[00:03] <smoser> utlemming, could you push on that mirror update ? looks like they're stale again.
[00:53] <tkeith> I apt-get installed python-django-doc, which automatically installed libpython2.6, python2.6, and python2.6-minimal. I then apt-get removed python-django-doc, and then ran apt-get autoremove. This removed libpython2.6, but not python2.6 or python2.6-minimal. Why aren't these other dependencies autoremoved?
[01:01] <rwilson> I wonder if google uses linux
[01:05] <twb> rwilson: what do you think android and chromeos are based on
[01:07] <rwilson> I was speacking in tearms of servers
[01:07] <rwilson> Yes they do http://en.wikipedia.org/wiki/Google_platform
[02:06] <utlemming> smoser: ack
[02:56] <hodgy> Is it possible to setup Ubuntu server to stream music and play it by SSHing into it through Putty and running cmus on it?
[02:57] <hodgy> Say the music is stored on the server, and I ssh on it to it with Putty from a windows machine. I run cmus on the server in the putty window. And decide to play music. How can I have that music stream and come through the speakers on the windows machine?
[03:01] <twb> Apart from cmus, that all sounds fine to me
[03:02] <twb> What I used to do was simply run an httpd (e.g. thttpd or busybox httpd) on the music directory, and then make a little m3u linking to all the tracks
[03:02] <twb> Then your client side just pulls them down magically as ordinary HTTP files (instead of a stream)
[03:02] <hodgy> There is no way I can make cmus do this? It's my player of choice.
[03:03] <twb> hodgy: I am simply not familiar with cmus
[03:03] <hodgy> I happen to love it.
[03:03] <twb> For all I know it works fine
[03:03] <twb> cmus appears to be a unix audio *client*, i.e. it can't serve music streams to a Windows consumer
[03:03] <hodgy> My idea is, music and software is on the server(no speakers/kb/screen), and the ssh session is from the laptop or desktop of choice, running cmus in the terminal window. Making audio play through the system that is remotely connecting
[03:04] <hodgy> The Linux server would actually be running cmus... Only thing active on the connecting system is an SSH window
[03:04] <twb> hodgy: are the speakers hooked up to the ubuntu server, or the windows desktop?
[03:04] <hodgy> Like right now I am on my windows desktop, in putty running IRSSI through SSH on the linux box.
[03:05] <twb> hodgy: where are the speakers
[03:05] <hodgy> Connected to the windows system.
[03:05] <twb> I don't think you can use cmus for this, then
[03:06] <hodgy> What about mpd?
[03:06] <twb> mpd and cmus both assume they run on the host where the speakers are
[03:06] <twb> you need more like icecast or thttpd on the server and winamp on the desktop
[03:06] <hodgy> I could try it I assume, transfer over one or two mp3's via ftp
[03:06] <hodgy> I want to try and avoid client side music players if possible.
[03:12] <hodgy> twb: what is DECnet?
[03:31] <twb> where did that question come from?
[03:45] <delinquentme> zookeeper?
[03:45] <twb> delinquentme: no where hodgy asked out of the blue about decnet
[03:45] <twb> Presumably he's reading a REALLY silly howto blog or something
[03:46] <delinquentme> twb, ? Im asking about its functionality
[03:46] <twb> delinquentme: ok, carry on then
[03:46] <delinquentme> O_o
[03:59] <geek0091> I have achieved true geek, a server rack in my room.
[06:12] <Guest32813> f
[06:21] <whoaski_> anybody awake?
[06:34] <greppy> whoaski_: nope.
[07:11] <whoaski_> thats what I thought
[07:14] <whoaski_> hey greppy if you were deploying a production ecommerce site what would you use?
[07:14] <whoaski_> right now I'm liking Ruby on Rais on ubuntu server but the rest is up in the air. I don't know whats best or fastest
[07:18] <greppy> whoaski_: it would depend on lots of things.
[07:18] <greppy> are you writing an ecommerce site? or just using something off the shelf?
[07:19] <whoaski_> yes I'm wantint currently getting my duck in a rowg to host it myself and all kind of a project of mine. jus
[07:20] <whoaski_> yes writing it
[07:22] <twb> That's probably not the best attitude for a production ecommerce site
[07:24] <whoaski_> would you say that b/c of the risk involved or.... my experience level?
[07:25] <twb> Because it's a hard problem and you're bound to fuck it up
[07:26] <twb> Frankly, the whole problem domain makes me naseous with its crawling salesmen and optimistic puppy developers
[07:26] <twb> *nauseous
[07:27] <whoaski_> I'm not tring to make millions I'm just tring to learn about the whole process
[07:27] <twb> Well then it's for pedagogy, not production
[07:27] <twb> And being a cowboy is acceptable
[07:28] <twb> As soon as it has actual peoples' actual money in it, you'd better be sure you have your shit squared away
[07:28] <whoaski_> even if I fuck it up,  karaoke bars still stay open even when people can't sing
[07:31] <whoaski_> twb i understand you and what your saying, it's just something I wanna do you know, after a few falls I might be good at it
[07:31] <whoaski_> who knows
[07:35] <whoaski_> I've looked at usps api's and I want to incorporate arduino's to help automate fulfillment, and currently tring to learn everything I can about ubuntu servers and hardening them
[07:35] <twb> fulfillment, eh?  This is some kind of humanist ecommerce site, I take it
[07:36] <whoaski_> no more like a giant vending machine
[07:38] <whoaski_> have you ever written your own kernel?
[07:41] <whoaski_> yes I want to sell things but I want to get it from the shelf to a place where a mailman can pick it up automatically
[07:45] <uksysadmin> morning all
[07:45] <uksysadmin> can someone let your ubuntu.com team know the download the final beta link is pointing to Beta1, not Beta2 page.
[07:47] <whoaski_> but my question is what do would you use for a secure webserver    I'm thinking ubuntu server, ruby on rails, apache2, mysql.    and what do you think about nginx?
[07:48] <twb> uksysadmin: #ubuntu-devel might be a better place to mention it
[07:48] <twb> Make that #ubuntu-release
[07:55] <jamespage> morning all
[07:58] <uksysadmin> cheers twb
[08:24] <lynxman> morning o/ morning jamespage
[08:39] <mrintegr1ty> hi, can anyone tell me what the number represents as prepended to various config files (any in /etc/apt/apt.conf.d for example)
[08:43] <xranby> jamespage: morning
[08:44] <jamespage> morning xranby!
[08:44] <mrintegr1ty> if it is a documented feature then i dont know how to find it!
[08:46] <greppy> mrintegr1ty: it's to control the order that they are executed in most of the time.
[08:46] <mrintegr1ty> greppy: but they are config files.. not executed
[08:46] <greppy> ok, fine, "read"
[08:47] <mrintegr1ty> hmm ok
[08:48] <mrintegr1ty> so for /etc/apt/apt.conf.d it contains 01autoremove and 01ubuntu.. does that mean that it doesn't matter which is read first or something else?
[08:48] <greppy> most likely, it will sort on the number first and then the alphas that follow.
[08:49] <mrintegr1ty> something doesn't quite sound right about that.. in the absense of a different explanation i will go with what you say though :)
[08:50] <mrintegr1ty> seems like an ultra basic dependency system
[08:50] <greppy> this is a bad thing? :)
[08:51] <greppy> why make it more complicated than it has to be?
[08:51] <mrintegr1ty> no, guess not
[08:51] <mrintegr1ty> should be documented somewhere though as it's not 100% obvious
[08:51] <mrintegr1ty> i guess udev does it too..
[08:52] <greppy> mrevell: man apt.conf
[08:53] <mrevell> mrintegr1ty, ^^
[08:53] <mrintegr1ty> greppy: hmm i was looking there.. must have missed it
[08:53] <greppy> mrintegr1ty: bah, man apt.conf
[08:53] <greppy> description, item 3
[08:53] <greppy> dangit, item 2
[08:53] <greppy> ( too much multitasking )
[08:56] <mrintegr1ty> greppy: i don't have an ubuntu system with man pages installed so reading this: http://linux.die.net/man/5/apt.conf and I can't see it
[08:57] <uksysadmin> hey all - anybody know if OpenStack Swift 1.4.8 is landing in 12.04? (ttx?)
[08:57] <greppy> mrintegr1ty: http://manpages.ubuntu.com/manpages/natty/man5/apt.conf.5.html
[08:57] <ttx> uksysadmin: err, should already be in there
[08:58] <ttx> uksysadmin: interesting
[08:58] <ttx> uksysadmin: Daviey should know more about that.
[08:59] <ttx> I no longer presides over the contents of Ubuntu Server :)
[08:59] <uksysadmin> cheers ttx - just did an apt-get update and still on 1.4.7
[08:59] <greppy> mrintegr1ty: is there a reason that you don't have man pages installed?  ( the irony of you complaining about a lack of documentation and not having man pages installed is... priceless )
[08:59] <uksysadmin> I'll wait for Daviey to respond
[08:59] <mrintegr1ty> greppy: fantastic, thanks
[09:00] <mrintegr1ty> greppy: haha. i looked extensively through the online docs, i do have manpages installed just not on any of the ubuntu servers. locally on my fedora desktop (which wouldn't help here obviosuly)
[09:01] <mrintegr1ty> greppy: reason being minimal virtual machines don't (usually) need them installed
[09:02] <greppy> mrintegr1ty: ah, I normally install them for the occaision when I don't have access to anything else and only have the man pages to check things against.
[09:02] <mrintegr1ty> also, i wasn't really complaining. i actually assumed that it was probably documented somewhere but i just didn't know what terms to search for
[09:02] <greppy> they don't take up that much space.
[09:02] <mrintegr1ty> true
[10:38] <iclebyte> where can I find v2.5 debs of puppet?
[10:42] <alamar> any ideas on what tool to use to generate "forged" ethernet packets in the shell (meaning i'd like to set src,dst,type etc. manually)
[11:29] <jibel> jamespage, I'm re-running raid1 i386. It failed on grub.
[11:30] <jamespage> jibel, oh - nice
[11:30] <jamespage> amd64 was OK?
[11:30] <jibel> jamespage, yes, same version of grub
[11:56] <dogmatic69> Hi all, I am looking for help in changing my dns config permanently. using /etc/resolv.conf is reset during startup.
[11:56] <dogmatic69> using http://tinyurl.com/cafkwbm in the gui seems to work, but I need to do it on a server.
[11:57] <dogmatic69> I cant seem to find what that is changing to replicate on the server. Any ideas? thanks
[12:00] <rbasak> dogmatic69: are you usingn dhcp on the server?
[12:01] <smb> dogmatic69, If you set static ip addresses in /etc/network/interfaces add dns-nameservers and dns-search to the interface stanca
[12:01] <dogmatic69> rbasak: the router is doing dhcp an I am not able to change it.
[12:01] <dogmatic69> smb: the thing is, that window I pasted does not do the config in /etc/network/interfaces
[12:03] <smb> dogmatic69, You are using dhcp anyway. Maybe somewhere under /etc/resolvconf/resolv.conf.d/
[12:04] <smb> or update.d
[12:04] <dogmatic69> smb: I have gone through all the files int /etc/resolvconf and /etc/dhcp and /etc/dhcp3
[12:04] <rbasak> dogmatic69: dhclient does it automatically.
[12:05] <dogmatic69> cant find anything with the ip I added
[12:05] <smb> rbasak, I think it needs to be different from what dhcp sets for some reason
[12:06] <rbasak> dogmatic69: on a really old server (hardy I think) I can override DNS using a file that does new_domain_name_servers="$old_domain_name_servers" in /etc/dhcp3/dhclient-enter-hooks.d but I don't know if that's still the recommended way and can't remember where it's documented
[12:06] <dogmatic69> smb: I set up a dev server with a dns, so I have *.dev for sites I build.
[12:06] <dogmatic69> running my desktop through the server
[12:07] <smb> I have not done that variation... so unlikely of much help... :/
[12:08] <rbasak> dogmatic69: see /sbin/dhclient-script - looks like that's still current
[12:09] <rbasak> dogmatic69: it allows for overrides
[12:09] <dogmatic69> I will eventually get a better router and that will be easier. Stupid internet provider locks you to their routers
[12:09] <rbasak> dogmatic69: that's the script that actually changes your resolv.conf (when not using resolvconf)
[12:10] <dogmatic69> thanks for the help any how
[12:10] <dogmatic69> better get back to actual work now
[12:11] <rbasak> dogmatic69: how about a cron job to replace your /etc/resolv.conf every minute? :-P
[12:11] <dogmatic69> rbasak: :/
[12:12] <dogmatic69> I cant believe this is so unknown. the amount of blog posts that say edit /etc/resolv.conf (which works, till reboot) and then comments on the post with doggy hacks like that :D
[12:14] <smb> There is a lot of new thinking needed now as the resolvconf replacing /etc/resolv.conf is new
[12:20] <rbasak> The simple reason is that DHCP isn't generally used on servers.
[12:20] <rbasak> (or even if it is, DHCP servers are expected to serve the right thing)
[12:21] <rbasak> And on the desktop, Network Manager allows you to set DNS overrides, and that covers most DHCP users.
[12:39] <zul> morning
[12:51] <Adri2000> zul: hi. why didn't you merge my branch for the horizon bug fix? you included the fix, that's ok, but now I've got a branch + an open merge request lying around in LP. if you really merged the branch, I guess those would have been marked "merged" accordingly?
[12:52] <zul> Adri2000: i did merge it its in the archive right now
[12:56] <jamespage> zul: can you give me an opinion on https://bugs.launchpad.net/ubuntu/+source/dovecot/+bug/970782
[12:57] <jamespage> I've checked the merge and fixed up another issue that the upstream author pointed out to me - mail-stack-delivery works OK
[12:57] <jamespage> but wanted a second opinion on the upstream stability of 2.0.x branch before I talk to the release team
[12:58] <Adri2000> zul: why is https://code.launchpad.net/~adri2000/horizon/rename-apache-config-file/+merge/99574 still open then? I suppose if you did "bzr merge" it would have updated that page, etc.
[12:58] <zul> Adri2000: i merged it against our ci branch, ill close it in a couple of minutes when im done
[12:58] <Adri2000> ok. that's not really important anyway, just wondering why you did it that way
[12:59] <zul> jamespage: is that the only fixes in 2.0.19?
[12:59] <zul> jamespage: and you scanned the ml to make sure there are no regressions?
[13:04] <zul> jamespage: im happy with 2.0.19 but i would talk to ivoks as well
[13:15] <jamespage> zul, lemme take a scan of the ML
[13:23] <jamespage> zul: nothing that I can see - I'd probably just push it up to 2.0.18 as thats been in Debian for a while now
[13:23] <zul> jamespage:  safer than sorry
[13:24] <jamespage> ivoks, if you are around would value your opinion on the above as well :-)
[13:26] <ivoks> jamespage: sure.. let me see what's that we are talking about :)
[13:26] <ivoks> jamespage: oh, dovecot
[13:29] <ivoks> jamespage: lucid's dovecot is of 1.4.x series; meaning that 'mbox: Fixed accessing Dovecot v1.x mbox index files without errors' is a - must have?
[13:30] <jamespage> ivoks, thats a good point
[13:31] <ivoks> that changelog doesn't tell much
[13:31] <ivoks> is that fixed or wasn't broken, but just printing errors? :)
[13:32] <ivoks> if it's just printing errors, while everything works, maybe we can stay with debian
[13:38] <jamespage> ivoks, looking at the commit does not tell much more - but it does do an explicit upgrade step to sort out old indexes
[13:40] <ivoks> sorry, i haven't looked at the code; give me a minute
[13:41] <ivoks> jamespage: right; it looks like it does an upgrade of the mailbox
[13:41] <ivoks> errr. index
[13:41] <jamespage> agreed
[13:42] <ivoks> so, that's a must have
[13:42] <jamespage> ivoks, ack - building now (I'd only merged 2.0.18 so far)
[13:42] <ivoks> it's also something that can be pushed later
[13:42] <ivoks> or just patched to 2.0.18
[13:42] <jamespage> ivoks, TBH the delta for 2.0.19 is quite small - I'd be more comfortable taking the whole release I think
[13:44] <ivoks> jamespage: ack
[13:44] <ivoks> i should find some time for that mail-stack stuff again :)
[13:45] <jamespage> ivoks, its been a bit neglected for the last couple of releases TBH
[13:45] <jamespage> I keep finding time about now in the cycle :-)
[13:45] <jamespage> which is a little late
[13:45] <ivoks> my plan is to work more on it in next releases
[13:46] <jamespage> the drac plugin is completely borked in the current package (Timo pointed this out to me...)
[13:46] <ivoks> too bad it's after lts
[13:47] <ivoks> jamespage: and it supports only ipv6
[13:47] <ivoks> typos, damn typos...
[13:48] <ivoks> jamespage: ipv4 of course
[13:51] <ivoks> oh, that reminds me...
[13:51] <ivoks> zul: around; when i install dashboard, i tries writting in /var/www/.novaclient/ as www-data user; is there a bug about that?
[13:52] <zul> ivoks: yeah fixed in rc3
[13:52] <ivoks> awesome
[13:52] <zul> which i just uploaded like 2 minutes ago
[13:52] <ivoks> zul: you rock :)
[13:53] <zul> ivoks: thats novaclient doing it
[13:53] <ivoks> oh, could be; i wasn't investigating
[14:26] <kirkland> smoser: do you have time to review https://code.launchpad.net/~kermit666/ubuntu/precise/ssh-import-id/newline-fix/+merge/100000 ?
[14:28] <Daviey> kirkland: you need to get the beers in for capturing review number 100000
[14:49] <smoser> kirkland, if i were going to touch that code one more time, i think i'd just remove all the checking.
[14:50] <smoser> you're trusting the other side of the https connection
[14:50] <smoser> the sanity checking has only hurt us
[15:06] <ivoks> win 15
[16:00] <cr3> hi folks, I'm getting this error when running virsh on a configuration containing <os><type>hvm</type></os>: unknown OS type hvm
[16:03] <jcastro> mrevell: hey can we talk maas docs real quick?
[16:04] <mrevell> jcastro, Sure. What works for you. Hangout?
[16:04] <jcastro> IRC is fine
[16:04] <jcastro> I just need to know the canonical location of the docs
[16:04] <jcastro> https://wiki.ubuntu.com/ServerTeam/MAAS
[16:04] <jcastro> and there's also, http://people.canonical.com/~gavin/docs/lp:maas/install.html
[16:06] <kirkland> smoser: yeah
[16:06] <kirkland> smoser: my original implementation had no checking
[16:06] <kirkland> smoser: and just hoped that you knew what you were doing when you put that key
[16:06] <kirkland> smoser: I think jdstrand added most of that checking, IIRC
[16:06] <mrevell> jcastro, The link on the Ubuntu wiki is what we're giving people. That'll always be either the docs themselves or have a link to them. Gavin's docs are generated from the restructured text in the MAAS branch. I'm going to pare back the docs in the branch and link out from them to the wiki. Within time we'll have a chapter in the server guide.
[16:06] <jamespage> SpamapS, OK if I pull your proposed upstream patch for http://code.google.com/p/memcached/issues/detail?id=252 into memcached for precise?  looks like we hit this on i386 during the most recent rebuild test...
[16:06] <mrevell> jcastro, So, ignore Gavin's docs unless you're interested in hacking MAAS.
[16:07] <jcastro> mrevell: ok awesome, that's all I needed, thanks!
[16:07] <mrevell> great :)
[16:10] <jcastro> mrevell: ok so one thing we're trying to figure out
[16:10] <jcastro> is where in the docs we tell people to test.
[16:11] <jcastro> so is it like "follow these 2 pages, and then run the checkbox tests" or are the tests done along with the instructions step by step
[16:11] <mrevell> jcastro, matsubara has been working with balloons to get some checkbox tests ready. matsubara is just at lunch. I'll get him to fill you in as soon as he's back.
[16:12] <jcastro> ok awesome, I'll just chill  until he returns
[16:12] <mrevell> cool
[16:23] <SpamapS> jamespage: totally ok. I talked to dormando about it the other day and he said that instead of 5, we should just make it 20
[16:23] <SpamapS> jamespage: I was looking at that FTBFS the other day actually, so glad you're picking it up. :)
[16:24] <jamespage> SpamapS, it should probably re-check the assertion after the loop as well so if it really does fail we pick it up on that test - I'll revise the patch and attach to the upstream bugtracker
[16:25] <SpamapS> jamespage: I talked to dormando about it.. the only reason I didn't merge it was that he was going to do exactly that, but he just hasn't gotten itto it yet
[16:26] <jamespage> SpamapS, ack - on it now
[16:30] <jcastro> jamespage: what flavor of hadoop is in our charm/hadoop PPA?
[16:30] <jamespage> jcastro, hadoop hadoop
[16:30] <jcastro> Apache Hadoop then?
[16:30] <jamespage> vanilla upstream 1.0.1 (will be 1.0.2 once it gets released)
[16:30] <jamespage> jcastro, spot on
[16:31] <jamespage> well almost vanilla - I had to pull in build system patches for precise and arm
[16:34] <MagicFab> If you use Zabbix or will be using it in Ubuntu 12.04 LTS, please mark this as "affects me too": https://bugs.launchpad.net/ubuntu/+source/zabbix/+bug/972881
[16:36] <EvilResistance> isn't 12.04's repos under freeze?
[16:48] <jamespage> MagicFab, EvilResistance: that does look like a bug fix only release so once its in Debian we should sync up - hopefully early next week looking at the DM's comment in the Debian bug report.
[16:48] <EvilResistance> you assume it'll be uploaded to Debian in time
[16:49] <EvilResistance> if its not in Debian by release, you'd have to have an update done which would dump it in precise-updates
[16:49] <EvilResistance> then everyone would need precise-updates repo added who wants that update
[16:50] <jamespage> EvilResistance, lets hope it is then :-)
[17:05] <zul> jdstrand: fyi https://launchpadlibrarian.net/99952594/buildlog_ubuntu-precise-i386.keystone_2012.1~rc2-0ubuntu1_BUILDING.txt.gz
[17:15] <jdstrand> zul: cool, thanks. am commenting in the bug on your latest
[17:15] <zul> jdstrand: cool
[17:21] <adam_g> jamespage: ping
[17:21] <dork> anyone know if it's possible to use compiz over vnc and a xen domu?
[17:23] <jdstrand> zul: commented
[17:23] <zul> jdstrand: thanks
[17:31] <Rapid2214> Hiya, does anyone have a working Bond (LACP) - Bridge with VLAN support?
[17:35] <patdk-wk> Rapid2214, sure, many
[17:36] <patdk-wk> but then, what do you mean by bridge?
[17:42] <zul> adam_g: i updated the changelog to get it ready for tomorrow
[17:42] <adam_g> zul: for what?
[17:42] <zul> for release tomorrow
[17:43] <adam_g> what pkg are you talking about?
[17:46] <matsubara> jcastro, hola!
[17:46] <matsubara> jcastro, mrevell asked me to join this channel as you're looking for me
[17:46] <jcastro> hey
[17:47] <jcastro> hey is there a maas channel or just this one?
[17:48] <jcastro> matsubara: https://wiki.ubuntu.com/ServerTeam/MAAS/Testing
[17:48] <jcastro> so I've been working on the wiki pages
[17:48] <jcastro> and the one thing left to do is fill in "Testing"
[17:48] <jcastro> which balloons will be filling out
[17:49] <matsubara> jcastro, AFAICT, there's no #maas channel here on freenode
[17:49] <jcastro> ok so this is the home channel then, good. :)
[17:49] <matsubara> I think this is the right place to talk about maas stuff
[17:49] <jcastro> matsubara: ok so I have an appointment so I have to step out for an hour, but my general idea is to build up this set of wiki pages
[17:50] <jcastro> with step by step instructions on what we want people testing
[17:50] <jcastro> and then we should be good to go
[17:50] <MagicFab> EvilResistance, yes, but I'm hoping to get a sync anyways.
[17:51] <matsubara> jcastro, how about the checkbox tests? would them help?
[17:51] <matsubara> they're fairly complete now
[17:53] <marcoceppi> Can I get some clarification on MAAS, the setup docs recommend having two servers for MAAS, is it that MAAS doesn't use virtualization and only uses the entire hardware or does it virtualize/create containers on the baremetal?
[17:57] <matsubara> marcoceppi, if you're using real hardware, then you'd need at least two machines. one for the maas server and another for the node (and if you add juju to the picture you'd need at least 2 nodes + the server)
[17:58] <matsubara> marcoceppi, what the setup guide doesn't cover yet is how to setup everything using virtual machines but I'm pretty sure there's some doc in the maas tree with that info
[17:58] <zul> jdstrand: we already supply a keystone.8 btw
[17:58] <marcoceppi> matsubara: So, in the case of Juju + maas, if I run a juju bootstrap and that uses an entire server for the bootstrap node?
[17:58] <jdstrand> zul: yes, I was saying add the lack of SSL to it :)
[17:59] <matsubara> another thing you can do is to install the maas server on the host and spin up a couple of virtual machines to act as the nodes. I'm trying that use case today and couldn't get it working just yet
[17:59] <matsubara> marcoceppi, yes
[17:59] <zul> jdstrand: ah ok
[18:00] <marcoceppi> matsubara: Ah, okay. I've got some beefy hardware. Using one of these machines for a bootstrap doesn't make sense. But if it's possible to do with virtual machines that might be better. So, I know this is all still new, just fishing, in the even of using maas with virtual machines as nodes, each node would have to exist prior to it being used, correct? Or would maas be able to create these nodes (
[18:00] <marcoceppi> doesn't seem like it would, given that I know what it does now)
[18:01] <hallyn> roaksoax: hey, would you mind adding guest uuid support to testdrive?
[18:01] <matsubara> yes, you'd need to setup your virtual machine to boot up from an precise-server iso and then choose the option to enlist into an existing MAAS
[18:01] <matsubara> marcoceppi, ^
[18:01] <marcoceppi> matsubara: cool, I look forward to playing with this more tonight
[18:02] <zul> jdstrand: done
[18:02] <matsubara> marcoceppi, cool! if you have any trouble setting it up, feel free to ping me
[18:02] <marcoceppi> will do, is there a preferred vm software to use?
[18:02] <claude2_> anyone here having issues with the megaraid_sas driver for a perc 5i with ubuntu 11.10?
[18:02] <marcoceppi> I guess, virtualizing method
[18:03] <matsubara> marcoceppi, I couldn't get enlistment working on my local machine but it might be a problem with my laptop. I'm debugging that
[18:03] <claude2_> the raid keeps coming up read only
[18:03] <hallyn> roaksoax: there's been a request to do it in kvm directly, but while "nothing is impossible" there is no good clean place to do that there.
[18:03] <matsubara> marcoceppi, I use virtual box
[18:03] <roaksoax> hallyn: sure, could you please file a bug with the details?
[18:03] <hallyn> roaksoax: I'd mark bug 959308 as affecting testdrive
[18:03] <matsubara> marcoceppi, but that's because I already had it set up. I think you can use any virtualization method you prefer as long as a) it can boot from the cd b) the guest can reach the host through the network
[18:04] <roaksoax> hallyn: cool thanks
[18:06] <hallyn> roaksoax: I don't know if it makes sense to specify a uuid in testdriverc, but it's also been suggested on qemu m-l that you could use the fs uuid of the root fs
[18:07] <marcoceppi> matsubara: I've got a spare proliant dl380 lying around, I may try to setup a few Xen machines. Would it be preferable for the maas server to be on the same server, or would it be fine to run it from say my desktop or laptop?
[18:08] <marcoceppi> Then treat each virtual machine on the 380 as a node
[18:08] <roaksoax> hallyn: cool, i'll look into it as soon as I have some free time
[18:09] <hallyn> roaksoax: thanks!
[18:09] <matsubara> marcoceppi, it should be fine to run on your desktop/laptop as long as they're in the same network
[18:09] <marcoceppi> matsubara: awesome, thanks. I'll play around this this more when I get home
[18:10] <matsubara> marcoceppi, ok. let me know how it goes. if i'm not around email the list ( maas-devel@lists.launchpad.net) and someone might be able to help you
[18:11] <matsubara> marcoceppi, and thanks for helping testing this!
[18:11] <marcoceppi> matsubara: will do!
[18:14]  * pehden is back (gone 00:34:57)
[18:24] <jamespage> adam_g, pong
[18:35] <adam_g> jamespage: hey about squid
[18:36] <adam_g> missed you yesterday
[18:37] <amarcolino> I have added my user to www:data and made /www writable to users within that group, however, I would like to know how I can make it that when I create file and folders the owner and group should be www:data instead of my user?
[19:02] <jamespage> adam_g, ah yes we did
[19:02] <jamespage> so whats you opinion on how to resolve this?
[19:03] <jamespage> I think we should not be aiming to transform the configuration - it risks to much in terms of edge cases and exposing thing accidentally.
[19:03] <adam_g> jamespage: i agree. i think the only thing we can really do is warn very loudly if we find that they are using a custom squid(v2).conf
[19:03] <jamespage> I'd rather place the users existing squid config somewhere accessible and let then know that they need to rationalise that into a squid 3 config
[19:03] <adam_g> jamespage: perhaps give an option to move it to the squid3 location with a warning that it may not parse or produce the same service
[19:04] <jamespage> adam_g: maybe
[19:04] <adam_g> jamespage: squid.conf (in both v2 and v3) is generated a build time, i went back to lucid and got the hashes for squid(v2): http://paste.ubuntu.com/915009/
[19:06] <jamespage> adam_g, I'll give it a bit of thinking over the next few days
[19:06] <jamespage> anyway have to dash now - later...
[19:07] <adam_g> jamespage: k
[19:08] <thys> hi
[19:08] <thys> where do I configure the  php limit in ubuntu?
[19:08] <thys> ubuntu server
[19:09] <thys> php memory limit
[19:09] <amarcolino> Hi I have added my user to www:data and made /www writable to users within that group, however, I would like to know how I can make it that when I create file and folders the owner and group should be www:data instead of my user?
[19:10] <amarcolino> thys,  /etc/php5/apache/php.ini
[19:11] <thys> I use lighttpd still the same place?
[19:12] <amarcolino> I actually wouldn't know
[19:13] <amarcolino> a quick google search gave this /etc/php5/cgi/php.ini
[19:46]  * pehden is away: I'm busy
[20:52] <Zaitzev> hello
[20:52] <Zaitzev> I just installed proftpd-basic and wonder what I can do about users, shares and whatnot
[20:52] <Zaitzev> anyone here that can help me out a little?
[21:02] <jcastro> matsubara: ok what do you think of this:
[21:02] <jcastro> https://wiki.ubuntu.com/ServerTeam/MAAS/Testing
[21:04] <matsubara> jcastro, it looks great to me. I'm really glad we won't have to duplicate the checkbox test in the wiki :-)
[21:05] <jcastro> is there a summary page with all the maas testing information?
[21:05] <jcastro> or is that on each person's hwdb-submissions launchpad page?
[21:08]  * robbiew just did a header cleanup of MAAS pages...fyi
[21:09] <jcastro> saw that, ta
[21:09] <jcastro> matsubara: what's the bare minimum # of machine to test maas?
[21:09] <jcastro> I'm thinking 3 right? one for maas and then 2 nodes for juju?
[21:10] <matsubara> jcastro, yep, for maas only, 2 is enough, so you can test that one node enlists into the server
[21:10] <matsubara> but if you want to use juju, then you'd need 3
[21:10] <matsubara> one for the server, and one for the bootstrap env and another for the service you're deploying
[21:11] <matsubara> jcastro, I don't think there's any other page with a summary page with all the maas testing info. I think we should use https://wiki.ubuntu.com/ServerTeam/MAAS/Testing as it looks good to me
[21:11] <jcastro> ok, doing the call for testers now then
[21:11] <matsubara> and we can improve from there, so let's make that the official page from now on :-)
[21:11] <jcastro> rock
[21:12] <matsubara> jcastro, great! thanks a lot for helping with this
[21:12] <matsubara> jcastro, also, you probably seem it, but Nicholas blogged about it here: http://www.theorangenotebook.com/2012/04/testing-maas-no-mas-no-poco.html and the blog post has all the contact info in case people get blocked
[21:12] <jcastro> I have 3 people already
[21:12] <matsubara> s/seem/seen/
[21:13] <jcastro> but with real hardware they can't just up and test right away, so it gave me some time to fix up the docs
[21:13] <jcastro> matsubara: yeah I am reblogging what he posted since he's not on planet yet.
[21:13] <matsubara> cool
[21:14] <matsubara> jcastro, and feel free to tell people to ping me here or email me if they need help. I'll help as much as can, and if I don't know I'll find someone who can help
[21:15] <jcastro> oh don't worry I got one guy who was so lost in the docs I am sure he'll have an earful for ya. :)
[21:15] <jcastro> though I do like how in one day we're already better than the orchestra docs ever were. *snicker*
[21:21] <matsubara> :)
[21:27] <jcastro> matsubara: hey so, I promise this isn't a trick question. But did we forget to announce MAAS on the server list?
[21:27] <matsubara> jcastro, don't know. mrevell will be able to help you with that as he's doing the announcement
[21:27] <sako> how can i check on which software raid# is configured?
[21:42] <jMCg> Hey folks o/~
[21:43] <jMCg> since I enabled ufw, I see tons olf loglines in syslog which look like this:
[21:43] <jMCg> Apr  4 21:43:09 monitoring kernel: [1539121.328202] [UFW BLOCK] IN=eth0 OUT= MAC=52:54:00:9f:f6:ef:fe:54:00:9f:f6:ef:08:00 SRC=176.9.39.37 DST=176.9.55.236 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=48470 DF PROTO=TCP SPT=879 DPT=40222 WINDOW=14600 RES=0x00 SYN URGP=0
[21:44] <jMCg> Yeah. Someone scanning for an exploit.. in.. exchange?
[21:45] <jMCg> Or something.
[21:45] <jMCg> Anyway. I think I should like to reduce the noise.
[21:46] <jMCg> Looks like I'll have to disable logging, because it already *is* on low
[21:51] <jMCg> I see I'm talking to myself again.
[22:36] <jbicha> hi, this is awfully late in the release cycle but would updating libvirt to 0.9.11 be considered?
[22:36] <jbicha> Debian just uploaded it to unstable today
[22:37] <jbicha> I was asking because gnome-boxes needs libvirt-glib 0.0.5 which needs libvirt 0.9.10
[22:38] <hallyn> zul: ^
[22:39] <hallyn> Daviey: ^
[22:39] <zul> ummm.....i would say no
[22:39] <hallyn> jbicha: we had a merge of 0.9.9 from unstable, but nacked it bc we had no compelling need
[22:40] <zul> what we are like a month a way from release arent we?
[22:40] <hallyn> less
[22:43] <zul> so yeah...nack from me
[22:44] <jbicha> ok, if it's a no, I think I'll backport the quirky libvirt (when it's ready) to the GNOME3 PPA then
[23:03] <jbicha> thanks, I was assuming it wouldn't happen but thought it was at least worth making sure! :)