--- Log opened czw maj 20 00:00:46 2004
00:02:45< taxcollector> I am male, if that is what you mean. :)
00:03:06< _id> nope - taxcollector you are a bot ;)
00:19:29< eyck> I thought I'm the only bot
00:22:17 * albeiro is bot also
00:22:39< albeiro> not to mention Bertl_oO :P
00:26:04>> taxcollector [~taxcollec@192.16.167.161] has quit [Remote host closed the connection]
00:43:21< Shotygun> What's the time in GMT/UTC at the moment?
00:45:53< eyck> 00:45
00:46:08< eyck> : Thursday May 20 2004 -- 00:45 +02:00
00:46:09< Shotygun> Thanks
00:47:32< Shotygun> Are you sure that it's 00:47 and not 23:47 ?
00:47:36>> Doener` [~doener@pD9588666.dip.t-dialin.net] has joined #vserver
00:49:22< Shotygun> 01:48:56 Thu May 20 2004 in Israel converts to
00:49:22< Shotygun> 22:48:56 Wed May 19 2004 in GMT 
00:50:23< eyck> positive.
00:50:42< eyck> UTC would be 22:45
00:51:33< Shotygun> Hmm, UTC is GMT I think..
00:51:34>> Doener is now known as Guest2
00:52:06>> Doener` is now known as Doener
00:54:49>> Guest2 [~doener@pD9E12688.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
01:02:34>> taxcollector [~taxcollec@192.16.167.161] has joined #vserver
01:03:15< taxcollector> Has anyone tried context file tagging? chxid and lsxid aren't behaving as I would expect them to.
01:06:36>> reverend [reverend@office.ozhosts.net] has joined #vserver
01:06:56< reverend> hi
01:07:22>> reverend is now known as Guest3
01:07:24< taxcollector> Howdy
01:08:37< Guest3> anyone been able to get vserver 1.9.0 working ?
01:08:40>> Guest3 is now known as reverend
01:09:03>> reverend is now known as rev
01:09:36< taxcollector>  What is not working for you?
01:14:57< rev> from memory, vservers wouldn't start 
01:15:21< rev> and the tools for creating the vservers didn't work either
01:15:23< taxcollector> Did you unhide the proc entries via /usr/lib/util-vserver/vprocunhide ?
01:15:47< rev> no i didn't
01:15:47< taxcollector> How did you create the vservers if the build tools didn't work. ;)
01:15:55< rev> using older versions of the tools
01:16:13< taxcollector> I don't have experience with the older set of tools so I can't comment on that.
01:16:59< rev> ok, so you have the 2.6.6 kernel version of the tools working fine ?
01:17:06< taxcollector> I have fedora core 2, using 1.9.0 and the 0.29.214 util-vserver tools
01:17:29< rev> excellent - that
01:17:36< taxcollector> I am having problems with context file tagging, but basic stuff is working
01:17:36< rev> excellent - that's the setup i'm trying to get working - fedora core 2
01:17:47< rev> any luck with quotas ?
01:17:56< rev> per context disk limits and per context quotas i should say
01:18:03< taxcollector> Haven't tried.
01:27:31< rev> hey taxcollector, do you think vps hosting with vserver is a viable option at this stage ?
01:28:20< taxcollector> Yes; there are several companies doing so, but I am not one of them.
01:31:00>> vserver-log [shotgun@212.199.206.70.forward.012.net.il] has joined #vserver
01:31:09< taxcollector> Not with the development version of course. ;)
01:41:08< rev> ahhh but the development version has all the nice features :D
01:48:32>> Bertl_oO is now known as Bertl
01:48:48< Bertl> evening folks!
01:49:05< taxcollector> Howdy!
01:49:16< Bertl> hey Ryan, how are you?
01:49:24< Shotygun> Bert, vserver-log.
01:49:58< Bertl> ah, excellent, thank you very much!
01:50:17< taxcollector> Pretty good; I went AWOL last week as I had family visiting...
01:50:23< Shotygun> Sure, np, I will pack you up the bot in a few days after we will see it's stable enough..
01:50:44< Shotygun> (For you will have it in case of whatever..)
01:50:47< taxcollector> I managed to get that 2.6.6 test server up -- I have 1.9.0.10 installed if there were particular things you wanted me to try
01:51:21< Bertl> well, you reported some issues with tagxid, if I read the history correctly?
01:51:47< Shotygun> Hey, would you like me to add 'seen' script to the bot?
01:52:25< taxcollector> Yes; lsxid gives me "!!ERR!!"
01:53:22< taxcollector> And chxid complains about an Invalid argument when I call it according to the usage spec...
01:53:29< taxcollector> but I haven't taken a look at the source code yet.
01:53:36< Bertl> Shotygun: might be interesting to get some kind of infobot combined with the logging, but I'd say that's a long term project ...
01:54:14< Bertl> taxcollector: first, you did mount the partition for your tests with the tagxid option, right?
01:54:29< taxcollector> doh! No. :)
01:54:54< Shotygun> Bertl: If you will aim me to what you want to be then I don't mind taking it on myself.
01:56:18< Bertl> Shotygun: could be interesting to 'flag' some kind of explanations something like /m bot memorize "Context Quota"
01:56:27< Bertl> bla bla bla ...
01:57:01< Bertl> and then if you do /m marvin explain context quota
01:57:17< Bertl> marvin would say: I'm so depressed 8-)
01:57:58< Shotygun> lol =)
01:58:10< Bertl> just an idea ... so don't take it too serious ...
01:58:11< Shotygun> Or better, we could import your evil on-join script to it
01:58:12< Shotygun> =D
01:58:22< Bertl> for example ... ;)
01:58:57>> marlow [~marlow@212.17.50.41] has joined #vserver
01:59:05< Bertl> evening marlow!
01:59:12< marlow> Bertl : hi ..
01:59:17< marlow> Bertl : got my last mail ?
01:59:18< Shotygun> Bertl: We can do it wiki style, that certain people (with access) will be able to add definitions to the bot, and will also be able to make the bot say it in the channel.
01:59:34< Bertl> marlow: am I mistaken, or is it quite a while since we met here?
01:59:52>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has joined #vserver
02:00:00< Napalm> hello everyone
02:00:05< marlow> Bertl : i'm too busy sometimes :)
02:00:15< Bertl> Shotygun: I really don't want to cause you too much trouble, so yes, please work on it 8-)
02:00:42< Bertl> evening Napalm, how is your vserver today?
02:00:43< Napalm> ensc: are you around?
02:00:44< marlow> Bertl : but thought, i might show up here for a change
02:00:52< Bertl> good idea!
02:00:55< ensc> Napalm: yep
02:00:57< Napalm> bert: its still sick
02:00:58< Shotygun> Bertl: Starting it right away
02:01:22< Napalm> ensc i got some problems with the util-scrips i think i know the problem, but im not sure
02:01:33< ensc> which problems?
02:01:39< Napalm> lots of error messages lol
02:02:13< rev> Bertl: are you aware of the issue with subscribing to the vserver mailling list ?
02:02:18< rev> python errors
02:02:26< Napalm> ensc: i cant copy them across to irc because they disappear off the screen?
02:02:30< marlow> rev: i'm hosting the list ..
02:02:34< marlow> rev: what errors ?
02:02:44< rev> marlow: when attempting to subscribe
02:02:44< ensc> Napalm: probably the first 5 lines are enough ;)
02:02:56< Bertl> marlow: that was the dude!
02:03:03< marlow> rev: drop me a mail with it and i've a look into it ..
02:03:07 * Bertl points at rev ;)
02:03:27< rev> marlow: email ?
02:03:33< Bertl> rev: no offense meant ...
02:03:36< marlow> rev: got that on private ..
02:03:44< rev> bertl: none taken :D
02:03:45< rev> Bug in Mailman version 2.1.4
02:03:45< rev> We're sorry, we hit a bug!
02:04:01< marlow> rev: yes ... 
02:04:07< marlow> rev: but have you tried today ?
02:05:01< marlow> Bertl / rev: that was the bug that caused the havock ..
02:05:30< marlow> Bertl / rev: it's actually a complete different list, that seems broken, but it takes the whole thing down .
02:05:35< marlow> Bertl / rev: not good
02:05:53< rev> marlow: i've been trying for a week ;\
02:06:12< marlow> rev: shouldn't have been a problem that long ..
02:06:22< marlow> rev: but i'll dig into it and fix it ..
02:06:31< rev> np
02:06:46 * marlow got an sms from rev :)
02:06:47< Bertl> I can confirm that he reported it some time ago, for sure 4 days, maybe even a week ...
02:06:51< rev> :D
02:07:11< rev> damn - my ipod's disk has died :(
02:07:47< marlow> Bertl : yep .. he should just ahve mailed me immidiatly .. 
02:07:51< marlow> Bertl : :o)
02:08:03< marlow> s/ahve/have/
02:08:15< Bertl> hmm, maybe we should add your mail address on the wiki? are you prepared for that?
02:08:16>> marlow is now known as Marlow
02:08:18< taxcollector> Bertl: took a look at sections E31 and E32 regarding context file tagging...
02:08:37< taxcollector> lsctx == lsxid?
02:09:01< Marlow> Bertl : hmm ... there is a special email-address for that ..
02:09:05< Marlow> Bertl : two secs .
02:09:12< taxcollector> and I don't get a a permission denied error for E32
02:09:16< Bertl> taxcollector: different name, similar tools, enrico can explain
02:09:45< taxcollector> I did mount with the tagxid option...
02:09:45< taxcollector> [root@fft7 root]# mount
02:09:54< ensc> Bertl: you wanted the s!ctx!xid! ;)
02:10:23< taxcollector> "/dev/sda2 on / type ext3 (rw, tagxid)"
02:10:46< Bertl> ensc: didn't say it was bad, did I?
02:10:46< rev> hrm - anyone played with virtuozzo recently ?
02:11:43< Bertl> taxcollector: hmm, which kernel/patch?
02:11:55< taxcollector> 2.6.6-vs1.9.0
02:11:57< taxcollector> ext3
02:12:02< taxcollector> 0.29.214
02:14:28< Bertl> hmm, 2.6.6-vs1.9.0.11, ext2, I do ...
02:14:33< Bertl> # chcontext --ctx 200 touch /mnt/100/file
02:14:33< Bertl> New security context is 200
02:14:33< Bertl> touch: /mnt/100/file: Permission denied
02:14:53< Bertl> let's check with ext3 now ... (btw no difference there to vs1.9.0)
02:16:01< Bertl> same permission denied with ext3 here ...
02:16:08< taxcollector> OK... my output from mount indicating tagxid means that I've passed the option correctly right?
02:16:44< Bertl> I'd say so, but wait, it's '/' and you can't mount that tagxid without advanced magic ...
02:17:04< Bertl> so where does the tagxid come from?
02:17:18< Bertl> did you do a remount on that?
02:17:26< taxcollector> I naively put it in fstab...
02:17:36< taxcollector> "/dev/sda2 / ext3 defaults,tagxid 1 1"
02:17:37< Bertl> ah, okay, so this is a fake one then ...
02:18:03< Bertl> it's neither easy nor a good idea to do this on the root fs ...
02:19:09< Bertl> (can't take credit for the former one)
02:19:33< taxcollector> Don't bother explaining the details, but could you indicate some keywords I could google to get a flavor for why it isn't easy or a good idea?
02:20:07>> kestrel [athomas@38.6.233.220.exetel.com.au] has quit [Quit: brb]
02:20:17< Bertl> hmm, well, you need a special kernel boot command option to mount the rootfs with anything not remountable
02:21:03< taxcollector> OK
02:21:22< Bertl> and it isn't a good idea, because if you managed to make '/' tagxid, you will be able to start a single vserver, after that one, the tools will have changed the ownership of /var in such way, that another server would be denied ...
02:21:49< Bertl> actually /var/vserver*
02:22:19< Bertl> so to make it short, tagxid is nothing for the rootfs at all ...
02:24:33< taxcollector> But if a filesystem isn't mounted tagxid then context file tagging shouldn't affect anything right? I'm experiencing a problem that I think is related...
02:25:14< taxcollector> no tag xid option, from host touch "/vservers/foo/tmp"...
02:25:29< taxcollector> er, touch "/vservers/foo/tmp/file"
02:25:49< taxcollector> from within vserver foo "ls /tmp" doesn't show any file
02:27:30< Shotygun> Bertl: Would you like to define the defintions in public or privately to the bot?
02:29:47< taxcollector> Shotygun: based on the previous IRC traffic, you are starting a new mechanism for logging the channel? So today's discussion is being logged and will be posted somewhere?
02:31:10< Shotygun> taxcollector: For now it only logs, but it will be posted on the web and will be searchable I believe.. ( sladen should take care of it [or me?] ); Right now it only logs and now I'm adding to it a wiki-style thingy.
02:32:18< taxcollector> OK -- thanks. Just wondering if I needed to jot down the details of what Bertl was telling me so I could write more formal docs.
02:32:33< Shotygun> You should copy them just in case..
02:32:50< Shotygun> I just brought it up and it seems stable, but can't tell within a hour of testing..
02:32:56< Shotygun> (10mins of testing, one hour running =P)
02:33:08>> fdlimit [bsdboy@smacked.org] has joined #vserver
02:33:18< Shotygun> Although it seems like it logs everything just fine..
02:33:19< Shotygun> hey fdlimit
02:33:36< fdlimit> hey sg
02:36:37< Shotygun> Bert found some timing to go away, bah.
02:36:40< Shotygun> =P
02:38:24>> kestrel [athomas@38.6.233.220.exetel.com.au] has joined #vserver
02:42:15< Bertl> hmm, seems I lost connectivity for some time ... catching up now ...
02:42:26< Shotygun> heh that's alright. Read up.
02:42:38< Shotygun> Finished the login system now making the definitions
02:43:49< Bertl> taxcollector: that is probably a side effect from the namespace stuff ..
02:44:15< Napalm> Bert, Ensc: i found my problem
02:44:16< Bertl> Shotygun: definitions in public of course ...
02:44:50< Bertl> evening fdlimit! may I add, good choice of nick! 
02:45:03< Shotygun> lol
02:45:11< Shotygun> fdlimit watch out, he got an evil auto greeting script.
02:45:38< Bertl> Napalm: what was it? and where did you find it?
02:46:01 * Bertl .oO( and do we have to solve it now? )
02:46:39< Napalm> Bert, Ensc: i slowly have gone thru newvserver script, the errors occur when it tries executing 'vbuild' as i installed RPM im guessing that i should recompile it, i would think that the vserver mount directory is hardcoded into it, as im using a custom directory
02:46:53< taxcollector> Bertl: Disabled namespace and problem went away. Thanks.
02:47:08< Bertl> taxcollector: one additional explanation regarding mount options, if you use mount, it just shows you the entries in /etc/mtab
02:47:45< Bertl> so if you add 'funny stuff' there, then mount will very likely show you 'funny stuff' ;)
02:48:04< Bertl> adding it to /etc/fstab 'normally' doesn't cause this, but the rootfs is an exception
02:49:08< Bertl> as it isn't mounted by the runlevel/init scripts, because they are on the rootfs and therefore it has to be mounted beforehand, so the entry is faked by some script magic, which makes you believe that it actually was mounted with those options 
02:49:09< taxcollector> OK; I did add it to fstab that time. :)
02:49:34< taxcollector> I appreciate you explaining the nuances.
02:49:53< Bertl> you can make sure by using /proc/self/mounts or /proc/mounts (which is a link to the former) ...
02:50:54< Bertl> Marlow: I'd say, send your proposal regarding whitelist to the ml, and we let the community decide on that, what do you think?
02:51:19< Marlow> yep .. probably the best ...
02:52:19< Bertl> taxcollector: you're welcome
02:52:21< Marlow> it's anyway a feature that individually can be enabled/disabled as soon as i'm finished with it.
02:52:33< Marlow> i mean pr. mailinglist
02:53:47< Bertl> taxcollector: IIRC you had quite some machine reserved/planned for testing, right?
02:53:56< taxcollector> I do
02:54:09< Bertl> actually there is something you could test for us ...
02:54:13< taxcollector> OK
02:54:32< Bertl> again a short introduction ... ;)
02:55:08< Bertl> as you might know, linux-vserver tries hard to keep the overhead low (very low as a matter of fact)
02:55:36< Bertl> this means overhead for the host system as well as overhead for the guests
02:56:59< Bertl> so I scheduled some tests on STP to put some numbers on that overhead ...
02:57:07< Napalm> Bert: i cant see to use Unified mode it seems to die with the errors, am looking into it futher
02:58:01< Bertl> Napalm: hmm, maybe you should rewrite that script from scratch ...
02:59:04< Napalm> Bert: now i understand how its functioning, it sounds like a better idea to just that, a couple of shell scripts and a php front end 
02:59:18< Bertl> and the results of the STP tests where very pleasing, as they showed (within the error caused by random noise) that linux-vserver kernels are better, regarding performance, than vanilla kernels ;)
02:59:27< Bertl> -h
02:59:36< taxcollector> Must make you proud. :)
03:00:24< Bertl> well, yes, it actually did, but, what we do not have numbers for, is the overhead for the guest, compared to the host ... or an unpatched kernel ...
03:00:36< Napalm> not sure exactly what that means, but to say linux-vserver kernel is better than the vanilla ones must make anyone pround
03:01:07< Shotygun> Bertl: To spare some bits of ram and cpu I am saving all terms in lower-case only, you don't mind do you?
03:01:15< Shotygun> (TCL arrays are slow)
03:01:20< Bertl> as I said, take it with a grain of salt, as the difference was within the expected noise ...
03:01:30< rev> anyone able to get to vserver.wireless-winds.de/ ?
03:01:39< Shotygun> (At least with 20k records that I verified)
03:01:40< Napalm> nope, its been down for sometime now
03:01:52< rev> anyone have a copy of the php scripts that were available there ?
03:01:52< Napalm> ive been trying to take a look at there php interface for sometime
03:02:11< Napalm> ive created my own if you want a copy, its a module for webmin
03:02:13< Bertl> Shotygun: hmm, not sure that this is such a good idea ...
03:02:20< Shotygun> Alright I will make it caps sensetive
03:02:22< rev> yeah, napalm, that sounds like a good idea 
03:02:31< rev> send by email ok ?
03:02:38< Shotygun> (The limition goes to the terms only, not to the definitions)
03:02:52< Shotygun> But will fix it, it's no biggy
03:02:55< Shotygun> Just being lazy overnight=)
03:03:04< Napalm> rev: yep
03:03:22< Bertl> hmm ... hmm ... maybe terms stored as mixed case, but recognized case insensitive?
03:04:23< Shotygun> Nah, it's about storing them, I will just add the mixed case term to the beginning of the definition and chop the first word from the string out of the definition when writing.
03:04:42< Shotygun> I doubt there would be even 1k definitions but still being a performance-whore =P
03:04:59< Shotygun> I had bad experience with some 20k DB =)
03:05:07< Shotygun> (Public seen script)
03:05:27< Bertl> taxcollector: so I'd appreciate some typical unix userspace test, maybe unixbench or kernlbench or something like this ... inside and outside a vserver/network context ... (maybe 3-5 times to get some statistics ...)
03:06:00< Bertl> do you think this would be doable?
03:06:21< Napalm> Rev: whats the email address?
03:07:07< Bertl> Marlow: wow 7 machines for distcc?
03:07:28< Bertl> sorry wrong nick, should be:
03:07:29< Shotygun> I want 7 machiens for distcc too!
03:07:34< Bertl> mcp: wow 7 machines for distcc?
03:07:35 * Shotygun is poor
03:10:24< taxcollector> Bertl: It is doable
03:10:56< Bertl> great!
03:14:16< rev> bertl: i have hardware you can test on
03:14:44< rev> i actually work for a webhosting company and i'm trying to evaluate the different vps systems around - i've got 3 boxes we can use
03:15:39< Bertl> sounds good, are you willing to do some testing on them?
03:16:19< Bertl> btw, what are the other vps systems around? 8-)
03:16:48< rev> there's virtuozzo
03:16:51< rev> freevds
03:16:55< rev> uml
03:16:57< rev> xen
03:17:14< rev> virtuozzo looks good, but it's $
03:17:32< taxcollector> Later everyone. Bertl: I'll have some unixbench results for you tomorrow.
03:17:35>> taxcollector [~taxcollec@192.16.167.161] has quit [Remote host closed the connection]
03:18:04< Bertl> rev: yeah, probably they use similar codebase ... although you won't get a source for virtuozzo ;)
03:18:34< Bertl> xen and uml are quite different IMHO ...
03:18:46< rev> bertl: that's true - i'm trying to hold off on us buying it in case a viable open source solution presents itself
03:19:06< rev> but from my understanding, we can't really set limits for host vps' with vserver?
03:19:16< rev> beside the per context disk limit
03:19:36< Bertl> really depends on the patch version/branch ...
03:19:41< kestrel> uml is way too slow for hosting, imho
03:20:01< Bertl> and it uses plenty of resources ... but it has some advantages too ...
03:20:25< Bertl> for example the separate kernel ...
03:20:31< kestrel> yeah
03:21:08< Bertl> rev: http://www.linux-vserver.org/index.php?page=Release+FAQ
03:21:32< kestrel> i wouldn't pay for a uml virtual server though
03:22:03< rev> bertl: what patch version/branch + tools version/branch would you suggest ?
03:22:51< rev> i couldn't get 1.9.0 working - but that was due more to the tools than anything else
03:22:56< Bertl> if you can live without the new features, then stable branch + util-vserver 0.29.5 or whatever is candidate for the next stable ...
03:23:38< Bertl> just because it is better tested and we (think) we know what it does ;)
03:24:10< Bertl> if you need hard scheduling (cpu limits) or memory limits, you have to go for development ...
03:24:16< rev> well, the new features are almost a requirement
03:24:19< rev> so development it is
03:24:45< Bertl> 2.6 (aka 1.9.x branch) has the advantage that it is faster and scales better than 2.4 ...
03:25:07< Bertl> but it is verry young compared to 1.3 for example ...
03:26:00< Bertl> I'm trying hard to make the patches simple and easy to understand, and splitting them up into small chunks to allow for source code verification ...
03:26:38< Bertl> that is something I was missing in the 'original' patches as well as in the freevps patches ...
03:29:26< rev> what i've found most confusing, as a newcomer to this project, is that there's no 'one stop' place for all the relevant information and patches
03:30:14< Bertl> well, it's a community project after all ...
03:30:34< Bertl> but feel free to 'organize' the stuff on linux-vserver.org ;)
03:31:09< rev> well, i'm planning on writing up my experiences and providing a central repository of patches, tools, etc
03:33:08< Bertl> any patches or tools you didn't find on 13thfloor.at?
03:34:19 * Bertl .oO( well probably enricos latest tools are still missing ;)
03:35:06< Shotygun> Ok
03:35:13< Shotygun> Now vserver-log is going to die alot so I better part it
03:35:30< Shotygun> Blah I set this channel as static, need to shoot it.
03:35:31>> vserver-log [shotgun@212.199.206.70.forward.012.net.il] has quit [Quit: Shotygun]
03:36:23< Shotygun> amazing, no typos, now that's rare for me..
03:36:48< Bertl> em oto!
03:36:58< Shotygun> =P
03:37:28< Shotygun> Oh, no wonder it didn't crash because of typos, forgot to load the script.
03:37:44< Shotygun> There, a crash =D
03:38:43< Bertl> rev: anyway, would be really interesting to get a comparison between freevps, virtuozzo, and linux-vserver (especially regarding performance and stability)
03:39:00< Shotygun> I would like to see it too
03:39:10< rev> i can't get freevps running - it requires jumping through so many hoops
03:39:11< Shotygun> Then maybe we will have a reason to complain to Bert =P
03:39:28< rev> we want to run fedora - not an option using freevps
03:39:30< Bertl> yeah, complaining is always good ... ;)
03:40:08< rev> virtuozzo - we've used it before, it has some great features, including a remote management console for win2k/xp, but it's pretty prohibitively priced
03:40:40 * Bertl .oO( hmm who would need a management console for win2k/xp? )
03:41:23< Marlow> not really .. and if needed it could be done in web .
03:41:33< Marlow> that would solve it universally :)
03:42:33< Bertl> rev: anyway, if the pricing should become important, we can charge you only a fraction of virtuozzo ;)
03:46:23< rev> :D
03:53:41< Bertl> ah, and I almost forgot, if you have any good suggestions, what is missing in linux-vserver, and maybe some ideas how to add it ... just let me know ...
03:54:44< rev> i think one thing that needs to be worked on, and i want to help in this regard, is documentation
03:55:35< Bertl> I totally agree with that, and it will take some time to improve on that ... 
03:56:14< Bertl> consider this (and similar) as a step in the right? direction ...
03:56:15< Bertl> http://vserver.13thfloor.at/Stuff/PAPER-05.4.txt
04:16:48< Napalm> byeee everyone
04:16:59< Napalm> i got to get me some sleep, its 3:18am
04:17:00< Napalm> lol
04:17:06< Napalm> i got to be up at 9
04:17:14>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has quit [Quit: cya all]
04:19:48< Bertl> okay, I guess I'll call it a day too ... so cya folks, and have a nice one ...
04:20:12>> Bertl is now known as Bertl_zZ
04:34:36< Marlow> i'm outta here, too ...
04:34:38>> Marlow [~marlow@212.17.50.41] has quit [Quit: bye]
04:51:46>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has quit [Ping timeout: 480 seconds]
05:09:51>> gilbert [gilbert@208-186-222-203.nrp4.brv.mn.frontiernet.net] has quit [Quit: ]
06:54:22>> rev [reverend@office.ozhosts.net] has quit [Read error: Connection reset by peer]
08:21:40>> serving [~serving@213.186.191.31] has quit [Ping timeout: 480 seconds]
08:22:29>> noel_ [~noel@pD9FFFC93.dip.t-dialin.net] has joined #vserver
08:29:44>> noel- [~noel@pD9FFEFD9.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
08:39:09>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has joined #vserver
09:34:29>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has joined #vserver
09:35:05< mhepp> hello
09:38:30< mhepp> i have trouble with new util-vserver. I have 0.29.213, and i found new configuration of virtual servers
09:38:40< mhepp> how can i conver it?
09:43:30< eyck> hmm, it should be compatible..
09:43:42< eyck> what exactly do you have problems with?
09:48:18< mhepp> new program vserver could not find configuration of any virtual server...
09:48:46< mhepp> 5[root]# vserver aptcacher enter
09:48:46< mhepp> Can not find vserver-setup; please make sure that the vserver configuration
09:48:46< mhepp> is located at /etc/vservers/aptcacher/.
09:48:46< mhepp> 5[root]#
09:49:36< mhepp> and in /etc/vservers/
09:49:55< mhepp> are aptcacher.conf
09:50:11< mhepp> and new_for_test directory
09:50:33< mhepp> with configuration build by new vserver-build 
10:00:23>> fdlimits [bsdboy@smacked.org] has joined #vserver
10:00:23>> fdlimit [bsdboy@smacked.org] has quit [Read error: Connection reset by peer]
10:00:41>> v00dY| is now known as v00dY
10:11:28>> JonB [~NoSuchUse@129.142.112.33.ip.tele2adsl.dk] has joined #vserver
10:13:42>> serving [~serving@213.186.191.31] has joined #vserver
10:48:57>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has quit [Read error: Connection reset by peer]
10:54:54>> infowolfe [~infowolfe@pcp04891550pcs.frnkmd01.md.comcast.net] has joined #vserver
11:28:34>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has quit [Remote host closed the connection]
12:52:41>> Sebastian [~cereal@pD9EAB7AC.dip.t-dialin.net] has joined #vserver
13:03:11>> Bertl_zZ is now known as Bertl
13:03:18< _id> moin Bertl
13:03:36< Bertl> morning everyone! moin _id!
13:04:31< Sebastian> moin Bertl
13:16:32>> fdlimits [bsdboy@smacked.org] has quit [Ping timeout: 480 seconds]
13:20:27< hiaslboy> good morning bertl
13:21:32< Bertl> morning matthias!
13:28:24< Doener> morning
13:28:42< Sebastian> good morning Doener
13:29:31< Bertl> a Doener without ` is like a ... ;)
13:29:51< Doener> hehe
13:29:57< Bertl> how are you?
13:30:21< Doener> quite alright :)
13:31:05< Doener> you?
13:31:19< Bertl> I'm fine, thanks ... just preparing 1.9.1 ...
14:50:45< JonB> Bertl: sorry for not fixing the testing machine
14:50:57< JonB> Bertl: what was it with it, can you reach it over the serial line ?
14:51:10< Bertl> hmm, thought you fixed it ;)
14:51:24< Bertl> well, it seems network connectivity restored itself then ;)
14:51:48< JonB> good
14:51:56< Bertl> but it seems it hung at a later point ;)
14:52:07< JonB> it is sitting on our dorms internet line
14:52:18< Bertl> didn't investigate it further ...
14:52:41< JonB> and one guy writes his master thesis about bandwith control
14:53:06< Bertl> that's probably the guilty one ... ;)
14:53:08< JonB> he tries to give us regular users liveable bandwith
14:53:14< JonB> and kill the p2p things
14:53:28< JonB> by automaticaly detecting them no matter what port the use
14:53:30< JonB> maybe
14:54:11< Bertl> how are you? everything fine?
14:54:37< JonB> i just turned in a project yesterday
14:54:42< JonB> now a new one awaits
14:54:48< JonB> a far more interesting one
14:55:11< JonB> the one where i write the specification from a BSD driver, and a friend writes the linux driver
14:55:36< JonB> today i think i'll just relax though
14:56:14< JonB> and fix the network if it still has trouble
14:57:47< Bertl> hmm, actually it seems like the dualmmx has some other troubles ...
14:57:54< JonB> yes ?
14:57:58< Bertl> (probably software related) 
14:58:11< JonB> kernel related ?
14:58:19< Bertl> but it would be fine if you could check them locally ...
14:58:34< JonB> i will in an hour or so
14:58:57< Bertl> no need to hurry, I probably won't use it before midnight CET
14:59:22< Bertl> would be nice to have the network script running though ...
14:59:32< JonB> that test thiongy ?
14:59:37< Bertl> yes (at midnight ;)
15:00:52< JonB> i just started it again
15:01:04< JonB> are you using the serial console to the dualmmx ?
15:01:38< Bertl> yep, currently I'm trying to reboot it to a known state ...
15:04:16< Bertl> okay, console is working, network connection not ...
15:04:46< JonB> Bertl: does dhcp give it an address ?
15:05:24< JonB> Bertl: that multi reset switch you wrote about on your webpage, does that really reboot all the machines if the master machine reboots?
15:05:56< Bertl> depends on how it is built ... and what the master does ...
15:06:40< Bertl> if you use two pins for each reset, it should be quite safe ...
15:06:57< JonB> like data0 and data1
15:07:02< JonB> and data2 and dat3 ?
15:07:09< Bertl> if you use only one, then the 'test' procedure done on kernel/system boot might cause unwanted resets
15:07:10< JonB> 4 and 5 for the next ...
15:07:24< Bertl> yep, exactly ...
15:07:53< JonB> and that teste procedure will not touch data(X) and data(X+1) ?
15:08:23< Bertl> it might but usually it doesn't at least not separately ...
15:09:05< JonB> Bertl: i have 3 linux machines, all with 2 serial consoles, but 2 upses connected to them
15:09:19< JonB> Bertl: can i run console on serial port
15:09:22< Bertl> the thing is, if you use two pins, you have one out of four states which will do a reset
15:09:31< JonB> and from the same machine connect that to a different host
15:09:52< JonB> what happens if 2 console on serial port is connected to each other ?
15:10:12< Bertl> that won't work well ...
15:10:21< Bertl> but you can daisy chain them ...
15:10:32< JonB> Bertl: i dont understand that line with the 4 states
15:10:53< Bertl> S2@a -> S1@b, S2@b -> S1@c ...
15:11:09< JonB> Bertl: how? i have 6 serial ports, and 2 taken by upses, that is 4 left
15:11:35< JonB> Bertl: the point is i would like to have a remote serial root option on all 3 machines
15:12:01< Bertl> not without adding serial ports ;)
15:12:05< JonB> hmm
15:12:19< JonB> allright
15:12:50< Bertl> ad 4 states: an opto coupler is a combination of an LED and a photo transistor/diode
15:13:06< JonB> yes, i know
15:13:27< Bertl> the LED requires a certain polarity to work ... +/-
15:13:32< JonB> ahhh
15:13:57< JonB> but setting 0-0 and 1-1 ... that shouldnt really work should it ?
15:14:02< Bertl> 0;0, 1;1, and 0;1 won't lit it 1;0 will
15:15:59< JonB> okay, but where are we going with this?
15:16:30< JonB> do i need to change the software? or wire them the right way?
15:16:38>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection]
15:16:57< Bertl> no, the software just sets or clears the bits/pins ...
15:17:28< JonB> yes, but how does the software know which pins to set, doesnt it just support one pin id ?
15:18:16< Bertl> well, that is enough, and no, basically you can set each pin independantly ...
15:18:52< JonB> so i have to try which settings reboot with machines?
15:20:58< Bertl> well, you can try, but you can also figure it out by thinking about it a little longer ...
15:21:49< JonB> does the ATX care which way the power is applied?
15:21:52< Bertl> what you should try is, which combination (pins to LED) do not trigger a reboot when the 'master' is rebooted
15:22:28< Bertl> hmm, ATX? power applied? please elaborate ;)
15:22:42< JonB> my dualmm machine is AT
15:22:48>> click [click@gonnamakeyou.com] has joined #vserver
15:23:07< JonB> meaning just the sligtest power across the reset pins will reset it
15:23:17< JonB> ATX requires 4 seconds?
15:23:30< Bertl> that can be configured in the bios ...
15:23:31>> netrose [netrose@24.207.228.55] has quit [Quit: ]
15:23:39< JonB> and i doubt the testing during boot takes 4 seconds
15:23:48< Bertl> but yes, it would be a good idea to enable this
15:23:49>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has joined #vserver
15:24:04< JonB> to enable what? the 4 seconds delay?
15:24:07< Bertl> yep
15:24:12< Bertl> greetings mhepp!
15:24:19< JonB> Bertl: as for the power direction
15:24:30< mhepp> hello
15:24:49< JonB> Bertl: suppose i connect the + from the optocoupler to the + on the mainboard... and - to -
15:25:00< JonB> and compare that with +- and -+
15:25:07< JonB> would both directions reset?
15:25:11< Bertl> well, there is no + and - on the transistor of the opto coupler ...
15:25:14< JonB> or just one of them work ?
15:25:19< JonB> okay
15:25:21< Bertl> there is an emitter and a collector ...
15:25:46< Bertl> which work like a diode ... so probably one one direction will reset it ... but you have to experiment with that
15:26:01< Bertl> s/one one/only one/
15:26:28< JonB> thought so
15:27:39< Bertl> most reset circuits have a power up resistor and are active low, so you usually connect the emittor to ground and the collector to the reset line 
15:32:13>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has joined #vserver
15:32:28< Napalm> hello everyone
15:32:33< JonB> Bertl: okay
15:32:36< Bertl> hi Napalm!
15:32:54>> Bertl changed the topic of #vserver to: http://linux-vserver.org/ | latest stable 1.27, devel 1.3.9, 1.9.1
15:33:03< mhepp> coool
15:33:32< Sebastian> can i use iptables on vserver?
15:33:51< Bertl> sure you can ...
15:34:18< Sebastian> ok :)
15:34:24< Bertl> on the host, and with the proper capability also inside a vserver ... but it isn't very secure to do so ...
15:35:19< Napalm> bert: what do you think if i was to install 2.6.6-vs1.9.0
15:35:45< Bertl> hmm, I'd say, why not 1.9.1?
15:36:35< Napalm> bert: well im not sure exectly which would best
15:37:11< Napalm> bert: i would like the option to hard limit things like memory and cpu, and i definatly need to hard set the disk space for each VDS
15:37:42< Napalm> bert: stability is my other concern as i will have customers on them
15:38:42< Napalm> bert: host will be a P4 2.8GHz, 512Mb DDR, 60Gb 7200 HDD
15:38:53< Napalm> bert: whats your suggestions?
15:39:35< Bertl> my suggestion: try vs1.27 and vs1.9.1 and see what you like better ...
15:39:53< Napalm> well im running vs1.27 at the moment
15:40:06< Napalm> but i cant seem to get quota support enabled
15:40:55< Bertl> quota support as in 'Disk Limits'?
15:41:25< Napalm> yes
15:46:25< Shotygun> Has anyone messed with linux clusters here before?
15:46:34< Napalm> bert: don't you need vroot device support in 1.9.0?
15:46:56< Napalm> bert: the table on this page seems to state otherwise http://www.linux-vserver.org/index.php?page=Release+FAQ
15:46:58< Bertl> Shotygun: like in *mosix or like in failover?
15:47:16< Shotygun> Bertl: *mosix
15:47:40< Bertl> Napalm: 1.9.x doesn't support vroot devices atm, maybe never will ...
15:48:00< Shotygun> I'm clueless about clusters and I need to set up one
15:48:18< Shotygun> (At least that's what the client ones, I think it's wrong)
15:48:22< Napalm> bert: so how do you setup disk limits on 1.9.(0|1)?
15:48:53< Shotygun> Do you know if Oracle can you two-way replication?
15:48:54< Bertl> Napalm: not at all, they are not part of 1.9.x yet ...
15:49:09< Bertl> Shotygun: I don't use oracle ...
15:49:47< Shotygun> Yeah me neitehr
15:50:38< Shotygun> I got an e-mail from third party for setting up two oracle servers that suppose to do 8500 queries per second, but they didn't detail what type of select/databases and what are the peaks and etc, 8500 sounds  a bit overhead
15:50:51< kestrel> oracle rac is active-active
15:51:05< kestrel> that's an awesome design specification
15:51:23< Shotygun> What do you mean?
15:51:37< SHu_here> Bertl is there a patch 1.9.0 to 1.9.1?
15:51:39< Napalm> bert: so what is my best choice of kernel patch as i need to set the "disk space allowance to each vserver"?
15:52:04>> mhepp [~mhepp@r72s22p13.home.nbox.cz] has quit [Remote host closed the connection]
15:52:05< Bertl> I'd say any db should be able to do 8500 q/s ... "select 1;"
15:52:39< Bertl> Napalm: in this case you limit yourself to 1.2x and 1.3.x
15:53:03< kestrel> i mean that is a crap specification :)
15:53:07< Napalm> bert: i suppose i could get away with creating a stats program to execute 'vdu' each server per night, if the server is over its set amount it gets stopped and other owner notified by email
15:53:12< kestrel> 8500 queries/second could mean anything
15:53:24< Napalm> bert: how does that idea sound?
15:53:35< Shotygun> Exactly
15:53:40< Shotygun> That's what I was saying
15:53:54< Shotygun> I got over my e-mail 8500 queries per second, I was like hmm.. now wtf that suppose to mean..
15:54:09< kestrel> heh
15:54:13< Bertl> Napalm: funny ... but I heard stranger ideas ....
15:54:18< Shotygun> Oh bert, The core of the bot is done, I need to add a function to do better data searching and split them into catagories.
15:55:01< Bertl> let me know when we can test it (on a separate channel maybe)
15:56:07< Bertl> SHu_here: sure there is ...
16:02:43< SHu_here> wherE
16:02:59< Bertl> hmm, where would you expect it?
16:03:19< SHu_here> linux-vserver.org/Experimental/
16:03:28< Bertl> a development release?
16:03:44< SHu_here> 1.9.0 to 1.9.1
16:04:24< Bertl> http://www.13thfloor.at/vserver/d_rel26/delta/
16:04:53< SHu_here> thanks Mr Bertl
16:05:00< Bertl> you're welcome ...
16:06:02< SHu_here> there was no problem with 1.9.0 for me
16:06:58< Napalm> ok im going for the gold 2.6.6-vs1.9.0
16:08:57>> Napalm [~napalm@host81-7-26-228.adsl.v21.co.uk] has quit [Quit: "be back later"]
16:14:27>> netrose [netrose@SP2-24.207.228.55.charter-stl.com] has joined #vserver
16:15:00< Bertl> hi bobi!
16:23:46>> SHu_here [~shushushu@cpu183.adsl.qc.bellglobal.com] has quit [Quit: http://base2091.com]
16:45:57>> Sebastian [~cereal@pD9EAB7AC.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
16:53:23< Bertl> okay, leaving now, cya all later ...
16:53:27>> Bertl is now known as Bertl_oO
17:36:28>> cereal [~cereal@pD9EA9F1C.dip.t-dialin.net] has joined #vserver
18:45:03>> chaosle [~yvan@port-212-202-168-55.dynamic.qsc.de] has joined #vserver
18:53:41>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
18:56:50>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
19:06:57>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
19:36:42>> _id [~id@p50835AA6.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
19:45:26>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
19:48:10>> _id [~id@pD951991D.dip.t-dialin.net] has joined #vserver
19:49:05>> hiaslboy [matthias@e-16.vc-graz.ac.at] has quit [Remote host closed the connection]
19:53:17>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection]
20:01:04>> click [click@gonnamakeyou.com] has joined #vserver
20:09:13>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection]
20:11:22>> JonB [~NoSuchUse@129.142.112.33.ip.tele2adsl.dk] has quit [Quit: Leaving]
20:14:14>> hiaslboy [matthias@e-16.vc-graz.ac.at] has joined #vserver
20:19:37>> cereal [~cereal@pD9EA9F1C.dip.t-dialin.net] has quit [Ping timeout: 480 seconds]
20:21:38>> cereal [~cereal@pD9EAAEAE.dip.t-dialin.net] has joined #vserver
20:28:10>> hiaslboy [matthias@e-16.vc-graz.ac.at] has quit [Ping timeout: 480 seconds]
20:34:18>> hiaslboy [matthias@e-16.vc-graz.ac.at] has joined #vserver
20:35:00>> hiaslboy [matthias@e-16.vc-graz.ac.at] has quit [Remote host closed the connection]
20:35:20>> hiaslboy [matthias@e-16.vc-graz.ac.at] has joined #vserver
20:37:24>> noel_ [~noel@pD9FFFC93.dip.t-dialin.net] has quit [Quit: Client exiting]
20:37:24< hiaslboy> at Bertl: the new patch has some troubles with modules: the unix module does not load correctly
20:38:04< hiaslboy> I will post dmesg if necessary :-)
20:43:09>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Quit: Client exiting]
20:43:40>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
20:56:30>> click [click@gonnamakeyou.com] has joined #vserver
21:13:49>> click [click@gonnamakeyou.com] has quit [Remote host closed the connection]
21:16:40>> Shotygun [shotgun@shotygun.com] has quit [Remote host closed the connection]
21:24:27>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
21:52:39>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
22:00:43>> taxcollector [~taxcollec@192.16.167.161] has joined #vserver
22:14:27>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
22:20:53>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
22:25:29>> taxcollector [~taxcollec@192.16.167.161] has quit [Remote host closed the connection]
22:36:51>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
22:41:20>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
22:57:08>> serving [~serving@213.186.191.31] has quit [Ping timeout: 480 seconds]
23:12:09>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has quit [Quit: Client exiting]
23:12:12>> yarihm [~yarihm@217-162-206-157.dclient.hispeed.ch] has joined #vserver
23:23:57>> rs [~rs@rs.admin.rhapsodyk.net] has joined #vserver
23:24:02< rs> hi
23:53:36>> rs_ [~rs@rs.admin.rhapsodyk.net] has joined #vserver
23:56:30>> rs [~rs@rs.admin.rhapsodyk.net] has quit [Ping timeout: 480 seconds]
23:57:06>> cereal [~cereal@pD9EAAEAE.dip.t-dialin.net] has quit [Quit: W)W)ip www.wwip.de | Hosting, Housing, CoLocation and VServer | Forum: http://forum.wwip.de]
23:57:52< hiaslboy> Bertl : May 20 20:03:36 e-16 vmunix: unix: Unknown symbol unhash_nx_info
23:58:32>> taxcollector [~taxcollec@192.16.167.161] has joined #vserver
23:58:43< hiaslboy> Bertl : May 20 20:03:36 e-16 vmunix: unix: Unknown symbol unhash_vx_info
--- Log closed pią maj 21 00:00:02 2004