17:00:26 <samkottler> #startmeeting cloud WG weekly meeting
17:00:26 <zodbot> Meeting started Wed Jan 22 17:00:26 2014 UTC.  The chair is samkottler. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:44 <mrunge> o/
17:00:49 <samkottler> #chair rbergeron number80 mattdm jzb mrunge
17:00:50 <zodbot> Current chairs: jzb mattdm mrunge number80 rbergeron samkottler
17:00:57 <samkottler> hey hey
17:01:15 <jzb> Good morning, afternoon, or evening as the situation dictates.
17:02:28 <samkottler> #topic rollcall
17:02:31 <samkottler> o/
17:03:53 <mrunge> o/
17:03:59 <jzb> o/
17:05:25 <samkottler> #topic PRD recap, remaining work, etc.
17:05:48 <samkottler> huge thanks for that last big push on the PRD
17:06:03 <samkottler> things are looking good from FESCo, already have gotten two +1's
17:06:22 <mrunge> oh yes, thanks to all contributors!
17:07:07 <samkottler> I don't know of anything that's left with it right now unless we get some negative feedback from fesco
17:07:18 <samkottler> mattdm: rbergeron: jzb: got anything on your mind regarding the PRD?
17:07:47 * mattdm shows up
17:07:54 <mattdm> hi sorry. losing track of time :)
17:08:04 <jzb> samkottler: not at the moment
17:08:05 <mattdm> double huge thanks to everyone
17:08:09 <jzb> samkottler: glad it's done :-)
17:08:13 <jzb> sorry I didn't do more personally
17:08:17 <samkottler> jzb: indeed :-)
17:08:23 <samkottler> re: glad it's done, not the latter :P
17:09:10 <mattdm> the user personas could still use prettying up....
17:10:21 <mattdm> but otherwise I don't think there's much to do immediately.
17:10:35 <jzb> so - next steps on PRD?
17:10:43 <jzb> we wait for feedback, or...?
17:11:34 <mattdm> yeah. although I don't expect a whole lot
17:11:42 <samkottler> I think we just wait
17:11:50 <samkottler> it's a perpetual thing best I can tell
17:11:56 <samkottler> we will need to change it as our goals change
17:12:07 <mattdm> So is it too soon to start thinking about who and how we'll implement?
17:12:40 <mattdm> compared to the other PRDs, we've got a of actually actionable things in the Requirements.
17:13:15 <samkottler> yeah, we've got a whole lot stuff to get done :)
17:13:28 <jzb> mattdm: clearly that was a mistake ;-)
17:15:07 <mattdm> um. as I'mr eading over this -- I'm going to downgrade the big data image from Must
17:15:13 <mattdm> probably to Nice to Have
17:15:21 <mattdm> because it doesn't really block anyhting else.
17:15:48 <mattdm> "Should. Very useful for our users, but does not block anything else."
17:17:31 <mattdm> also we should change the wording on the kernel bit to reflect reality -- not really a separate kernel
17:19:29 <mattdm> since no one else is talking I just went ahead and did that :)
17:19:43 <samkottler> cool, well I haven't got anything else PRD related
17:19:53 <rbergeron> hi. sorry
17:19:55 <samkottler> let's perhaps wait for it to get approved before we start with assignment of work?
17:19:57 <samkottler> heyya rbergeron
17:20:03 <rbergeron> internets being cruel
17:20:12 <rbergeron> typing on phone :(
17:21:08 <jzb> samkottler: +1
17:21:18 <mattdm> yeah. next? :)
17:22:04 <samkottler> #topic GCE image status update
17:22:22 <mattdm> i've seen some interest about this on the mailing list
17:22:25 <samkottler> someone's been working on getting GCE images built for themselves and it seems like there's been success
17:22:37 <samkottler> the packaging and build stuff needs more work, but it's mostly there
17:23:57 <mattdm> i just pinged zooz
17:24:01 <zooz> hi people
17:24:15 <mattdm> hi! want to give a super-quick summary of the state of gce?
17:25:12 <zooz> sure. so as I mentioned on the mailing list, pretty much everything is in a good shape. I have got a working GCE image which is not much different that the current EC2/OpenStack image
17:25:59 <mattdm> i saw the thing with dhclient.
17:26:22 <zooz> right the fix is in updates-testing already
17:27:02 <mattdm> cool. maybe we should start making the changes in the rawhide kickstart so that the nightly images could be used?
17:27:13 <zooz> the only outstanding problem is how we setup ssh keys, users and forwarded IPs on google cloud when instance starts
17:28:06 <mattdm> can we bend cloud-init to this task or does it need something else?
17:28:28 <zooz> I have started working on google-compute-daemon package (have got it packaged already), but cloud-init could work too
17:28:58 <zooz> I am not very familiar with cloud-init code, but I use it a lot in production
17:29:44 * samkottler votes for keeping with cloud-init
17:30:10 <rbergeron> i think that seems sane.
17:30:19 <mattdm> there was some earlier discussion (irc? mailing list -- sorry i forget) about a patch from google to enable gce support
17:30:33 <mattdm> they sent it to debian directly and for some reason not to cloud-init upstream
17:30:36 <mattdm> not sure of the state of that.
17:30:38 <samkottler> mattdm: yeah the one debian committed
17:31:09 <zooz> if we stay with cloud-init (which I totally agree with) then the current kickstart files don't have to change at all AFAIK
17:31:22 <mattdm> okay that's pretty awesome.
17:32:05 <samkottler> sweeeet
17:32:06 <mattdm> I guess the next thing then is to check with fedora legal about the agreement we need to upload images?
17:32:59 <samkottler> mattdm: already done. for non-RH people they can choose to sign it if they want
17:34:21 <mattdm> zooz, the f20 image won't work out of the box without the dhclient update, right?
17:34:35 <zooz> mattdm, that's correct
17:34:55 <mattdm> so that means we would need to ask release engineering to spin an image with updates
17:35:00 <mattdm> and to upload that
17:35:29 <mattdm> which is going to be sigificantly more than just 'please upload the existing f20 image to gce'
17:35:32 <mattdm> unfortunately.
17:35:42 <zooz> if current cloud-init cannot setup ssh keys and users on GCE, then we need to solve that problem first
17:35:50 <mattdm> oh yeah that too. :)
17:36:17 <mattdm> do we have a tracking ticket for this?
17:36:36 <zooz> I am not sure, I was not following on that, sorry
17:37:31 * mattdm looks
17:37:35 <mattdm> no. let's make one.
17:38:35 <mattdm> https://fedorahosted.org/cloud/ticket/5
17:38:37 <mattdm> there. :)
17:39:31 <mattdm> #info put GCE needs / progress into https://fedorahosted.org/cloud/ticket/5
17:40:02 <mattdm> samkottler can we make all trac tickets automatically also go to the mailing list?
17:40:18 <samkottler> mattdm: sure - does that require an infrastructure ticket to be raised?
17:40:26 <mattdm> i dunno
17:40:34 <mattdm> i hoped you knew :)
17:41:19 <rbergeron> yes just in case? better safe than sorry? :)
17:41:28 <rbergeron> oh i misread
17:41:31 <samkottler> dunno
17:41:32 <samkottler> nirik: ^^
17:42:25 <nirik> sure you can. no ticket needed. Go to admin -> notifications and add the list to 'always cc' or 'always bcc'. done
17:43:23 <mattdm> and add an accept filter for the address to the moderation
17:43:25 <mattdm> i assume
17:45:32 <mattdm> done. :)
17:46:20 <samkottler> mattdm++
17:46:21 <mattdm> this is a Working Group so I just do things instead of making todo items :)
17:46:50 <mattdm> so, next? :)
17:47:09 <jzb> meetup planning?
17:47:35 <jzb> A lot of us should be in the general vicinity of FOSDEM shortly
17:47:41 <jzb> should we plan a meetup?
17:48:00 <mrunge> jzb +1
17:48:17 <mrunge> I'd assume, most of us will show up in virtualization devroom
17:48:24 <mrunge> or in Brno as well
17:48:36 <jzb> mrunge: I'll be in the distributions devroom
17:48:43 <jzb> mrunge: also in Brno
17:48:55 <mrunge> I'll be switching between rooms
17:49:13 <mrunge> at FOSDEM, and will attend devconf, too
17:49:14 <mattdm> I will be whereever people tell me to go. :)
17:49:37 <jzb> mattdm: ....
17:50:07 <mattdm> there will be a session at devconf specifically covering fedora wg status
17:50:15 <mattdm> that's sunday morning, I think.
17:50:21 <mattdm> nothing but the best timeslots for fedora!
17:50:43 <samkottler> heh
17:50:45 * mattdm is joking. scheduling was not meant to be malicious.
17:50:51 <jzb> OK, so Sunday at devconf?
17:50:52 <mattdm> (for the record.)
17:51:02 <mrunge> sounds good to me
17:51:19 <jzb> I have one item for open floor if that's decided...
17:51:55 <samkottler> #topic openfloor
17:51:58 <samkottler> jzb: hit us
17:52:16 <jzb> so, we are doing or proposing doing a lot of work for cloud stuff
17:52:33 <jzb> what is missing is a plan for informing the rest of the world outside Fedora about this.
17:52:36 <jzb> e.g., the new cloud images
17:52:54 <jzb> We should be coordinating with the marketing group to get a promotion plan in place for these things.
17:53:13 <mattdm> jzb +1
17:53:14 <jzb> I can help coordinate there
17:53:16 * stahnma reads the scrollback
17:53:18 <mattdm> that would be awesome.
17:53:23 <jzb> but might need some help with actual content.
17:53:43 * mattdm has already promised marketing group content that he hasn't delivered
17:54:02 * stahnma needs bette znc settings to see all scrollback :(
17:54:36 <mattdm> stahnma meetbot to the rescue http://meetbot.fedoraproject.org/fedora-meeting-1/2014-01-22/fedora-meeting-1.2014-01-22-17.00.log.txt
17:55:13 <jzb> OK, I'll start a thread on that on marketing
17:55:26 <jzb> or cloud - then carry over to marketing
17:55:35 <jzb> that's all I've got EFO
17:55:38 <jzb> er, EOF
17:56:02 <mattdm> jzb thanks for gettin that going
17:56:12 <mattdm> getting.
17:56:24 <mattdm> i really do talk that way, but in this case it's just my keyboard is finicky
17:58:21 <samkottler> anything else before we finish up?
17:58:45 <mattdm> i'll report back if there's anything exciting from fesco
17:58:58 <samkottler> please do :-)
17:59:01 <samkottler> #endmeeting