19:00:47 #startmeeting Cloud SIG 19:00:47 Meeting started Fri Dec 2 19:00:47 2011 UTC. The chair is robynna. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:47 Useful Commands: #action #agreed #halp #info #idea #link #topic. 19:00:52 #meetingname Cloud SIG 19:00:52 The meeting name has been set to 'cloud_sig' 19:01:01 #topic YO YO, who's here 19:01:07 chirp, chirp 19:01:11 * jforbes might be 19:01:12 * russellb hops 19:01:41 is it spring? hopping and birds chirping? 19:01:44 * robynna checks the weather 19:01:59 Am here 19:02:08 * ke4qqq shows up for once 19:02:15 robynna: Decided to change things up on IRC? 19:02:25 gholms: it's more awesome than that. 19:02:31 traylenator: welcome 19:02:32 LOL 19:02:40 gholms: i ... yeah, it's too mortifying. i'll tell you later 19:02:46 ;) 19:02:52 * robynna moves onwards 19:02:55 mr. darcy, welcome 19:03:02 Hey, ke4qqq 19:03:22 kewhoqqwho? 19:03:26 hi gholms 19:03:29 ;) 19:03:33 #topic Agenda for today 19:03:45 We're going to play a new game. Since we have all these new people hanging out and whatnot. :) 19:03:47 here 19:03:59 Normally I'd talk about ec2 stuff, but, well, that's less pressing. My, how things have changed. 19:04:04 thermonuclear war? 19:04:15 Hehe 19:04:21 Well, there's that. 19:04:35 So I'm looking for YOUR TOPICS today. 19:04:36 The only winning move is not to play 19:04:45 * robynna suggests cloudstack and eucalyptus packaging 19:04:55 :D 19:05:35 now now, people 19:05:39 I know. 19:05:44 #topic Gluster/HekaFS 19:05:50 since johnmark is the last to arrive. 19:05:58 lol... thanks 19:06:01 I hear there's a BoF this week at LISA. 19:06:10 so you're encouraging me to arrive late every time? ;) 19:06:13 anything else exciting to announce or share? 19:06:17 robynna: why yes! yes there is! 19:06:17 Rumor has it that I'm supposed to attend. 19:06:22 * robynna looks at jdarcy and kkeithley 19:06:22 jdarcy: rumor has it... 19:06:25 and kkeithley 19:06:31 indeed 19:06:41 so yes, there's a BoF at LISA on 12/7 at 8:30 pm 19:06:50 any report on multisite gluster? 19:06:50 if you're in the neighborhood... 19:07:05 sdake: I'm hopeful that kkeithley and/or jdarcy can speak to that 19:07:09 which is why I'm inviting them :) 19:07:40 speak to what? 19:07:57 I can babble about it. 19:07:57 kkeithley: sdake's question - which I took to refer to multi-tenancy 19:08:11 running gluster wide area 19:08:12 johnmark: I think he was asking about wide-area replication, actually. 19:08:20 ah, yes. I suppose I can babble about it too, al little bit 19:08:25 * gregdek lurks. 19:08:28 sdake: ah! well, we do have geo-replication, if that's what you're asking 19:09:08 jdarcy: your babbling is other people's best coherent statement 19:09:34 johnmark: Truly, I have a dizzying intellect? 19:09:37 well, someone feel free to babble 19:09:56 sdake: what do you want to know, specifically? 19:09:59 No meeting is complete without a Princess Bride reference. 19:10:23 jdarcy had indicated there were some problems with wide area replication - thought that was what purpose of hekafs was 19:10:46 sdake: Long term, yes. Just barely getting started on that part, though. :( 19:10:53 sdake: hekafs was mostly abotu multi-tenancy and SSL 19:11:14 ie. making glusterfs safer for hosted service providers 19:11:21 got that 19:11:24 jdarcy: if I mangled that, feel free to correct me :) 19:11:43 ok well i'll take a look at georep, jdarcy sounds like georep isn't sufficient for wide area? 19:12:11 johnmark: Only by necessity, i.e. those features were missing from current offerings (not just GlusterFS). It just had to get done first, but the "vision" that I started with had more to do with wide-area multi-master etc. etc. etc. replication. 19:13:03 sdake: Georep will do wide-area, but single-master with limited failover (and even less failback), plus it's essentially unordered. 19:13:26 sdake: It handles an important disaster-recovery use case, but beyond that it kind of fails. 19:13:27 the writes are unordered you mean 19:13:47 sdake: Epochs are ordered, writes within an epoch aren't. 19:14:03 that should be ok for HA case where only 1 vm is ever writing 19:14:07 ok thanks for update 19:14:18 I should also note that it's a fairly recent feature 19:14:24 as in March, 2011 19:14:35 nothing like a green machine to test on :) 19:14:44 sdake: awesome :) 19:14:56 jdarcy or kkeithley or johnmark: any other updates before i mustard on: 19:14:57 ? 19:15:02 mustard? 19:15:05 progress 19:15:09 I like it brown and spicy 19:15:14 johnmark: TMI 19:15:17 LOL 19:15:30 I'll take that as a no. 19:15:37 #topic OpenStack 19:15:47 oh- just that you can expect a stead stream of announcements re: glusterfs project wrangingly 19:15:53 * robynna looks aroud for anyone having anything to discuss here 19:15:54 starting next week 19:16:13 * robynna looks at russellb, ke4qqq 19:16:32 ie. advisory board, project governance, feature priority, release schedule, etc. 19:16:41 #undo 19:16:41 Removing item from minutes: 19:17:04 #info expect a steady stream of announcements re: glusterfs project wrangling; ie, advisory board, project governannce, feature priority, schedule, etc. 19:17:05 johnmark: is that stuff being discussed anywhere publicly - or just being decided by fiat? 19:17:51 ke4qqq: the discussion starts next week 19:17:53 no fiat 19:18:11 ahhh /me assumed by announcements that decisions had been made - my bad 19:18:25 i think advisory board may be the key word here, perhaps 19:18:30 some decisions have been made - as in, who we're inviting to be on the board 19:19:31 .... anything else? 19:19:35 but the announcements will be centered on how we're moving the project from essentially a wholy-owned open core thing 19:19:38 to something better 19:19:40 ok, done :) 19:19:43 * robynna is really moving on now 19:19:47 #topic openstack 19:19:58 lol 19:20:33 * robynna looks around 19:20:33 asalkeld and I have got openstack integration with pacemaker cloud (HA project) 19:20:51 and some fedora folks are making good progress getting involved upstream 19:20:52 if you want to try it out, use git master ;) 19:21:04 #info sdake and asalkeld have openstack integration with pacemaker cloud (HA project) 19:21:06 sdake: link 19:21:07 ? 19:21:08 russellb: that's awesome news 19:21:11 markmc got nominated into the nova-core group (able to approve patches and such), so that's pretty cool 19:21:25 russellb: which parts upstream? nova? glance? 19:21:26 Nice 19:21:36 mainly nova upstream so far 19:21:37 cool 19:21:49 #info markmc got nominated into nova-core group (able to approve patches, etc.), more involvement in general upstream (mostly nova thus far) from fedora folks 19:22:24 link -> http://www.pacemaker-cloud..org 19:22:31 one dot 19:22:32 ;) 19:22:33 markmc also has some more detailed notes about status in the last week on his user page: https://fedoraproject.org/wiki/User:Markmc 19:22:56 sdake: didn't angus write up a blog post on it too? 19:22:59 #info status from past week in https://fedoraproject.org/wiki/User:Markmc 19:23:24 russellb that was the mashup, we have actually got it functional now in last few days with restart of vms and resources 19:23:31 ah, great 19:23:34 without 100 manual steps 19:23:45 sounds like an improvement :) 19:23:59 :) 19:24:05 mustard ftw 19:24:25 * robynna applauds appropriate use of mustard 19:24:29 okay, anything else in openstack-land? 19:24:36 and there was a proposal for an openstack hackfest at FUDCon 19:24:40 oh, yes 19:24:48 didn't see any responses to it other than me, heh 19:24:52 #info proposal for openstack hackfest at FUDCon, see mailing list for info 19:25:14 though i'm curious who is planning to be there from the Cloud SIG, not just openstack 19:25:25 #link http://www.fedoraproject.org/wiki/FUDCon:Blacksburg_2012 19:25:32 let's see. 19:25:35 I am. 19:25:41 maybe we could do a cloudy bof or something 19:26:00 russellb: I'll be there 19:26:04 I believe ke4qqq, johnmark, gregdek, hopefully gholms, ... mull? yes? ... mdomsch, russellb 19:26:06 as will I 19:26:11 cool 19:26:17 I hope so. 19:26:45 * gregdek yes. 19:26:59 okay, onwards! 19:27:02 :) 19:27:04 #topic Eucalyptus 19:27:10 * mull hides 19:27:15 now that i have mull and gregdek and gholms all paying attention 19:27:24 D: 19:27:29 ? 19:27:40 LOL 19:27:55 * gregdek notes that many of us are in a #eucalyptus-meeting at the same time. 19:27:56 Hi guys! 19:27:59 yar. 19:28:05 Southern_Gentlem: go ahead 19:28:07 haven't had a chance to dirve deep into eucalyptus yet, is there an imaging service? 19:28:10 I think I'll have more to say on this topic next week 19:28:24 robynna, pm 19:28:27 mull: okay. just giving you the opportunity to troll/beg for reviews or etc if needed 19:28:34 kinda busy prepping something right now. :-) 19:28:45 I did get three or four new packages approved last week 19:28:59 mull: gotcha 19:29:05 sdake: Join #eucalyptus-meeting where we're talking about that project right now. 19:29:10 * robynna looks at sdake's question and wonders if there is an answer 19:29:13 oh, and there is the answer 19:29:29 I still need to respond to spot for my review. 19:30:15 #info please reserve your rooms for fudcon before the block runs out, as in, ASAP please :) 19:30:24 robynna: thanks for the tip 19:30:44 but who else would possibly be going to blacksburg in january? 19:31:11 johnmark: well, there are 127 people signed up 19:31:20 ;) 19:31:26 how many rooms in the block? 19:31:43 50 and 34 have been reserved 19:31:46 russellb: right now, 50, but i am assuming that number can be bumped upwards if we meet this number 19:31:53 * sdake selling soul for travel budget 19:31:56 * russellb is 1 of 34 \o/ 19:32:08 sdake: haha, yeah, would be cool if you could be there 19:33:12 okay, i'm moving onwards from euca 19:33:17 #topic cloudstack 19:33:19 also if you have a room and are looking for a roomamate let me know 19:33:20 ke4qqq: beep beep 19:33:28 Heh 19:33:31 what am i talking about? 19:33:38 ke4qqq: whatever you want :) 19:33:48 * gholms recommends bacon 19:33:51 ke4qqq: you can start with the weather 19:33:55 we are packaging things - targetting F17 - and we'll ship with bacon 19:34:03 Yay! 19:34:04 nom 19:34:13 ke4qqq: beefy miracle doesn't like bacon 19:34:16 ke4qqq cloudstack have an imaging service? 19:34:16 though slower than intended at the moment. 19:34:38 sdake: not really - but an intermediary that deals with images, but doesn't create them itself. 19:35:09 cloudstack have a mechanism to provide custom image uploads? 19:35:14 sdake: I"m curious why you keep asking about imaging services 19:35:28 sdake: yes 19:35:34 probably looking for what supports the building blocks for pacemaker-cloud? 19:35:37 my take is that if you have a decent filesystem, an imaging service is almost, but not quite, moot 19:35:51 let me rephrase that - a filesystem that supports multiple means of access 19:35:53 pacemaker cloud test harness creates images using oz and uploads them to various cloud platforms 19:36:02 sdake: ah, ok 19:36:15 sdake: so we are odd in that we require an url to pull rather than accept upload 19:36:18 no imaging servce, no way to have appropriate test harness 19:36:26 sdake: got it, ok 19:37:03 thanks ke4qqq i'll dig into the docs 19:38:25 ke4qqq: have you been able to get with the java sig crew with your package java dramz? 19:38:36 robynna: some of them 19:40:13 ke4qqq: cool 19:40:17 okay 19:40:31 * robynna is going to move onwards, and slightly backwards, to another topic 19:41:05 #topic Eucalyptus, round 2, plus image building, I think 19:41:09 #chair gregdek 19:41:09 Current chairs: gregdek robynna 19:41:27 Doh. 19:41:28 :) 19:41:41 yeah. 19:41:44 Talk. 19:41:58 OK. 19:42:27 * gregdek thinks a bit. 19:42:41 We've got this concept in Euca-land. 19:42:45 * robynna gets out the wd-40 19:42:57 That we're working on right now in a meeting in #eucalytpus-meeting. :) 19:43:07 so we should just go there? 19:43:14 It's basically a store of images and/or recipes that are Euca-ready. 19:43:31 johnmark: yes, if you want to, but not required. They're talking about details. 19:43:35 This is a good place for a summary. 19:43:47 (and trolling for help from a fedora-angle) 19:43:55 LOL, yes. 19:44:07 In a nutshell, Euca needs images to run. 19:44:12 We have a set of starter images. 19:44:20 Beyond that, we want to make it as easy as possible to: 19:44:41 1. Allow users/partners/etc. to contribute their own images; 19:44:49 2. Make it easy for Euca users to find those images; 19:45:05 3. Make it easy to pull Amazon images in. 19:45:27 So we're working on those problems right now, as we speak. 19:45:42 And Fedora obviously falls into that category. 19:45:55 Everything that is a "spin" is also a potential image for Euca. 19:46:45 #info Summary: Euca needs images to run. They have starter images, but want to (1) allow users/partners/etc. to contribute their own images; (2) make it easy for euca users to find those images; (3) make it easy to pull amazon images in. 19:46:51 Are there things in Fedora-land that can make that easier? 19:47:03 It appears that obino (Graziano Obertelli from Eucalyptus) has worked with mgoldmann to set up Euca as a target platform for Boxgrinder, so we can certainly play with that. 19:47:28 If there's a standard process that Fedora is going through for its "official AMIs", leveraging that process might also be useful. 19:47:38 Which, AIUI, is Koji-based right now. 19:47:50 Right. Unless I'm oblivious. 19:47:53 * robynna looks at jforbes 19:47:57 we have a project at https://projects.eucalyptus.com/redmine/projects/images/wiki that explains some of the issue and progress 19:48:15 gregdek: awesome stuff. I'd like to make it easier for your users to deploy Gluster, so would like to know more about generating images 19:48:16 robynna: Yes, it's Koji-based right now 19:48:16 * obino ops just behind the conversation 19:48:41 well, standard process is releng generated by koji 19:48:43 johnmark: is Gluster just an RPM? 19:48:45 gregdek: you prolly want to look at boxgrinder instead 19:48:53 gregdek: yes 19:48:57 If we want additional image formats, we beg Red Hat Engineering Services to add the functionality to Koji 19:49:29 jsmith: not true 19:49:36 dgilmore: Oh? 19:50:14 johnmark: are you working on a Fedora+Gluster "spin/whatever" to show off what Gluster can do? 19:50:21 jsmith: well some of it can be done via kickstart, or someother way 19:50:39 gregdek: that's what I have in mind, but not just limited to fedora 19:50:50 * gregdek nods. 19:50:55 jsmith: it depends on the change 19:51:16 gregdek: and I'm currently creating an open req for someone to do cool shit like that :) 19:51:18 * robynna thinks that many of us want to solve the same problem (how to build an easy way to show off multiple types of cloudy infrastructure things) 19:51:20 just fyi... 19:51:38 * robynna wonders if that's up ke4qqq's alley as well 19:51:38 johnmark: So I'm unsure of details of config et al, but it would certainly be interesting to have a Fedora+Gluster image in the Euca store. 19:51:39 robynna: ah geez, are you suggesting we work together??? 19:51:51 johnmark yes the provisioning toolchain problem is being solved by 10+ different projects 19:51:54 johnmark: i would never recommend such insanity ;) 19:51:57 sdake: wow, ok 19:52:01 robynna: lol :) 19:52:02 sdake: I think this is lower level than what you are thinking 19:52:03 what is up my alley? 19:52:11 robynna: well, it *is* a good idea 19:52:18 unless we're talking about multiple projects here 19:52:28 sdake: is that all - I thought there were 10+ inside rht alone 19:52:46 iirc, the euca thing *started* with the idea of making a live CD that can be a nice demo 19:52:48 ke4qqq probably more then 10 ;) 19:52:58 i'm not sure if that's grown to a larger idea, or if we're talking about multiple ideas 19:53:00 There's a lot of tools for building images. We just need help getting those tools targeting Euca images. 19:53:03 different ideas 19:53:13 Gluster is just a package, submitted bug 19:53:45 So I guess the thing is, are there specific actions we want to identify that we can/should take? 19:53:51 To make it start post instalation, it does not currently but could i expect. 19:54:09 gregdek: my repeated stance is that bg is still the best tool out there for taking images and targeting diverse platforms. 19:54:27 traylenator: it could with an appropriate spin/remix, i would expect 19:54:51 OK. We've got an action item in the other meeting for one of our guys to explore BG. 19:55:05 So maybe we should just start playing with BG and circle back. 19:55:14 Do you have a bunch of BG recipes for Fedora images? 19:55:33 traylenator: right 19:55:38 gregdek: a few here and there - I am using BG to spin up a bunch of instances for a packaging class at lisa and using that. 19:55:50 gregdek: have you looked at it at all? 19:55:54 Any good place to share BG recipes? 19:55:58 and it's super EASY 19:56:00 * robynna notes that mgoldmann is a huge source of info, but is in europe 19:56:00 robynna: Enough to understand the gist. 19:56:01 gregdek: not yet 19:56:09 ke4qqq: Maybe we should build that. 19:56:11 HEY MGOLDMANN 19:56:14 morning! 19:56:16 Hey mgoldmann :) 19:56:23 someone hailed you.l 19:56:25 I have to think. 19:56:26 gregdek: sure - if you're at lisa - lets talk this week 19:56:27 I'm late to the party, right? 19:56:32 ke4qqq: I am not. 19:56:35 anything left to drink? 19:56:35 I'm in SB. 19:56:38 Hi, mgoldmann! 19:56:53 hello mgoldmann 19:56:54 mgoldmann: very - I've had to play mgoldmann in the meeting and promote BG - and I am a terrible actor. :) 19:56:59 mgoldmann: gregdek was talking about the ability of boxgrinder to maybe kick out eucalyptus images... and other stuff 19:57:07 robynna: gregdek: obino: gholms: ke4qqq: morning! 19:57:10 over my head but I'm here to play cheerleader 19:57:15 rah rah cloud! 19:57:23 * robynna shakes her pompoms 19:57:24 mgoldmann: the more interesting thing - a place to share aapls 19:57:27 ke4qqq: I'll read the logs! :) 19:57:44 lol 19:57:47 mgoldmann: howdy :) 19:57:52 YAY CLOUD! 19:57:54 mgoldmann: isn't that one of the things torquebox might be able to eventually do? 19:57:56 johnmark: hiya! 19:58:06 aapl: is that the name of a bg recipe? 19:58:08 be a backend to store / share? 19:58:10 robynna: we can use TB to do this ,sure 19:58:15 but TB is just a platform 19:58:17 aapl is apple stock ticker, iirc :) 19:58:24 Yes, that's what I thought :) 19:58:33 which could allow us to write the web stuff in rails for example 19:58:38 feel free to share your aapls with me, if you have any 19:58:47 gregdek: yes or maybe it's appl - I keep them different 19:59:10 ke4qqq: is it as simple as "this is a single recipe text" that can just be slapped up on github somewhere? 19:59:16 gregdek: yes 19:59:17 gregdek: i think the short answer is, there's no place that already has appliances from a variety of people, templated 19:59:21 simpler than kickstarts 19:59:22 but it's pretty simple 19:59:24 ke4qqq: wow 19:59:27 one sec 19:59:35 * gregdek waits for the expert to weigh in :) 19:59:36 It's a single yaml file that more or less emulates a kickstart. 19:59:43 https://github.com/boxgrinder/boxgrinder-appliances 19:59:52 So it does exist already. 19:59:54 https://github.com/boxgrinder/boxgrinder-appliances/blob/master/BoxGrinder/boxgrinder-meta.appl 19:59:55 * gregdek looks. 19:59:56 (Though they can inherit from one another) 20:00:22 https://github.com/boxgrinder/boxgrinder-appliances/blob/master/CentOS/CentOS-lamp.appl this is for example community submitted 20:00:23 http://boxgrinder.org/tutorials/appliance-definition/ <-- sample at bottom of this page, too 20:00:39 yay CentOS!~ 20:00:43 http://boxgrinder.org/ < sample in the middle of the page :) 20:00:46 mgoldmann: ahhh interesting - I was unaware of the existence. 20:01:15 * gregdek click whirr click click whirr... 20:01:20 :) 20:01:23 * robynna passes gregdek the wd40 again 20:01:53 * ke4qqq notes that he feels like the BG evangelist in NA sometimes - :) 20:01:58 So why do we not have scripts that auto-build these recipes in the Eucalyptus Community Cloud? 20:02:08 * mgoldmann will buy some beers ke4qqq 20:02:20 * robynna will take those beers on ke4qqq's behalf 20:02:30 gregdek: you don't have delivery plugin for euca yet 20:02:34 i will just... pick them up in brno or something 20:02:37 meet you halfway 20:02:39 gregdek: in general - we don't have support for eucalyptus now in BG 20:02:50 and might want to do some platform specific stuff. 20:02:53 mgoldmann: ok. So we need to work on that. 20:03:06 last time I checked there were issues with uploading stuff to the eucalyptus cloud 20:03:11 we have a draft of the plugin ready 20:03:28 I filled a few bugs and some of them should be resolved now 20:03:36 so I need to check this one more time 20:03:40 Oh yeah, those bugs. 20:03:50 sorry for not being specific, it was some time ago :) 20:03:53 * robynna wonders if those bugs have links or this draft has a location :) 20:03:55 Something tells me they were "fixed in eucalyptus 3." 20:04:03 gregdek: me too 20:04:41 robynna: looking for it 20:04:58 * gregdek is shifting between 5 different tasks, sirry 20:05:00 sorry 20:05:08 OK. 20:05:10 gholms: do we have the patch? Can we backport them to 2.x? 20:05:44 https://github.com/boxgrinder/boxgrinder-build/commit/5b2e35aef5f528d1d5a593096b7702276892d560 20:06:01 but it was some time ago and it's most probably not sync with master now 20:06:16 * gregdek nods. 20:06:21 obino: I would need to take a look what's the current state of it 20:06:22 obino: At least some of his bugs were long enough ago that they should be in devel. 20:07:40 * robynna notes that boxgrinder has a mailing list that might be helpful for gregdek to be on as well 20:07:43 and an irc channel 20:07:50 aye! 20:07:58 http://boxgrinder.org/community/ 20:08:09 gholms: let;s look at them later on if you have time 20:08:16 gregdek: AND STICKERS WITH MONKEYS 20:08:19 * gholms headasplodes 20:08:28 awesome :) 20:08:39 * mgoldmann wonder wtf happened to robynna's nick 20:08:45 wonders* 20:08:52 mgoldmann: we will not talk about right now 20:08:53 :) 20:08:54 robynna: We can has storytime later? 20:09:13 * mgoldmann is eagerly waiting for open floor 20:09:15 gholms: the machine where my irc is at is... um... well, unavailable 20:09:18 OK, I've got enough to melt my brain for now. 20:09:20 * gregdek is done. 20:09:23 because I forgot my password 20:09:25 and i locked myself out 20:09:30 is about the bottom line 20:09:30 :D 20:09:31 :) 20:09:41 i'm one of those users, today 20:09:45 anyway 20:10:01 self-ownage 20:10:03 #topic Open Floor 20:10:32 HELLO. Does anyone else have topics they would like to discuss?? 20:10:44 mgoldmann: are you comign to fudcon? 20:10:54 robynna: that's a damn good question! 20:10:56 :) 20:11:03 maybe 20:11:07 mgoldmann no ubuntu plugin for boxgrinder? 20:11:15 sdake: there is one :) 20:11:22 oh - webpage must be out of date 20:11:35 nope, it's written by someone from the community 20:11:41 lemme find the link 20:11:51 http://boxgrinder.org/tutorials/boxgrinder-build-plugins/ 20:11:55 debain/ubuntu not listed 20:11:56 robynna: if I get my budget approved - I'll be there 20:12:09 sdake: that's true - it's not part of BG 20:12:26 sdake: https://github.com/rubiojr/boxgrinder-ubuntu-plugin 20:12:44 mgoldmann: why not? 20:13:02 mgoldmann: i saw that bob is coming ;) 20:13:02 should get some fedora packaging on that 20:13:03 ke4qqq: exactly what I was wondering 20:13:10 ke4qqq: because I need to sort out one important thing :) 20:13:24 currently you cannot build ubuntu on fedora 20:13:28 and vice versa 20:13:50 odd - oz builds ubuntu on fedora 20:13:51 so, submitting BG to ubuntu would make sense with ubuntu plugin only, for example 20:13:57 is there some technical limitation? 20:14:06 sdake: sure, but it uses completely different method 20:14:37 sdake: oz uses KVM - I don't want to use KVM because it's dead slow on VM's => image in 7 hours? 20:14:54 ya it is a bit slow about 5 mins on my laptop to generate a jeos 20:14:58 i'll give bg a try 20:15:05 sdake: on bare metal :) 20:15:13 BG deployments are mostly on vms 20:15:14 bg will never be able to generate ubuntu images on fedora then? 20:15:20 correct 20:15:38 unless we find a nice way to work around the limitations 20:15:48 but there is no plan for it now 20:15:48 Well, fedora could grow apt/deb support... :P 20:16:00 something like this ^ :) 20:16:06 there are apt/deb tools that were at least up for review at one point 20:16:09 and an asteroid could fall to earth in my lifeitme ;) 20:16:16 Yeah, and that stalled out. 20:16:24 2012 is close 20:16:26 hey, those guys on meterorite men make a crapton of money off asteroids falling to earth 20:16:27 yeah - I was disappointed 20:16:32 or i gues those are meteorites 20:16:34 never mind 20:16:44 * gholms wanted to be able to run pbuilder on fedora 20:16:45 meteorwrong! 20:16:49 * robynna scowls at sdake laughing at her 20:17:38 well, meteorites are just small asteroids 20:17:52 btw, there is no reason why we couldn't add support for oz in BG :) 20:17:55 or big ones 20:18:29 mgoldmann: i don't follow 20:18:53 oz is just a builder, BG is a whole suite with conversion and delivery 20:18:58 OZ can be BG back-end 20:19:08 like appliance-creator is now 20:19:09 kkeithley: yeah - meteorites are a subset of meteors which are a subset of asteroids. 20:19:15 mgoldmann: interesting 20:19:35 we just need to convert .appl into file eatable by OZ and run it 20:19:42 sdake: who works on Oz? 20:19:52 chris lalancette 20:19:53 clalance ;) 20:19:55 clalance and imcled 20:20:04 clalance is gone 20:20:07 ian mcloud 20:20:07 yeah 20:20:12 but still alive! 20:20:14 pl 20:20:17 gah... ok 20:20:18 i don't think it's spelled that way (re: ian) 20:20:22 mcleod 20:20:24 gone from rht but still with the community :) 20:20:37 sdake: got it. cool 20:20:53 * ke4qqq apologizes, but needs to depart 20:21:03 have fun at LISA 20:21:12 * gholms notes we are way overtime 20:21:15 gholms: indeed 20:21:21 robynna: OH WE WILL 20:21:21 i blame sdake 20:21:28 it turned into the cloud sig0 meeting today 20:21:29 robynna: ftfy: "have fun with LISA" 20:21:53 mgoldmann: ........ 20:22:03 see ya 20:22:11 mgoldmann: it was directed at ke4qqq 20:22:18 robynna: http://instantrimshot.com/ 20:22:24 gholms: thanks 20:22:26 LOL 20:22:44 * gholms needs to leave now 20:23:01 * robynna is going to end meeting 20:23:06 Thanks for the discussion, everyone. Hopefully there will be good news in euca-land next week. 20:23:08 thanks for playing today, kids :) 20:23:15 gholms: keep an eye on gregdek's brain 20:23:19 #endmeeting