[00:00] <donspaulding> stgraber: ok, that all worked, succesfully resized and turned on the swap.
[00:01] <donspaulding> stgraber: is there an easy command to "grow" my root partition to all available free space?
[00:02] <stgraber> great, then you can try: lvextend -l +100%FREE files/root
[00:04] <donspaulding> stgraber: ok, it says it was successfully resized.
[00:04] <stgraber> if it works, then you can do: resize2fs /dev/mapper/files-root
[00:04] <stgraber> that'll make your root filesystem grow to use that newly available space
[00:05] <donspaulding> it says it's currently mounted, so it's doing an on-line resizing…. awesome.
[00:05] <stgraber> yeah, I only did it once but it's very interesting :)
[00:05] <stgraber> if you check: df -h
[00:05] <stgraber> you'll actually see it grow
[00:06] <donspaulding> hah!
[00:06] <donspaulding> that's freaking sweet.
[00:06] <donspaulding> ok, so once that's done, I'm golden right?  It doesn't even look like I need to reboot.
[00:07] <stgraber> yep, everything should work just fine after that
[00:07] <donspaulding> great, thanks for the help stgraber.
[00:07] <stgraber> np
[00:10] <xperia> hello to all. i have problems with receiving mails on my postfix server. the mail was send with thunderbird for some reason sucessfull but
[00:10] <xperia> when i do "cat /var/log/mail.log | tail" there is no info that a mail was received or handled.
[00:10] <xperia> can somebody maybe help me with finding out the problem ?
[00:11] <alex_joni> did you try to send one by hand?
[00:11] <alex_joni> telnet server 25
[00:11] <alex_joni> HELO whatever.com
[00:11] <alex_joni> MAIL FROM:<mail@mail.com>
[00:12] <alex_joni> RCPT TO:<dest@email.com>
[00:12] <alex_joni> DATA
[00:12] <alex_joni> Test
[00:12] <alex_joni> .
[00:12] <xperia> on the server where postfix run i am able to send emails with no problems. but when i try to receive a mail on my server it fails
[00:12] <alex_joni> try the above from a box that is on the outside
[00:12] <I-Blocklist069> hi, using... find /home/samba/backups/ -name "*.txt" -type f -mtime +30 -exec gzip {} \; which .gz's all the .txt files, how can i zip all the txt files individually?
[00:12] <alex_joni> maybe you're blocking 25
[00:13] <xperia> alex_joni: hmmm will try
[00:14] <qman__> I-Blocklist069, do you mean using zip instead of gzip?
[00:15] <I-Blocklist069> yea
[00:15] <I-Blocklist069> i tired just zip instead of gzip but it thinks im trying to do .zip files
[00:15] <I-Blocklist069> insted of zipping each .txt file
[00:16] <qman__> zip needs a target name
[00:16] <qman__> err, archive name
[00:16] <I-Blocklist069> anyway of making it just do it the same as the input name but with .zip at the end
[00:16] <qman__> gzip defaults to just compressing the given file, but zip doesn't
[00:16] <I-Blocklist069> i.e. like gzip does
[00:16] <qman__> I don't know if/how to do that, but reading the manual should help
[00:17] <I-Blocklist069> i tried -@ on the zip command but that didnt do anything
[00:17] <alex_joni> zip - doesn't
[00:18] <xperia> alex_joni: really great howto. i am getting this error after the RCPT TO:<dest@email.com> "554 5.7.1 <root@myserver.com>: Relay access denied"
[00:18] <alex_joni> it takes the file from input
[00:18] <qman__> -@ is not relevant to what you want
[00:18] <alex_joni> you can also use -print from find
[00:18] <alex_joni> to pass the filename to zip
[00:18] <qman__> yeah, if you just pass the filename twice, it will work
[00:18] <alex_joni> xperia: sounds like config issues ;)
[00:20] <xperia> alex_joni: yeahh this configs of the software drives me crazy :-)
[00:21] <alex_joni> xperia: the 4th time or so will be easier
[00:22] <alex_joni> good night & good luck
[00:22] <xperia> okay thanks a lot for the debugging of the problem
[00:22] <xperia> and good night alex_joni
[00:26] <henriquev> On Amazon Ec2 I launched an Ubuntu instance, played with it a little and now I'd like to save my own image. How do I do that?
[00:29] <erichammond> henriquev: Here are some notes I wrote on the subject: http://alestic.com/2009/06/ec2-ami-bundle
[00:29] <henriquev> erichammond: thank you
[00:31] <erichammond> There are a lot of other tutorials if you do a search.  Here's Amazon's: http://ec2gsg-creating.notlong.com
[00:41] <I-Blocklist069> any her einstalled xz-utils ?
[01:01] <UnixDawg> hey guys having a issue
[01:02] <UnixDawg> we installed ubuntu 9.10 server
[01:02] <UnixDawg> and did apt-get update upgrade
[01:02] <UnixDawg> then we tru ti install apache and deps fail
[01:03] <UnixDawg> is there a step we are missing ?
[01:04] <unit3> nope, that should work fine.
[01:04] <erichammond> UnixDawg: This should work: apt-get update && apt-get upgrade && apt-get install apache2
[01:05] <erichammond> What are you doing differently?
[01:05] <UnixDawg> not working
[01:05] <unit3> can you pastebin the error?
[01:05] <UnixDawg> it says it has unmet deps
[01:05] <unit3> the full error, including what's unmet.
[01:05] <unit3> into a pastebin, plz.
[01:06] <UnixDawg> http://pastebin.com/m281ed8af
[01:07] <unit3> can you pastebin your /etc/apt/sources.list?
[01:07] <valindil89> can someone help me setup apache, mysql, and php?
[01:08] <unit3> valindil89: can you be more specific about what you need help with?
[01:08] <UnixDawg> http://pastebin.com/m198a8fa8
[01:08] <valindil89> unit3, I just installed ubuntu  and I need php, apache, and mysql to be able to run on the computer.
[01:09] <unit3> UnixDawg: there's your problem, you only have the security repo. You should also have the normal karmic repo, and the karmic-updates repo enabled.
[01:09] <UnixDawg> ok what lines
[01:10] <unit3> UnixDawg: let me make you a more condensed sources.list, I'll pastebin it in a second.
[01:10] <UnixDawg> ok
[01:11] <valindil89> unit3, I guess I need help installing those.
[01:11] <unit3> UnixDawg: http://pastebin.com/m42bbf103
[01:11] <unit3> valindil89: so, you're having problems with apt-get, then?
[01:12] <unit3> UnixDawg: note that I changed the PPA line there too, since for some reason you had it pulling packages for hardy. it may be that that PPA isn't even needed on karmic, it might already have the drivers in there.
[01:12] <valindil89> unit3, I tried to figure out how to do it. But I just cant find what to do. I dont know what apt-get is.. I just started using ubuntu today.
[01:12] <valindil89> unit3, I am a newb in linux.
[01:13] <unit3> valindil89: and you decided to jump right into configured a web and database server, hey? well... that's quite the plunge into the deep end, good luck.
[01:13] <unit3> you may wish to start on the Ubuntu wiki then.
[01:13] <unit3> https://wiki.ubuntu.com/
[01:13] <UnixDawg> ok
[01:13] <unit3> and the official ubuntu docs site: https://help.ubuntu.com/
[01:14] <unit3> valindil89: use those to get aquainted with Ubuntu, and then once you're more familiar with how things work, see if the docs on there about configuring the server software make sense.
[01:14] <unit3> if not, come back and ask some specific questions, and we'd be glad to help. ;)
[01:14] <unit3> UnixDawg: let me know if that works out for you, you shouldn't have any problems once you use that sources.list. :)
[01:17] <UnixDawg> thanks
[01:17] <UnixDawg> thats working
[01:17] <unit3> awesome. :)
[01:18] <qman__> valindil89, sudo tasksel install lamp
[01:18] <qman__> but you definitely need to read up on it before you run a production server
[01:18] <unit3> alright, this day is finally over. time to get home.
[01:19] <qman__> the ubuntu server guide tells you how to install most of the things you'll want, so read through that as well
[01:20] <UnixDawg> will do
[01:23] <valindil89> qman__, I am a programmer, and I am getting on an airplane tomorrow. so if I have it installed I can work on programs while not online :-)
[01:24] <valindil89> It is the only reason I run those software on my laptop
[01:27] <UnixDawg> almost  ahve freeswitch installed
[01:27] <qman__> valindil89, if it's just for private use, and not hosting to the internet, a default setup should be just fine
[01:28] <qman__> it creates a default site located at /var/www
[01:29] <valindil89> ok, then only thing I ever do is after I setup these things. Normally in windows I have my firewall block it completely from going outside my computer, Is there a way to do that in linux?
[01:30] <qman__> valindil89, yes, but it would be easier to configure apache to listen on localhost only
[01:31] <valindil89> ok
[01:31] <valindil89> brb
[01:31] <qman__> valindil89, see /etc/apache2/ports.conf
[01:31] <qman__> change Listen 80 to Listen 127.0.0.1:80
[01:32] <qman__> there are other ways to do it too
[01:33] <valindil89> that file path is not there
[01:34] <qman__> it must have changed, that was on a hardy server
[01:34] <qman__> grep -R Listen /etc/apache2
[01:34] <qman__> that will show you where the Listen directive is
[01:35] <valindil89> apache2 wasn't there. so I am goign to install them separately.. for some reason lamp didn't install when I did that command. I am getting it now.
[01:38] <UnixDawg> thanks again freeswitch apache and all needed apps installed.
[01:38] <UnixDawg> this is why I am liking ubuntu server next to using bsd  server install
[01:38] <UnixDawg> both small and easy to install
[01:41] <UnixDawg> frick I forgot the svn brb
[01:55] <valindil89> qman__, I ran this and it works like a charm... "sudo apt-get install lamp-server^"
[02:09] <UnixDawg> ok I found a issue
[02:09] <UnixDawg> it seems the apache2 install is installing a blank  httpd.conf
[02:09] <nat2610> hey, if I want a user1  (not root) to be able to do sudo su user2 (not root) what am I supposed to add to visudo ?
[02:10] <UnixDawg> it there one that can be installed ?
[02:12] <UnixDawg> ?
[02:15] <UnixDawg> anyone have a default httpd.conf ?
[02:16] <UnixDawg> is there a pkg I missed ?
[02:23] <ajmitch> UnixDawg: see apache2.conf instead
[02:24] <UnixDawg> ok
[02:25] <UnixDawg> now I need to add php and a site alias
[02:42] <UnixDawg> got it
[03:25] <pipedream> .
[04:05] <FuzzyKittens-TC> my server seems to take forever to reboot. it seems to get stuck on "saving the system clock" any tips on speeding up this process?
[04:27] <qman__> well it must not have been that big a deal, since he couldn't wait more than five minutes for an answer
[04:29] <sub> it never is, qman__
[06:21] <trimeta> Apparently apparmor and glib just had security updates. Two questions: 1. I didn't have /boot mounted when updating; I mounted it afterwards and ran update-initramfs -u, will that do all the necessary work? 2. Should I reboot to get the security updates to apply? There's nothing in USN about this update.
[06:24] <ScottK> The USN doesn't get published until after all the updates are out, so it's not unusual for them to lag a bit.
[06:24] <trimeta> I'll see if they come up in a day or two, but in the past I've had updates which seem to never have been published in USN.
[06:34] <ScottK> From -security if it's in Main, it gets a USN.
[06:35] <ScottK> In -updates there are non-security bug fixes that don't get a USN.  -security fixes get copied to -updates, so you may get them from there.
[06:35] <trimeta> I'm not exactly sure where they came from...I mean, I don't have any unusual repositories included here, but I guess both Main and Security are default.
[06:38] <ScottK> Yes
[06:39] <trimeta> Hmm, they all came from http://us.archive.ubuntu.com karmic-updates/main, so I guess that means they're not security updates?
[06:40] <ScottK> Unless they were copied from -security before you updated.
[06:40] <ScottK> You can look them up on Launchpad and see.
[06:41] <ScottK> Apparmor was not security, it was a bugfix: https://launchpad.net/ubuntu/+source/apparmor/2.3.1+1403-0ubuntu27.3
[06:43] <ScottK> Same for glib: https://launchpad.net/ubuntu/+source/glib2.0/2.22.3-0ubuntu1
[06:44] <trimeta> OK. In general, I find those by looking for the package on Launchpad and then looking for the changelogs?
[06:47] <ScottK> You can also look on packages.ubuntu.com and it can be easier because it's arranged by binary package and LP is arranged by source package.
[06:47] <ScottK> The only problem with packages.ubuntu.com is the data tends to lag a little bit.
[06:49] <twb> ScottK: only by a day or so, surely?
[06:50] <ScottK> I think so, but if you're wondering about this new update you got, that's likely a problem.
[06:50] <twb> You can also hit "C" in aptitude's GUI to fetch changelogs
[06:50] <twb> And IIRC there's a tool to show changelogs as you install packages
[06:50] <trimeta> twb: GUI? This is a headless server.
[06:50] <twb> trimeta: aptitude's GUI doesn't use X
[06:51] <trimeta> So you're referring to an ncurses-based thing, then?
[06:51] <twb> That's still a GUI
[06:51] <trimeta> Debatable, but whatever.
[06:51] <twb> There seems to be a /de facto/ retronymming to "TUI", which I don't hold with.  As you say, "whatever".
[06:54] <twb> Of course, aptitude 0.6 includes a GTK UI...
[07:20] <trimeta> Hmm, "aptitude changelog <packagename>" apparently gives me the changelog, though the interface is more like less than it is an ncurses-based interface. Still, that's exactly what I was looking for, and I'll try to remember it in the future.
[07:31] <twb> There's a perl library that it (can) use to make it put the newer changelog entries in bold, too
[07:31] <twb> Oh, but "aptitude changelog" uses $PAGER instead of the internal aptitude pager.
[07:31] <twb> So it probably *is* using less, unless you've customized your pager
[07:33] <trimeta> Which is fine by me; I like less.
[07:34] <twb> trimeta: less is ncurses-based, too... :-)
[07:34] <trimeta> I guess in the sense that it's taking over the whole screen rather than showing stuff line-by-line, but the interactivity consists of scrolling up and down and searching for stuff. Which is nice, but I certainly wouldn't call it a GUI.
[07:35] <trimeta> I think of make menuconfig as being the quintessential ncurses-based program.
[07:35] <trimeta> Or maybe the installer on the ubuntu-server CD. That level of interactivity.
[07:37] <twb> That's the debian-installer, FYI.
[07:38] <twb> It also has a GTK version, but I don't think Ubuntu provides it
[07:39] <trimeta> The closest I've come to using Debian (aside from Ubuntu, of course) is using Sidux...I don't remember what that installer was like, though.
[07:52] <alkisg> Hi, I found the Ubuntu server guide at 2 different pages, and their contents are not syncronized. Which one would be the correct one?
[07:52] <alkisg> https://help.ubuntu.com/9.10/serverguide/C/index.html
[07:52] <alkisg> http://doc.ubuntu.com/ubuntu/serverguide/C/
[07:53] <twb> Well, obviously the former applies to 9.10
[07:53] <twb> Perhaps the latter applies to ubuntu+1?
[07:53] <trimeta> The latter has a large Draft watermark, though I'm not sure what that means in this context.
[07:54] <alkisg> They're both for 9.10
[07:54] <alkisg> And they're quite similar, but not exactly syncronized
[07:55] <twb> alkisg: the best version is the one you get via apt-get, I think
[07:55] <twb> Since that should be the latest stable build that applies to your current installed version
[07:55]  * trimeta reads the package management section of that document
[07:55] <alkisg> Heh, right :) I just thought I'd notify you people in case one of them needed to be taken off.
[07:55] <trimeta> Wow, I never knew that just typing "sudo aptitude" launched an ncurses-based interface.
[07:55] <twb> alkisg: notification should probably be done as a LP bug report
[07:56] <twb> alkisg: most of us aren't Ubuntu developers
[07:56] <alkisg> I'm not sure if that's a bug, so how can I file one?
[07:56] <alkisg> I tried asking in #ubuntu-doc, but I got no answer yet...
[07:56] <twb> IIRC, "apport <package name>"
[07:56] <twb> Yeah, #ubuntu-doc is usually quiet
[07:56] <twb> Developers hate writing documentation :-)
[07:56] <ewook> didn't know they knew how to :P
[07:57] <alkisg> Heh.... sure but I'd expect people in ubuntu-doc *not* to be devs :D
[07:57] <alkisg> Anyway thank you guys, /me goes back to trying LDAP for the first time...
[07:58] <twb> alkisg: which release are you doing that on?
[07:58] <alkisg> I'm on 9.10
[07:59] <alkisg> sudo apt-get install ubuntu-serverguide :)
[07:59] <twb> Righto.
[08:01] <alkisg> Bah.. that's the same as the first link, and it misses the "configuring ldap backend" part :(
[08:03] <twb> OK, so what makes you think the newer link applies to 9.10?
[08:04] <alkisg> Because it writes 9.10 on top
[08:04] <twb> Fair enough
[08:04] <alkisg> "Ubuntu Documentation > Ubuntu 9.10 > Ubuntu Server Guide"
[08:04] <twb> The "DRAFT" in CSS makes me suspect it's an unreleased version or something
[08:05] <twb> And since 9.10 has been released, I was (am?) assuming that it can only apply to the next release
[08:05] <twb> The sources are available in a (bzr?) repo of docbook files, if you want the utter bleeding edge
[08:05] <alkisg> Maybe it's how they prepare the next server docs... but then it shouldn't be in that URL imho
[08:06] <alkisg> Nah, I just thought I'd ask, no need for such extreme measures :)
[08:28] <tarsman> what is the solution for not detected NIC RTL8102EL for 8.04? thanks.
[08:29] <arj> loading a driver
[08:32] <jmarsden> tarsman: Is any of the detail in bug #326891 or bug #240470 relevant to your issue?
[08:36] <tarsman> jmarsden: I'm not sure. I have not read bugs #326891 or bug #240470. I'll check it.
[08:40] <tarsman> thank you guys for your help. I'll check those links you gave.
[08:44] <twb> Huh, there's a udeb for qnx filesystems
[09:16] <alvin> Will the new version of mountall be released for Jaunty?
[09:18] <alvin> It would be nice to be able to boot ubuntu-server without worries, like the previous versions.
[09:19] <twb> Unless it fixes a critical bug or security vulnerability, new stuff is not usually backported to existing releases.
[09:20] <alvin> I'd say booting is pretty critical
[09:20] <twb> Depends on how widespread it is
[09:21] <alvin> On all my servers (about 10), I see the same issues. Of course, these all use NFS. If you don't use NFS mounts, booting probably works
[09:21] <twb> alvin: is NM installed?
[09:21] <twb> I have found that NM doesn't play nice with NIS/NFS
[09:21] <alvin> What is NM?
[09:22] <twb> NetworkManager
[09:22] <alvin> There is an other issue, but I've only seen it on 5 servers: bug 460914 This one actually bugs me most
[09:22] <twb> It's installed by default if you don't use the server media to do the install
[09:22] <alvin> Ah, no, it isn't
[09:22] <alvin> All those are just default installes. CD installed, and sometimes ubuntu-vm-server (and nfs-common). That's it.
[09:23] <twb> These are virtual machines?
[09:23] <alvin> No, they aren't
[09:23] <alvin> but sometimes they are virtual hosts
[09:24] <alvin> Virtual machines have the same problems booting wen NFS is used. Having /home on NFS is tricky now
[09:24] <alvin> (This all worked very well on previous Ubuntu versions.)
[09:24] <alvin> I did use the ubuntu-server CD. No network-manager.
[09:25] <twb> I don't know what's going on there.  I use 8.04, with NFS and md RAID and LVM, and it doesn't exhibit behaviour like that unless the array is degraded (i.e. if I remove on of the nodes in the array).
[09:25] <twb> s/on/one/
[09:25] <alvin> Problem is now: I rebooted my remote headless fileserver and bug 460914 is probably active now, so I can't do anything about it. I ordered a KVM switch for that
[09:26] <alvin> Ubuntu 8.04 is working fine with NFS mounts
[09:29] <alvin> I posted a message on the list with my experiences, but no reactions so far: https://lists.ubuntu.com/archives/ubuntu-server/2009-December/003574.html I can't possibly be the only one with all those problems unless I really do the installations wrong. (I doubt that. It isn't so hard after all.)
[09:31] <alvin> (ok, I did get one personal reaction with congratulations. At least one other person must have the same experiences.)
[09:33] <alvin> RedHat doesn't have those problems, but I know Ubuntu far better and I'd rather aid in making Ubuntu work than learning another distribution. There IS being worked on most of those bugs, but Lucid is a bit far off.
[09:40] <twb> You could be the only one if something obscure is happening, e.g. your network is negotiating 10baseT and thus it's taking much longer than normal to bring up NFS
[09:41] <alvin> Then why is it working for Solaris and all versions of Ubuntu, except karmic? No, I'm sure there are a whole list of new bugs that where introduced with upstart/mountall/grub2
[09:42] <alvin> Karmic beta wasn't booting at all. The final version was, but there are problems on every server. I fear for every reboot.
[09:43] <alvin> A strange thing is: it looks like (difficult to prove) karmic has more problems finding the network 'in time'.
[09:45] <alvin> There are still Jaunty machines in this network, and they boot a lot faster. They brin up the network, mount the NFS shares and finish bootiing. Karmic tries to brung up the network, then tries to find the NFS server, doesn't find it, throws an error, retries, goes on, might mount the drive, (might not), gives a prompt. You log in and... stil no network. If you wait a bit, the network will come up. The whole process takes a lot longer than Jaunty.
[09:46] <tenach> What would be different about using Unreal instead of ircd-hybrid or any other ircd?
[09:47] <alvin> On top of that (most of the time, NFS shares do get mounted, but never without errors), the 'not-finding-the-root-drive' from time to time worries me the most. Jaunty could always find root.
[09:55] <twb> tenach: they all have different hacks on them
[09:56] <twb> tenach: for private use, and old ircd will suffice
[09:57] <tenach> Alright.
[09:57] <tenach> Thanks twb.
[09:57] <tenach> I think I'll just go with the ircd-hybrid, as it's the one that IrcServer in Ubuntu help focuses on
[10:04] <imagodei> I have 9.04 Server, I manually configured DHCP on two interfaces. One is plugged to network, the other not. I get this message (reloading /etc/samba/smb.conf smbd only) every 5 minutes or so. Is there any way to stop smbd from reloading every time dhcp tries to get IP? Or at least decrease the interval? There was a discussion about this issue at http://ubuntuforums.org/showthread.php?t=1140094 but
[10:04] <imagodei>  no solution...
[10:05] <fahadsadah> imagodei: put interfaces = eth0 in smb.conf
[10:05] <fahadsadah> Assuming eth0 is the one that is facing the network
[10:06] <twb> imagodei: remove it from /etc/dhcp3/dhclient...hooks.d/
[10:06] <twb> fahadsadah: ah, that's a better idea
[10:07] <imagodei> fahadsadah: OK, will try, brb
[10:14] <imagodei> OK, I should say I'm pretty much n00b... I have a line in the smb.conf, says";    interfaces = 127.0.0.0/8 eth0" The connected interface is eth1
[10:15] <fahadsadah> interfaces = eth1, it should say
[10:15] <fahadsadah> Leave the 127.0.0.0/8 bit, actually
[10:15] <fahadsadah> So interfaces = 127.0.0.0/8 eth1
[10:15] <imagodei> do I ommit semicolon?
[10:16] <fahadsadah> Aye
[10:19] <imagodei> then /etc/init.d/samba restart ?
[10:19] <fahadsadah> Yes
[10:20] <imagodei> OK... no message so far at a restart
[10:20] <imagodei> just wanna wait for a few minutes :)
[10:21] <imagodei> :(
[10:22] <imagodei> no joy
[10:22] <imagodei> the message's back
[10:24] <fahadsadah> Hmm
[10:24] <fahadsadah> What are you trying to do?
[10:26] <imagodei> Generally? I want the message out of console, cause it bugs me there while typing. The link http://ubuntuforums.org/showthread.php?t=1140094 says that there are also network outages every time this happens. Don't wanna have them
[10:30] <imagodei> I don't understand... is this normal behavior? I have little experience with linux, but I never had this issue. Is this because of dynamic IP config on server?
[11:36] <frojnd> Hello there
[11:36] <frojnd> I'm trying ot install ampache
[11:37] <frojnd> I've downloaded tar.gz extracted it to /var/www/ampache/
[11:37] <frojnd> as root
[11:37] <frojnd> but now I have a problem
[11:37] <frojnd> I think something with file persmissions
[11:37] <frojnd> like websesrver wouldn't allowe ampache to write or change things
[11:37] <frojnd> :S
[11:37] <frojnd> http://90.157.178.175/ampache/test.php
[11:38] <frojnd> This attempts to read /config/ampache.cfg.php If this fails either the ampache.cfg.php is not in the correct locations or it is not currently readable by your webserver.
[11:38] <frojnd> ...it is not currently readable by your webserver
[11:38] <frojnd> what does that mean?
[11:48] <kervel> hi
[11:48] <kervel> i'd like to know how i can create a jeOS AMI with some custom stuff installed
[11:48] <kervel> i don't want to use the standard AMI, because i heard you loose all your configuration when your instance terminates
[11:49] <kervel> i searched the web for how to create custom AMI but i couldn't find anything recent on ubuntu
[12:02] <mathiaz> kervel: you'll loose all your configuration when your instance terminates no matter what
[12:03] <mathiaz> kervel: I'd suggest to look into rebundling your instance to create your own AMI
[12:03] <kervel> mathiaz: yeah, but if i have a preconfigured AMI then not
[12:03] <kervel> mathiaz: ah, i tought it was not possible to do that
[12:03] <mathiaz> kervel: start from one of the official AMI image, log into the running instance, configure it the way you want, then rebundle it
[12:04] <mathiaz> kervel: rebundled AMI are stored into your own S3 bucket and can be started whenever you want
[12:04] <kervel> ah, is it really that simple ? great
[12:05] <kervel> btw mathiaz do you know how frequently an instance gets "terminated" because of external problems ?
[12:05] <kervel> i mean amazon terminates your instance because of hardware problems / allocation conflicts or so
[12:05] <mathiaz> kervel: not really
[12:05] <mathiaz> kervel: you should plan accordingly though
[12:06] <mathiaz> kervel: ie create an infrastructure where instances can go away at any time
[12:06] <mathiaz> kervel: and make sure your overall infrastructure can cope with that
[12:06] <kervel> mathiaz: and that it auto-launches a new instance i guess when something like that happens
[12:07] <mathiaz> kervel: probably - that includes some monitoring services
[12:07] <mathiaz> kervel: amazon autoscaling feature may be helpful there
[12:08] <kervel> mathiaz: the problem is that i'm going to host a complex webapp with complex caching (not only read caching but write caching too)
[12:08] <kervel> and distributing state over different machines is impossible now
[12:08] <kervel> eg the webapp doesn't support running on 2 containers
[12:08] <mathiaz> kervel: well - some applications haven't been designed to run on a cloud
[12:09] <mathiaz> kervel: with the assumption that instances can go away anytime
[12:09] <mathiaz> kervel: these apps are not necessarly the best candidates to be run on  a cloud
[12:10] <kervel> mathiaz: i was thinking the same way.. thing is, i get these requests all the time, a lot of people see a cloud as a really big vmware
[12:10] <kervel> so i wanted to find out what it takes to get a reliable hosting on EC2
[12:10] <mathiaz> kervel: it depends how you define reliable
[12:10] <kervel> and i found out that i can have EBS block storage that doesn't go away
[12:11] <kervel> mathiaz: we offer SLA
[12:11] <mathiaz> kervel: hosting on EC2 can be very very reliable
[12:11] <kervel> mathiaz: i can imagine that if your app is loadbalanced on 5 instances it can be very reliable
[12:11] <kervel> but i wonder if i can get a "single instance setup" also reliable
[12:12] <mathiaz> kervel: well - depends how you define reliable - how much downtime can you afford
[12:12] <kervel> so my current setup ID would be: postgres database on EBS / application state on EBS / basic operating system and application code in AMI
[12:12] <mathiaz> kervel: you can restart instances in EC2 and monitor whether they go away
[12:13] <kervel> mathiaz: do you know how long it takes to start an instance ? and do you think my setup with 1 instance and state on EBS makes sense ?
[12:13] <mathiaz> kervel: it can take a few minutes
[12:13] <mathiaz> kervel: you usage of EBS makes sense
[12:13] <mathiaz> kervel: it's one of the use case for EBS.
[12:13] <kervel> and fast enough for databases ?
[12:14] <mathiaz> kervel: sure
[12:14] <kervel> hmm then i'm going to try it
[12:14] <mathiaz> kervel: you should definetaly test it before going to production
[12:15] <mathiaz> kervel: it's very application specific
[12:15] <kervel> yeah
[12:15] <mathiaz> kervel: and make sure that you web application works correclty (ie data are stored safely) on EC2.
[12:16] <mathiaz> kervel: and plan for instance termination accordingly
[12:16] <kervel> ok that seems possible
[12:16] <kervel> now one thing: i'd like to develop certain features in our webapp that makes it more suitable. eg we have a built-in media asset manager (files/ videos/ pictures) and it would be cool to use S3 for them instead of the filesystem
[12:17] <kervel> i wonder how to setup a development environment for that
[12:17] <kervel> because having to deploy my application on EC2 every time i want to do a small test will slow things down
[12:30] <kervel> anyway mathiaz thanks a lot for your time, i'll figure out i guess
[12:30] <kervel> you gave me the pointers i needed :)
[12:32] <mathiaz> kervel: for testing purposes, I'd suggest to look into UEC
[12:32] <mathiaz> kervel: which is a way to deploy an EC2-compatible cloud on your own servers
[12:33] <mathiaz> kervel: http://www.ubuntu.com/cloud/private
[12:35] <mathiaz> ttx: is there a server team meeting tomorrow?
[12:44] <kervel> just launched my instance ... can't ssh into it, even tho i selected the right key. going to try karmic instead of hardy
[12:46] <ttx> mathiaz: yes, but it should be fast
[12:46] <mathiaz> ttx: who's gonna be around? you, me, ??
[12:46] <ttx> smoser ?
[12:47] <mathiaz> ttx: I don't think so
[12:47] <ttx> hmm
[12:47] <ttx> mathiaz: let me think about it then :)
[12:48] <mathiaz> ttx: well - we could have a meeting in french!
[13:16] <mathiaz> ttx: how do you de-register an kernel/image/ramdisk?
[13:17] <ttx> mathiaz: never tried
[13:44] <Jare> can i set per user bandwidth limits in ubuntu server? For example one user would have max. 30Mbps/20Mbps and the other 5Mbps/10Mbps
[13:46] <Jeeves_> Jare: Not that I know off
[13:46] <ttx> mathiaz: so the current UEC/ISO start and runs. I fail to run an instance, I get a libvirt error.
[13:46] <ttx> mathiaz: will file bugs
[13:47] <mathiaz> ttx: ok - I'll stick to 1.6.1 then
[13:47] <mathiaz> ttx: when doing the stress testing
[13:47] <teddymills> http://manpages.ubuntu.com/manpages/karmic/man8/tc.8.html
[13:47] <ttx> mathiaz: yes
[13:48] <Jeeves_> Jare: http://www.mastershaper.org/index.php/Main_Page maybe?
[14:04] <ttx> mathiaz: err... the latest from bzr upstream is even more broken, beh
[14:24] <mathiaz> ttx: that always happen when you're in *bug* *fixing* mode
[14:24] <ttx> mathiaz: in fact it's not /more broken/.
[14:25] <ttx> it's just a bug that only affects upgraded installs.
[14:25] <ttx> mathiaz: https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+bug/499480
[14:31] <mathiaz> ttx: could this be related to https://bugs.launchpad.net/bugs/437012?
[14:33] <ttx> mathiaz: yes, that's a regression introduced in there
[14:33] <ttx> mathiaz: I just fixed it
[14:33] <siretart`> hi
[14:33] <siretart`> how to enable serial console access in grub2?
[14:56] <mathiaz> smoser: hi - it seems that the latest lucid UEC image requires at least 400M of Ram
[14:56] <mathiaz> ttx: ^^ have you seen something similar?
[14:56] <mathiaz> ttx: I can't run the lucid image with c1.medium anymore
[14:57] <ttx> mathiaz: haven't tried recently
[14:57] <ttx> or... rather... I can't get to that point anymore :)
[15:05] <ttx> mathiaz: here he comes again ^
[15:05] <frojnd> What is the http user?
[15:06] <frojnd> I need granted the http user to be able to write into /var/log/ampache
[15:07] <mathiaz> frojnd: www-data
[15:08] <frojnd> so if www-data needs to write into /var/log/ampache
[15:08] <frojnd> how can I do that?
[15:09] <smoser> good morning
[15:10] <mathiaz> smoser: o/
[15:10] <frojnd> any ideas?
[15:11] <frojnd> how can I grant http user in this case www-data to be able to write rad /var/log/amoache?
[15:11] <mathiaz> frojnd: it depends on the permissions of /var/log/ampache
[15:12] <mathiaz> frojnd: ls -l /var/log/ampache will give you the permission
[15:12] <mathiaz> frojnd: and then you'd probably add the www-data user to the correct group, and make sure the group can write to /var/log/ampache
[15:13] <frojnd> mathiaz: so -rw-r--r-- 1 root   root  25497 2009-12-21 16:41 Xorg.0.log
[15:13] <frojnd> drwxr-s--- 2 mysql  adm    4096 2009-12-21 01:47 mysql
[15:13] <frojnd> just chown -R www-data:www-data /var/log/ampache?
[15:14] <frojnd> now it's drwxr-xr-x 2 root   root   4096 2009-12-22 01:55 ampache
[15:14] <frojnd> but I manually crated it /var/log/ampache
[15:15] <mathiaz> frojnd: right chown -R ... should work
[15:18] <mathiaz> ttx: so you've merged the latest eucalyptus?
[15:18] <ttx> mathiaz: yes
[15:18] <mathiaz> ttx: is the package available in the archive?
[15:18] <ttx> mathiaz: accpeted, building
[15:18] <mathiaz> ttx: I could test them on some of the machines I have access to
[15:18] <mathiaz> ttx: ok - I'll monitor it then
[15:18] <ttx> mathiaz: it's not very different from last week code drop
[15:19] <ttx> 1.6.2~bzr1103 -> 1.6.2~bzr1113
[15:19] <mathiaz> ttx: well - I haven't tested any 1.6.2
[15:19] <ttx> heh
[15:19] <penta>  hi guys, anybody use scgi module with apache here?
[15:19] <mathiaz> ttx: I only tested 1.6.1
[15:20] <ttx> mathiaz: kirkland did, and it worked for him, though I really wonder how it could work. He should have run into the things I had to fix recently
[15:20] <mathiaz> ttx: I'll give a try to 1.6.2 - as you seemed to be blocked
[15:20] <ttx> mathiaz: I've to isolate which version broke it, or if it's an issue in my test setup
[15:20] <ttx> (which is brand new :)
[15:21] <penta> why it just doesn't wor with my apache? http://pastebin.com/d753c1d07
[15:53] <mathiaz> smoser: hey - have you looked into the increased memory footprint for the uec-images?
[15:53] <mathiaz> smoser: it seems that they can't be run as c1.medium on UEC anymore
[15:54] <smoser> i ran the alpha1 on c1.medium, but i did have issues on c1.small
[15:54] <smoser> you think that is memory related?
[15:54] <mathiaz> smoser: http://paste.ubuntu.com/344786/ <- this is the console output of today's image
[15:54] <mathiaz> smoser: it's running on a c1.medium
[15:55] <smoser> how much memory do you give? i'm guessing this isn't at all uec-only
[15:55] <mathiaz> smoser: by default c1.medium is 256M
[15:55] <mathiaz> smoser: I don't think you'd see that on EC2 - the smallest amount of memory you get there is >1G
[15:55] <smoser> do you have reason to believe something is uec specific?
[15:55] <smoser> rather than generic ubuntu
[15:56] <mathiaz> smoser: well - I think it's related to the uec images we published
[15:56] <mathiaz> smoser: IIRC the default -server install uses less memory
[15:56] <mathiaz> smoser: though I had to move to 256M of RAM for my default server installs last release cycle (I was at 128M)
[15:58] <smoser> yes, its obviously related to uec, but do you thikn that a "normal -server install" will not give the same issue on kvm (or real hardware)
[15:58] <mathiaz> smoser: under memory pressure, yes it will do the same thing
[15:58] <mathiaz> smoser: another particularity of uec images is that they don't have swap
[15:59] <mathiaz> smoser: I'm gonna check how much RAM is used by a default uec image
[16:00] <smoser> ah. thats it. swap.
[16:00] <smoser> so
[16:01] <smoser> above "under memory pressure" i think will occur in "normal -server install" with 256M of memory, i do not think that the -uec images are any more memory hungry than -server in general
[16:01] <smoser> except, for they do load python early in boot
[16:01] <smoser> second
[16:01] <smoser> i recently turned off swap
[16:01] <smoser> :)
[16:01] <smoser> which is probably what is causing your problem
[16:02] <smoser> the goal is to turn it back on if swap is available.
[16:02] <smoser> so, thats the primary difference i guess.
[16:02]  * mathiaz nods
[16:03] <mathiaz> smoser: http://paste.ubuntu.com/344792/
[16:03] <smoser> yeah, i'm aware
[16:03] <mathiaz> smoser: ^^ this is the memory on the latest uec image
[16:03] <smoser> explicitly i took out the /dev/sda3 (i think) entry for swap
[16:03] <smoser> you probably have a swap partition there
[16:04] <smoser> can you do me a favor, since you apparently have a uec up ?
[16:04] <mathiaz> smoser: right - sda3 is a swap partition
[16:04] <mathiaz> smoser: sure!
[16:06] <smoser> can you run ec2-get-info --block-device
[16:07] <mathiaz> ttx: smoser: I don't see a good reason to have such small instances defined in UEC if the default UEC image we provide requires at least 300 M of RAM to run correctly
[16:07] <smoser> mathiaz, its broke, we'll fix it
[16:07] <mathiaz> smoser: you mean reanable swap?
[16:07] <smoser> or are you saying that even with swap you think it doesn't make sense.
[16:07] <smoser> yes, i was saying re-enable swap
[16:08] <mathiaz> smoser: even with swap it doesn't really make sense
[16:08] <smoser> fair
[16:08] <smoser> its possible that it only sees the memory pressure early in boot
[16:08] <mathiaz> smoser: you boot a default uec image on m1.small or c1.medium and you're swapping right away
[16:09] <smoser> some of the new boot stuff is i think fairly memory hungry
[16:09] <mathiaz> smoser: well - I don't think that the system should be swapping right after boot
[16:09] <smoser> well, if it swaps on boot, yes, that sucks, but then if it stops swapping, then its not that big of a deal
[16:09] <mathiaz> smoser: http://paste.ubuntu.com/344792/ <- this is just after boot
[16:10] <smoser> ie, if it ballooned to 512M during the first 10 seconds of its life, but then trimmed to 128, who cares. thats what swap is for.
[16:10] <mathiaz> smoser: true
[16:10] <kirkland`> ttx: howdy
[16:10] <ttx> kirkland`: yo
[16:10] <ttx> kirkland: just reading your email
[16:10] <smoser> mathiaz, can you run the above ? the ec2-get-info
[16:10] <sub> +1, agreed. swapping is natural, normal behavior. it's thrashing that's bad D:
[16:11] <kirkland> ttx: <kirkland> saw you had some trouble with the nc upstart script
 strange, worked perfectly for me
 my testing was from an upgraded system
[16:11] <kirkland> ttx: also, nurmi and I did go over that diff
[16:11] <mathiaz> smoser: hm - I don't have the command ec2-get-info
[16:11] <smoser> i want to know if uec is correctly showing that there is a swap partition, so we could notice that and enable swap
[16:11] <kirkland> ttx: it's mostly branding and euca_rootwrap that are the diffs
[16:11] <smoser> hmm...
[16:11] <smoser> ok
[16:12] <mathiaz> smoser: this is whith ec2-api-tools 1.3.46266-0ubuntu1
[16:12] <smoser> its from ec2-init
[16:12] <smoser> but maybe its not packaged. i was n't lloking at an image, but rather source tree.
[16:12] <smoser> hold on.
[16:14] <ttx> kirkland: I think we should have WI covering their transformation into separate patches
[16:15] <kirkland> ttx: really?  we want separate patches?
[16:15] <kirkland> ttx: i particularly like using bzr merging
[16:15] <ttx> hmm
[16:15] <kirkland> ttx: it's *far* cleaner to me than fuzzing patches
[16:16] <kirkland> ttx: i dare say that's the power of using a shared bzr source
[16:16] <mathiaz> kirkland: ttx: you may wanna look at bzr looms if you want both have bzr merge goodness and separate patch
[16:16] <ttx> kirkland: I agree with that, just thought that wasn't what the spec said
[16:17] <ttx> kirkland: spec says "Keep debian/patches for ubuntu-specific non-upstreamable patches, merge all others"
[16:18] <ttx> kirkland: that's why I was wondering
[16:18] <kirkland> ttx: hrm
[16:20] <kirkland> ttx: having done the merge many times in the last few weeks, i find it far easier to fix conflicts in the bzr managed source, than re-fuzzing patches
[16:20] <ttx> kirkland: I agree with that. As long as we don't keep the patchsystem around
[16:21] <kirkland> ttx: true; shall we try to get rid of the patch system entirely, in favor of inline code changes?
[16:21] <ttx> kirkland: ...and we don't lose track of what the diff contains
[16:21] <kirkland> ttx: right, to me, that's the one advantage of debian/patches here
[16:21] <ttx> kirkland: I think the mix is confusing
[16:21] <ttx> At least it confused me
[16:21] <kirkland> ttx: we have a very specific list of what the patches are (and in my ideal world, documentation about the patch above it in the file)
[16:21] <kirkland> ttx: i agree;  the WSDL issue is still open, though
[16:22] <ttx> ok.
[16:22] <kirkland> ttx: we haven't agreed with upstream on the best way to solve that
[16:22] <kirkland> ttx: well, we sort of have
[16:22] <kirkland> ttx: there's some build dependency, that if promoted to Main, could generate the diff on the fly during the build
[16:22] <ttx> kirkland: I think they underestimate the work in there
[16:22] <ttx> I asked a question on that bug, unasnwered
[16:23] <kirkland> ttx: i have it in the bug, don't remember it off the top of my head
[16:23] <mathiaz> smoser: hm - TBH I don't think the guest is just swapping during boot
[16:23] <ttx> kirkland: I'll reply to thread to make sure jos is in the loop
[16:23] <ttx> kirkland: no need to interrupt your vacation :)
[16:23] <kirkland> ttx: heh, well, i just wanted to say hi :-)
[16:23] <smoser> mathiaz, if thats true, we have generic issues that should be discussed.
[16:24] <mathiaz> smoser: otherwise why would the OOM killer kick in *after* the instance has successfully booted (http://paste.ubuntu.com/344786/)
[16:24] <kirkland> ttx: i didn't want you blocking on me, or anything I did (or didn't do)
[16:24] <ttx> kirkland: it affects a spec contents that he approved. I don't think he would mind, but hey
[16:24] <smoser> ie, if we care that server doesn't boot in 256M
[16:24] <smoser> or s/boot/boot and be useful/
[16:24] <ttx> kirkland: you managed to run an instance ?
[16:24] <smoser> mathiaz, can you run the following in your instance:
[16:24] <smoser>  b=http://169.254.169.254/latest/meta-data/block-device-mapping/; for x in $(wget "$b" -O - -q); do v=$(wget "${b}/${x}" -q -O - ) ; echo "${x}:${v}"; done
[16:24] <ttx> kirkland: with 1.6.2~bzr1103 ?
[16:25] <smoser> on small on ec2, that gives me:
[16:25] <smoser> ami:sda1
[16:25] <smoser> ephemeral0:sda2
[16:25] <smoser> root:/dev/sda1
[16:25] <smoser> swap:sda3
[16:25] <mathiaz> smoser: sure - let me boot a new instance
[16:25] <kirkland> ttx: yes, absolutley
[16:25] <ttx> kirkland: beh, couldn't make it run
[16:26] <kirkland> ttx: well, timeout ...
[16:26] <kirkland> ttx: with the build i did locally
[16:26] <ttx> maybe another upgrade vs newinstall
[16:26] <kirkland> ttx: i ran several instances
[16:26] <kirkland> ttx: i did not test the soyuz-build deb's
[16:26] <ttx> issue
[16:26] <kirkland> ttx: mine were upgrades
[16:26] <ttx> I'll test again
[16:26] <kirkland> ttx: i did have to bundle a new image, though
[16:27] <kirkland> ttx: something about my images were out of sync, or not fully registered or something; there's an open bug on it
[16:27] <ttx> kirkland: anyway, it's almost working for me :)
[16:28] <kirkland> ttx: almost?
[16:29] <ttx> kirkland: well, when the NC is running libvirt, you're pretty close to the end of the test
[16:29] <ttx> https://bugs.launchpad.net/eucalyptus/+bug/499491
[16:35] <mathiaz> smoser: http://paste2.org/p/573930
[16:35] <mathiaz> smoser: there isn't any information there :/
[16:35] <smoser> can you append a slash
[16:35] <smoser> hold on
[16:36] <smoser> b=http://169.254.169.254/latest/meta-data/block-device-mapping/; for x in $(wget "$b" -O - -q); do v=$(wget -q "${b}/${x}/" -O - ) ; echo "${x}:${v}"; done
[16:36] <smoser> (i added '/' to the wget inside the loop. crossing fingers)
[16:36] <ttx> kirkland: see https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+bug/487270/comments/2
[16:36] <mathiaz> smoser: nope - same result
[16:37] <ttx> kirkland: it's either too much work or no work at all, based on what dan meant
[16:37] <mathiaz> smoser: well - I'm not sure if it's relevant to rely on the meta-data service to provide that information
[16:37] <mathiaz> smoser: IIUC there is a swap partition on /dev/sda
[16:38] <kirkland> ttx: yeah, i'll touch base with him after the holiday's on this
[16:38] <mathiaz> smoser: and it's the disk that is mounted as sda
[16:38] <smoser> there is on i386 small
[16:38] <kirkland> ttx: this is the biggest piece of work i think i have with that spec
[16:38] <smoser> there is not on a large
[16:38] <kirkland> ttx: your new comments not withstanding
[16:38] <smoser> its dynamic
[16:38] <ttx> kirkland: new comments ?
[16:38] <mathiaz> smoser: ah - so on EC2 you may have a swap partition or not
[16:38] <kirkland> ttx: diff vs. patches
[16:38] <smoser> right. it changes based on some things
[16:39] <ttx> kirkland: right
[16:39] <smoser> ideally the metadata service tells me what to do with things
[16:39] <kirkland> ttx: it would be nice if you could get clarification on that ;-)
[16:39] <smoser> but i can instead look at available devices and if one is type swap, then use it
[16:39] <mathiaz> smoser: it seems that UEC doesn't provide extra devices that could be used as swap
[16:39] <ttx> kirkland: will try to ping them on it
[16:39] <smoser> the metadata service is providing more explicit info
[16:39] <smoser> mathiaz, uec is strnage.
[16:40] <smoser> :)
[16:40] <ttx> smoser: UEC is good
[16:40] <smoser> well, ec2 is strange, and uec is strange differently
[16:40] <smoser> diversity :)
[16:40] <mathiaz> kirkland: is there a reason why uec is still using scsi instead of virtio devices?
[16:41] <smoser> mathiaz, the reason was related to hotplug of ebs volumes
[16:41] <mathiaz> ah yes - I remember - for compatibility with EC2
[16:41] <smoser> no, not that
[16:41] <mathiaz> virtio devices show up as /dev/vd*
[16:42] <smoser> that may have been a concern. at one point virtio hotplug was broken.  i asked nurmi this once.
[16:42] <mathiaz> smoser: well - fstab says to mount /dev/sda1 on /
[16:42] <mathiaz> smoser: ah right.
[16:42] <smoser> it does, but it doesn't matter.
[16:42] <smoser> because the kernel or ramdisk is reading the parameter from cmdline
[16:42] <smoser> so fstab doesn't matter for /
[16:43] <smoser> which is why i decided to only really put / in fstab, and dynamically write the rest
[16:43] <kirkland> mathiaz: compatibility with Amazon, I think
[16:45] <mathiaz> smoser: where should I file bugs related to UEC images?
[16:45] <mathiaz> smoser: I'd like to keep track of this memory issue
[16:45] <smoser> ideally from an instance
[16:45] <smoser> ubuntu-bug
[16:46] <smoser> will tag it as ec2-images although i dont know if it does that correctly for uec
[16:46] <smoser> but just file it against ec2-init for now, but really, i dont think its a UEC thing
[16:46] <smoser> our -server distro will fail to run in 256 M at the moment is my assertion
[16:46] <smoser> so however you would normally open such a bug
[16:47] <mathiaz> smoser: right - -server on 256M will be swapping after boot
[16:47] <smoser> so file that bug, and tag it uec-images ec2-images
[16:50] <ttx> smoser: are you around tomorrow ?
[16:50] <smoser> no
[16:51] <ttx> hmm
[16:51] <smoser> i could call some where, but i will not be near computer for most of day
[16:51] <ttx> smoser: no, I have to see if we should cancel the meeting
[16:54] <ttx> mathiaz: we'll see if anyone else shows up
[16:55] <ttx> mathiaz: if not, we'll cancel it. 2 is not quorum :)
[17:37] <kervel> i wonder if it is possible to run an ubuntu enterprise cloud on a single server (documentation mentions you need at least 2 servers) , for testing  / development purposes
[17:37] <kervel> eg can the node controller and the cluster controller run on the same machine ?
[18:01] <smoser> mathiaz, if you're still around and have uec up, could you verify that:
[18:01] <smoser> for x in $(blkid -t TYPE=swap -o device); do sudo swapon ${x}; done
[18:01] <smoser> will get your swap partitions mounted
[18:01]  * mathiaz tries
[18:02] <mathiaz> kervel: nope
[18:05] <mathiaz> smoser: \o/ - need to run blkid as root though
[18:05] <smoser> ah. yeah.
[18:05] <smoser> duh.
[18:06] <mathiaz> smoser: http://paste.ubuntu.com/344858/
[18:07] <smoser> so, the goal will be to look at metadata if available, if not, use blkid
[18:07] <smoser> assuming (possibly hazardly that if blkid finds something it thinks is swap that it is)
[18:07] <smoser> that is kind of scary
[18:08] <smoser> mathiaz, what is the default small? m1.small on uec
[18:08] <smoser> is it 128 ?
[18:09] <smoser> we used to run on whatever it was (karmic) but in lucid it crashes horribly. and very hard to debug because you can't get console logs because UEC doesn't cache them. so once the thing is dead the log is gone
[18:16] <mathiaz> smoser: m1.small gives 128 M for ram
[18:51] <alkisg> I'm having problems setting up my first LDAP. Why would `ldapsearch -x admin` succeed, and `ldapid admin` fail? http://pastebin.com/f5aaa1ee0
[18:57] <alkisg> Also, `sudo ldapaddgroup george` gives me "ldap_bind: Invalid credentials (49)" in the log file. I'm following this: http://doc.ubuntu.com/ubuntu/serverguide/C/openldap-server.html and I don't think I've missed any required steps... :(
[19:33] <teddymills> can someone paste bin the partition map of a default ubuntu 8X install?
[19:34] <teddymills> I am trying to recover my partition map, but I do not know what to put it back to..it was the default install...with ext3 mostly I need to know where root is and what partition number...
[19:36] <alkisg> teddymills: partition map?! what do you mean? grub device map?
[19:38] <teddymills> fdisk -l /dev/sda
[19:40] <smoser> teddymills, its going to be specific to the size of the disk at very least
[19:41] <teddymills> i dont need the partition sizes..i need the partition types and numbers...
[19:41] <alkisg> Here's one from Lucid: http://pastebin.com/f5e89205d
[19:42] <alkisg> But the extended partition has its own partition table, so I wonder how that would help you...
[19:42] <teddymills> thx lucid that should help...
[20:38] <gkahla> i'm trying to get fetchmail to pull my mail down onto my laptop, and fetchmail tells me port 25 isn't responding on localhost. Do I *really* need to set up postfix to get rid of this problem? Advice appreciated... destination machine is my laptop - NOT a server.
[21:11] <ziesemer> I replaced my dead motherboard.  Still an Intel x64, etc., but different on-board network controller.  Still my network (eth0) is working as expected.  However, my VLAN, eth0.5 is sending but not receiving packets.  What do I need to check / change?
[21:12] <ziesemer> Attempting to ping from this box to another box, the managed network switch shows the packets being sent from local to remote, then the remote to local, but "ifconfig eth0.5" shows only the transmitted packets, with 0 bytes received.
[21:16] <ziesemer> (I had already removed the existing entries from /etc/udev/rules.d/70-persistent-net.rules , so that the new interface still came up as the old eth0, rather than a new eth1.)
[23:24] <unit3> what's the best practises way under karmic to adjust which services get started at boot?
[23:24] <unit3> "update-rc.d cman enable" doesn't seem to get cman to start at boot.
[23:30] <unit3> hm... looks like "update-rc.d -f cman remove && update-rc.d cman defaults" does what I want. confused as to why enable/disable exist.
[23:30] <unit3> oh well.