21:00:06 <notmyname> #startmeeting swift
21:00:07 <openstack> Meeting started Wed Sep  7 21:00:06 2016 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:11 <openstack> The meeting name has been set to 'swift'
21:00:22 <notmyname> who's here for the swift team meeting?
21:00:24 <mattoliverau> o/
21:00:26 <m_kazuhiro> o/
21:00:28 <jrichli> o/
21:00:30 <mathiasb> o/
21:00:32 <acoles> here
21:00:32 <nadeem_> o/
21:00:34 <kota_> hello
21:00:35 <hosanai> o/
21:00:37 <ntata> HELLO
21:00:37 <cutforth> o/
21:00:38 <pdardeau> hi
21:00:42 <timburke> hi
21:00:45 <mattoliverau> acoles is back \o/
21:00:54 <ntata> s/HELLO/hello
21:00:54 <jrichli> yay!
21:00:58 <torgomatic> .
21:01:04 <mattoliverau> ntata list to shout
21:01:10 <mattoliverau> *likes
21:01:10 <bkeller`> o/
21:01:15 <ntata> :D
21:01:17 <joeljwright> hey
21:01:22 <notmyname> torgomatic has gone from emoji hellos to simply a dot
21:01:36 <notmyname> welcome everyone
21:01:36 <torgomatic> next up is Unicode whitespace characters
21:01:57 <notmyname> torgomatic: let's find out who's irc client doesn't do unicode! ;-)
21:02:00 <bkeller`> then unicode zero-width non-joiner characters
21:02:01 <acoles> mattoliverau: i just missed you all so much ;)
21:02:01 <cschwede> hello!
21:02:11 <notmyname> agenda for this week is at
21:02:13 <notmyname> #link https://wiki.openstack.org/wiki/Meetings/Swift
21:02:20 <mattoliverau> of course you did acoles we're lovely :)
21:02:27 <notmyname> mostly some FYI things to go over this week
21:02:39 <notmyname> #topic newton release
21:03:08 <notmyname> the newton release is approaching, which means we need a release to include
21:03:10 <notmyname> #link https://releases.openstack.org/newton/schedule.html
21:03:14 <notmyname> there's the overall schedule
21:03:25 <notmyname> tl;dr is that we need a release at the end of the month
21:03:28 <notmyname> so...
21:03:47 <notmyname> that means that I'll be starting to build a list of in-progress stuff that we can land before then
21:04:23 <notmyname> I'll use https://wiki.openstack.org/wiki/Swift/PriorityReviews like we've done in the past (don't bother clicking--it's not updated)
21:04:43 <notmyname> so expect to see that, and please bug me about patches that need to be on the list
21:05:04 <acoles> notmyname: shall we also mark bugs critical that we'd like fixed for the release?
21:05:06 <notmyname> the things I'll be looking at first are the patches I've starred in gerrit
21:05:11 <notmyname> acoles: yes, that would be helpful too
21:05:13 <notmyname> thanks
21:05:44 <notmyname> my starred patches (more than just swift) are https://review.openstack.org/#/q/starredby:notmyname
21:05:46 <acoles> notmyname: ok I will mark a couple of bugs, you can always downgrade them if you disagree
21:05:54 <notmyname> acoles: perfect
21:06:39 <notmyname> any questions about the release or what that's going to look like? any of the logistics?
21:06:48 <cschwede> i should have another look at my unicode patches…
21:07:01 <notmyname> you mean ünicode?
21:07:09 <cschwede> yeah, thät one!
21:07:23 <notmyname> (you have more keys on your keyboard than me)
21:07:31 <mattoliverau> \o/ mu irc client is unicode
21:07:41 <mattoliverau> *my
21:07:48 <acoles> mattoliverau: lucky yu
21:07:55 <torgomatic> I � Unicode
21:08:00 <notmyname> #topic golang status
21:08:01 * mattoliverau its obviously too early for me to actually be able to type
21:08:20 <notmyname> nadeem_: you've proposed (and landed) the replicator repconn change
21:08:34 <notmyname> current golang work is focused on https://review.openstack.org/#/c/365849/
21:08:34 <patchbot> patch 365849 - swift (feature/hummingbird) - go: fix replicator policy bugs / big refactor
21:08:36 <notmyname> right?
21:08:40 <nadeem_> true
21:09:16 <notmyname> here's what I'd like to see: a golang patch done and ready to land by the time we get to barcelona (meaning, the friday before the summit)
21:09:41 <notmyname> there will be a lot of discussions about golang there, I'm sure, and I'm also pretty sure they will be "intense"
21:10:01 <nadeem_> ready to land in feature/repconn or master?
21:10:07 <notmyname> instead of saying "we will do something", I'd *much* prefer to be able to say "here's a thing, where does it go?"
21:10:29 <nadeem_> okay
21:10:48 <notmyname> nadeem_: in your opinion, what are the chances of meeting that deadline?
21:11:23 <notmyname> or rather, what do you need to make that happen?
21:12:32 <nadeem_> I think most of the work would be on deciding how could we split golang code which is easier to deploy for everyone
21:13:19 <nadeem_> I think redbo is slightly hesitant in splitting the golang repository.
21:13:31 <notmyname> ok
21:13:37 <notmyname> I can understand that
21:14:11 <notmyname> it's not a split IMO. it's a gradual migration to the mainline location of where it will live long-term for the community
21:14:43 <joeljwright> nice spin :)
21:14:45 <nadeem_> true...it is just that we don't know how long that interim period would be
21:15:33 <notmyname> nadeem_: it is taking longer than I initially expected, but that's because of the extra non-technical side of things that has gotten rather complicated
21:15:33 <nadeem_> may be we should have a video call and discuss more about it
21:15:42 <notmyname> I think that would be perfect
21:15:55 <notmyname> I can set one up for later this week or early next week
21:16:14 <notmyname> tdasilva: your video system still available to use?
21:16:31 <notmyname> well, the RAX one is pretty good too, if nadeem_ hosts it
21:16:44 <notmyname> I can work on setting that up
21:16:48 <mattoliverau> nadeem_: can do that thing
21:16:51 <nadeem_> yep our video system is awesome
21:16:57 * notmyname adds a note in his todo app
21:17:19 <mattoliverau> nadeem_: sarcasm doesn't come across in irc well :P
21:17:30 <nadeem_> :D
21:18:07 <notmyname> mattoliverau: if you have problems with it, its probably just because of the state of the aussie internet ;-)
21:18:25 <notmyname> anything else about golang? any other questions from anyone?
21:18:27 <mattoliverau> lol, its true, its true
21:19:15 <notmyname> ok, let's move on then
21:19:21 <notmyname> #topic the new PTG
21:19:43 <notmyname> openstack has invented a new conference: the Project Team Gathering (PTG)
21:19:58 <notmyname> this is designed to replace (and co-locate) all project midcycles
21:20:12 <notmyname> lots of info is in this link (or linked to in it)
21:20:13 <notmyname> #link http://lists.openstack.org/pipermail/openstack-dev/2016-September/102981.html
21:20:33 <notmyname> the important thing is that it's the week of February 20 in Atlanta GA USA
21:21:13 <notmyname> this will be in place of any midcycle hackathon we would self-organize between barcelona and boston summits
21:22:06 <notmyname> so that's mostly an FYI--I don't really know much else except what's in the email
21:22:21 <notmyname> oh! that's the other thing.
21:22:30 <notmyname> with this PTG, the Ocata release cycle will be shorter
21:22:50 <notmyname> and the PTG will kick off the Pike release cycle
21:22:59 <notmyname> (instead of it being kicked off in boston)
21:23:23 <notmyname> mostly, this schedule change will not affect us, since we do the "cycle with intermediary" releases
21:23:25 <mattoliverau> oh wow ok.. good to know
21:24:05 <kota_> good point, that's what I was  surprised at
21:24:22 <notmyname> the email has a nicely-formatted chart of how it works. if you like gantt charts, you'll love it ;-)
21:24:51 <notmyname> any questions about the PTG?
21:24:59 <notmyname> good, no questions (because I don't know)
21:25:00 <notmyname> ;-)
21:25:15 <notmyname> (but seriously, any questions?)
21:25:26 <mattoliverau> can I order pepsi in Atlanta.. I know you cant but its fun when you do ;)
21:25:29 <notmyname> if I don't know, I'll find the answers for you
21:25:31 <nadeem_> will it be free for everyone?
21:25:54 <joeljwright> will this completely remove the design element of the summits?
21:25:56 <notmyname> nadeem_: IIRC there will be "a nominal fee" and you'll likely get a discount for boston by going to the PTG
21:26:11 <notmyname> joeljwright: no, not quite. but sortof. it depends (mostly on who you ask)
21:26:19 <mattoliverau> lol
21:26:20 <joeljwright> :(
21:26:33 <jrichli> will it for swift?
21:26:37 <notmyname> it does seem like there will be some dev availability at the summit (renamed the "forum")
21:26:55 <notmyname> and perhaps some space for a midcycle-style event
21:27:52 <notmyname> if there is space available for project teams to meet at a summit, I think we should ask for that in boston. but I don't know any details about that yet (I'm not sure anyone does, yet)
21:28:17 <clayg> acoles: is back!
21:28:33 <acoles> clayg: is late! ;)
21:28:46 <kota_> lol
21:28:51 <notmyname> the forum event in boston will be a good place to hear from users and ops, so I'd still encourage people to attend. it's hard to predict what it will be like at this point, though
21:29:16 <mattoliverau> will the ptg ever be outside the US?
21:29:20 <notmyname> clayg: next week's meeting starts at 2030UTC
21:29:27 <notmyname> everyone else: meeting starts at 2100UTC
21:29:32 <acoles> lol
21:29:34 <notmyname> ;-)
21:29:47 <mattoliverau> I know that's a maybe, not sure, no one knows, but yes maybe
21:29:56 <notmyname> mattoliverau: good question. it's been asked and the only answer I've seen is "maybe, but the first ones will very likely be in the US)
21:29:56 <patchbot> Error: No closing quotation
21:30:04 <notmyname> patchbot: "
21:30:04 <patchbot> Error: No closing quotation
21:30:17 <notmyname> patchbot: I'm going to rewrite you
21:30:46 <pdardeau> notmyname: you should hook up siri and patchbot. not really, but it would be fun to watch.
21:31:06 <jrichli> lol
21:31:22 <notmyname> any other questions about the PTG?
21:31:26 <acoles> notmyname: if there is space in Boston we should ask for it, otherwise we drop from ~14 face to face days per year to 6
21:31:35 <notmyname> acoles: right!
21:32:23 <mathiasb> and if there's no design/dev activity at the forum, it might be trickier getting travel approval
21:32:36 <notmyname> mathiasb: indeed
21:32:40 <acoles> mathiasb: agree
21:32:40 <mattoliverau> no we just do our own midcycle or do we call it a start/end cycle, now I'm getting confused :P
21:32:45 <joeljwright> I'm worried that the hackathons will lose the focus we've achieved in the past, and if there's no design at the summits… we'll lose a lot
21:33:11 <mattoliverau> yeah especially from the other side of the world
21:33:30 <acoles> joeljwright: yes, I think many of us share that concern
21:33:40 <notmyname> yeah, it's different, and it will take some getting used to. but I don't want to dismiss the new thing before it's even happened (also, it's happening no matter what)
21:33:56 <mattoliverau> fair point :)
21:33:58 <joeljwright> *sigh*
21:34:24 <clayg> it's this kind of pragmatism that makes me glad we hve notmyname I don't want to dismiss the new thing before it's even happened (also, it's happening no matter what)
21:34:41 <acoles> notmyname: yep, not dis-ing the new thing, just keen that we look out for other opportunities
21:35:04 <timburke> something tells me we'll find a way to do design work, though it may or may not be officially sanctioned
21:35:08 <notmyname> yeah, we should definitely focus on when and where we can have face-to-face time to make good progress in a focused environment
21:35:09 <mattoliverau> LCA is a great conference in Australia, you should just come to that ;)
21:35:16 <jrichli> take a whiteboard with us to the bars?
21:35:27 <timburke> jrichli knows what's up
21:35:28 <clayg> jrichli: you know i'd love it
21:35:36 <acoles> mattoliverau: I'm still waiting for yuo to invite us all to your beach house ;)
21:35:37 <joeljwright> we could all accidentally get the wrong week for the summits ;)
21:35:49 <clayg> oh dear
21:36:01 <mattoliverau> acoles: your always invited, I have the guest room made up for you
21:36:04 <notmyname> barca is going to be great and we'll have space in atlanta too. we'll figure out boston, but that's a *long* time from now
21:36:06 <acoles> jrichli: we'll end up leaving a whiteboard in every bar in town
21:36:11 <jrichli> lol
21:36:41 <acoles> mattoliverau: sweet
21:36:52 <clayg> notmyname: you're always talking about having someone shadow you on release work
21:37:03 <clayg> notmyname: did anyone ever *do* tath?
21:37:15 <notmyname> ok, one other thing on the agenda to discuss this week
21:37:22 <notmyname> clayg: interesting segue
21:37:28 <notmyname> #topic release stewards
21:37:29 <tdasilva> hello, sorry i'm late
21:37:41 <notmyname> so this is a new thing announced just today
21:37:46 <notmyname> #link http://lists.openstack.org/pipermail/openstack-dev/2016-September/103115.html
21:37:47 <clayg> tdasilva: ha!  i wsan't last, here you go -> http://eavesdrop.openstack.org/irclogs/%23openstack-meeting/latest.log.html
21:37:53 <mattoliverau> tdasilva: it's ok, we have volunteered you to do all the things
21:38:07 <tdasilva> good bye ;)
21:38:22 <notmyname> there's a proposed new role for projects: a release steward
21:38:26 <notmyname> the thinking is this:
21:38:59 <notmyname> for projects that pretty strictly follow the milestones and cycle release pattern, there's a good bit of overlap between finishing one release and prepping for the work on the next one
21:39:39 <notmyname> so instead of formally putting all that work on one person (the PTL), designate a release steward or two to handle those things
21:39:53 * clayg read this email acctually - but I htink it was titled something about elections so I feel snookered
21:40:22 <cschwede> it seems more important for projects that change PTLs quite often
21:40:31 <notmyname> or rather, the release steward would watch over the release, starting before the N-1 one is released
21:41:19 <notmyname> cschwede: I think it's more for projects that have a lot of last milestone stuff and a lot of requirements/prioritization work. currently that kinda hits all at the same time
21:42:37 <notmyname> the new thing raises questions for us to consider
21:43:05 <mattoliverau> and back ports, so they're essentially in change of the release for its entire life I gather.
21:43:11 <notmyname> ...trying to figure out phrasing...
21:43:28 <clayg> like what?  you should keep trying to get someone to help with release related stuff - i'm guessing as a shadow at first - unless you're feeling particularlly overburdended?
21:43:40 <clayg> I never got the feeling that *this* was the thing that stessed you out about being PTL
21:43:56 <notmyname> that is correct
21:43:58 <clayg> releases are supposed to be easy if you're doing it right - I like to call it the "pick a sha" releasee strategy
21:44:44 <mattoliverau> almost sounds like a pokemon. "pick a sha"
21:45:14 <notmyname> I'm not feeling particularly overburdened by the release tasks. also, I know that pretty much any one of you who do stuff in swift would happily do this if you were asked, regardless of seeking it out or not
21:45:50 <clayg> notmyname: so that *does* mean no one has volunteered for the extra work that you normally handle for us?
21:45:55 <notmyname> but if it is something that you'd like to do, even as a chance to take on more responsibility and work with other teams, please let me know
21:46:22 <clayg> I think mattoliverau tdasilva timburke acoles would all be great at this
21:46:37 <clayg> I think aside from redbo I would be the worst at this
21:46:41 <notmyname> I think most people in the swift community would be great at it
21:47:07 <clayg> torgomatic: do you agree that we a non-representative minority?
21:47:09 <notmyname> but that doesn't mean it needs to be forced on people or that it needs to be "withheld" either
21:47:53 <notmyname> all i'm trying to say is that if it sounds interesting to you, please let me know and we'll figure it out. if it's not something that is particularly interesting to you, that's cool too
21:48:00 <notmyname> does that make sense?
21:48:15 <torgomatic> clayg: I agree that I, too, would be terrible at this
21:48:49 <notmyname> are there any questions about this new release steward thing?
21:48:57 <clayg> gah!  cschwede - cschwede would great at this... who else in here - notmyname is right
21:48:58 <acoles> notmyname: I read the email and thought it was a useful process/structure that we could use if we needed but that we probably don't need to right now. That said, when you need help you shuold ask us.
21:49:20 <notmyname> (ok, so the worst part of it so far is that I haven't been able to type "steward" correctly on the first try yet. yes, including this time)
21:49:29 <clayg> notmyname: can we trust you to ask for help when it's needed without needlessly burdening yourself with busy work that should/could be easily distributed?
21:49:33 <acoles> clayg: what is a sha? (and did I manage to disqualify myself?)
21:49:36 <mattoliverau> lol
21:49:48 <joeljwright> :D
21:50:59 <notmyname> I don't want people to necessarily jump up and say something here, nor do I think it's particularly good t speculate on who'd be good or bad at it. many people don't like jumping up in IRC. that being said, I want to make sure people know that it's a thing and I'd love to talk to you about it if you're interested
21:51:38 <tdasilva> notmyname: so will this be another thing that if swift doesn't have one 'swift is different bla bla bla' ??
21:51:39 * mattoliverau will read more about it :) I haven't really read the email yet
21:51:44 <notmyname> it's an interesting opportunity, I think, more from the community leadership development perspective than from actual technical need at this moment in time
21:51:53 <clayg> that's fair - and i'm sorry for calling people out - but I'm also trying voice the opinion that we're probably all quite happy to just let you do it?
21:52:23 <clayg> i say this because - you've called out for this before - and you didn't say that anyone jumped on it?
21:52:27 <notmyname> clayg: I know :-) and yes, I'm happy to do it too (I currently am, and it's not the hard part of being the swift ptl)
21:53:04 <notmyname> but this time it's more of an official OpenStack thing, and as tdasilva said, people are watching ;-)
21:53:06 <clayg> not that it's becoming an official thing - i think it opens the door to question if this would could/should be easily distributed and we can as a group encourage people to try and get involved - i might not be great at it - but i'd step up if ment you burn out less?
21:53:40 <clayg> if we know that you'll say uncle when it's needed - we're *probably* just going to take the easy road - or at least I am
21:53:42 <notmyname> tdasilva: no, I don't think swift having one or not will be a thing. at least based on the language that's been used so far about it
21:54:39 <clayg> ok, so we let it stew and see what happens - unless someone reaches out to notmyname which is great too
21:54:49 <notmyname> tdasilva: sure, somebody might try to criticize us for not having one, but it would be hard to make a good argument for that IMO
21:54:51 <pdardeau> notmyname: how about we hold a bake sale and then hire a scrum master with proceeds? ;-)
21:55:12 <clayg> pdardeau: lol
21:55:14 <notmyname> pdardeau: that sounds....terrible. except for the cake. cake is good ;-)
21:55:15 <acoles> notmyname: tdasilva or for the record we can have one but its one and the same person
21:55:25 <notmyname> acoles: right
21:55:43 <notmyname> #topic open discussion
21:55:49 <notmyname> anything else to bring up this week?
21:55:49 <acoles> pdardeau: if you invite me to the cake sale you'll be able to afford two scrum masters :D
21:55:59 <clayg> i wanted to go back to setting expectations for the feature/repconn!
21:56:01 <notmyname> ntata: you're looking at oslo.config? how's that going
21:56:12 <notmyname> clayg: yeah, we covered that in the first 30 minutes ;-)
21:56:16 <clayg> ntata: OMG that's awesome! I hadn't heard - ugs
21:56:18 <timburke> speaking of release-y things, we might want to go land these two https://review.openstack.org/#/q/status:open+project:openstack/swift+branch:stable/mitaka+topic:bug/1583305
21:56:23 <clayg> notmyname: that's why I said "go back" :P
21:56:29 <notmyname> clayg: we're going to set up a video call about repconn stuff
21:56:36 <clayg> oh right...
21:56:41 <clayg> yeah i saw that too - that's fine
21:57:05 <notmyname> timburke: thanks. I'll take a look
21:57:13 <clayg> timburke: yeah wfm - i'll star those - can we release them anytime?
21:57:21 <joeljwright> I know everyone's busy, but I'd love comments on this before Barcelona: https://review.openstack.org/#/c/365371/
21:57:21 <patchbot> patch 365371 - swift - WIP: Add Tar Large Object Middleware
21:57:32 <clayg> aww man - matt's all over it
21:57:59 <ntata> Well, I have the same list of questions you raised, and have been taking them around some OSIC folks (who's worked on oslo config) and then planning to talking to some contacts from oslo team
21:58:09 <notmyname> yeah, it's easy to land the backports. AIUI, it's the stable team that does releases from there
21:58:10 <tdasilva> notmyname: I could help setting up the video call if needed
21:58:15 <notmyname> tdasilva: thanks
21:58:27 <ntata> also, I stumble dupon something called oslo middleware for the middleware stuff we discussed
21:58:44 <notmyname> ntata: awesome! can you give us a report in next week's meeting?
21:58:57 <notmyname> ntata: but definitely let's talk ing -swift before then! ;-)
21:59:19 <ntata> okay, sure.. I can do that :)
21:59:23 <notmyname> thanks
21:59:29 <notmyname> one minute left. anything elsE?
21:59:37 <notmyname> torgomatic: sorry. meeting took the whole time ;-)
21:59:55 <torgomatic> bah ;)
22:00:10 <notmyname> thanks, everyone, for working on swift. it's a pleasure to work with you
22:00:14 <notmyname> #endmeeting