17:11:26 <decause> #startmeeting FedoraCouncil Report from WG's
17:11:26 <zodbot> Meeting started Mon Jul 13 17:11:26 2015 UTC.  The chair is decause. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:11:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:11:36 <decause> #topic workstation report
17:11:55 <decause> #info decause missed the intro for #startmeeting :P
17:12:10 <decause> #link https://fedoraproject.org/wiki/Workstation/2015_report_to_Council
17:12:17 <decause> This is the report to council
17:12:29 <decause> I will go quickly, as everyone can read the URL, but have some narrative
17:12:32 <decause> some ideas are:
17:12:39 <decause> 1) Multi-release deliverables
17:12:43 <decause> will take some policy, and some tech
17:12:58 <decause> some of the tech we worked on recently, like xdgapp, wayland, kdebus, are beign driven by WG Group
17:13:03 <decause> and are fundatmentally solving problems
17:13:13 <decause> those account for much of the PRD specific progress we've made over past releases
17:13:24 <decause> otehr than wayland, the softare tool has support now for disabled repos
17:13:36 <decause> wayland continues to move forward
17:13:40 <decause> default for log-in screen
17:13:46 <decause> handling bugs quickly with it now
17:13:51 <decause> devassistant tool has come along too
17:14:07 <decause> to produce a workstation tha thas a smooth look/feel regardless of waht toolkits user prefers
17:14:21 <decause> worked with kernel team to get kdebus in rawhide, to test into upstream accpetance ahead
17:14:27 <decause> there is another link on the page, hope you can see it
17:14:34 <decause> this lnk toward the top, workstation tasks
17:15:14 <decause> #link https://fedoraproject.org/wiki/Workstation/Tasklist/Archive
17:15:32 <decause> You can find a proposal for buliding workstaitons as atomic releases
17:15:42 <decause> which has implications for making life easier for users
17:16:00 <decause> #topic Workstation Future
17:16:16 <decause> We're working with Anaconda team to potimtimaze/streamline installer
17:16:43 <decause> we're also working on the concept for a design a "Fedora Upgrade" app
17:16:52 <decause> this isn't as much about how it looks to users, but how it works on the backend
17:16:57 <decause> affected by things like dnf
17:17:01 <decause> movving away from fedup
17:17:13 <decause> will be affected by deliver, and atomic, drastically
17:17:33 <decause> there are other deps ahead too, which rely on working with dev, rel-eng, and atomic, that rely on rpm ostree
17:17:36 <decause> containers
17:17:43 <decause> and working on definign what content
17:17:57 <decause> it feels to me, that f24 and beyond will be where those atomic roadmpa items land
17:18:03 <decause> we're on the cusp of f23 alpha
17:18:11 <decause> one last thing, the task list
17:18:22 * decause request someone else post the link please
17:18:36 <decause> this is list is relatively current, and you can find details
17:18:42 <decause> with that, concluding the report on workstation
17:18:45 <decause> any questions?
17:19:10 <mcatanzaro> #link https://fedoraproject.org/wiki/Workstation/Tasklist
17:19:13 <sgallagh> https://fedoraproject.org/wiki/Workstation/Tasklist
17:19:16 <decause> ok, turning over to kushal, to talk about cloud wg
17:19:19 <decause> #topic Cloud WG
17:19:27 <decause> #info kushal is speaking
17:19:27 <sgallagh> mcatanzaro: So close ;-)
17:19:32 <langdon> #link http://kushaldas.in/talks/fedora_cloud_wg_update
17:20:03 <decause> #topic F22 Updates
17:20:07 <decause> Vagrant box
17:20:15 <decause> we e an official vagrant box for Fedora
17:20:50 <decause> another big change is Fedora Docker Files
17:20:53 <decause> there is a github repo
17:21:07 <decause> it is available as an rpm, or on github Feodora-Cloud account
17:21:08 <langdon> #link https://github.com/fedora-cloud/Fedora-Dockerfiles/
17:21:12 <decause> you can build your own containers on top
17:21:21 <decause> there are over 40 examples on the docker files repo
17:21:23 <decause> langdon++
17:21:35 <decause> Tunir
17:21:47 <decause> a simple UI to test fedora cloud images on their machine
17:21:56 <decause> you can use a small json script to configure the whole system and use it
17:22:03 <decause> we have plans to use with taskotron
17:22:04 <langdon> #link https://getfedora.org/en/cloud/download/
17:22:13 <decause> after that, we can use it for dev systems on laptop or desktop
17:22:17 <langdon> #info vagrant images are about 1/2 way down the cloud download page
17:22:25 <decause> 3) Atomic IMprovements
17:22:42 <decause> 4) Updated Docker Images on Docker Hub
17:22:54 <decause> after the official release, we upgraded to the F22 version
17:22:56 * stickster notes that he opened the Q&A toolbox in the Hangout On Air -- so questions submitted there should appear in my queue
17:23:28 <decause> Q: Are docker images going up to the hub automatically?
17:23:32 <decause> A: No, manually
17:23:47 <decause> A: One update is "layered Images" to be used with f23 in the future
17:23:58 <decause> Q: So, will fedora host their own images, or will they go onto th ehub?
17:24:03 <decause> A: We will decide in the near future
17:24:09 <decause> #topic F23 Changes
17:24:17 <decause> 1) Layered Docker Image Build Service
17:24:36 <decause> we can search for Fedora image on docker hub, or publish to the hub, or go to something else
17:24:49 <decause> we ahve this change, and we'll be working on it, where we can maintain our own files, and then build images
17:24:53 <decause> 2) testcloud
17:25:02 <decause> a tool to run cloud images on local machine
17:25:12 <decause> used to be a "set of commands" to boot a cloud images
17:25:22 <decause> now there is a small tool, written in python, to boot up on fedora
17:25:30 <decause> 3) systemd-networkd
17:25:38 <decause> we want to replace scripts for networking in cloud image
17:25:42 <decause> so we want to use networkd
17:25:48 <decause> we will work with upstream dev to make it happen
17:25:56 <decause> we know other cloud images started using networkd this way already
17:26:01 <decause> we want this as part of f23
17:26:05 <decause> 4) Cloud MOTD
17:26:16 <decause> in the future, when someone logs into a runnin instance
17:26:28 <decause> they should get a message with udpates, so they can update thier systems
17:26:44 <decause> q: can we get IP address for itself in there too?
17:27:02 <decause> often, you connect through a console, so it would be good to broadcast that so you wouln't have to dig
17:27:09 <decause> a: that could be a good feature, sure
17:27:20 <decause> a: often, there is a private IP, not a public IP
17:27:28 <decause> q: Sure, it dpeends on the deployment enviornment
17:27:30 <decause> A: yes, thanks
17:27:48 <decause> sgallagh: I'll talk about htis a bit in the server status update, but we just gained that capabliity ourselves
17:27:56 <decause> kushal: nice! which? The updates, or the IP?
17:28:13 <decause> sgallagh: the IP. There is a patch by steph walter(sp) about htat
17:28:21 <decause> 5) two Week Atomic Images
17:28:35 <decause> this moves from 6 month release sched, to a 2 week schedule
17:28:38 <decause> this is a work in progress
17:28:49 <decause> we want to have our test sytem in place before rleaseing to public
17:28:55 <decause> #topic Q&A
17:29:26 <decause> #topic ServerWG
17:29:31 <decause> #info sgallagh is talking now
17:29:39 <decause> Good morning/evenign/afternoon eveyrone
17:29:48 <langdon> was all excited that he was going to be using paper slides!
17:29:49 <decause> I don't have fancy slides or wiki pages, but I have a verbal report
17:29:57 <decause> #topic High Level Goals for ServerWG
17:30:28 <decause> 1) Primary Goals were to provid ea platform for handling critical infra (dns, identity, filesharing, etc...) more of on a "pet" model.
17:31:06 <decause> in additiona to that, having a platform for maintaining legacy or "classic" applications--that thing tha thas been running for 10 years, that no one understands or updates.
17:31:12 <decause> Fedora Server is likely the right place fo rhtat
17:31:19 <decause> 2) We want to part of the new DevOps world
17:31:41 <decause> which means, having somehting somewhere we can use to host an IaaS (infra as a service) to build out atomic/openshift deployments
17:31:52 <decause> 3) Manageability
17:32:26 <decause> LInux is not always easy for managing a lot fo things. It is better at having lots of tools, but requires much "black magic" that makes it difficult to adapt to from non-linux enivornment
17:32:52 <decause> Cockpit, and the openLMI project, an ansible we are working with to dpeloy large amounts of systsms and still understand wha tis going on within them
17:33:00 <decause> to do so, we've developed new tools and concepts
17:33:06 <decause> 4) Server "roles"
17:33:14 <decause> this is the largest focus area thusfar
17:33:26 <decause> minimal effort, make a machine/vm, that "becomes" this service
17:33:38 <decause> not necessarily "one-click" but the minimum amount of effort
17:34:20 <decause> a reasonably jr individual should be able to roll out a domain controller, to provide  backend for dienttiites for a webapp
17:34:34 <decause> that reads from a database server,a nd this caching server, which all will be rolled out too
17:34:52 <decause> many indivdiauls do this on a regular basis, but it can take *weeks* of learning and tweaking to do so
17:35:05 <decause> since there is an ecosystem, there are various tools
17:35:19 <decause> there isn't a common language, so after you've done one, it doesn't translate well to others
17:35:42 <decause> Server Roles, will help to identify these different roles, and the cockpit project should provide a graphical env to ask questions, and then roll out after
17:35:47 <decause> That is a good high-level view
17:35:52 <decause> #topic So Far
17:35:57 <decause> in F21, we had one role
17:36:05 <decause> domain controller, based on FreeIPA
17:36:32 <decause> there are many questions it asks, but really, you only need one "what is the admin password"
17:36:43 <decause> in F22, this has been expanded to a postgresql database server
17:36:51 <decause> you can do the same thing now, by just providing a name
17:37:23 <decause> in F23, we're working on a memecache server role, which is a backend service used by many webapps to reduce the time it takes to load commonly requested resources
17:37:51 <decause> aligned with our efforst to work with CloudWG, we'll be basing this on Nulecule--which is the basis behind the Fedora Atomic Project
17:37:56 <langdon> #link https://github.com/projectatomic/nulecule
17:38:07 <langdon> #link https://github.com/projectatomic/atomicapp
17:38:14 <decause> we want people to bootstrap with Feodra Server, make a few important services, and then trivially migrate to an atomic platform to scale as needed
17:38:28 <decause> #info we want server to be stepping stone between "classic" and "new hottness"
17:38:34 <decause> langdon++
17:38:45 <decause> we want to make it not quite as inimidating to the new designs
17:38:54 <decause> that is our big push, project atomic integration type things
17:39:06 <decause> we want to make it easier to deploy new roles, and new contribs to deploy new roles
17:39:13 <decause> I'll be presenting at FLOCk in August
17:39:24 <decause> #link https://register.flocktofedora.org
17:39:37 <decause> I'd like to post tutorials and wiki pages before then too
17:39:56 <decause> we're a bit stalled at the moment with plumbing that needs to be done, but we're looking for storage expertise
17:40:11 <decause> #help ServerWG is looking for folks with Storage Expertise
17:40:36 <decause> we're working to get that "live IP" into an MOTD, to access the admin console
17:40:55 <decause> primary tenet: Fedora Server must operate Headless
17:41:05 <decause> you can look forward to that in F24
17:41:08 <decause> err, F23
17:41:13 <decause> #topic Q&A
17:41:20 <decause> Q: Vagrant releases?
17:41:39 <decause> A: NO formal process for that, but, rolekit, has a vagrant file upstream that takes the cloud image and makes it a server image
17:41:47 <decause> we can likely take that and generalize it
17:42:11 <decause> Q: I'd like to see vagrant for all variants/editions. That may be me not konwing how to switch
17:42:29 <decause> A: kushal can talk more to this, and cloud is s aspecial case. They have a tool for converting from Cloud to Server
17:42:40 <decause> there would need to be developed a new tool for going from Workstation to Server though
17:42:47 <decause> any other questions?
17:42:48 <decause> going once
17:42:50 <decause> going twice
17:42:51 <decause> going thrice
17:42:53 <decause> ok
17:43:06 <decause> #topic Rel-Eng
17:43:11 <decause> #info dgilmore is now speaking
17:43:34 <decause> Fedora Rel-eng in the past few months has been reveamping our tooling
17:43:47 <decause> at the moment, we've been hittin bug-after-bug, but we're getting through it
17:43:53 <decause> our tooling is different now
17:43:58 <decause> there are Koji Tasks
17:44:32 <decause> we want to do more as Koji Tasks because it is more transparent, and can break down the process into manageable pieces so new contirbutors can get involved
17:44:43 <decause> a big monolithic environemnt is not ideal
17:44:57 <decause> test suite takes 3 hours to get to the point where we know if something fails or not
17:45:05 <decause> we want to be able to test components more easily
17:45:24 <decause> we are working to unify the tooling we use, with the tooling inside RHT, to make the different pieces
17:45:38 <decause> we want to provide a good foundation where Fedora Doesn't have to maintain the tooling a process alone
17:45:55 <decause> to be able to spread worload, and get more usecases and needs, to make more robust tools
17:45:59 <decause> we're tyring to be more open too
17:46:14 <decause> we've been under pressure to get things done, so we are working ot be more approachable
17:46:21 <decause> come by and say Hello :)
17:46:41 * decause requests someone post the link to the releng fad wiki page
17:46:50 <decause> #info branching tomorrow for F23
17:47:05 <decause> Questions?
17:47:21 <kushal> https://fedoraproject.org/wiki/FAD_Release_Tools_and_Infrastructure_2015
17:47:24 <decause> #link Rel-Eng FAD Etherpad link: http://etherpad.osuosl.org/fedora-releng-060515
17:47:24 <kushal> decause, ^^
17:47:36 <decause> #link //fedoraproject.org/wiki/FAD_Release_Tools_and_Infrastructure_2015
17:47:54 <decause> #topic BaseWG
17:47:58 <decause> #info nirik is talking
17:48:12 <decause> IN addtion to the processes, there are 4 groups:
17:48:14 <dgilmore> it is infra not BaseWG
17:48:14 <kushal> decause, just pasting the link is enough, meetbot picks them up as #link :)
17:48:18 <stickster> decause: I think Kevin is talking about Infrastructure
17:48:20 <decause> now
17:48:22 <decause> fedora 23
17:48:23 <decause> fedora 24
17:48:28 <decause> longer term
17:48:33 <decause> #topic ASAP things
17:48:36 <decause> 1) Bodhi 2
17:48:42 <decause> we're hoping to roll it out before F23 Alpha
17:48:49 * decause can confirm lmacken has been cranking
17:48:58 <decause> we have a staging instace, and we're working on bugs
17:49:06 <decause> 2) Mailman3
17:49:16 <decause> we started migrating lists a few weeks ago
17:49:20 <decause> we worked through a few bugs
17:49:33 <decause> abompard has been working on this, and is taking well deserved vacation now
17:49:41 <decause> when he's back, we wanna get everything rolled over
17:49:48 <decause> we're really excited, and ahve been ready for a long time
17:49:54 <decause> #topic Fedora 23 cycle
17:50:03 <decause> we're working with releng to pull in seondary architecture stuff
17:50:13 <decause> they've manageed their own koji hubs/builders, and we'd like to unify that
17:50:20 <decause> that way, everone is on the same page with same code/process
17:50:25 <decause> tha tis proceding along
17:50:36 <decause> 2) New opstack cloud instance is up an functional
17:50:45 <decause> we're moving our stuff over, and repurposing the compute nodes
17:51:02 <decause> once we retire old cloud, we should have a fair bit of extra resource, and we can tlak to the community to find out what we can use that best for
17:51:20 <decause> dev isntances, copr builds, ar ecurrently the things we use it for, but we can find out what else in the next cycle
17:51:25 <decause> 3) New Mediawiki
17:51:36 <decause> there is a long-term support version that we'll be switching to
17:51:48 <decause> wer're going to switch to OpenID, so we don't need to maintain a FAS module
17:51:53 <decause> 4) Migration from Puppet
17:52:03 <decause> hoping to have an "ansible migration" cake ready at FLOCk
17:52:17 <decause> #action decause work with nirik to get an ansible cake for FLOCK
17:52:38 <decause> 5) Look at new tech--Atomic, Docker, etc...-- and see if we can use them better
17:52:54 <decause> we're conservative, we want to be sure there is stabliity and secuirty, but we'd like to do a pilot application or two
17:53:09 <decause> we have some good ideas, had a meeting where we discussed thigns to use it for, and want to explore
17:53:17 <decause> #topic Fedora 24
17:53:27 <decause> 1) Looking into Red Hat Storage
17:53:44 <decause> would like to do something more open, like gluster, or something else open
17:53:53 <decause> 2) Finish migrations from RHEL6 to RHEL7
17:54:00 <decause> 3) FAS3
17:54:05 <decause> rolling out later this year
17:54:25 <decause> looking at moving to FreeIPA in the future, and see what features we're missing
17:54:30 <decause> #topic Longer-Term
17:54:36 <decause> 1) More 2 factor auth all around
17:54:41 <decause> we need it enabled in our webapps too
17:54:59 <decause> with things like FreeOTP, and yubikeys being more common, we could maybe roll out for other apps
17:55:13 <decause> 2) Rumors about moving our datacenter
17:55:27 <decause> we want to make sure we're ready for that, so we need to have plans in place for how we would do that
17:55:46 <decause> making that plan probably won't happen til next year, or the year after, but we want to prepare
17:55:51 <decause> #topic Q&A
17:55:53 <decause> questions?
17:55:57 <decause> going once
17:55:58 <decause> going twice
17:56:01 <decause> going thrice
17:56:12 <decause> Q: Is the "hubs" stuff on your radar?
17:56:18 <decause> A: Yes, absolutely
17:56:33 <decause> A: Hubs is def in there, and being worked on. WE'd like to see a prototype later this year
17:56:41 <decause> there are plans for updating other existing apps
17:56:50 <decause> packages app is showing it's age, and we are looking at revamping
17:56:56 <decause> there is a wiki page for application plans
17:57:07 <stickster> #link https://fedoraproject.org/wiki/Fedora_Engineering/FY16_Plan
17:57:14 <decause> stickster++
17:57:20 <decause> Q: Where do new people come to find out?
17:57:33 <decause> A: We have IRC meetings, and we're happy to take new folks
17:57:46 <decause> #link http://whatcanidoforfedora.org
17:57:53 <decause> langdon++
17:57:59 <decause> any other questions?
17:58:05 <decause> on Base, or on any other WG?
17:58:10 <decause> #topic Open Q&A
17:58:21 <langdon> just like to note that Fedora needs help from web app developers ... not just OS-type people :)
17:58:27 <striker> http://whatcanidoforfedora.org/
17:58:36 <stickster> langdon: I'll add that in a moment, thank you!
17:59:04 <decause> Q: Quick question for nirik. At some point, the mirror system is no longer oging ot work for us. Have we put any thought into what infra we can run to enable people to get Fedora without using the mirror network?
17:59:12 <decause> A: content delivery type thing?
17:59:15 <stickster> langdon: Oh, I was going to say something verbally on the HOI
17:59:17 <stickster> er, HOA
17:59:25 <langdon> stickster, ack
17:59:38 <decause> A: There has been some talk, but it is still very early stage. There was talk about geogrpahically distributed systems, and running our own network.
17:59:46 <decause> you're right that'd give us more flexibility
17:59:54 <decause> A: Yes, finding more ways to deliver fedora
18:00:27 <decause> stickster: we want to identify current risks, and what we'd like tos olve. If we do need a solution more like a big content delivery, what the state of FOSS Software for doing that is
18:00:54 <decause> stickster: also wanted to point out that Fedora is always looking for webapp developers, not just OS or console developers
18:01:04 <decause> if you are a webdev, we have teams to work with
18:01:19 <decause> there is much code flying around, lots of python and cool frameworks, and interesting problems to work on
18:01:49 <decause> langdon: adding to that, I think it is commmon for webdevs and app devs, to think there isn't much for them to contribute to a distro, and to them I say there is.
18:02:01 <decause> fedmsg is really cool for big-data and real-time stuff
18:02:10 <decause> hubs is a cool UX problem and community problem
18:02:28 <decause> the mirror problem looks like an interesting bittorrent/sync/torrent model, could be cool machine learning perspective there
18:02:40 <decause> there are places for folks who aren't doing OS things here, so reach out!
18:03:00 <decause> #info #fedora-apps and #fedora-admin are channels where new contribs can come to get info
18:03:01 <dustymabe> langdon++ for a bitorrent model
18:03:01 <zodbot> dustymabe: Karma for langdon changed to 3:  https://badges.fedoraproject.org/tags/cookie/any
18:03:38 <dustymabe> is there a link for the "frontend" development volunteer work?
18:03:47 <dustymabe> I have someone I'd like to send some information to
18:04:17 <stickster> dustymabe: http://whatcanidoforfedora.org is a great place to start
18:04:18 <decause> dustymabe: you can come hang out in #fedora-apps, where most webdev happens
18:04:22 <decause> on Freenode
18:04:24 <nirik> http://whatcanidoforfedora.org/en/
18:04:26 <stickster> Also #fedora-apps on IRC
18:04:26 <dustymabe> stickster: thanks
18:05:08 <decause> dustymabe: there is a github repo too where you can find our bootstrap fork, and other front-endy things:
18:05:17 <decause> #link http://github.com/feodra-infra
18:05:20 <dustymabe> decause: sweet
18:05:23 <dustymabe> thanks
18:05:26 <decause> s/repo/organization
18:05:30 <langdon> stickster, i have a question too
18:06:03 <decause> dgilmore: http://webchat.freenode.net
18:06:03 <kushal> We also need to repackage mumble for Fedora.
18:06:08 <decause> hubs will get there
18:06:14 <kushal> dgilmore, try.waartaa.com
18:06:49 <dustymabe> https://try.waartaa.com/chat/server/Freenode/channel/fedora-cloud
18:06:57 <decause> http://webchat.freenode.net/?&channels=fedora-apps
18:07:00 <dustymabe> kushal: beat me :(
18:07:18 <decause> dgilmore++
18:07:29 <decause> #agreed
18:07:31 <decause> langdon: lol
18:07:48 <decause> anyone know if firefox hello will go multi-user anytime in the future?
18:07:55 <stickster> decause: No idea.
18:08:01 <kushal> decause, no clue.
18:08:09 <decause> it is good for 1x1, but not multiple folks yet
18:08:17 <stickster> decause: I can't imagine they're not thinking of it, but the infrastructure required to make it work well resiliently is *extensive*
18:08:39 * decause +1's the Engineering report meetings
18:08:50 <decause> stickster: makes sense
18:09:10 <decause> #topic closing thoughts
18:09:31 <decause> stickster: having a video, and a status meeting, where you don't have to subscribe to multiple lists, is very useful
18:09:49 <decause> stickster: hopefully you can see some more "behind the scenes" info, and on your own time
18:09:52 <decause> stickster: I'd be happy to help again
18:09:59 * decause would def transcribe again too :)
18:10:34 <decause> langdon: maybe after a release, we do one of these that is used to describe what was launched as a marketing tool
18:10:47 <decause> stickster: we have some vehicles for that, but maybe we're not getting all the content we want
18:10:51 <decause> langdon: yeah
18:11:03 <decause> we are a bit over time, and in the nterests of time, we're goin to close out
18:11:07 <decause> thanks everyone for being here
18:11:08 <decause> council
18:11:10 <decause> working groups
18:11:16 <decause> this is recorded, and will be available
18:11:29 * decause will get the link for the hangout and post here before clsing meeting
18:11:46 <decause> thanks mattdm for letting me help with the council activity
18:11:48 <decause> stickster++
18:12:02 <decause> Hope you have a great week whereever you are in the world!
18:12:11 <kushal> decause, stickster Thanks :)
18:12:24 <decause> kushal: np
18:12:55 <kushal> decause, pm
18:13:19 <langdon> decause++
18:13:19 <zodbot> langdon: Karma for decause changed to 12:  https://badges.fedoraproject.org/tags/cookie/any
18:13:43 <stickster> Ha, for a moment I looked up and thought that said langdon's karma was decreased by someone
18:13:55 <langdon> stickster, :(
18:13:56 <stickster> I was like "wait, wha?!"
18:13:57 <kushal> :)
18:14:07 <kushal> kushal--
18:14:10 <stickster> decause++
18:14:15 <kushal> stickster, does not work :)
18:14:21 <decause> #link https://www.youtube.com/watch?v=552P7TXRt7w
18:14:24 <stickster> Oops, I guess I already cookie'd Remy
18:14:30 <gholms> kushal: Well, you also can't give yourself karma.  ;)
18:14:30 <decause> #endmeeting