directory with ulimits. Possible resources are cpu, data, fsize, locks, memlock, nofile, nproc, rss and/or stack.
17:33:54< SHu_here> resource
17:34:10< Brucey> I have one little compilation problem with 2.4.26 patched with Trond's NFS_ALL and the vs1.27 patch
17:34:23< SHu_here> what is the syntax
17:34:31< Brucey> one of the NFS_ALL patches adds:
17:34:47< Brucey>                 if (IS_IMMUTABLE(inode))
17:34:47< Brucey>                         return -EACCES;
17:34:58< Brucey> to fs/nfs/dir.c
17:35:02< Bertl> SHu_here: sorry, but you have to ask enrico about the details ...
17:35:19< Brucey> I'm wondering if simply replacing it by if (IS_IMMUTABLE_FILE(inode)) will do
17:35:34< Bertl> will probably do fine ...
17:36:00< Bertl> if it is _too_ restrictive, then replacing it with IS_IMMUTABLE_LINK() is the other option
17:36:30< Brucey> the comment just over it says:
17:36:31< Brucey>                * Nobody gets write access to an immutable file.
17:37:52< Brucey> so I guess it won't be too restrictive :)
17:38:18< Bertl> depends on what write access here means, but probably you are on the safe side with IS_IMMUTABLE_FILE()
17:38:57>> nalfein [~gaertner@212.68.83.129] has quit [Remote host closed the connection]
17:39:27< Brucey> I'll try that... if I see anything "weird", I'll use IS_IMMUTABLE_LINK()
17:39:28< Brucey> thanks
17:39:37< Bertl> np
17:42:02< Bertl> hmm, nap attack ... will be back later ...
17:42:09>> Bertl is now known as Bertl_zZ
17:54:43>> gst [~gst@eris.sysfrog.org] has quit [Quit: Client exiting]
17:56:07>> Napalm [~napalm@host81-7-20-232.adsl.v21.co.uk] has quit [Quit: ]
19:19:18>> Brucey [cinacio@snape.aspic.com] has quit [Quit: "War doesn't determine who's right, war determines who's left."]
19:20:42>> Bertl_zZ is now known as Bertl
19:20:59< Bertl> okay ... I'm awake again ...
19:29:02>> _id [~id@pD95E9332.dip.t-dialin.net] has joined #vserver
19:29:49< Bertl> hey _id!
19:30:22>> rs [rs@ice.aspic.com] has quit [Quit: leaving]
19:52:32< SHu_here> lol
19:52:53< SHu_here> you sleep only 2 hours!
19:53:11< Bertl> yep, was a short nap attack ... actually longer than expected ;)
20:02:37>> nalfein [~gaertner@212.68.83.129] has joined #vserver
20:02:51< Bertl> greetings nalfein?!
20:03:12< nalfein> bertl ? hi
20:03:31< Bertl> wasn't sure if this was a real join ;)
20:03:57< nalfein> it was ;) bnc ws down duing network failures ...
20:04:08< nalfein> <-tippgicht
20:06:10< Bertl> so what's up aside from your network issues?
20:06:52>> pflanze [~chris@gate.wyona.com] has joined #vserver
20:07:24< Bertl> greetings chris?
20:07:37< nalfein> nothing bertL ;)
20:07:41< nalfein> hi chris
20:07:54< Bertl> this _is_ probably an autojoin ;)
20:08:14< nalfein> yes ;)
20:09:00< Bertl> so interested in testing linux-vserver?
20:09:58< nalfein> yes i have all installed on a debian host .. in hannover
20:10:31< Bertl> 2.6 branch?
20:11:22< nalfein> latest stable with 2.4 kernel
20:11:45< Bertl> ah I see, and how is it working for you?
20:11:49>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has joined #vserver
20:11:57< Bertl> evening mhepp!
20:12:39< nalfein> last to do is get the quota to work, all the other works perfect. i will use it do isloate shell acounts in a vserver, and some other things
20:14:03< Bertl> quota is easy, if your setup is quota friendly (separate partition, ext2/ext3, recent tools)
20:14:30< nalfein> yes each vserver has a partition of his own
20:14:34>> nIx0n [nixus@reserved-99.netplus.ro] has joined #vserver
20:15:18< Bertl> nalfein: well, so you 'just' need to setup the vroot device then ...
20:15:21< Bertl> hi nIx0n!
20:15:28< nIx0n> hi
20:16:05< nalfein> i know , we talked about few days ago, and i read the doku on vserver side
20:16:16< nalfein> hi nixon
20:16:57< nIx0n> wsup
20:18:34< Bertl> nIx0n: what is your opinion about linux-vserver?
20:18:54< nIx0n> i didn't tryed yet ..
20:19:00< nIx0n> just read some docs
20:19:13< Bertl> and? no opinion?
20:19:22< nIx0n> nothing yet
20:19:22< nIx0n> ;)
20:19:55< Bertl> any questions then?
20:20:19< nIx0n> not yet
20:20:20< nIx0n> ;)
20:20:29< Bertl> okay, enjoy your visit then ...
20:20:33< nIx0n> yup
20:20:49< nalfein> bertl i reggd linux-vserver.info should i redirect it to your content ?
20:21:23< Bertl> hmm, interesting, if you like, I could setup nameservices for that?
20:22:04< nalfein> i have nameservers but i could update the zone to yours if you like
20:22:12>> nIx0n [nixus@reserved-99.netplus.ro] has quit [Quit: Fear The Inferator!]
20:24:52< nalfein> tell me to where the A record has to point and i update it , or tell me the names of your dns servers
20:25:09< Bertl> apollo:
20:25:14< Bertl> sorry ;)
20:27:37< Bertl> nalfein: the nameservers can be listed with 'dig linux-vserver.org NS'
20:27:55>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has quit [Remote host closed the connection]
20:28:02< mhepp> coooool!!! 2.6 stuff is not experimental?
20:28:54< nalfein> ok bertl then create the zone an i make an update for the info NIC
20:29:06< Bertl> primary nameservices for linux-vserver .org/.com are at 212.16.62.60  ...
20:29:28< Bertl> you can add one of you nameservers as secondary zone if you like?
20:31:28< nalfein> ok i create zones for linux-vserver.org / com / info on ns1.msi2000.com as secondary
20:32:07< nalfein> as primary for the .info ns1.13thfloor.at , ok ?
20:32:45< Bertl> and also .net ;) yep
20:37:10< nalfein> ok, done ..
20:37:32< nalfein> if you create the .info zone i will update the NIC
20:39:16< Bertl> okay, transfer should now be possible ...
20:39:31< Bertl> only .info has the sixth nameserver yet ...
20:39:55< Bertl> hmm, ns1.msi2000.com.	172800	IN	A	81.3.29.3
20:40:07< Bertl> but you are requesting from 81.3.29.2
20:40:32< Bertl> are your nameservices outdated?
20:40:39< Bertl> or did I miss something?
20:40:58< nalfein> hmm i should fix this ..
20:41:09< Bertl> mail.msi2000.com. is 81.3.29.2
20:43:34< nalfein> ok i think i fix it
20:43:43< nalfein> 81.3.29.3 it right
20:44:09< nalfein> the .info zone is created ?
20:44:34< Bertl> yep, your? 81.3.29.2 is already requesting for some time now ;)
20:45:22< nalfein> grmpf ..
20:46:43>> eyck [~eyck@62.233.189.138] has quit [Ping timeout: 480 seconds]
20:47:00< nalfein> };
20:47:00< nalfein> query-source address 81.3.29.3;
20:47:00< nalfein> };
20:47:56< Bertl> May 13 20:47:47 ns1 named[30742]: client 81.3.29.2#47346: zone transfer 'linux-vserver.net/IN' denied
20:48:36< Bertl> a vserver bug?
20:53:50< nalfein> no ;) allow transfer from .. fixed
20:54:50< Bertl> hmm, so I should add mail.msi2000.com. to the transfer list, right?
20:55:13< nalfein> ok
20:55:26< nalfein> zone linux-vserver.net/IN: refresh: non-authoritative answer from master 207.253.4.241#53
20:55:26< nalfein> zone linux-vserver.info/IN: refresh: non-authoritative answer from master 207.253.4.241#5
20:55:26< nalfein> zone linux-vserver.net/IN: refresh: non-authoritative answer from master 207.253.4.2#53
20:55:26< nalfein> zone linux-vserver.info/IN: refresh: non-authoritative answer from master 207.253.4.2#53
20:55:35< nalfein> should be better
20:58:08< Bertl> seems your nameserver has some kind of split personality ;)
20:59:42< nalfein> lol .. could be
21:00:20< Bertl> you should add the folowing:
21:00:53< Bertl> listen-on { 81.3.29.3; };
21:01:23< nalfein> ok thx bertl
21:01:50< Bertl> and transfer-source { 81.3.29.3; };
21:03:23< Bertl> this should fix those issues ...
21:05:11>> eyck [~eyck@62.233.189.138] has joined #vserver
21:10:43< Doener`> re
21:10:57< Doener`> that nap was longer than what i planned...
21:11:07< eyck> hehehe
21:11:15< eyck> you're all so lucky..
21:11:19< eyck> napping all the time
21:12:35< Bertl> Doener`: had a nap too, almost said the same ;)
21:13:09< Doener`> hehe :)
21:13:16< Doener`> so the patch is ready?
21:13:37< Bertl> yeah, but it doesn't make much sense testing it atm ... because I can crash it here :(
21:13:54< Doener`> ah, i see... 
21:14:05< Bertl> (and unfortunately I have no idea where or why ...)
21:17:30< Bertl> anybody here with some kernel debugging expertise? 8-)
21:18:28< Doener`> hmm.. we used to rely on some austrian guy ;)
21:19:00< Bertl> damn! I knew it! *G*
21:19:34< nalfein> bertl : domain update @ NIC ist ok
21:22:16< Bertl> okay, thanks, Thomas?
21:23:06< nalfein> yes thomas in RL
21:28:06>> Shotygun [shotgun@shotygun.com] has joined #vserver
21:28:40< Bertl> hello Shotygun! this is your first and only greeting for today, as you selected ;)
21:29:52< Shotygun> =D How you doing Bert ?
21:30:16< Bertl> fine thanks (aside from a bug I do not see) ... and you?
21:30:48< Shotygun> Fine fine, fine fine..
21:32:10>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has joined #vserver
21:32:20< Bertl> evening infowolfe!
21:32:26< infowolfe> lol
21:32:31< infowolfe> good afternoon Bertl
21:32:40< Bertl> ah, that's the spirit!
21:32:52< Bertl> good afternoon to you too ... ;)
21:33:00< Shotygun> Hey infowolfe
21:33:12 * Shotygun is hungry as well, hmm, perhaps a pizza will do the trick..
21:33:31< Bertl> hint: try to heat it up first 8-)
21:33:55< infowolfe> Hey Shotygun, how ya doin?
21:34:16< Shotygun> =P Do any of you know any free billing systems ? need to implement to someone and maybe even myself..
21:34:22< Shotygun> Fine thank you infowolfe =)
21:34:29< infowolfe> just wonderful
21:37:16< SHu_here> wich?
21:37:45< SHu_here> what is wonderful:P
21:43:29< Doener`> hmm... that's a nice error message 'The following errors may help:' :)
21:44:07< Bertl> well, errors could prevent further usage, so they might actually help
21:44:49< Doener`> well, that was the complete message
22:02:57>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has quit [Remote host closed the connection]
23:17:18>> pflanze [~chris@gate.wyona.com] has quit [Quit: nigh]
23:21:46>> Apollo [~Apollo@caracal.norcomcable.ca] has left #vserver []
23:24:05>> hiaslboy [matthias@e-16.vc-graz.ac.at] has joined #vserver
23:24:22< Bertl> evening matthias!
23:27:50< Bertl> Doener`: still around/awake/interested in testing?
23:28:28< Doener`> around: yes, awake: yes, interested: in about 15 minutes, have to complete some mysql stuff...
23:28:44< Bertl> okay, let me know ...
23:29:15< Doener`> hmm... i could compile the kernel in the meantime 
23:30:15< Bertl> okay, I'll upload the patch ...
23:41:29< Bertl> http://vserver.13thfloor.at/Experimental/patch-vs1.9.0-rcu0.3.diff
23:41:30< Doener`> ok, i'm done with my stuff
23:41:38< Bertl> okay, this is ontop of vs1.9.0
23:42:05< Doener`> 2.6.6?
23:42:07< Shotygun> What you folks testing?
23:42:17< Bertl> Doener`: yep 2.6.6
23:42:40< Bertl> Shotygun: I'm currently 'improving' vs1.9.0 scalability
23:42:58< Bertl> and Doener is volunteering to test the 'new' approach ...
23:43:41< Bertl> Doener`: make sure to set spin lock debugging, debug info and sleep inside spinlock, as well as magic sys-req ...
23:44:24< Shotygun> Hmm I need a dns backup and maybe mx backup, anyone willing to help or know any cheap services?
23:45:01< Bertl> would you trust your dns data to a script?
23:45:08< Shotygun> lol =)
23:45:45< Shotygun> I need to find some overseas backup
23:45:57< Shotygun> In case my new-to-be ISP won't be reliable enough..
23:46:49< Bertl> oversea in what direction?
23:46:53< Shotygun> Any.
23:47:00< Shotygun> I'm in israel.
23:47:38< Shotygun> I'm going to buy a cabinet soon over some ISP and move out my servers that spreaded around into this one, so I will need to get a backup outside of this ISP.
23:47:42< Bertl> .oO( hmm, what sea is between israel and austria? )
23:47:45< Doener`> Bertl: what about the VServer kernel options? anything required/suggested? didn't use 1.9.0 before...
23:48:09< Bertl> Doener`: I'd suggest disabling the vproc security, and take the defaults for the rest
23:48:13< Shotygun> Mediterranean sea
23:49:04< Doener`> hmm... legacy stuff is still required?
23:49:10< hiaslboy> hi Bertl :-)
23:49:13< Bertl> Shotygun: so I could backup your dns services .. but I do not do mail buffering ...
23:49:27< Shotygun> dns would be great, thanks bert, I will keep you informed.
23:49:29< Bertl> Doener`: yes
23:56:30< Bertl> Doener`: ah, and almost forgot, PREEMPTION please! ;)
23:57:00< Doener`> you told me some hours ago ;) and i always have that one enabled (desktop box)
23:57:18< Bertl> good ;)
--- Log closed pią maj 14 00:00:09 2004