[01:05] <harushimo> hey question for you
[01:05] <harushimo> I'm using ubuntu desktop version but I want to install maas on it
[01:05] <harushimo> I want to mess around with cloud computing for ubuntu
[01:05] <harushimo> can I use it?
[01:06] <bigjools> yes, you just need to boot the cloud nodes with the server image
[01:06] <harushimo> I need ubuntu server iso also then?
[01:07] <harushimo> I'm learning some server side right now. I just want to do this right
[01:07] <harushimo> especially with the cloud
[01:07] <bigjools> read this: https://wiki.ubuntu.com/ServerTeam/MAAS
[01:07] <harushimo> I currently installed maas on my desktop version
[01:08] <harushimo> I'll get the server image cd
[01:08] <harushimo> that was easy
[01:08] <harushimo> thank you
[01:08] <bigjools> np
[01:10] <harushimo> maas detected my router ip address not the main one
[01:10] <harushimo> would that be a problem for the cloud
[01:12] <harushimo> is it better use maas with the ip address of my service provider?
[01:13] <harushimo> does it take long to import maas profile? I'm running the script right now
[01:35] <yaboo> want to setup ufw, but have three interfaces, how do I allow all ports on e.g. eth3,4 and block ppp0
[01:36] <twb> Not sure if you can in ufw
[01:38] <twb> Based on ten seconds looking at the manpage, I would guess: ufw default deny; ufw allow on eth3; ufw allow on eth4
[01:38] <yaboo> two, thanks missed it
[04:32] <twb> OK so in lucid, when booting a server without the "quiet splash" options, when init prints "fsck of /srv/backup/boot failed with exit status 1" (or thereabouts), and then JUST HANGS FOREVER, refusing to continue booting, what am I supposed to do?
[04:33] <twb> Because AFAICT the only way to rescue it is to boot a live image and manually fsck it clean.
[04:33] <twb> Which is bullshit, because the /srv/backup/boot filesystem isn't needed to bootstrap :-/
[04:34] <henkjan> try S for skipping mounting
[04:35] <henkjan> plymouth :( i don't understand why its in the server version
[04:36] <qman__> I haven't seen an fsck on my file server in years because of it
[04:36] <qman__> just have to hope it's working
[04:40] <twb> henkjan: I booted without splash
[04:40] <twb> henkjan: there is no prompt
[04:42] <twb> I think this also happens if you boot with splash, but you don't have the fancy plymouth-...-theme pacakge installed, but only the unremovable plymouth deps
[04:44] <twb> Yeah, I just checked and same symptoms with "single splash", there is no splash at all even though plymouth is installed
[04:45] <twb> It does respond to ctrl-alt-del event though, not totally hung.  Just upstart sitting there being useless as ever
[04:46] <twb> And this mobo is one of those stupid UEFI hybrid ones, and it thinks my USB rescue keys are UEFI bootable, so it can't boot them correctly
[05:54] <henkjan> twb: try booting in grub with init=/bin/bash
[05:55] <twb> henkjan: well of course that will work
[05:56] <twb> But then I have to fiddle-fart around assembling the array and lvs and so on, and AFAICT the actual problem won't be fixed
[05:57] <henkjan> no, mdadm an lvm should be working correct with init=/bin/bash
[06:02] <twb> Oh, yeah, I was thinking break
[06:02] <veebull> installed squid, then squid-deb-proxy on 12.04 server (VM)... installed squid-deb-proxy-client on a Lubuntu 12.04 VM... squid doesn't appear to be working on the server for some reason?
[06:05] <twb> henkjan: FWIW init=/bin/sh also hangs same way
[06:06] <twb> It boots correctly if I comment out all the HDD's filesystems (i.e. all of /srv/backup), so now I'm trying it with them enabled, but with their fsck pass field set to 0
[06:06] <twb> (i.e. skip fsck of them on boot)
[06:06] <Techie> is there any way to have ubuntu-server automatically transcode videos that appear in one directory, put the output in a seperate directory and delete the source?
[06:07] <Techie> and do this on a cronjob without trying to transcode if its already transcoding something
[06:07] <twb> http://paste.debian.net/168684/
[06:08] <twb> Ooooh, I wonder if those LVs are set to be read-only
[06:08] <twb> I *know* under that circumstance, mountall jsut completely shits itself.  ISTR it does so with a much clearer error message, though...
[06:08] <twb> Could be all the ureadahead-other lines are getting in the way of seeing it.
[06:09] <twb> Fuck, I worked out what it actually is.  Some of these LVs no longer exist, but fstab still expects them
[06:10] <twb> So mountall is just sitting there waiting for them to be inserted or something equally inane
[06:10] <twb> !&^@#&^%!@#^!@#
[06:11] <twb> Some debugging messages would've been nicer
[06:11] <henkjan> twb: i'm so fed up with plymouth and upstart :(
[06:11] <henkjan> missing sane logging
[06:11] <twb> I've been fed up with it for three years
[06:12] <twb> It was only managable in hardy because it was handballing all the work to a proper init
[08:05] <Daviey> Goooooooooooooooooooooooood Morning
[08:14] <linocisco> hi all
[08:15] <linocisco> i want to be ubuntu partner. but nobody here will buy  ubuntu commercial edition. how could I do?
[08:15] <linocisco> they will just ask for support
[08:15] <linocisco> preloaded mail server, proxy server will be popular
[08:19] <ikonia> linocisco: what are you on about
[08:20] <Daviey> linocisco: There is no such thing as Ubuntu commericial edition.
[08:20] <linocisco> actually, I want to start a new business like tech support or preloaded ubuntu servers(mail,squid,firewall, dual wan loadbalancing) with system unit together. But I would like to be acknowledged by ubuntu
[08:21] <linocisco> actually, I want to start a new business like tech support or sell  preloaded ubuntu servers(mail,squid,firewall, dual wan loadbalancing) with system unit together. But I would like to be acknowledged by ubuntu
[08:25] <Daviey> linocisco: you probably want to get on the Ubuntu Marketplace, http://www.ubunut.com/partners/programme
[08:26] <linocisco> Daviey, thanks. let me check out
[08:34] <linocisco> askubuntu.com vs ubuntuforums.org ??
[08:34] <linocisco> for server related questions, which one is more interactive?
[08:40] <bioman> Hello
[08:42] <bioman> Got a problem with networking :/ I use eth0 and eth0:1. "service networking status" returns "networking stop/waiting" but it works infact...
[08:44] <twb> bioman: that's an artefact of how upstart reports things
[08:46] <bioman> when I try with the deprecated method "/etc/init.d/networking restart" it says "RTNETLINK answers: File exists" and "Failed to bring up eth0:1"
[08:46] <bioman> but both appear with ifconfig...
[08:48] <soren> You probably created eth0:1 manually and have now added it to /etc/network/interfaces?
[08:49] <twb> :1 is gibberish
[08:50] <twb> It's only applicable to ifconfig and only because ifconfig is too stupid to understand that an iface can have zero or more addresses
[08:50] <bioman> soren: no, I've created it in /etc/network/interfaces
[08:50] <twb> bioman: pastebin your /etc/network/interfaces
[08:53] <bioman> twb: a moment please. I'm used to use Archlinux, Ubuntu is at work. Which tool can I use please ?
[08:53] <twb> apt-get install pastebinit; pastebin -u http://paste.debian.net/ /etc/network/interfaces, IIRC
[08:53] <maxb> /sbin/ip is generally regarded as the modern replacement for ifconfig
[08:54] <twb> maxb: AIUI ifupdown 0.7 uses ip internally
[08:54] <twb> And I thought ifupdwn 0.7 was rolled out a few releases back, but I'm still on lucid so I dunno
[08:56] <bioman> twb: http://paste.ubuntu.com/986831
[08:57] <twb> bioman: what version of ifupdown do you have
[08:57] <twb> If it's a second *card* it should be iface eth1 not eth0:1
[08:57] <twb> (Or whatever the iface is named; "ip link" will list all known ifaces.)
[08:58] <bioman> twb: It's for a DRBL. It will be installed on a laptop which has only one physical network card
[08:58] <twb> I don't know what a DRBL is
[08:59] <bioman> twb: drbl.sourceforge.net :)
[09:00] <twb> What, PXE?
[09:01] <bioman> yep
[09:01] <twb> You know live-initramfs can PXE boot off an arbitrary http URL, right?
[09:01] <twb> So you can PXE the ramdisk, and that can pull down the .iso or .img from the internet even
[09:02] <bioman> twb: here it will be used for the clients to install a syspreped Windows XP Pro
[09:03] <twb> Lame
[09:03] <soren> twb: Oneiric is the first release with a 0.7 ifupdown.
[09:03] <twb> soren: thanks
[09:03] <twb> bioman: what ubuntu release are you running?
[09:05] <bioman> twb: 12.04 LTS (need "latest" kernel, for hardware recognizing). Here we have one in 11.10, but it does not work with recent PC's because their hardware is too recent
[09:06] <soren> twb: Are you familiar with rmadison?
[09:07] <twb> soren: yeah I was just lazy -- sorry
[09:08] <soren> twb: No problem at all. Just wanted to make sure. It's quite handy :)(
[09:08] <soren> :)
[09:11] <bioman> Ah, another thing : Now when I boot, there is a message "Waiting for network configuration..." for more than 40s, then "waiting up to more 60 seconds for network configuration..." which is really annoying. How to fix this please ?
[09:18] <trapni> does keystone (still) need "passlib"? is it a hard dependency? (I am to write packages for Gentoo, currently, and feel a bit clumsy about OpenStacks dependencies :)
[09:18] <twb> bioman: I *strongly* recommend you investigate live-initramfs (and maybe live-config and live-build).
[09:19] <twb> bioman: they are a central effort to solve this class of booting, once and for all
[09:19] <twb> The Ubuntu effort is called "casper" but AFAIK it only tries to solve Canonical's problems, and isn't really set up for end users to use
[09:23] <bioman> twb: I come from Archlinux world, dunno what all you tell me is for...
[09:47] <bioman> hey people don't tell me you dunno how to get rid of those annoying messages ?
[09:53] <twb> bioman: what annoying messages
[10:02] <bioman> twb: at boot, "Waiting for network configuration..." (nearly 40 seconds) and "Waiting up to 60 more seconds for network configuration..."
[10:05] <bioman> twb: found those messages in /etc/init/failsafe.conf
[10:08] <twb> Just comment them out
[10:12] <bioman> twb: before doing this, can you explain me what is this file for please ? using the sleep command make the PC in stand mode isn't it ?
[10:14] <twb> Sorry, I'm busy
[11:47] <Techie> if anybody in here is after automated transcoding of videos inside a folder, http://pastebin.com/wWYDpcvg
[11:48] <Techie> it uses HandBrakeCLI   and doesnt play nice with sub folders
[12:02] <ikonia> Techie: handbrake has a command line tool that works fine
[12:02] <ikonia> no need for anything else
[12:04] <Techie> ikonia, its for automation
[12:04] <ikonia> yes, you can automate with handbrake
[12:04] <Techie> ikonia, when hooked up as a cron
[12:04] <ikonia> just feed things into the queue and click "go"
[12:04] <Techie> it checks to see whether anythings converting
[12:05] <Techie> and if nothings converting it converts everything in a specified folder
[12:05] <ikonia> handbrake does that
[12:05] <Techie> so rather than having to que things up, you just dump them in a folder
[12:05] <ikonia> it converts to what you want and checks comatability before processing
[12:06] <Techie> did you even look at the script, or are you just bashing me because you dont understand it?
[12:06] <ikonia> I'm not bashing anything
[12:06] <Techie> then did you atleast look at it
[12:07] <ikonia> I glanced at it
[12:07] <ikonia> didn't really follow it through as the description you've given explained the idea
[12:07] <Techie> then you should have noticed that it doesnt require the user to queue anything or invoke it manually
[12:07] <ikonia> it does
[12:07] <ikonia> it requires users to put things in the right directories
[12:07] <ikonia> that's the same as adding to a queue
[12:08] <Techie> not when your using server
[12:08] <ikonia> you may as well just add to the queue in handbrake a proven and supported solution
[12:08] <ikonia> Techie: of course when you are using a server
[12:08] <Techie> when server adding to a folder is simple
[12:08] <ikonia> the command line tool has add to queue functionality
[12:08] <ikonia> Techie: adding to the queue is simple
[12:08] <ikonia> and you don't need to worry about arranging subdirectories
[12:09] <Techie> unfortunately the queue function requires too much reading for my taste, so i created an alternative based on what i need, and i felt that it may be suitable for others
[12:09] <ikonia> so rather than read a few lines you've created an unsupported shell script wrapper
[12:10] <Techie> yep
[12:10] <Techie> coz thats how i roll
[12:10] <ikonia> people could just read about 10 lines of docs for the cli synatax
[12:10] <ikonia> seems silly to offer this to people as a solution when an out of the box product does it better with more functionality
[12:12] <Techie> im not saying its a replacement for the proper method
[12:12] <ikonia> you're offering as a solution
[12:13] <Techie> a solution thats easily configuable
[12:13] <ikonia> which in my view seems pointless as the official tool your wrapping around works fine without this script/need for specific layout and offers more options without your wrapper
[12:13] <Techie> thats the only reason im offering it
[12:13] <ikonia> easier to just use the correct tool without the wrapper
[12:13] <ikonia> more options, no specific file system layout needed etc
[12:13] <ikonia> but a good effort on your part
[12:14] <Techie> i find that for most users, more options = more confusion
[12:14] <Techie> including me
[12:14] <ikonia> Techie: but you've hardcoded it to settings you like
[12:14] <ikonia> and you've stipulated a specific directory layout
[12:14] <ikonia> that seems harder than the official tool
[12:14] <Techie> that being said, this is my first version
[12:14] <ikonia> more so as there is an unofficial ncurses wrapper for it (if you want a menu driven interface without the gui)
[12:15] <Techie> i havent had years to refine it
[12:15] <ikonia> you shouldn't need years
[12:15] <ikonia> it's a 30 line shell script calling a program that exists and is well documented
[12:15] <ikonia> I did something very similar with mplayer and mencode, then realised handbrake did it better and easier
[12:15] <ikonia> mencoder even
[12:16] <Techie> but you still did it
[12:16] <ikonia> I had it doing subdirectories, chapter seperation/naming etc etc, then realised it was a pointless effort
[12:16] <ikonia> Techie: yes, because mencoder didn't have a tool to do what I want
[12:16] <ikonia> and I didn't offer my solution to people
[12:16] <Techie> the same with this
[12:17] <ikonia> you're aware of the tool handbrake as you actually call it and you're chosing to make it more complex/rigid by wrapping it in a shell script
[12:17] <Techie> i didnt realised that i could control the CLI version to the extent that i could control the GUI
[12:17] <ikonia> Techie: I'd say %85 maybe a little more is available in the cli
[12:17] <Techie> thats why i created this wrapper
[12:17] <ikonia> the profile options are very useful
[12:18] <Techie> not because i wanted to change the way things worked, but to get things to work how i wanted, with the knowledge i already have
[12:18] <ikonia> but you'll find it easier/better to probably just use the tool
[12:18] <ikonia> especially if you created a profile for the settings (encoding) you like
[12:19] <Techie> maybe further down the line i will redo it from scratch to work compltely wiht every option handbrake offers
[12:19] <ikonia> why do it at all ?
[12:19] <ikonia> just call the handbrake binaries
[12:19] <Techie> because then i have to call them each time
[12:19] <ikonia> you have to call this script each time
[12:19] <Techie> this was created so i could just dump files onto the server via samba and have them convert
[12:19] <ikonia> I don't see the difference
[12:19] <Techie> also so my dad could drop files
[12:20] <ikonia> sorry, I don't see any value to it, but if you find it useful, good on you
[12:20] <Techie> okay
[12:20] <Techie> anyway im off, i have work tomorrow.    night ikonia
[12:20] <ikonia> good night
[12:33] <zul> good morning
[12:47] <gary_poster> apw, hi.  IS has some concerns about implementing the work-around to bug 944386 in the data center (comment #4 from kees disabling protected_nonaccess_hardlinks).  If I were able to beg borrow or steal escalation privileges from someone, how likely is it that we'd be able to get a Precise release of a fix for this?
[12:48] <gary_poster> and what might the timeline be?
[12:51] <apw> gary_poster, i thought we expected that to be fixed upstream, so working now in quantal, if so then it may well be an easy job.  if so then 3 or so weeks to an official kernel a few days perhaps to get something you can test in the interim?
[12:52] <gary_poster> apw, ok, fantastic, thank you.
[12:52] <apw> gary_poster, am off today will make a note to poke that in the am
[12:52] <gary_poster> apw, understood, thank you
[14:08] <ruben23> hi guys i have uname -r ---> 2.6.18-308.el5.028stab099.3 an i would liek to install kernel headers but only see this version available ---> http://pastebin.com/rPqHjgav ---> it does not fit on its version and it wont install, any idea how to make this work..? to have same exact version..?
[14:09] <ruben23> help please how to make it worked..?
[14:18] <nathwill> ruben, that looks like a rhel kernel...
[14:21] <filo1234> hi all
[14:22] <filo1234> I'm looking for a way to start recovery ( when I choice it ) in netroot by default... without display choice menu options
[14:36] <ruben23> nathwill: this is ubuntu server 10.034 LTS instaled on a VPS server..
[14:37] <nathwill> ah. i'm used to seeing the el business on my rhel machines
[14:38] <RoyK> ruben23: the 'el' in the kernel name there, is "enterprise linux", meaning redhat
[14:38] <_ruben> that's no ubuntu kernel, so good luck finding the kernel headers, i'd ask the vps provider
[14:39] <RoyK> ruben23: it might be openvz, though
[14:39] <RoyK> ubuntu running under redhat with openvz, although it looks hairy
[14:40] <ruben23> yes its openvz..sorry guys
[14:44] <nathwill> wonder if the linux-headers-generic package would work for you.
[14:50] <zul> Daviey: ping does maas setup the dnsmasq to do the tftp?
[14:51] <ahxcjb> RoyK: OpenVZ is unsupported on RHEL
[14:51] <ahxcjb> it will invalidate Red Hat's ability to support the RHEL server
[14:51] <ahxcjb> given it replaces glibc iirc
[14:57] <koolhead17> zul, am not correct but currently its cobbler which creates profiles 4 the deplyment
[14:57] <zul> k
[14:57] <RoyK> ahxcjb: I didn't know that, but the kernel version he posted, 2.6.18-308.el5.028stab099.3, looks like RHEL to me
[14:58] <ikonia> it's going to be some junk virtualzation
[14:58] <RoyK> might be completely wrong, though :)
[14:58] <koolhead17> *sure
[14:58] <RoyK> ikonia: openvz doesn't virtualize the kernel, does it?
[14:58] <ikonia> it's a RHEL 5.5 customised kernel
[14:58] <ikonia> it's not even a stock RHEL one
[14:58] <ikonia> RoyK: no, it uses the hosts
[14:58] <ikonia> the 028stab099.3 big suggests someone's modified it beyond redhat's though
[14:59] <ikonia> "bit"
[15:08] <Daviey> zul: maas sets up tftp by default, but doesn't touch dhcp, unless you install maas-dhcp
[15:08] <zul> Daviey: gotcha but maas-dhcp is dnsmasq right?
[15:08] <Daviey> yah
[15:51] <LemU_> Hi! I would need help... I was following this guide https://help.ubuntu.com/community/UbuntuCloudInfrastructure#Deploying_Ubuntu_Cloud_Infrastructure_with_Juju     Im almost end of guide but cloud-publish-tarball ubuntu-11.10-beta1-server-cloudimg-amd64.tar.gz images ain't working.
[15:51] <LemU_> It just gives me: Unable to run euca--describe-images.  Is environment for euca- set up?
[15:52] <LemU_> So I was wondering if there is someone who understands eucatools better than I am and sees why i can't use it
[15:54] <alex88> hello guys, someone tried chef deploy tool?
[16:09] <JC617> Can anyone verify if scp and rsync come pre-installed (shipped) with Ubuntu Server 10.04 LTS?
[16:31] <bioman> Hi
[16:43] <Guest20494> I really need help here: https://answers.launchpad.net/ubuntu/+question/197184
[16:47] <Guest20494> If someone is familiar with eucatools+juju+maas combination, I would love to have opinion what I am doing wrong
[16:47] <bioman> What's going on with the "service" command ? "service networking restart" says "stop: Unknown instance:". service start networking says "networking stop/waiting"
[17:05]  * bioman is thinking...
[17:08] <pdtpatrick> has anyone installed Landscape Dedicated server ?
[17:12] <KM0201> doesn't that cost money?
[17:14] <pdtpatrick> right - I'm trying to figure out whether it is just to license the landscape software or will it require a license for each server it manages (that'll suck).
[17:14] <KM0201> no idea.
[17:16] <Jarmo> Anyone familiar with setting up Cloud with MAAS + juju?
[17:17] <pdtpatrick> Jarmo: i believe there's a #juju room
[17:17] <Jarmo> there they did sent me elsewhere, because they thought it has something to do with eucatools... (or something like that)
[17:18] <Jarmo> but I'll try juju room too :)
[17:19] <pdtpatrick> seems like the juju technology is quite new so i've been struggling to find help myself or get a solid understanding of how the inner parts work. For instance, it'll be nice to see a decent guide on openstack
[17:22] <Jarmo> we are settnign up openstack :)
[17:22] <pdtpatrick> ha! is that what you're trying to find help on ?
[17:22] <Jarmo> and actually we did it up with maas + juju, but we cant add images
[17:22] <Jarmo> *get it up
[17:23] <pdtpatrick> was there a guide you followed?
[17:23] <pdtpatrick> i got all way to the dashboard
[17:23] <Jarmo> yes
[17:23] <pdtpatrick> was a PITA -- their documentation is LONNNNGGG
[17:23] <Jarmo> sec..
[17:23] <Jarmo> https://help.ubuntu.com/community/UbuntuCloudInfrastructure
[17:24] <Jarmo> easy & good guide... there is couple of bugs wich we did find
[17:25] <Jarmo> just make sure your bios time is same on all machines  & If machines wont get RUNNING state: do this on MAAS server machine: sudo ufd disable, then detroy environment and put it up again... Seems your firewall is blocking connections between zookeeper & machines!  Note for myself: You can delete installed "charm" without destroying whole enviroment.... just cant remember command atm...
[17:25] <Jarmo> and it works
[17:25] <pdtpatrick> i tried that .. that openstack.cfg never worked. each time i typed "juju deploy --config=openstack.cfg keystone"
[17:25] <pdtpatrick> it'
[17:25] <pdtpatrick> it would say charm not found
[17:26] <Jarmo> i have solution for that!
[17:26] <alex88> i installed openstack from official guide and i can confirm it's a PITA! :) it also have some errors
[17:27] <Jarmo> or in that document it is phase 2.... then instead "juju deploy --config=openstack.cfg keystone" you type juju --config=openstack.cfg --repository=. local:keystone
[17:28] <pdtpatrick> yeah their documentation is not clear. And all their usernames and id field -- better hope you have a large paste buffer or else you'll find yourself having to start over. Especially because you are not warned that you'll need each of those "id" fields it keeps generating
[17:28] <Jarmo> everything else is working, but just can't add images there..... I hate when i cant find solution :P
[17:28] <pdtpatrick> Jarmo:  so you copied the repo to ur local box ?
[17:29] <Jarmo> yeah and to ROOT
[17:29] <pdtpatrick> Jarmo: did you turn on logs for juju and see what happens when u click to make an image ?
[17:29] <Jarmo> but you cant be root users when giving that command, but u must be at root, not at that  folder you craeted
[17:31] <pdtpatrick> Jarmo:  at a different terminal, in ur bootstrapped environment. Type - juju debug-log
[17:31] <pdtpatrick> or u can do it in your current
[17:31] <Jarmo> there is no click to make image :(  it is like uec-publish-tarball ./ubuntu-11.10-beta1-server-cloudimg-amd64.tar.gz images gives me nothing, because i dont have access to eucatools or something :/
[17:31] <pdtpatrick> interesting
[17:32] <pdtpatrick> I'm sure well let me know if the guys in #juju help as i'll be going through this myself soon :(
[17:32] <bioman> Got a problem : After a reboot, resolv.conf is empty so I cannot apt-get install...
[17:32] <Jarmo> uec-publish-tarball isn't right btw it is cloud-publish-tarball
[17:32] <Jarmo> It gives me this output: Unable to run euca--describe-images. Is environment for euca- set up?
[17:32] <pdtpatrick> bioman: are you using static entries in /etc/network/interfaces? if so then add dns-nameservers and dns-search as part of you interfaces config
[17:33] <Jarmo> And if I undrestand right following guide, using those CREDS should give me right access to right place :P but there happens something what i just dont understand
[17:34] <bioman> pdtpatrick: thank you :) Yep static... so for example I add "dns-nameservers 8.8.8.8" ?
[17:35] <pdtpatrick> yeah its a bit baffling right as as its a new technology. Im hoping it'll get automated soon where the .deb package actually walks you through setting it up or they create some frontend to configure it.
[17:35] <pdtpatrick> bioman: yes.
[17:41] <bioman> pdtpatrick: thank you. Maybe you could help me again... I have installed DRBL. The problem is that when I boot PXE on a machine I want to get, I have "PXE-32: TFTP open timeout" :/
[17:43] <pdtpatrick> on the machine running tftp .. do you see anything in the logs?
[17:43] <bioman> I've installed tftp-hpa but done no configuration...
[17:44] <pdtpatrick> you should probably configure that and ensure that your DHCP server has the correct pxelinux.0 settings. So that when the clients boot up, they know which server to contact to the files
[17:46] <bioman> pdtpatrick: I'm new to Ubuntu Server. Any wiki please ?
[17:46] <pdtpatrick> one second.
[17:46] <pdtpatrick> https://help.ubuntu.com/community/PXEInstallServer
[17:48] <Jarmo> It is funny when I throw my question at #juju they say come here, and when i throw it here people tell me to go #juju :P
[17:50] <bioman> pdtpatrick: thanks
[17:50] <pdtpatrick> np
[17:52] <bioman> pdtpatrick: it going the wrong way. No /etc/init.d/tftpd-hpa :(
[17:52] <pdtpatrick> you'd got to install the package
[17:52] <pdtpatrick> you've*
[17:53] <pdtpatrick> you can check for the tftp package "apt-cache search tftpd-hpa"
[17:53] <bioman> pdtpatrick: I've installed the wrong one ;)
[17:53] <pdtpatrick> and install it with "apt-get install <packagename>"
[17:53] <pdtpatrick> hehe well there u go :)
[17:58] <bioman> pdtpatrick: at the dhcpd config, here it goes :
[17:58] <bioman> filename "pxelinux.0"; next-server <pxe host>;
[17:58] <pdtpatrick> right
[17:58] <pdtpatrick> so that's what you'll fill out. That needs to point to whatever server has tftp running and listening
[17:59] <bioman> where do I get pxelinux.0 ? <pxe host> : what should I write here ?
[17:59] <pdtpatrick> see where it says "Configure dhcpd"
[17:59] <pdtpatrick> in /etc/dhcp3/dhcpd.conf
[17:59] <pdtpatrick> you've also have to have dhcp3 installed
[18:00] <pdtpatrick> sudo apt-get install dhcp3-server
[18:00] <pdtpatrick> and if your server is at 10.0.1.2 then your next-server would be "next-server 10.0.1.2"
[18:01] <bioman> pdtpatrick: installed yet ;) OK for next-server. But where do I find the file pxelinux.0 ?
[18:01] <pdtpatrick> that's also on the tftp server
[18:02] <bioman> So I only enter the line and that's all ?
[18:02] <pdtpatrick> yup
[18:02] <pdtpatrick> http://www.youtube.com/watch?v=q-MX2B4OpO8
[18:03] <bioman> thank you :)
[18:03] <pdtpatrick> watch that video and see if it explains things a bit
[18:03] <pdtpatrick> here's an extended video -- its for CentOS but same concept
[18:03] <pdtpatrick> http://www.youtube.com/watch?v=XxULo0FLDR8
[18:20] <pdtpatrick> Has anyone used this ?
[18:20] <pdtpatrick> http://www.drbd.org/mc/management-console/
[18:35] <Jarmo> has anyone been able to do something like this : https://help.ubuntu.com/community/UbuntuCloudInfrastructure and making it working? i'm having problems with eucatools part at the end... and would love to have tips... :D
[19:02] <alex88> Jarmo, page isn't loading here :/ is that openstack?
[19:02] <Jarmo> alex88 yes
[19:03] <pdtpatrick> Jarmo:  the logs didn't show anything ?
[19:03] <Jarmo> alex88 It aint loading here either atm :/
[19:03] <pdtpatrick> i remember during my install, you had the option of using ec2 or eucal cloud or creating your own local
[19:04] <pdtpatrick> maybe u choose the former and now it is looking for such service ?
[19:04] <Jarmo> pdtpatrick  not those logs wich I was watching... but im not 100% sure where to look... thats why im having really hard time to find where it goes wrong..
[19:05] <pdtpatrick> in the same terminal you're working on
[19:05] <pdtpatrick> juju debug-log
[19:05] <pdtpatrick> it would do something like a tail -f
[19:05] <pdtpatrick> and then try navigating around and see if u get any logs
[19:05] <pdtpatrick> it might tell u what it is trying to do and failing
[19:05] <Jarmo> But im installing it from charms, so i didnt have to choose device...
[19:06] <pdtpatrick> i guess the question then is - do the charms default to local or some other cloud technology
[19:07] <Jarmo> hmmm, i think I have to try that tomorrow when I'm at my office again... at some other channel ppl told me that those creds wich i did use should give me ability to use eucatools at any computer with is at same network
[19:07] <pdtpatrick> that sounds like u should already have euca server
[19:08] <pdtpatrick> unless the charms auto deploy one for you
[19:08] <pdtpatrick> dammit - - we need the person who wrote that charm haha
[19:08] <Jarmo> For that guide, wich we can't reach atm... (damn) i think charms did install it (or atleast should have done that)
[19:09] <Jarmo> this is what that guide deploys juju deploy mysql juju deploy rabbitmq juju deploy --config=openstack.cfg keystone juju deploy --config=openstack.cfg nova-cloud-controller juju deploy --config=openstack.cfg nova-volume juju deploy nova-compute juju deploy glance juju deploy openstack-dashboard
[19:10] <pdtpatrick> i don't know much about the euca itself so maybe try to see without openstack if you can get to it and make some changes. If it continues to complain then it could be that it has wrong permissions. Considering it does not run as root (maybe the files/service it is accessing is controlled by root currently? )
[19:11] <pdtpatrick> yeah when i was installing mine, i don't recall seeing eucatools n e where.  so that's why I was wondering whether it is using local storage or some other cloud
[19:11] <pdtpatrick> i know it tells you that you can use KVM, or XEN, or EC2 with amazon or euca
[19:11] <pdtpatrick> but u have to have already set them up otherwise u cannot proceed.
[19:11] <pdtpatrick> at least that was my understanding when i read the docs
[19:12] <Jarmo> hmmm, tbh that makes sense... because before I used eucatools, i had to be root, now that guide doesn't say anything about beeing root (even tho I did try it, and didnt help)
[19:12] <pdtpatrick> right so maybe the logs would say permission denied for <filename> and that would be ur canary in the coal mine?
[19:13] <Jarmo> well that is atleast best hint where to look it :)
[19:13] <pdtpatrick> yup
[19:13] <Jarmo> this is how guide says it: EC2 API  To begin using the EC2 API, select Settings-> EC2 Credentials -> Download EC2 Credentials in the Openstack dashboard. Save the file (eg, /home/adam/openstack/"). We can then unzip these and begin using our cloud:  cd /home/adam/openstack unzip 247e0c5de9dc415a8e411da643308eb6d-x509.zip . ec2rc.sh wget http://uec-images.ubuntu.com/server/server/releases/oneiric/beta-1/ubuntu-11.10-beta1-server
[19:14] <Jarmo> uec-publish-tarball ./ubuntu-11.10-beta1-server-cloudimg-amd64.tar.gz images euca-add-keypair adam >adam.pk euca-run-instances -k adam ami-00000002
[19:15] <Jarmo> and uec-publish-tarball  is where my problems begins.. (yeah I did realize that it is cloud-publish-tarball now)
[19:16] <Jarmo> and BTW when i give command "wrong" way it realises what im doing and helping me..... until I give right command and then it tells: Unable to run euca--describe-images. Is environment for euca- set up?
[19:16] <Jarmo> *"wrongway"
[19:17] <pdtpatrick> yup so looks like ur going to have to setup euca and then continue with openstack
[19:20] <Jarmo> well, normally it wouldnt be problem, but Im not exactly sure how I can add it to maas + juju environment :O
[19:20] <pdtpatrick> when it said get ec2 credentials
[19:20] <pdtpatrick> u already had a working AWS account ?
[19:21] <Jarmo> yeah
[19:21] <broder> out of curiosity, what does the ec2 ami build process look like? trying to just do some basic auditing - i.e. where does it run, how controlled is the access, etc
[19:21] <Jarmo> or wait...
[19:22] <Jarmo> i had installed those charms i did mention before, i had created new project, added user to it (via dashboard), the logged in as that user, and there was putton export credentials
[19:23] <Jarmo> now it works again!! :https://help.ubuntu.com/community/UbuntuCloudInfrastructure
[19:23] <pdtpatrick> Jarmo:  but can u for instance log into AWS console and create an EC2 instance? do u have that ability?
[19:23] <pdtpatrick> wait what? it works now? what changed?
[19:23] <Jarmo> i mean that guide page :D
[19:24] <Jarmo> remind me AWS stands for?
[19:24] <pdtpatrick> http://docs.amazonwebservices.com/AWSEC2/latest/UserGuide/using-credentials.html
[19:27] <Jarmo> hmmm... now I'm really confused... I dont think I had to use AWS before, and didnt think I have to use it now....
[19:27] <pdtpatrick> yeah looks like that guide is using their API
[19:27] <pdtpatrick> to create an EC2 instance
[19:28] <pdtpatrick> so u'll need an AWS account
[19:28] <pdtpatrick> to be able to create/launch the instances
[19:28] <pdtpatrick> they should probably be more clear on that documentation
[19:28] <Jarmo> agree!
[19:29] <Jarmo> but that explains why I'm having  problems!
[19:29] <pdtpatrick> But the good news is u can sign up for the AWS account and you get a free instance for 90days i think -- i forgot what the promo was but Amazon will tell u when the trial ends either way
[19:30] <pdtpatrick> so once u do that and then u import ur credentials - u should be able to launch those images and u'll be less stressed :)
[19:30] <Jarmo> thank you very much, i have been wrestling with these images whole day! :)
[19:30] <pdtpatrick> and maybe u can blog about it with screenshots hehe. I might add that to mine after i finish my installation. http://thinkfirstblindsecond.com
[19:30] <pdtpatrick> Jarmo: no problem
[19:31] <pdtpatrick> let me know how it goes
[19:31] <Jarmo> i will, after that i think I have to start looking way to deployin own system so wotn have to use AWS :D
[19:32] <pdtpatrick> yeah :) . If u look on the openstack docs, it shows you how to use local. One second i think i have that guide
[19:32] <resno> pdtpatrick: you use openstack?
[19:33] <Jarmo> It is just wierd that guide dont way any word about AWS...
[19:33] <pdtpatrick> resno: trying to -- still new to it myself. I'm just learning on the fly
[19:33] <Jarmo> *way  = say
[19:34] <pdtpatrick> Jarmo:  so in your environments.yaml file
[19:34] <pdtpatrick> is where you'll do that
[19:34] <pdtpatrick> https://juju.ubuntu.com/docs/getting-started.html
[19:34] <pdtpatrick> see where it says ec2
[19:34] <pdtpatrick> i believed i changed that to local for me
[19:34] <pdtpatrick> scroll down a bit - it should show u an example
[19:35] <Jarmo> ok, I dont know way to show how thankful I am at this moment :D
[19:36] <pdtpatrick> no worries - we all help each other right? :) I'm sure I'm going to run into a PITA problem soon. like DRBD cluster!
[19:36] <resno> pdtpatrick: are you using it in "enterprise" or just home use?
[19:36] <pdtpatrick> resno: home. Will be testing it in enterprise in couple of weeks when I'm done with my current project.
[19:37] <resno> pdtpatrick: im on the cusp on setting up our office system and im not sure which of the multitudes of optoins to go with
[19:37] <pdtpatrick> resno: i keep hoping that the process is more streamlined but you've gotta jump through some hoops for now to get it working but once u do - you're quite pleased.
[19:37] <resno> pdtpatrick: you said, yaml file, is it ruby based?
[19:39] <pdtpatrick> resno: well the whole juju idea is quite fantastic actually. Two commands and you have an entire service up is pretty lovely but its very nice. So currently I'm using ESXi cluster. But i'm going to continue testing until i feel comfortable enough to switch. I would have for it to start misbehaving and i'll have to spend hours troubleshooting while servers are down. That would be horrible
[19:39] <pdtpatrick> resno: yaml does not necessarily mean ruby. I believe most of openstack is python
[19:39] <pdtpatrick> it just uses the yaml format to import the configs. They could have easily used JSON but yaml is easy to read
[19:40] <pdtpatrick> so u'll put that in your ~/.juju/environments.yaml
[19:40] <veebull> How can I tell if squid-deb-proxy is working?
[19:43] <pdtpatrick> veebull: did u check the .conf files to see if it listens on some port?
[19:43] <pdtpatrick> veebull: sounds like the squid-deb-proxy is similar to apt-cacher-ng which listens on port 3142
[19:45] <veebull> I know its *running*... I see it in the list of services @ startup, and I can see it running in htop.
[19:45] <veebull> when client machines run updates, the squid process for squid-deb-proxy goes active...
[19:45] <pdtpatrick> does it web interface to show u the hits and misses? if not -- i guess the way to check is to do "apt-get update"
[19:46] <pdtpatrick> the first time it would pull the packages
[19:46] <veebull> but I'm not seeing anything cached in /var/spool/squid3
[19:46] <pdtpatrick> and then run it again - it should be significantly faster since it does not have to out to the web
[19:47] <veebull> where should I be seeing the cache growing at?
[19:52] <pdtpatrick> veebull: okay i just wrote
[19:52] <pdtpatrick> this
[19:52] <pdtpatrick> http://www.thinkfirstblinksecond.com/2012/05/14/apt-cache-ng-on-ubuntu/
[19:52] <pdtpatrick> that should help you set up an apt-proxy
[19:53] <pdtpatrick> This way you can actually see the hits and misses
[19:59] <veebull> I'll keep that in mind... I kind of wanted to find out what was wrong with the setup I have - especially since its supposed to be 'zeroconf'... ;)
[20:00] <veebull> install squid-deb-proxy, avahi-tools, and squid-deb-proxy-client on the server
[20:00] <pdtpatrick> right they want u to use the client
[20:00] <pdtpatrick> so whenever the server downloads a file, it also caches that
[20:00] <veebull> and install squid-deb-proxy-client on the clients, and voila, it *should* be working
[20:00] <pdtpatrick> but you can also see what files the proxy package installed
[20:01] <pdtpatrick> dpkg -L <packagename>
[20:01] <pdtpatrick> then check those configs to see where it puts things
[20:02] <pdtpatrick> for instance - u can tail -f the logs on the server or you can do a watch -d  on the directory where squid should be writing and see if any new files are modified or created
[20:03] <pdtpatrick> but it doesn't seem like it has a web ui - so i'll suggest using apt-cacher-ng as i posted on the blog which gives u the hits and misses so u at least know what's going on
[20:04] <pdtpatrick> and if u wanna go a step further - u can repackage the .deb file so it automatically installs the client 03proxy file. But that's if u wanna geek it out
[20:16] <nyr0x> so, i'm trying to deploy a bunch of server with maas, is there any way to define custom host names that actually work? i can specify them in the web front end but cobbler doesn't pass them to dnsmasq
[20:18] <pdtpatrick> nyr0x: does clobber use dhcp server? i think it gives u the option to have it use its own dhcp server or use one that you already have right?
[20:18] <guntbert> Guest74162: connection problems? :-)
[20:19] <Guest74162> guntbert, yeah :)
[20:20] <pdtpatrick> nyr0x: so if ur using your own dhcp server .. then adding something like this would give your machine a hostname
[20:20] <pdtpatrick> http://pastie.org/3911799
[20:21] <guntbert> Ursinha:  may I suggest you fix them where you are not cluttering this channel?
[20:21] <nyr0x> pdtpatrick: the only running dns and dhpc server is dnsmasq setup with 'sudo aptitude install maas-dhpc' the node is available with 'ubuntu.domain' but not with the hostname i specified in the maas web front end
[20:21] <Ursinha> guntbert, sorry, they are fixed now
[20:22] <guntbert> Ursinha: glad you made it :-)
[20:22] <veebull> ack.  Found it.  /var/cache/squid-deb-proxy => 346MB :) :) :)
[20:22] <pdtpatrick> nyr0x: check and see what files were installed by mass-dhpc and then see where it keeps the dhcp file. Once found, check the file and see if it added your host correctly.
[20:24] <pdtpatrick> veebull: :)
[20:39] <nyr0x> pdtpatrick: apt-flie doesn't show any files, the package only sets the range gateway etc. with orchestra cobbler wrote a host file based on the know systems that would be included in the dnsmasq config. using the maas web front end this is not longer happening, further more it looks like something goes wrong when cobbler installs server because cobbler sets a wrong hostname
[20:39] <pdtpatrick> nyr0x: "dpkg -L mass-dhcp"
[20:41] <pdtpatrick> nyr0x: i'm too well versed on what cobbler should do behind the scenes. I'm using foreman + puppet to provision my servers.
[20:41] <pdtpatrick> http://theforeman.org/projects/foreman/wiki/Screenshots
[20:43] <nyr0x> pdtpatrick: the only file that is written is '/usr/share/lintian/overrides/maas-dhpc' with the content 'maas-dhcp: postinst-uses-db-input'
[20:47] <brainysmurf> Can I setup an ldap that populates the directory with the posix accounts and passwords?
[20:51] <pdtpatrick> nyr0x: sorry i cannot be of further help with cobbler. Haven't used it much myself.
[20:52] <brainysmurf> I know openldap needs a config file, but I was wondering if there wasn't a way to do so automatically?
[21:45] <ARTSIOM> on 12.04 I have created a new user. why it is not picking up path from /etc/environment?
[21:53] <ARTSIOM> I thought variables added to /etc/environment are global...
[21:57] <nathwill> ARTSIOM, does the bashrc in /etc/skel redefine the PATH?
[21:57] <nathwill> ah
[21:57] <nathwill> nm, other variables
[21:57] <nathwill> probably not path
[22:03] <ARTSIOM> anyway I have checked it and scripts in /etc/skel are just trying to enhance PATH with ~/bin folder
[22:04] <JonEdney> I was reading in the Ubuntu Server book, that  you could install an OEM version of the server O/S, but it didn't say what the difference was between the OEM, and going about it the normal way.
[22:04] <ARTSIOM> maybe newly created user should be in some special group to see variables in /etc/environment?
[22:14] <ARTSIOM> one more question: I have added conf file to /etc/init folder. Now I can stop/start my application using service comand, but my application is not starting automatically on boot. how can I configure start on boot?
[22:28] <pdtpatrick> ARTSIOM: if ur using upstart then "start on runlevel [23]"
[22:29] <pdtpatrick> if ur using SysV then "update-rc.d <servicename> defaults"
[22:30] <ARTSIOM> pdtpatrick: I am using upstart. what is the difference between runlevels 2 and 3?
[22:31] <pdtpatrick> http://www.debianadmin.com/debian-and-ubuntu-linux-run-levels.html
[22:31] <ARTSIOM> pdtpatrick: thanks!
[22:31] <pdtpatrick> np
[22:43] <dotnetted> Hey all - Is it possible to get a 12.04 MAAS server on the same box as a node? (Similar to CC & NC on same box w/ UEC) - I'm limited to one physical box and I'd get a single-node cloud running for some general testing. Thanks.
[22:43] <ARTSIOM> back to my variables problem: can someone please explaine the difference between "su user" and "su - user". especially the diffrenece in loading environment variables
[22:45] <dotnetted> I tried installing 12.04 as a MAAS node inside VirtualBox on the same machine as the MAAS server -- The install picked up the MAAS server but shutdown the installation on the next step (And I doubt nested virtualization would be very fun - even if it worked)
[23:09] <koolhead17> dotnetted, even i would love to see if MAAS works in virtualbox
[23:10] <dotnetted> Is it possible to install the maas node-specific stuff after a regular 12.04 server install?
[23:11] <dotnetted> that way I could install server 12.04 on vbox and have log files to tell me when it can't register with the maas server -- instead of the installer crashing on me ;)