15:33:45 <dgilmore> #startmeeting RELENG (2015-05-18)
15:33:45 <zodbot> Meeting started Mon May 18 15:33:45 2015 UTC.  The chair is dgilmore. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:33:45 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:33:52 <dgilmore> #meetingname releng
15:33:52 <zodbot> The meeting name has been set to 'releng'
15:33:52 <dgilmore> #chair dgilmore nirik tyll sharkcz bochecha masta pbrobinson pingou maxamillion
15:33:52 <zodbot> Current chairs: bochecha dgilmore masta maxamillion nirik pbrobinson pingou sharkcz tyll
15:33:55 <dgilmore> #topic init process
15:33:58 <dgilmore> who all is here
15:34:00 <nirik> morning.
15:34:08 <threebean> hey hey
15:34:10 * bochecha is sort of here
15:34:47 * maxamillion is here
15:35:03 * nirik is gonna go grab more coffee, back in a minute or two
15:35:09 <tyll> hi
15:35:42 <dgilmore> #topic #6158 Request to discuss Rel-Eng Project Planning Proposal
15:35:47 <dgilmore> https://fedorahosted.org/rel-eng/ticket/6158
15:36:04 <dgilmore> maxamillion: want to give an update from the meeting last week
15:36:34 <maxamillion> certainly
15:37:51 <maxamillion> last week we met to discuss project planning (https://fedoraproject.org/wiki/Infrastructure/Meetings/ProjectPlanProposa), I have and will continue to update that wiki page as progress is made.... we agreed upon giving Taiga.io a try, I have a test (throw away) instance of taiga mostly up and running, there's an error still on-going that I'm attempting to diagnose (I con'd threebean into helpi
15:37:57 <maxamillion> ng me today)
15:38:10 <dgilmore> #link https://fedoraproject.org/wiki/Infrastructure/Meetings/ProjectPlanProposa
15:38:10 * threebean waves
15:38:27 * masta is here
15:38:29 <maxamillion> #link https://fedoraproject.org/wiki/Infrastructure/Meetings/ProjectPlanProposal
15:38:39 <maxamillion> dgilmore: missing the trailing 'l' ... my fault
15:38:45 <dgilmore> maxamillion: :)
15:38:54 <dgilmore> #undo
15:38:54 <zodbot> Removing item from minutes: <MeetBot.items.Link object at 0x2bdd2650>
15:38:54 <maxamillion> from there I plan to scope out the packaging work and then try and chew through that this week at least in a COPR
15:38:56 <dgilmore> #undo
15:38:56 <zodbot> Removing item from minutes: <MeetBot.items.Link object at 0x23d49b90>
15:39:02 <dgilmore> #link https://fedoraproject.org/wiki/Infrastructure/Meetings/ProjectPlanProposal
15:39:05 <maxamillion> +1
15:39:12 <dgilmore> okay
15:39:40 <maxamillion> as soon as the packaging work is in a COPR, I hope to deploy a test instance to gtd.stage.fedoraproject.org (iirc that the was URL that was decided upon)
15:39:46 * pbrobinson is here
15:40:00 <dgilmore> cool
15:40:08 <dgilmore> where is the test instance?
15:40:39 <dgilmore> so people can poke at it
15:40:56 * nirik thinks that url is confusing, and would prefer just testing it on a cloud instance.
15:41:11 <maxamillion> dgilmore: it's in the OS1 OpenStack environment, it's hosted internal but to the best of my knowledge is publicly accessible but it doesn't work and this instance exists just to figure out what we need to package, I didn't intend for people to poke at it here
15:41:26 <dgilmore> maxamillion: okay
15:41:30 <maxamillion> dgilmore: it currently is at http://10.3.13.43
15:41:51 <dgilmore> maxamillion: :( that's not very useful
15:41:59 <dgilmore> people can not get at it
15:42:01 <dgilmore> me included
15:42:39 <dgilmore> I need to get on a different machine to access it and get on the vpn
15:42:46 <maxamillion> the Fedora Cloud networking wasn't working last week so nobody could get to that either, this one at least let me get started working ... I'll nuke this and redeploy out somewhere useful
15:43:01 <dgilmore> please do
15:43:15 <dgilmore> anything else you want to point out?
15:43:23 <maxamillion> not that I can think of
15:43:29 <nirik> threebean was also spinning up an instance for this?
15:43:30 <maxamillion> at least not related to that
15:44:22 <dgilmore> #topic #6159 Request to discuss Git workflow for rel-eng tooling
15:44:29 <dgilmore> https://fedorahosted.org/rel-eng/ticket/6159
15:44:41 <dgilmore> so we moved everything to pagure.io last week
15:44:55 <dgilmore> I am pushing changes to both places for now.
15:45:03 <maxamillion> +1
15:45:17 <maxamillion> I've started using pagure.io, everything seems really good
15:45:20 <dgilmore> need to update the cron jobs for rawhide and branched to pull from pagure.io and not fedorahosted
15:45:36 <dgilmore> and eventually will update the fedorahosted repos to point people to pagure.io
15:45:48 <dgilmore> I do have a few RFE's to file
15:45:58 <nirik> will need a freeze break and will have to be very carefull we don't mess up the release.
15:46:14 <dgilmore> nirik: i am going to wait until after f22 is out
15:46:21 <nirik> good plan
15:46:35 <dgilmore> just something that will need to be done
15:47:01 <tyll> IMHO we should wait till pagure provides https secured git access before adjusting URLs
15:47:50 <tyll> otherwise we need to adjust the GIT URLs twice (and are less secure in the mean time)
15:48:02 <dgilmore> tyll: it does afaik
15:48:47 <nirik> doesn't seem to, but should be easy to fix.
15:48:54 <tyll> dgilmore: no, I openend a ticket for it
15:49:04 <tyll> https://pagure.io/pagure/issue/139
15:49:57 <dgilmore> I honestly do not see it as a big issue
15:50:03 <dgilmore> but that is just me
15:51:05 <tyll> the git protocol is unauthenticated, therefore man-in-the-middle attackers can intercept connections and insert arbitrary contents if a git repo is cloned
15:51:37 <dgilmore> we use it now
15:51:42 <dgilmore> it would be no different
15:52:03 <tyll> afaik it is https currently
15:52:09 <dgilmore> no it is not
15:52:27 <dgilmore> 15 7 * * * masher TMPDIR=`mktemp -d /tmp/branched.XXXXXX` && cd $TMPDIR && git clone -n git://git.fedorahosted.org/releng && cd releng && git checkout -b stable rawhide-stable && LANG=en_US.UTF-8 ./scripts/buildbranched $(date "+\%Y\%m\%d") && sudo -u ftpsync /usr/local/bin/update-fullfilelist fedora
15:52:32 <dgilmore> thats from the cron file
15:52:45 <dgilmore> that is what gets run to build branched nightly
15:53:13 * nirik is definitely in favor of changing it to https...
15:53:21 <dgilmore> anyway that is off topic
15:53:21 <threebean> yeah, tyll has a point.  let's change that.
15:53:24 <nirik> but not sure what else there's to discuss in this meeting on it.
15:53:29 <tyll> I see, I must have missed it, because it was not in the releng repo
15:53:54 * tyll updated the rawhide/branched build scripts to use https a while ago
15:54:08 <tyll> but we can discuss this later
15:54:14 <dgilmore> #topic #6164 bodhi2 status update requested
15:54:19 <dgilmore> https://fedorahosted.org/rel-eng/ticket/6164
15:54:27 <dgilmore> lmacken: did you get bodhi2 in stg ?
15:56:11 <nirik> I don't think so, but we should start trying to push to do so...
15:56:52 <dgilmore> at last weeks meeting he said it would be done during the week
15:57:13 <nirik> could be. ;) I dunno...
15:57:42 <dgilmore> it is why we have the item on the agenda still
15:57:52 <dgilmore> we need it to be done
15:58:14 <dgilmore> #info lmacken unavailble for status update
15:58:24 <dgilmore> #topic Secondary Architectures updates
15:58:24 <dgilmore> #topic Secondary Architectures update - ppc
15:58:33 <dgilmore> pbrobinson: how was ppc last week?
15:59:20 <pbrobinson> good!
15:59:25 <pbrobinson> we have cloud images!
15:59:38 <dgilmore> woohoo
15:59:39 <pbrobinson> a little bit manual at the moment, I have some patches to send upstream
15:59:51 <pbrobinson> but overall good
15:59:53 <dgilmore> pbrobinson: upstream for which piece?
16:00:05 <dgilmore> koji, oz, imagefactory, libvirt, etc?
16:00:07 <pbrobinson> for oz/imagefactory and I think possibly koji too
16:00:11 <dgilmore> okay
16:00:20 <pbrobinson> I've got patches for most of them
16:00:31 <dgilmore> I have some patches for koji for 32 bit arm images
16:00:34 <pbrobinson> I need to just verify a few bits, and do a few minor cleanups
16:01:04 <dgilmore> cool. let me know if you want patch review
16:01:05 <pbrobinson> and I've got some patches for aarch64 but it's still not quite there yet, on the todo for tomorrow
16:01:30 <dgilmore> #info ppc cloud images exist
16:01:34 <dgilmore> #topic Secondary Architectures update - s390
16:01:44 <dgilmore> no sharckcz :(
16:01:52 <pbrobinson> sharkcz is on public holiday
16:02:07 <dgilmore> #info sharkcz off on public holiday
16:02:09 <pbrobinson> I believe it's OK there, I think he's go
16:02:16 <dgilmore> cool
16:02:18 <pbrobinson> got a TC, or almost got one
16:02:22 <dgilmore> #topic Secondary Architectures update - arm
16:02:30 <dgilmore> how is aarch64
16:02:31 <pbrobinson> we did a TC last week
16:02:37 <dgilmore> excellent
16:02:46 <dgilmore> #info final TC done last week
16:02:50 <pbrobinson> it's looking OK
16:03:13 <pbrobinson> a few bits I need to cleanup this week, and koji-shadow has issues on rawhide I need to dig into further
16:03:26 * pbrobinson is quickly getting better at python ;-)
16:03:28 <dgilmore> pbrobinson: I was asked last week on the plans for when aarch64 will be proposed as a primary arch.
16:03:49 <dgilmore> the person was thinking it would be a f23 thing I suggested f24 at the earliest
16:04:04 <pbrobinson> well at the moment it might be F-24, but possibility even F-25
16:04:12 <dgilmore> but we should try come up witha  plan and start getting people on board with the idea
16:04:17 <pbrobinson> dgilmore: was it email or some public list?
16:04:24 <dgilmore> pbrobinson: it was public irc
16:04:46 <pbrobinson> dgilmore: OK, feel free to poke me as necessary to reply
16:05:04 <pbrobinson> the major issue IMO is the availability of widely available HW
16:05:10 <dgilmore> right
16:05:19 <dgilmore> hopefully 96boards will help a bit there
16:05:23 <pbrobinson> at the moment there's less available than there was for PPC when it was demoted ;-)
16:05:28 <dgilmore> though that is a different kettle of pain
16:05:34 <pbrobinson> LOL
16:05:44 <pbrobinson> yes, pain is the right word!
16:06:07 <maxamillion> yeah, I was going to ask if 96boards being available would be good enough to promote aarch64 to primary or if there needs to be at least $X number publicly available hardware vendors, etc etc
16:06:32 <nirik> there also needs to hopefully be good enterprise stuff too.
16:07:02 <masta> Any opinion on the number of available boards before we can consider the push to primary?
16:07:18 <masta> say 5 or 10?
16:07:18 <dgilmore> maxamillion: biggest issue is we can not distribute compiled uefi and the boards out there need kernel patches not upstream
16:07:18 <pbrobinson> maxamillion: well when 96boards pulls their finger out, 3 announced, not a single one with upstream support
16:07:22 <maxamillion> dgilmore: +1
16:07:41 <pbrobinson> nirik: there already is that if you have budget! And that's going to improve shortly
16:07:46 <maxamillion> nirik: yeah, I think the availability of hardware to put in the datacenter for koji purposes is likely important too
16:07:51 <nirik> yep. just not in hand yet.
16:07:57 <pbrobinson> nirik: agreed
16:07:59 <nirik> but hopefully soon
16:07:59 <maxamillion> speaking of ... what happens if/when the Calxeda hardware dies?
16:08:14 <maxamillion> pbrobinson: "fun"
16:08:16 <pbrobinson> it's all being worked on, I'm hoping we can do F-24 but I'm still a little on the fence
16:08:24 <dgilmore> I think the quality of the port is ready for primary, the missing thing is widely available end user hardware
16:08:34 <pbrobinson> maxamillion: we have a number of spare chassis
16:08:42 <maxamillion> pbrobinson: ah ok
16:08:50 * maxamillion was sad to see them go under
16:08:50 <pbrobinson> maxamillion: and with F-23 we'll be able to do armv7 on aarch64
16:08:55 <dgilmore> maxamillion: we are using 48 nodes, we have 96
16:08:58 <maxamillion> pbrobinson: oh, swank
16:09:10 <pbrobinson> maxamillion: they've actually been bought out and we might see some more of them soon!
16:09:49 <maxamillion> pbrobinson: oh? good news, I'll probably nag you offline about some more info on that if you have any .... just don't want to derail the meeting more than I already have :/
16:09:53 <pbrobinson> dgilmore: I agree on the quality..... primarily because I've been sitting on it and keeping it in order ;-)
16:10:05 <pbrobinson> maxamillion: feel free to ping me
16:10:17 <dgilmore> pbrobinson: :)
16:10:49 <dgilmore> anything else arm related anyone want to discuss?
16:10:53 <maxamillion> pbrobinson: check pm pls
16:11:12 <dgilmore> #topic Open Floor
16:11:18 <pbrobinson> I think we're really close with it and there is a lot of "almost there" and ultimately a lot will change quick;y in the next few months
16:11:22 <dgilmore> does anyone have anything for open floor?
16:11:39 <dgilmore> pbrobinson: indeed
16:11:58 * nirik hopes for a final rc soon and a smooth non slipping release.
16:12:07 <dgilmore> nirik: indeed
16:12:15 <threebean> is there anything further to discuss for the upcoming FAD?
16:12:32 <dgilmore> threebean: we do need to agree on the deliverables
16:12:34 <pbrobinson> I'm just going to note I've been helping the mono4 people get it all landed in rawhide on a side tag, should re-tag it all later today
16:12:43 <dgilmore> and we do not have to wait for the FAD to start on them
16:12:49 <nirik> pbrobinson: thanks for helping them out. ;)
16:12:53 <maxamillion> threebean: +1
16:13:14 <dgilmore> pbrobinson: :) hopefully they maintain mono better than it has been previously
16:13:33 <threebean> dgilmore: how do you feel about agreeing to deliver some plans for composedb at the FAD?
16:13:35 <pbrobinson> dgilmore: yes, I've made that point
16:13:42 <dgilmore> threebean: sure
16:13:59 <dgilmore> threebean: It has been something I have long wanted
16:14:46 <dgilmore> threebean: there is some internal things that may fill the roll, or some of it that afaik are in the process of being opened
16:14:58 <threebean> cool, sounds good.  I think that's my main concern.
16:15:09 <dgilmore> the wider a community we can have to help develop and maintain things the better
16:15:54 * nirik nods
16:16:18 <dgilmore> we should probably engage with some CentOS folks to at least make sure they know what we are doing and get on board. I need to get the internal releng guys to be more open, which is hard to do :(
16:16:32 <maxamillion> dgilmore: +1
16:16:43 <maxamillion> I think at least one or two of the CentOS folks will be coming to the FAD
16:17:10 <dgilmore> I need to follow up with the Scientifc Linux guys, I know they used pungi for Scientific Linux 7
16:17:22 <dgilmore> maxamillion: not that I know of
16:17:52 <dgilmore> maxamillion: the guys that do releng for CentOS are in OSAS
16:18:15 <dgilmore> maxamillion: mikem will be there, he is on the CentOS board, but afaik doesnt do any of the work in CentOS
16:18:22 <maxamillion> dgilmore: ah ok, for some reason I thought one or two were signed up on the wiki page but they don't appear to be ... I was mistaken
16:19:51 <dgilmore> threebean: so yes adding at the least a definition of what composedb should be, as a fAD deliverable seems wise
16:20:24 <dgilmore> does anyone disagree with any of the thigs I listed as a deliverable?
16:20:31 <maxamillion> not I
16:22:01 <threebean> not I.  the more the merrier on a composedb doc from my perspective.
16:22:22 <dgilmore> shoudl we add the list of deliverables https://fedoraproject.org/wiki/FAD_Release_Tools_and_Infrastructure_2015/Deliverables
16:22:28 <dgilmore> or somewhere else
16:22:34 <dgilmore> in the main page?
16:23:23 <maxamillion> either way
16:23:24 <threebean> main page works
16:23:26 <threebean> easier to fine
16:23:36 <maxamillion> I don't have a hard preference ... I think a table on the main page is probably fine
16:23:41 <maxamillion> though*
16:24:31 <dgilmore> sure
16:25:45 <dgilmore> #info dgilmore to add the list of FAD deliverables to the wiki
16:27:28 <dgilmore> anything else for open floor?
16:27:34 <maxamillion> nothing here
16:28:11 <dgilmore> #endmeeting