[00:00] <mathiaz> kirkland: is the good one?
[00:00] <kirkland> mathiaz: i don't know yet
[00:00] <kirkland> mathiaz: testing
[00:00] <kirkland> mathiaz: i'm looking at each of about a dozen functional commits
[00:18] <kirkland> mathiaz: i think that last git patch is the winner
[00:19] <mathiaz> kirkland: oh - good news. Now does the last patch rely on changes made by previous patches?
[00:19] <kirkland> mathiaz: no ...
[00:19] <kirkland> mathiaz: i applied just that little patch
[00:20] <kirkland> mathiaz: i'm pushing to p.u.c
[00:21] <mathiaz> kirkland: great - I'll test it
[00:22] <kirkland> mathiaz: dget http://people.ubuntu.com/~kirkland/mdadm/new/mdadm_2.6.7.1-1ubuntu8.dsc
[00:22]  * kirkland crosses his fingers
[00:33] <mathiaz> kirkland: hm - :/
[00:33] <mathiaz> kirkland: ubuntu8 doesn't fix the problem.
[00:33] <mathiaz> kirkland: booting from a degraded RAID1 array still fails.
[00:35] <Stargazer> How can i get ubuntu server to auto-mount 'new' devices that connect(physically).
[00:35] <Stargazer> *?
[00:36] <kirkland> mathiaz: wtf
[00:36] <kirkland> mathiaz: okay, we have some disconnect between your testing and mine
[00:37] <kirkland> mathiaz: this is probably back to why 2.6.7.2 works for me and not for you
[00:37] <mathiaz> kirkland: this time I made sure that both disks were in sync before unplugging one of them
[00:38] <kirkland> mathiaz: so both disks were sync
[00:38] <kirkland> mathiaz: you installed the newly built package
[00:38] <kirkland> mathiaz: it updated the initramfs
[00:38] <giovani3> Stargazer: usb devices? try looking into usbmount
[00:38] <kirkland> mathiaz: you powered off
[00:38] <kirkland> mathiaz: booted up with only 1 of the 2 devices
[00:38] <giovani3> Stargazer: http://usbmount.alioth.debian.org/
[00:39] <kirkland> mathiaz: and it dropped you to initramfs?
[00:39] <mathiaz> kirkland: yes
[00:39] <Stargazer> Does that package generally mount all devices, like SD cards and external HDDs(via USB)
[00:40] <Stargazer> *?
[00:40] <giovani3> Stargazer: well, considering SD cards are never directly connected to the system ... yes
[00:40] <giovani3> SD cards are almost always fed through a USB SD reader
[00:40] <mathiaz> kirkland: let me rerun dpkg-reconfigure mdadm - to make sure that initramfs is updated
[00:40] <kirkland> mathiaz: k
[00:40] <Stargazer> Giovani3, my laptop has an SD port.
[00:40] <giovani3> Stargazer: however, servers don't typically have usb/sd devices plugged into them often
[00:40] <Stargazer> What does Ubuntu Desktop use then ?
[00:41] <giovani3> Stargazer: why would you be using laptop hardware to run a "server"?
[00:41] <giovani3> I believe it uses usbmount
[00:41] <giovani3> why?
[00:41] <Stargazer> Giovani3: it's light, doesn't distract me with GUIs and... i can concentrate!
[00:41] <giovani3> Stargazer: do a lot of text web browsing then? heh
[00:42] <Stargazer> No.
[00:42] <Stargazer> Typically: i don't web browse using Lynx/other.
[00:42] <Stargazer> I may chat... if i have a connection.
[00:43] <mathiaz> kirkland: http://paste.ubuntu.com/136364/ <- this is the kvm command line used to start the vm by libvirt
[00:43] <giovani3> I wouldn't find non-gui terminals very useful on a physical machine
[00:43] <giovani3> but to each their own
[00:43] <mathiaz> kirkland: and yes - I can confirm that it cannot boot from a degraded RAID1 array.
[00:43] <giovani3> servers, should definitely never have guis on them -- but I don't usually sit in front of keyboards connected to servers :)
[00:43] <kirkland> mathiaz: erg
[00:43] <mathiaz> kirkland: I'm still dropped to the busybox shell even if I ask to boot from a degraded array.
[00:43] <kirkland> mathiaz: my test env must be polluted in some way
[00:44] <kirkland> mathiaz: i'm going to create some virgin images
[00:45] <kirkland> mathiaz: okay, i gotta run for a bit
[00:45] <kirkland> mathiaz: i'm going to drop for a while
[00:45] <kirkland> mathiaz: how big is the .img file you're working with?
[00:45] <kirkland> mathiaz: any chance you can set the password to 'ubuntu' or something and upload it to p.u.c ?
[00:45] <kirkland> mathiaz: and i'll test against it?
[00:47] <mathiaz> kirkland: hm - mdadm 2.6.8 also fails now
[00:47] <mathiaz> kirkland: the images are around 1 Gb.
[00:48] <mathiaz> kirkland: we can continue debugging this issue tomorrow.
[00:48] <mathiaz> kirkland: I'll have fresh new images anyway since the -server iso are respun.
[00:55] <Stargazer> Does server use X ?
[01:00] <mathiaz> Stargazer: no
[01:02] <Stargazer> Ty.
[01:21] <{bosco}> anyone know of a good website tracker????
[01:21] <{bosco}> for ubuntu server
[01:22] <twb> wget.
[01:23] <{bosco}> huh wget
[01:23] <JanC> what do you mean by "website tracker" ?
[01:24] <{bosco}> like webalizer ???
[01:24] <{bosco}> JanC,
[01:27] <{bosco}> and awstats
[01:27] <{bosco}> need to know the best one
[01:27] <{bosco}> !info awstats
[01:27] <{bosco}> !info webalizer
[01:28] <JanC> {bosco}: I suggest you look for log analyzer or such, and webalizer & awstats are probably fine
[01:28] <{bosco}> that is what i thought i would install awstats
[01:29] <JanC> if you are used to one of these, then use it for now
[01:29] <renatokrause> Good night
[01:30] <{bosco}> JanC, i am more used to webalizer
[01:32] <giovani3> or you can use google
[01:32] <JanC> then use it, and maybe check the others in a test environment to investigate if they might be better in the future
[01:33] <JanC> (better for you)
[01:33] <{bosco}> giovani3, lol
[01:33] <giovani3> {bosco}: ?
[01:33] <giovani3> google's web stats are good
[01:36] <{bosco}> giovani3, i know just thought it was kinda funny to hear you say that i guess
[01:36] <{bosco}> they are good
[02:03] <TwoToneSpirit> Hello.  Is there a graphical FTP server?
[02:04] <giovani3> TwoToneSpirit: GUIs are not supported in #ubuntu-server
[02:04] <{bosco}> net2ftp TwoToneSpirit is that what yo umean
[02:04] <giovani3> I don't know why you'd run linux and then want a "graphical ftr server" anyhow
[02:04] <{bosco}> something like that
[02:04] <{bosco}> i know lol
[02:05] <giovani3> net2ftp seems to be a web-based ftp client ...
[02:05] <scopecreep> can you guys recommend anything more detailed than phpsysinfo to monitor my server remotely?
[02:05] <giovani3> scopecreep: ssh, and learning the shell
[02:05] <shrimp> question:  is 2mb upstream plenty "in your opinion" to run a decent server?
[02:05] <TwoToneSpirit> giovani3:  Well some people don't want to use the terminal, but still want to be able to grab a list of people logged into their FTP.
[02:05] <giovani3> htop is your friend
[02:05] <giovani3> TwoToneSpirit: well, I suggest you learn, really
[02:06] <{bosco}> look into net2ftp
[02:06] <TwoToneSpirit> giovani3:  Come on now, that's not a good attitude.  Of course I can deal with the command line, but as you understand I'm sure, there are many people out there who aren't computer people, but still want to have working features.
[02:06] <giovani3> TwoToneSpirit: it's an appropriate attitude, I think -- I don't know of a single "graphical" ftp server for linux -- it's possible they exist -- but if that's the type of software you desire, you're unlikely to be happy with linux in general
[02:07] <scopecreep> i can use the shell but i dont want to have to ssh in just to see cpu load and whatnot
[02:07] <mattmitchell> could someone give me some advice on whether or not i should setup my own email server (no experience) OR just use something like google apps?
[02:08] <giovani3> mattmitchell: it all depends on your aptitude, and willingness to try things out
[02:08] <TwoToneSpirit> giovani3:  Well, the point is that we need to develop a solution so that people who want a GUI environment *are* happy.  I have turned many people on to Ubuntu who don't know or care about the command line, but unfortunately, if they want to run an FTP server, I have to tell them to wait.  And I don't like making people wait when it means continuing to run non-free software.
[02:08] <mattmitchell> giovani3: i see. so, what I'm worried about is getting hacked, or spam. i don't mind spending time setting things up and learning though.
[02:09] <giovani3> TwoToneSpirit: I do not agree -- first of all, ubuntu doesn't write software, second of all -- GUIs on servers should never, and hopefully never will be supported by ubuntu -- different distributions have different opinions on this
[02:09] <giovani3> TwoToneSpirit: if you want GUIs, linux is probably not for you
[02:09] <giovani3> if you think it is -- RedHat is probably more your style
[02:09] <giovani3> mattmitchell: postfix, by default, is secure -- spam is totally separate, and is difficult for anyone to battle
[02:10] <mattmitchell> giovani3: ok that's a good start though. do you have any good references for dealing with spam when using postfix?
[02:10] <giovani3> mattmitchell: I'd recommend hosting your mail with google -- and playing around on a test machine until you feel comfortable
[02:10] <giovani3> mattmitchell: yep, the postfix website has lots of links for that -- look into greylisting, RBLs, and spamassassin
[02:10] <mattmitchell> giovani3: yeah, that sounds like a good idea!
[02:11] <shrimp> <---just signed up for roadrunner business class account, 15mb downstream and 2mb upstream, will this be plenty for a decent server, or will I require more?
[02:11] <mattmitchell> giovani3: cool. thank you :)
[02:11] <giovani3> shrimp: there are no guidelines -- it just depends on what you need to use the server for
[02:12] <giovani3> clearly, as long as you need less than 2Mbps ... you'll be covered -- however, roadrunner has pretty crappy SLAs, and, I wouldn't consider it very reliable for something mission-critical, if you plan to host mail on that connection, I'd make sure you have a backup mx provider elsewhere on the internet
[02:12] <shrimp> not running a dns server, go through dyndns
[02:13] <TwoToneSpirit> giovani3: The phrase "linux is probably not for you" needs to be phased out of our vocabulary.  The slogan "Linux for human beings" must begin to mean something.
[02:13] <giovani3> TwoToneSpirit: I don't represent Canonical, therefore, my views don't have to be, and really, shouldn't be in line with theirs
[02:13] <giovani3> linux is an open place -- everyone is welcome to their view
[02:14] <giovani3> it's an official (so I've been told) policy of this support channel that GUI-related requests are not to be handled here
[02:14] <shrimp> have experiance with rr giovani?
[02:14] <giovani3> shrimp: yep
[02:14] <shrimp> what happened?
[02:15] <shrimp> I hear nothing but good around here
[02:15] <giovani3> shrimp: they're a low-class isp in my experience, but, my requirements may be different from yours
[02:15] <shrimp> maybe a difference in managment?
[02:15] <TwoToneSpirit> giovani3:  No, of course not - I didn't mean to say that you need to conform in any way.  All I was saying is that it behooves all of us to have a gentle approach and a warm atmosphere - "linux is not for you" will obviously scare some people away.
[02:16] <shrimp> it's just to start anyway, a community based website with joomla, I shouldnt have any problems with it just as a starting point I wouldnt think
[02:16] <giovani3> TwoToneSpirit: I don't believe that everyone needs to be welcomed with open arms no matter what the request -- there are idealogical differences between different distributions, and operating systems, GUIs on servers is a pretty clear-cut one, most people in the linux community would be horrified to see one on a server, and to have you insist on one, is not gonna get you far with most linux folk
[02:17] <shrimp> X on a server, lmao NEVER
[02:20] <shrimp> if a person doesnt know how to ssh into a server, or install and use Linux without a GUI, they really dont need to be setting up a server "just my opinion of course"
[02:26] <{bosco}> www.boscoslife.com permission denied www.boscoslife.info works fine both on same server both used to load fine can someone help me :??????
[02:44] <shrimp> anyone had any GOOD experiences with rr business class?
[03:00] <orudie> is there a good free ssh client for blackberry ?
[03:01] <shrimp> no
[03:01] <Votan> ello
[03:26] <shrimp> no
[04:32] <nrich> how do I install the xen kernel for ubuntu, so that it can be paravirtualized ?
[05:23] <shrimp> man, ubuntu doesnt include the ROUTE target
[05:24] <shrimp> man, ubuntu doesnt include the ROUTE target
[05:25] <shrimp> yea I know, isnt that a shame?
[05:25] <shrimp> yes it is
[05:25] <shrimp> I wanted it, so I didnt have to recompile
[05:25] <shrimp> I know!
[05:40] <{bosco}> anyone here awstats problems premission problems when sudo -u www-data /usr/bin/perl /usr/lib/cgi-bin/awstats.pl -update -config=www.boscoslfe.com error message here http://pastebin.com/m49a8d708
[05:48] <BentFranklin> I installed postfix on my Kubuntu 8.10.  I think it is configured correctly (maybe).  I'd like to test it.  Now I believe I need a MUA to compose email.  Is that correct?
[05:55] <jmarsden> BentFranklin: Well, you could telnet localhost 25 and type in SMTP commands to do it, if you know enough SMTP.  But for most normal humans, using an MUA is preferred :)
[05:57] <BentFranklin> So, Thunderbird is an MUA right?  Is MUA roughly equivalent to "email client"?
[05:58] <BentFranklin> Are you aware of any MUA's that support command line composition (for shell scripts)?
[05:58] <jmarsden> You can't run Thunderbird on Ubuntu Server... the server has no GUI.  Command line MUAs include mutt and bsd-mailx
[05:59] <jmarsden> Example shell command to send email:   echo "This is my message" | mail -s "my subject" someone@example.com
[06:00] <BentFranklin> Sweet. Just what I need.
[06:01] <BentFranklin> In your example, 'mail' might be replace by 'mutt'?
[06:01] <jmarsden> No, use mail.  mutt is a client that uses the whole screen , mail (really /usr/bin/bsd-mailx underneath) can be non-interactive.
[06:03] <BentFranklin> Ok, I'll apt-get bsd-mailx.  Thanks a lot jmarsden!
[06:03] <twb> "the server has no GUI" is utter bollocks.  An X client and/or X servers can be installed on an Ubuntu Server system just as easily as on any other Ubuntu system.  Whether it's *advisable* to do so is another matter.
[06:03] <BentFranklin> Anyway I have Kubuntu 8.10
[06:04] <BentFranklin> I asked in here because they told me to the other day on #kubuntu
[06:04] <twb> You could say "no GUI is installed by default" or "ubuntu server has no GUI by default".
[06:05] <jmarsden> This IRC channel is for support of Ubuntu Server.  Which comes with no GUI, and questions about GUIs and GUI applications or installing GUIs on servers are unlikely to be well received here, in general... that is not what this channel is for.
[06:07] <BentFranklin> How do people traditionally configure their ubuntu servers - with console emacs maybe?
[06:08] <twb> BentFranklin: IMO it is silly to install an fullscreen editor on a server, when cat and ed work perfectly well.  Or using Emacs' TRAMP from a remote machine.
[06:09] <jmarsden> With whatever text editor they choose, yes.  Emacs, vi, nano, whatever.
[06:09] <twb> BentFranklin: certainly most servers I know run without a screen or keyboard attached.
[06:09] <BentFranklin> I see, you run your editor on a remote machine
[06:10] <BentFranklin> I installed bsd-mailx and it installed a bunch of exim stuff.  Also it said "* Starting MTA ...done."  Will this conflict with postfix?
[06:12] <jmarsden> That's a little odd.  Did it uninstall postfix??  if not you should be fine.
[06:12] <jmarsden> You can always do   telnet localhost 25    and see what the banner line from the MTA says, if you are not sure whether postfox or exim is running on your machine.
[06:14] <BentFranklin> It says exim
[06:14] <BentFranklin> Oops! My bad.  I got my hosts confused.
[06:15] <jmarsden> You might want to put the hostname in your shell prompt to help avoid that... PS1='\u@\h:\w\$ '   in bash, for example.
[06:16] <BentFranklin> Actually I have that already...
[06:17] <jmarsden> Ah... you installed bsd-mailx on some other machine that did not have postix on it... right? Hence the accidental exim installation?
[06:17] <BentFranklin> On the correct system it did not install exim (because there was already an MTA I assume)
[06:17] <jmarsden> Right.
[06:17] <twb> I have noticed that various packages in Ubuntu inconsistently recommend exim4 | m-t-a, postfix | m-t-a, or even just m-t-a (which defaults to cyrus, ffs)
[06:17] <twb> I don't think any MTA is installed by default.
[06:18] <twb> Anyway, I recommend you use postfix instead of exim4.
[06:18] <p_quarles> twb: well, you're wrong
[06:24] <BentFranklin> It works perfectly.  Thank you sirs!
[06:28] <p_quarles> twb: anyway, exim4 is installed by default in any version of Ubuntu; and an MTA is a fundamental element of any Linux distro
[06:28] <twb> p_quarles: that's not true.  On desktop Ubuntu, exim4 is *definitely* not installed.
[06:29] <twb> p_quarles: this is different to Debian, which now installs exim4 but leaves it unconfigured -- so the mail is accepted but never goes anyway.
[06:29] <twb> s/anyway/anywhere/
[06:30] <twb> That's assuming you don't uncheck the "standard" task on Debian, of course.
[06:33] <infinity> p_quarles: "You're wrong" isn't the best way to jump into a conversation.
[06:33] <infinity> p_quarles: And, for the record, we've not installed any MTA by default on Ubuntu for years now, and back when we did, it was postfix, not exim.
[06:34] <infinity> p_quarles: (And this isn't a religious issue, it's fact.  For the sake of religion, I'll state that I'm an exim user)
[06:35] <twb> The main reason I avoid exim4 isn't because it's crap, but because I don't know anybody that uses it -- apart from a small subset of hard-core Debian users.
[06:36] <twb> Most of the Debian *and* CentOS users I know use postfix, so I can easily discuss problems and learn good ideas about it.
[06:36] <infinity> twb: It's got a pretty massive install base, actually.  But I tell people to stick with that they know.  MTAs are confusing beasts, and changing to the flavour of the month so the cool kids will like you isn't all that sane.
[06:36] <twb> infinity: install base of people who got it because it was the default, or installed base of people who actually understand it, more or less?
[06:37] <infinity> twb: A massive install base of real sysadmins, not accidental users, yes. :)
[06:37] <twb> Fair enough.
[06:37] <infinity> twb: And I'd argue that a large portion of postfix's install base is accidental users too, since it's a common recommendation for people to install for a "just works" MTA.
[06:37] <infinity> (Hence why it's really hard to find someone who can answer difficult postfix questions... And why I never let lamont out of my sight)
[06:38] <twb> Shrug.
[06:38] <infinity> But, yeah.  It'a all about what you're comfy with, IMO.
[06:38] <infinity> They're both powerful enough to do whatever you want, if you know how to make 'em do it.
[06:38] <infinity> *shrug*
[06:41] <twb> infinity: is there an exim channel as there is #postfix?
[06:46] <{bosco}> can anyone help me out never had to use sudo to change anything in root http://pastebin.com/m7e329d86
[06:50] <{bosco}> ::::""""???\
[06:50] <infinity> twb: Probably.  I don't tend to look for support for it.  The exim spec (the manual) is pretty comprehensive.
[07:03] <{bosco}> i run a vps
[08:16] <kraut> moin
[08:21] <Skaag> moin
[08:22] <Jeeves_> OLA!
[08:25] <_ruben> hm .. icecream? :)
[09:03] <AlexC_> g'morning
[09:04] <AlexC_> yesterday I was working on expanding an EXT3 partition as it was only 260mb in size, all was working fine until I have resized the partition with 'resize2fs', I restarted and now upon boot GRUB is spitting out error 22
[09:07] <AlexC_> basically, sda3 was 260 mb and sda2 (which I did not need) was 2gb. Talking with someone here I 'dd' sda3 to sda3, deleted the partitions and re-created a partition starting where sda2 started, then end where sda3 ended. This seemed to work, and IIRC I managed to reboot just fine. Issue seems to be after I ran resize2fs to expand the 260mb filesystem to fill up the now 2gb size of sda2
[09:09] <AlexC_> I do have console redirection for this server provided via DRAC, so at least I can poke it from here =) Any help would be greatly appreciated
[09:12] <soren> AlexC_: ..so you have access to grub?
[09:13] <AlexC_> soren, well I thought so - however I was not promoted to press 'e' or what ever it is to edit grub, before it attempts to load (ie, it goes stage 1.5 straight to error 22)
[09:14] <soren> Ah.
[09:16] <AlexC_> looking though this RAC we have, I may be able to virtually add a media device, so a live 'CD' should do the trick
[09:21] <AlexC_> when I get into a live cd, what could the issue be?
[09:24] <soren> I think the problem is that the partition where it goes looking for the stage2 loader and the configuration file is encoded into the stage1.5 loader.
[09:24] <soren> Let me think for a second..
[09:25] <AlexC_> I still have a backup of the original partition, and an iso of the partition before I did resize2fs btw, in case that will be of any help
[09:42] <soren> AlexC_: I don't think that will help much.
[09:45] <soren> AlexC_: Did you boot the live cd yet?
[09:46] <simplexio> AlexC_: iso , do you mean dd image ?
[09:46] <AlexC_> soren, not yet, got some issues with installing the browser plugin to let me do it
[09:46] <simplexio> AlexC_: created using dd if=/dev/sd? of=image.img
[09:47] <AlexC_> simplexio, yes
[09:48] <simplexio> AlexC_: you can restore data using dd
[09:50] <AlexC_> mmhmm, trouble is I am locked out currently =)
[10:07] <simplexio> AlexC_: need livecd because your grub is broken
[10:07] <AlexC_> I know
[10:07] <AlexC_> just getting that Live CD working is tricky when Dell decide that everyone should be using 32bit Windows XP and MSIE6
[10:08] <simplexio> AlexC_: you probably deleted /boot/ dir or something .. if i rcall right installed grub only need data in mbr + /boot/grub/ directory stuff
[10:32] <AlexC_> urgh, why do companies spit out software that simply does not work is beyond me
[10:34] <AlexC_> especially on something so vital as this
[10:52] <AnRkey> how can i send a test job to lpd remotely?
[10:52] <AnRkey> i know the queue and ip
[10:53]  * AnRkey will hug whoever knows
[10:55] <AlexC_> soren, ah, finally - I am in the live cd
[10:55] <AlexC_> I had to install Firefox 2, manually install the Dell Firefox extensions and run the browser as root! How insanely insecure and stupid is that
[11:08] <simplexio> damn www-proxy is broken.. have to use links2 to browsing from server :)
[11:08] <simplexio> nice day in work
[11:29] <embrik> I want my ubuntu hardy server to be a domain controller and that my pupils can log on to every workstations at school with a roaming profile - I haven't found any howto on this - does anyone know?
[11:30] <embrik> someone has told me it is difficult. I use debian-edu (skolelinux) today, but it doesn't appeal to the users
[11:37] <AlexC_> is it possible to edit the grub menu before Error 22 kicks in? I thought it was possible to do
[11:37] <AlexC_> just this live cd is incredibly slow over a network
[11:39] <soren> AlexC_: No, I'm afraid not.
[11:39] <soren> AlexC_: Error 22 is about grub not being able to find the partition with the stage 2 loader on it.
[11:40] <soren> ...and that's what has the menu and all that.
[11:40] <AlexC_> oh, cool. Though it would be possible to re-install GRUB if/when I get into this live cd?
[11:41] <soren> Yes.
[11:42] <AlexC_> what would be  the most lightweight/smallest live CD to boot from to do this? Net connection here isn't the best, currently got the System Rescue CD running, though yet to get into it due to sheer size
[11:42] <AlexC_> well, not live cd - just to get a console
[11:43] <XiXaQ> any experiences regarding jaunty server? Is it mostly stable now, or will there still be radical changes? I'm not going to use it for a stable production system, but I have a slow connection at home, so I'd prefer to avoid downloading so many updates.
[11:47] <AlexC_> XiXaQ, it's not at a stable release, I think that answers the question
[11:47] <rst-uanic> XiXaQ: if you do not want many updates you should wait for couple of weeks after release
[11:48] <XiXaQ> but the feature freeze is in effect after alpha 6?
[11:49] <AlexC_> XiXaQ, which in theory means it will be at its buggiest, as they will now focus on bug fixes
[11:49] <rst-uanic> XiXaQ: FF was in the February according to https://wiki.ubuntu.com/JauntyReleaseSchedule
[11:50] <XiXaQ> yes, I was looking at it.
[11:51] <rst-uanic> XiXaQ: and what do you mean by a slow connection?
[11:53] <XiXaQ> a 3g connection with bad reception; 200Kbps in average.
[11:53] <rst-uanic> oh
[11:54] <rst-uanic> but you can download iso, install it and never update until final release
[11:54] <XiXaQ> so, if I get 300MBs of updates daily, then it really isn't worth it. :)
[11:54] <rst-uanic> XiXaQ: yeah)
[11:55] <XiXaQ> what I want to do, is explore it and its new features so that I'm ready to start working seriously with it when it's released and stable.
[11:56] <rst-uanic> XiXaQ: wait for beta (two days I think) and try it then
[12:12] <rgreening> ScottK-desktop: ping
[12:22] <rgreening> ping ScottK2
[12:24] <rgreening> I need a FFe for python-django-tagging. Current package is broken. New svn snapshot corrects this.
[12:28] <rgreening> bug/347882
[12:28] <rgreening> bug 347882
[12:28] <rgreening> I have a package prepared for upload
[12:40] <ScottK-desktop> rgreening: I'd say fix it then (that's an FFe with my server guy I can so OK by myself hat on).
[12:45] <XiXaQ> does Darwin Calendar Server in jaunty support LDAP or PAM for users and groups?
[12:57] <rgreening> ScottK-desktop: ok. I'll upload
[13:02] <rgreening> ScottK-desktop: uploaded
[13:02] <rgreening> yay. first upload as motu :P
[13:20] <rgreening> ScottK-desktop: ping
[13:27] <kirkland> mathiaz: i just install a raid1 from today's iso
[13:27] <kirkland> mathiaz: boot-degraded magically works
[13:27] <kirkland> mathiaz: can you test it out, and see if i'm completely insane?
[13:31] <mathiaz> kirkland: and will test it out and confirm whether you're insance or not
[13:31] <mathiaz> kirkland: and will test it out and confirm whether you're *insane* or not
[13:32] <kirkland> mathiaz: :-)
[13:32] <kirkland> mathiaz: i installed from today's server iso
[13:32] <kirkland> mathiaz: i think that's important
[13:33] <mathiaz> kirkland: 20090324.2?
[13:37] <kirkland> mathiaz: 7f2d68febda3d67c9dbae8b6d3011281  jaunty-server-amd64.iso
[13:38] <mathiaz> kirkland: ok - I'm using the same iso
[13:42] <kraut> are there any best practises how to use iscsi?
[14:02] <kirkland> mathiaz: i'm going to test it one more time, with a different partitioning scheme
[14:04] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid1.preseed
[14:04] <mathiaz> kirkland: ^^ this the preseed I'm using for installing the system
[14:05] <kirkland> mathiaz: thanks, reading....
[14:05] <mathiaz> kirkland: look at expert_recipe to get the exact partition scheme I'm using
[14:06] <kirkland> mathiaz: okay, so no separate /boot partition
[14:06] <kirkland> mathiaz: i'm going to go test that now
[14:16] <ivoks> hello
[14:18] <mathiaz> ivoks: hi
[14:19] <ivoks> mathiaz: do you have 5 minutes?
[14:22] <AlexC_> I need a lightweight cd that either a) will let me install grub b) uses grub to boot an existing partition. Anyone have any ideas what I can use?
[14:22] <mathiaz> ivoks: sure
[14:22] <mathiaz> ivoks: I won't be able to attend the meeting as I have an appointment at the same time
[14:22] <ivoks> mathiaz: regarding dovecot-postfix
[14:23] <ivoks> mathiaz: ok, then it's even better to talk about this before the meeting :)
[14:23] <mathiaz> ivoks: yes :)
[14:23] <ivoks> mathiaz: bottom line, the question is: what to do when user removes dovecot-postifx, and what to do when user purges it?
[14:24] <acicula> AlexC_: boot an ubuntu cd in live mode or it's grub menu to boot the disk directly?
[14:24] <ivoks> should the functionality be there if the package is removed?
[14:25] <ivoks> should we revert to state before dovecot-postfix is installed?
[14:25] <ivoks> (on purge)
[14:25] <mathiaz> ivoks: I would guess so
[14:25] <AlexC_> acicula, ideally I would, however it needs to be extremly light weight. I have no physical access to the server, but I do have console redirection and virtual media that I can use. Net connection is not that good to boot a Ubuntu live CD, I pretty much just need grub or minimum to get me grub
[14:25] <ivoks> then we have to backup postfix's configuration
[14:25] <ivoks> this wouldn't be too hard
[14:26] <acicula> AlexC_: could try the netboot iso
[14:26] <ivoks> as far as purging goes, i thing it's sane to assume that user wants everything that camed with that package - removed
[14:26] <mathiaz> ivoks: right - couldn't we just undo the changes we made to the postfix configuration?
[14:26] <acicula> AlexC_: not sure if it comes with grub, but you can get a console that way
[14:26] <AlexC_> acicula, I'll try
[14:26] <mathiaz> ivoks: well the thing here is that the package is only about configuration
[14:26] <ivoks> mathiaz: yes, postconf is great for that
[14:26] <mathiaz> ivoks: so I wonder if removing == purging in that case
[14:26] <ivoks> i would agree that removing and purging is the same
[14:27] <acicula> AlexC_: though it'll take longer to get the console then booting a livecd in console mode i think
[14:27] <ivoks> then we break policy - not removing configuration on remove :)
[14:27] <ivoks> should we rename it to dovecot-postfix.conf.bak?
[14:28] <ivoks> cjwatson: ping
[14:28] <mathiaz> ivoks: that would be a good compromise
[14:28] <ivoks> cjwatson: would it be too bad to purge package even though user requested remove?
[14:28] <mathiaz> ivoks: move dovecot-postfix.conf to .conf.bak on remove and delete the file on purge
[14:28] <ivoks> ok
[14:29] <ivoks> mathiaz: why didn't we use update-alternative in the first place? :)
[14:30] <cjwatson> ivoks: yes
[14:30] <cjwatson> if the package actually includes configuration, removing it on remove rather than on purge is a bug
[14:30] <ivoks> cjwatson: how about renaming config to .bak?
[14:31] <cjwatson> that would probably be fine as long as you clean up entirely on purge
[14:31] <ivoks> since in this case, if the package is removed and configuration untouched, user won't actually remove dovecot-postfix integration
[14:31] <ivoks> ok
[14:31] <cjwatson> right, you just need to be careful not to trash user adjustments if they removed it by accident
[14:31] <cjwatson> you might want to consider whether you should move .bak back if they reinstall straight away
[14:32] <ivoks> hm... good idea
[14:32] <cjwatson> AlexC_: the server install CD has a rescue mode that permits reinstalling grub
[14:32] <cjwatson> AlexC_: you can do this from the netboot image by adding the boot parameter rescue/enable=true
[14:32] <ivoks> cjwatson mathiaz thank you guys
[14:34] <kirkland> mathiaz: results of your raid test?
[14:34] <AlexC_> cjwatson, awesome, I shall try that as well
[14:34] <mathiaz> kirkland: not finished yet.
[14:34] <mathiaz> kirkland: the other installations are still running
[14:34] <mathiaz> kirkland: it should take another hour or so to complete
[14:41] <stickystyle> Does anyone know if there is a way to do client side LUN masking with open-scsi?  My Dell MD3000i has a LUN that is presented as a 20MB disk.  Dell support tells me it's a special disk for managing the unit and if I'm not managing the unit from the the client it's not needed (the windows drivers let you hide it).
[14:42] <soren> I'm not sure what purpose that serves..
[14:42] <soren> stickystyle: Why would you bother hiding it?
[14:42] <stickystyle> the reason i ask is because every time i boot the machine I get "Buffer I/O error on device sdc, logical block 4" spewed across my tty and I'm a little worried that my linux box might be touching that lun in a way that may break the target.
[14:43] <soren> sdc is an iscsi share?
[14:43] <soren> That sounds odd.
[14:44] <stickystyle> It's just this weird 20MB FAT drive that each machine that logs into the box gets by default.
[14:45] <stickystyle> I have not explicitly told the array to export this LUN to my clients, thats just its design :-/
[14:53] <stickystyle> I guess while there is someone around that is familiar with iscsi, soren do you have any advice on bug 345957 ?
[14:54] <stickystyle> *Should* the box be able to connect to iscsi shares prior to network being configured?
[14:54] <stickystyle> Mathias' response didn't quite make sense to me.
[14:58] <LHC> hey
[14:58] <LHC> anyone know anything about webhosts
[14:59] <stickystyle> LHC: as in hosting companies, or hosting your own website?
[15:00] <LHC> hostin companies
[15:00] <LHC> I started up a reseller site but its not complete yet
[15:00] <ttx> It's time for the weekly meeting of the Ubuntu Server team, on this channel.
[15:00] <LHC> its something to bring in money on the side hah
[15:00] <ttx> Who's here for the server team meeting ?
[15:01] <LHC> Im only selling dedicated atm but I was wondering about webhosting (lots of users on one server)
[15:01] <kirkland> ttx: o/
[15:01] <stickystyle> ttx: I came to listen (metaphorically)
[15:01] <ivoks> isn't the meeting now?
[15:02] <ttx> ivoks: that's what I was trying to say, yes.
[15:02] <ttx> ok, let's pretend we have a Mootbot.
[15:02] <ttx> #startmeeting
[15:02] <ivoks> oh, tb meeting in -meeting :/
[15:02] <ttx> Agenda: https://wiki.ubuntu.com/ServerTeam/Meeting
[15:02] <ttx> Last week minutes: https://wiki.ubuntu.com/MeetingLogs/Server/20090317
[15:03] <ttx> Does anyone have anything to add to the agenda ?
[15:04] <ttx> ... I'll suppose nobody has.
[15:04] <ivoks> :)
[15:04] <ttx> [TOPIC] Review ACTION points from previous meeting
[15:04] <ttx> kirkland to write a blog post asking for testing of kvm 84 backport to hardy with specific instructions on how-to setup kvm 84 from the ubuntu-virt PPA
[15:04] <kirkland> ttx: done.
[15:04] <ttx> yes, I saw that one.
[15:04] <kirkland> ttx: http://blog.dustinkirkland.com/2009/03/ubuntu-server-kvm-call-for-testing.html
[15:05] <kirkland> ttx: some feedback
[15:05] <kirkland> ttx: it appears that this might require backporting libvirt as well, possibly
[15:06] <kirkland> ttx: anything more you want to know about this?
[15:06] <ttx> kirkland: the libvirt backport is to fix additional bugs ? Or for kvm-84 compatibility ?
[15:06] <kirkland> ttx: one user reported that some of his vm's would not boot, until he installed a backported libvirt as well
[15:06] <kirkland> ttx: i haven't investigated the issue yet
[15:07] <ttx> ok.
[15:07] <ivoks> and on still doesn't boot
[15:07] <ivoks> one
[15:08] <ttx> anything else to add on the KVM topic ?
[15:08] <kirkland> ttx: i don't think so ...  maybe more in the coming weeks
[15:08] <ttx> [TOPIC] Other progress reports on Roadmap items
[15:08] <kirkland> ttx: please keep testing kvm-84 on hardy, if you have the ability to do so!
[15:09] <ttx> anyone has anything to report on the https://wiki.ubuntu.com/ServerTeam/Roadmap items ?
[15:10] <sommer> there's new sections on eucalyptus and opennebula in the serverguide :)
[15:10] <kirkland> \o/
[15:10] <sommer> reviews welcome
[15:10]  * ttx has a look
[15:11] <ttx> sommer: i'll be in contact with you this week for adaptations to the Likewise-Open chapter to cover the likewise-open5 option
[15:12] <sommer> ttx: cool, but fyi SF is the 26th
[15:12] <ttx> sommer: that I'll be in contact with you... tomorrow ?
[15:12] <sommer> ttx: coolio :)
[15:13] <ttx> that brings us to another milestone...
[15:13] <ttx> [TOPIC] Beta release coming up
[15:14] <ttx> We should be busy with ISO testing soon. We need all the help we can to test on various hardware
[15:14] <twitzel> Does anybody have time to discuss multipath/udev issues ?
[15:14]  * sommer should be able to help with ISO testing
[15:15] <twitzel> on 8,10
[15:15] <ttx> so if you have some free time and some server box lying around, please give it a shot
[15:15] <ttx> http://iso.qa.ubuntu.com/
[15:15] <ivoks> ok
[15:16]  * lamont considers again _not_ highlighting on 'postfix'
[15:16] <_ruben> hehe
[15:17] <ttx> We usually test a lot in VMs but not so much on real iron.
[15:17] <ttx> [TOPIC] Open Discussion
[15:18] <ttx> Anything/anyone ? Or else i'll break the shortest meeting in history record.
[15:19] <ttx> Any regression that you've seen ignored and you would like to bump ?
[15:20] <ttx> OK, we enter the final month so everyone is already busy, I see :)
[15:21] <ivoks> :)
[15:21] <zul> just a tad
[15:21]  * sommer is sleepy
[15:21] <ttx> [TOPIC] Agree on next meeting date and time
[15:22] <ttx> so I think Europe is catching up DST this next week-end
[15:22] <kirkland> woohoo
[15:22] <AlexC_> soren, just want to say thank you very much for your help over the past day, been highly appreciated. I have it all up and working now =)
[15:23] <ttx> Should we adjust the meeting time accordingly
[15:25] <sommer> sure
[15:26]  * ttx doesn't remember the usual meeting time
[15:26] <bittin_> 15:00
[15:26] <bittin_> and i tought i was to late :p
[15:26] <ttx> then we just keep 1500 UTC, right ?
[15:27] <sommer> works for me
[15:27] <ttx> OK, and we should get back our slot on #ubuntu-meeting
[15:28] <ttx> so next week, 1500 UTC, #ubuntu-meeting
[15:28] <bittin_> sounds great
[15:28] <ttx> #endmeeting, then
[15:29] <ttx> have a great week, and happy Beta ISO testing.
[15:52] <Asad2005> I have installed zfs-fuse on ubuntu hardy server but unable to export NFS from mounted zfs "/tank does not support NFS export" Is there a solution to this without compiling kernel ?
[16:10] <nealmcb> mathiaz: I missed the server team meeting, having forgetfully assumed that the info on the fridge was right (meet at 16:00 utc in u-meeting....) - how should that be updated?
[16:11] <mathiaz> nealmcb: well the fridge has been wrong for some time.
[16:11] <mathiaz> nealmcb: things should be back in order soon I think.
[16:11] <nealmcb> mathiaz: right - I just replayed my thinking from last week when I figured I'd just stay there until 16:00, but then saw the announcement there that the fridge was wrong....
[16:24] <kshbu0> hello
[16:26] <LHC> hey
[16:29] <tinjaw> Today I am going to be researching the conflicts under a default install between DHCP's dynamic updating, Bind9, and AppArmor on 8.10.
[16:30] <tinjaw> First task is to find out "who" should be the one to set best practices. Bind9 or AppArmor.
[16:31] <jdstrand> tinjaw: the default apparmor configuration is supposed to support standard application configurations
[16:31] <kshbu0> something more elementary... Why defaults a server install to dhcp?
[16:32] <Jeeves_> kshbu0: It does in the installer
[16:32] <tinjaw> kshbu0: not default install of server, but of dhcpd
[16:32] <Jeeves_> you can always 'go back' and do manual configuratino
[16:32] <kshbu0> That is what i mean: the installer should give me the choice.
[16:33] <tinjaw> jdstrand: understand, but default install of Bind9 makes directories in /etc/bind not /var/lib/bind
[16:33] <kirkland> mathiaz: sorry to nag, but did you finish the raid1 test yet?
[16:34] <jdstrand> tinjaw:   /var/lib/bind/** rw,
[16:34] <jdstrand>   /var/lib/bind/ rw,
[16:34] <jdstrand> tinjaw: already there
[16:34] <mathiaz> kirkland: doing it now.
[16:34] <twitzel> sorry, to repeat myself. Can someone please help me with a multipath problem in 8.10 ?
[16:35] <jdstrand> tinjaw: that is a standard configuration
[16:35] <jdstrand> (though not default)
[16:35] <tinjaw> first research question I am googling now: should I just move /etc/bind wholesale to /var/lib/bind? and if so, why isn't that the default install config.
[16:35] <jdstrand> /usr/share/doc/bind9/README.Debian.gz
[16:37] <twitzel> Specifically its a multipath/udev problem
[16:37] <tinjaw> jdstrand: understood. trying to determine if default package install of Bind9 under U8.10 should be changed via bug report.
[16:37] <kshbu0> bind9 same story. I've always thought that installation of bind should be in a chrooted jail.
[16:37] <jdstrand> the chroot breaks things and with apparmor is is not needed
[16:37] <kshbu0> Why is this not standard procedure on UB-server? Is Debian not doing this either?
[16:37] <kshbu0> OK. Coming from Mandriva yesterday, I've no complete picture of AppArmor
[16:39] <kshbu0> Going thru a steep learning curve... I have been running Mandriva for years...
[16:39] <ahasenack> I asked about the bind thing in the ubuntu server mailing list, you may find it in the archives
[16:39] <kshbu0> now moving to Ubuntu on both laptop,desktop and server.
[16:39] <ahasenack> I always thought it weird that postfix is chrooted by default, but not bind
[16:40] <tinjaw> ahasenack: thx, searching now
[16:40] <ahasenack> http://osdir.com/ml/linux.ubuntu.server/2008-07/msg00037.html
[16:40] <kshbu0> will hav a look in the archives. Posted the same question yesterday.
[16:40] <mathiaz> kirkland: still not working.
[16:40] <mathiaz> kirkland: I'm dropped to a busybox shell when I choose to boot from a degraded RAID1 array.
[16:40] <ahasenack> kshbu0: ^^^
[16:41] <kirkland> mathiaz: holy smokes
[16:41] <kirkland> mathiaz: um, okay, we have diverged
[16:41] <lamont> ahasenack: that's because bind had an installed base
[16:41] <kirkland> mathiaz: i'm booting fine
[16:41] <lamont> (before I inherited it)
[16:41] <ahasenack> lamont: you mean, it was too late to chroot it
[16:42] <kshbu0> According to the latest server manual/faq dropping into busybox should not happen if you don't wantthat.
[16:42] <kirkland> mathiaz: how big is your .img file?
[16:42] <kirkland> mathiaz: just one disk
[16:42] <Deeps> kshbu0: they're testing jaunty i believe
[16:42] <mathiaz> kirkland: haaaaa
[16:42] <kirkland> mathiaz: ?
[16:42] <mathiaz> kirkland: which drive do you disable?
[16:43] <mathiaz> kirkland: the first one or the second one?
[16:43] <kirkland> mathiaz: the 2nd
[16:43] <kirkland> mathiaz: you?
[16:43] <mathiaz> kirkland: If I disable the 1st one it fails to boot
[16:43] <kirkland> mathiaz: i boot without the 2nd one
[16:43] <mathiaz> kirkland: however disabling the 2nd one make the system boot
[16:43] <kirkland> mathiaz: oh
[16:43] <kirkland> mathiaz: let me try that
[16:43] <lamont> ahasenack: I mean that I haven't bothered to start chrooting new installs of it, while leaving the installed base alone, nor have I given much thought to converting the installed base
[16:44] <lamont> ahasenack: especially since everyone and their brother does it differently already
[16:44] <Deeps> kirkland / mathiaz: bios configuration set to only boot from first HD, and not looking to 2nd/3rd/etc.?
[16:44] <mathiaz> Deeps: it's in kvm.
[16:44] <Deeps> hah, nm then :)
[16:44] <mathiaz> kirkland: also - I'm using libvirt.
[16:44] <LHC> anyone know about webhosts
[16:45] <lamont> ahasenack: there is at least one bug in the debian bts about chrooting bind9, I'd be happy with a patch that did it in a manner that (1) doesn't break the installed base and (2) is sane.
[16:45] <mathiaz> kirkland: when I state disable the first disk, I meant I comment out the disk definition in the guest xml definition.
[16:45] <kirkland> mathiaz: i either launch with:
[16:45] <kirkland> kvm -hda foo1.img -hdb foo2.img
[16:45] <kirkland> or
[16:45] <kirkland> kvm -hda foo1.img
[16:47] <kshbu0> its nearing 18:00 here. I'm going dowstais o eat some thing.. Bye
[16:48] <mathiaz> kirkland: http://paste.ubuntu.com/136859/
[16:49] <mathiaz> kirkland: ^ these are the two kvm command lines launched by libvirt
[16:50] <viezerd> http://paste.ubuntu.com/136858/  <<<-- dmesg output: does this mean that my ethernet-card is broken ?
[16:50] <mathiaz> kirkland: hm - so I think what this does is that the second disk is actually switched in place of the first one when I uncomment the first disk
[16:50] <viezerd> Because when there is heavy networkload, I loose connection
[16:55] <viezerd> earlier this week I found this --> http://bugzilla.kernel.org/show_bug.cgi?id=12500 with a 2.6.28.1 kernel, So I asked my hoster for a different kernel but I got the same problem now
[16:59] <Asad2005> I have installed zfs-fuse on ubuntu hardy server but unable to export NFS from mounted zfs "/tank does not support NFS export" Is there a solution to this without compiling kernel ?
[17:00] <kirkland> mathiaz: do you have a screen shot of your error again?
[17:00] <kirkland> mathiaz: i want to check something ...
[17:04] <kirkland> mathiaz: and can you pastebin the output of `sudo fdisk -l` ?
[17:05] <mathiaz> kirkland: hold on - I need to be able to reproduce this.
[17:05] <tinjaw> I'm wondering if Bind9 should run chrooted and under AppArmor. Maybe that is overkill, but wouldn't be anymore work that only using one.
[17:06] <kirkland> mathiaz: btw ... acpid is on today's iso's ;-)
[17:09] <mathiaz> kirkland: yop - o3
[17:10] <kirkland> o3 = ?
[17:10] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid1_boot_failure_1.png
[17:10] <mathiaz> kirkland: it looks like a high-five to me - but nobody else seems to agree with me.
[17:11] <kirkland> mathiaz: looks like a little head and big boobs
[17:11] <kirkland> mathiaz: o/*\o  is a high five
[17:11] <mathiaz> kirkland: right - but that involves two people.
[17:12] <Deeps> lol
[17:12] <mathiaz> kirkland: my message is that I want to high-five you - whether you answer or not is not in my realm of control
[17:12] <kirkland> mathiaz: and how do i respond, should i choose to accept your high five?
[17:12] <mathiaz> kirkland: well - you high-five me back
[17:13] <mathiaz> kirkland: like it real life
[17:13] <mathiaz> kirkland: o/*
[17:13] <kirkland> o3
[17:13] <mathiaz> kirkland: o3
[17:13] <kirkland> mathiaz:   *\o
[17:13] <dantalizing> that looks like a hobo with a sack
[17:14] <dantalizing> is that jazz hands going on there?
[17:14] <mathiaz> kirkland: jdstrand has another way to high-five in the online world - but it takes two lines
[17:14] <kirkland> pom poms
[17:14] <jdstrand>  5
[17:14] <jdstrand> o/
[17:14] <ttx>   \o
[17:15] <mathiaz> ttx: you're just raising your arm
[17:15] <jdstrand> (I use an alias though, so I only type one line)
[17:15] <mathiaz> ttx: you're *not* high-fiving
[17:15] <ttx> I hardly reach jdstrand shoulder.
[17:15] <Deeps> <o/
[17:15]  * jdstrand waves to kees
[17:16] <kirkland> mathiaz: okay, so i've uploaded a new image to http://people.ubuntu.com/~kirkland/jaunty-server1.img
[17:16] <kirkland> mathiaz: this one does exhibit the problem
[17:16] <kirkland> mathiaz: 05cb71e60a534c5ff1a00c5d12d77b7d  /tmp/jaunty-server1.img
[17:16] <kirkland> mathiaz: i've passed it on to apw on the kernel side
[17:16] <mathiaz> jdstrand: ^5
[17:17] <mathiaz> jdstrand: that's a onliner for high-fiving people in the online world
[17:17] <mathiaz> kirkland: ^5
[17:17] <jdstrand> mathiaz: it doesn't have near the *pizazz* though ;)
[17:18] <mathiaz> kirkland: have you been able to reproduce the bug?
[17:18]  * kirkland high fives mathiaz...yeah!
[17:18] <mathiaz> \o/
[17:18] <kirkland> mathiaz: yes, and no.  i'm thinking this is non-deterministic
[17:18] <kirkland> mathiaz: one install this morning ...  no issues booting degraded raid
[17:18] <kirkland> mathiaz: second install, issues
[17:18] <kirkland> mathiaz: which is the image i uploaded to p.u.c
[17:19] <kirkland> mathiaz: non-determinism is the only explanation for the difference in behavior that you and i are seeing
[17:19] <incorrect> I am debating about backporting kernel 2.6.28 to hardy,
[17:19] <mathiaz> kirkland: right. I've seen one instance where the system booted while I expected not to.
[17:19] <kirkland> mathiaz: i think the upgrades of the mdadm userspace were red herrings
[17:19] <mathiaz> kirkland: but I can reproduce it 90% of the times
[17:20] <kirkland> mathiaz: i think it was "working" for other reasons in some of those cases
[17:20] <kirkland> mathiaz: agreed.
[17:23] <theunixgeek> Does Ubuntu Server run well under a virtual machine?
[17:23] <stickystyle> mathiaz: In bug 345957 you commented that _netdev is not honored in jaunty, I'm a bit confused as to how then iscsi targets are supposed to get mounted when networking has not been configured at that point (I'm the bug reporter).  Is it that I'm doing something wrong?
[17:23] <stickystyle> theunixgeek: Ubuntu Server works great as a virtual machine.
[17:23] <tinjaw> theunixgeek: umm, yes? is that a trick question?
[17:24] <stickystyle> theunixgeek: Ubuntu JeOS works even better http://www.ubuntu.com/products/whatisubuntu/serveredition/jeos
[17:24]  * kees waves to jdstrand 
[17:24] <kees>      5
[17:24] <kees>      \o
[17:24] <sbeattie> kirkland: in your attempts to fix raid degraded mode, can you keep an eye towards fixing bug 316670 as well, as raid10 is just broken in jaunty.
[17:25] <kirkland> sbeattie: thanks for the reminder, yeah, i've been planning on moving on from this one, to that one at some point
[17:25] <kirkland> sbeattie: i keep trying to put this one to bed
[17:27] <sbeattie> kirkland: thanks, appreciated.
[17:27] <kirkland> sbeattie: so you've reproduced that one?
[17:27] <theunixgeek> tinjaw: just wanted to make sure ;)
[17:28] <theunixgeek> stickystyle: ah, thanks
[17:28] <theunixgeek> that does seem to be better
[17:28] <sbeattie> yeah, it's easy to do so; you can do it on a regular jaunty system with loopback mounts. (I put a recipe in the description of that bug)
[17:28] <stickystyle> theunixgeek: no problem.
[17:28] <kirkland> sbeattie: thanks
[17:28] <theunixgeek> what "unecessary drivers" are left out of JeOS?
[17:31] <stickystyle> I'm not sure of the exact list, but most of the video cards, IDE/SCSI/RAID controllers, network drivers, etc. that will never be present in a VM (since you have a fixed list of 'virtual HW') they are removed.
[17:40] <mathiaz> stickystyle: iscsi devices should be mounted when the network interfaces are brought up by udev
[17:40] <mathiaz> stickystyle: not mounted - the block devices are created when the network interfaces are brought up by udev
[17:41] <mathiaz> stickystyle: that way iscsi devices are just regular block devices (treated like local ones)
[17:41]  * theunixgeek is downloading ubuntu server
[17:42] <theunixgeek> 210 kpbs, not bad
[17:42] <mathiaz> stickystyle: this is why the iscsi init script runs at level S20 (IIRC) *before* local file systems are mounted.
[17:42] <mathiaz> stickystyle: the init script is just waiting for all automatic targets to be logged in.
[17:44] <stickystyle> mathiaz: Okay, I get what your saying now.  So I probably need another init script to actually mount the device after networking has come up correct?
[17:46] <mathiaz> stickystyle: no - you just need to put your iscsi devices in fstab as you would do for normal local block devices
[17:47]  * stickystyle is reversing his changes, perhaps he missed something when he first installed.
[17:48] <mathiaz> stickystyle: iscsi support in the installer hasn't been really tested in jaunty
[17:49] <mathiaz> stickystyle: it may not work correclty.
[17:56] <theunixgeek> Who here has experience deploying Ubuntu Server on virtual machines?
[17:56] <theunixgeek> (I prefer Fedora as a workstation OS, but Ubuntu as a server OS; this is my first time trying it on a VM)
[17:57] <theunixgeek> I know, it should be the other way around.... :P
[17:57] <stickystyle> theunixgeek: It's not the 'other way around' in this channel ;-)
[17:57] <theunixgeek> stickystyle: ah, good point
[17:57] <theunixgeek> ;)
[17:58] <theunixgeek> stickystyle: sorry; I don't mean to start flame wars of any kind
[17:58] <stickystyle> No worries, just playin with you.
[17:59] <stickystyle> Did you have any specific questions about Ubuntu in VM?
[18:06] <centaur5> Could anybody help me with forwarding VNC using iptables I've tried multiple commands I've found online and still can't get it to work?
[18:10] <twitzel> Sorry to repeat myself, don't want to bother anyone. Can someone please help me with multipath/udev ?
[18:20] <|max|> there are fr channel ? for ubuntu server ?
[18:21] <twitzel> Where do I go for help with udev/multipath in ubuntu ? Is there a mailing list ?
[18:29] <theunixgeek> |max|: je crois... attend un moment
[18:29] <theunixgeek> !fr
[18:29] <|max|> mouai :p
[18:29] <|max|> pas serveur
[18:29] <theunixgeek> |max|: essaye #ubuntu-server-fr
[18:30] <theunixgeek> |max|: je crois que les gens dans #ubuntu-fr peuvent t'aider
[18:30] <mathiaz> |max|: il n'y a pas de channel francophone dedie a ubuntu-server
[18:30] <|max|> oki
[18:31] <|max|> en fait j ai 3 carte reseaux et j aimerai gerer deux connections internet
[18:31] <|max|> et en faire que une
[18:33] <|max|> je sais que sous windows on peut le faire
[18:36] <andol> mathiaz: Noticed that you just recently marked all the Ubuntu i386 server images (jaunty rc beta) as tested. Are there any use in having me continuing test the ones I yet haven't tested?
[18:38] <mathiaz> andol: what kind of hardware are you using?
[18:38] <mathiaz> andol: well - since it's software testing it doesn't really matter.
[18:38] <andol> mathiaz: yeah, that's what I kind of figured
[18:38] <andol> (just an old dell optiflex gx260 I have standing around in the office.)
[18:38] <mathiaz> andol: more hardware testing is always welcome. Otherwise you can focus on other isos to test.
[18:39] <mathiaz> andol: right. So if you're up for more testing you can try the desktop one.
[18:39] <andol> mathiaz: Regarding the pure hardware testing the default installation should be "enough"?
[18:39] <mathiaz> andol: you could also drop by #ubuntu-testing and ask there what is the most needed.
[18:39] <mathiaz> andol: yes.
[18:40] <mathiaz> andol: default install for hardware testing is enough.
[18:40] <andol> mathiaz: I'll do that tomorrow. By then I'll also see if we have any "real" servers idle, which I can steal for some testing.
[18:41] <mathiaz> andol: awesome. Thank you very much for your help!
[18:43] <andol> Is the comment field a good place to mention on what hardware the install is tested on?
[18:43] <andol> (I'll save the rest of my potential questions for #ubuntu-testing :)
[18:43] <scope006> has anyone here used the tuning-primer.sh script on ubuntu server for analyzing mysql?
[18:47] <theunixgeek> Does Ubuntu Server come with everything preinstalled for installing Joomla
[18:47] <theunixgeek> ?
[18:49] <theunixgeek> let me reword that
[18:49] <theunixgeek> Does Ubuntu Server come with everything ready for installing Joomla?
[18:50] <twitzel> Can anyone see me ?
[18:51] <acicula> yup
[18:52] <andol> theunixgeek: Perhaps not preinstalled, but you should find what you need in the repositories.
[18:52] <twitzel> Okay, quick question. I found a bug in ubuntu-bugs thats virtually identical to my problem. The bug was closed with "fixed in jaunty". How do I use that ?
[18:52] <andol> theunixgeek: Doing the LAMP-install should give you most of what you need. To that you might need to install an extra php module or two.
[18:53] <acicula> twitzel: depends on your bug i suppose
[18:55] <theunixgeek> andol: cool, thanks
[18:55] <twitzel> As described here before, udev doesn't generate devices for any partitions on my multipathed device
[18:56] <theunixgeek> twitzel: if it's "fixed in jaunty" that means it's fixed in Jaunty Jackalope
[18:56] <theunixgeek> twitzel: "fixed in Jaunty" means it's already fixed for the next release
[18:56] <twitzel> Can I install this ? apt-get doesn't show anything
[18:57] <twitzel> Basically I'm  stuck now because I had promoted the switch from RHEL to ubuntu server and now 2x12TB of storage have been down since 2 weeks since we switched
[18:58] <twitzel> So I really need a solution asap, and if this "jaunty" thing fixes it, I'd like to use it
[18:58] <acicula> jaunty is still alpha, beta in a few days
[18:58] <twitzel> So going back to RHEL is the best option right now ?
[19:00] <twitzel> I opened a new bug
[19:00] <twitzel> Maybe I can argue there that jaunty doesn't help if its not available
[19:01] <acicula> if the problem is backported(able) into intrepid maybe you can do that, but if it's a new feature it seems you just have to wait for jaunty to get stable, or run the alpha/beta or go back to what worked before
[19:02] <twitzel> I believe its basically just the udev rules files and their order
[19:03] <twitzel> no binary stuff at all
[19:05] <twitzel> How do I merge from debian unstable ?
[19:05] <twitzel> Actually found this: This bug was fixed in the package multipath-tools - 0.4.8-14ubuntu1
[19:05] <twitzel> ---------------
[19:05] <twitzel> multipath-tools (0.4.8-14ubuntu1) jaunty; urgency=low
[19:05] <twitzel>   * Merge from debian unstable, remaining changes:
[19:05] <twitzel>     - control:
[19:05] <twitzel>      + Bump debhelper dependency to install udev rules to
[19:05] <twitzel>       /lib/udev/rules.d, bump udev dependencies as well.
[19:05] <twitzel>      + multipath-udeb must not depend on multipath-modules,
[19:05] <twitzel>       because the Ubuntu kernel already includes the modules and no
[19:05] <twitzel>       package provides m-m.
[19:05] <twitzel>     - initramfs/hooks: Install rules from /lib/udev/rules.d.
[19:05] <twitzel>     - kpartx.udev, multipath-tools-boot.init:
[19:05] <twitzel>       Replace multipath-tools-boot init script with udev rules.
[19:06] <twitzel>     - multipath-tools.preinst: Fix preinst script to modprobe dm-multipath.
[19:06] <dguitar> pastebin....
[19:06] <twitzel>       This will make sure that multipathd will be able to start.
[19:06] <twitzel>     - multipath-tools.prerm: Add prerm script to not fail when upgrading
[19:06] <twitzel>       from a broken version of multipathd.
[19:06] <twitzel>     - patches/1000--set-umask-in-multipathd.patch: Set umask in multipathd.
[19:06] <twitzel>     - rules: Move udev rules to priority 95, because rules that load
[19:06] <twitzel>       modules should be >90.
[19:06] <twitzel>   * Fix bugs (LP: #316851, #320156)
[19:06] <twitzel>   * multipath-tools.prerm/preinst: Remove checks for old ubuntu
[19:06] <twitzel>     versions, we don't support upgrading from them anymore.
[19:10] <kraut> twitzel: don't know if i'll get kickeassed for that, but yes, i would use redhat or centos for that.
[19:13] <twitzel> Thats actually double sad. First that ubuntu has difficulty dealing with multipathed SANs (not sooo rare in the server world) and then that even when it works under RH etc. its a major PITA. No wonder people shell out loads of money to managed storage ala Equallogic etc.
[19:14] <racecar56> i want to use a static ip address, but i only know how to do that on desktop, what do i do to make ubuntu server have a static ip?
[19:17] <twitzel> racecar56: do you mean set it without using the GUI ?
[19:17] <twitzel> You ought to change the file /etc/network/interfaces
[19:18] <twitzel> racecar56: man interfaces will tell you all you need to know about that
[19:54] <yeason1> this is a noob question but I can't remember... where are the kernel source files stored?
[19:55] <_ruben> /usr/src/
[19:55] <yeason1> ah, thnx
[20:14] <jurism> Hi! I have problems with restarting apache2. It says something about unable to open logs. I am 90% sure it is php5-ffmpeg problem, but I need this extension very much. Can You tell me where should I add 'pkill -9 apache2' in /etc/init.d/apache2 so server could restart with no problems? Thank You!
[20:21] <racecar-56> this place is silent
[20:23] <oruwork> can i get apache to host 2 separate websites on one ubuntu server box ? i mean set it up the way that each website would have its own www directory
[20:24] <Vog-work> yes oruwork
[20:24] <Vog-work> I have to go so I can't help with it though
[20:28] <macd> oruwork, should be pretty self explanitory, look in /etc/apache2/sites-available/ glance at default, use it for a template to make a new one, then you can "load" the sitre with "sudo a2ensite whatyoucalledit"
[20:28] <macd> oruwork, refer to: http://trent.instantspot.com/blog/2009/02/11/Configuring-Apache-to-run-virtual-hosts-on-Ubuntu
[20:39] <Elite> Hey guys
[21:53] <nat2610> how do you know if selinux is activated or not on a system ? there isn't any command setenforce ?
[22:51] <nat2610> hey I'd need a little bit of help. I'm trying to install bind9 packages, it worked in the past, them I messed up some stuff and now when I try to reinstall it I get dpkg: error processing bind9 (--configure): subprocess post-installation script returned error exit status 100 and looking at my messages log I have Mar 24 15:46:29 nat32 kernel: [  369.095590] type=1505 audit(1237934789.179:11): operation="profile_replac
[22:51] <nat2610> e" name="/usr/sbin/named" name2="default" pid=14034
[22:51] <nat2610> I don't know what to do