19:01:05 <notmyname> #startmeeting swift
19:01:06 <openstack> Meeting started Wed Mar 18 19:01:05 2015 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:01:07 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:01:09 <openstack> The meeting name has been set to 'swift'
19:01:22 <notmyname> who's here for the swift team meeting?
19:01:24 <mattoliverau> o/
19:01:33 <tdasilva> hello
19:01:39 <kota_> o/
19:01:42 <ho> hi
19:01:47 <cutforth> o/
19:01:53 <peluse> hi
19:02:12 <acoles> here
19:02:14 <notmyname> kota_: ho: it's 4am in Japan! I'm amazed you're here
19:02:20 <torgomatic> 👋
19:02:23 <kota_> :)
19:02:30 <jrichli_> here
19:02:38 <ho> ;)
19:03:03 <notmyname> ok, but topic this week is EC. we're getting close to kilo, and we've got a lot to do :-)
19:03:07 <notmyname> #topic EC status
19:03:41 <notmyname> if you haven't figured out by now, EC (beta) for kilo is pretty much taking up about 90% of my brain right now
19:03:59 <notmyname> so here's where we are
19:04:09 <notmyname> the OpenStack kilo release date is April 30
19:04:17 <notmyname> that date is set in stone
19:04:46 <notmyname> backing up from that date, there will need to be a release candidate (and time for a 2nd if needed)
19:04:57 <notmyname> so that puts us to the vicinity of April 10
19:05:23 <notmyname> to allow for testing of swift and for overall integration testing for the entire openstack release
19:05:36 <notmyname> we've agreed to a beta of EC in the kilo release
19:05:59 <notmyname> so backing up from April 10 and allowing for a 2 week merge window, that gets up to March 27
19:06:03 <notmyname> which is 1.5 weeks from today
19:06:04 <notmyname> fun!
19:06:17 <mattoliverau> wow, that sneaks up on ya
19:06:40 <notmyname> the march 27 and april 10 dates are somewhat arbitrary (ie not completely fixed), but you see where they come from
19:07:03 <notmyname> so, looking at what we have outstanding on the feature/ec branch, here's where we are:
19:07:19 <notmyname> there are 4 patch chains right now that need to be reviewed and landed on feature/ec
19:07:27 <notmyname> they are listed on https://wiki.openstack.org/wiki/Swift/PriorityReviews
19:07:33 <notmyname> https://review.openstack.org/164380 - start of chain for refactors and reconstructor
19:07:52 <notmyname> this is what clayg and acoles are working on, and peluse will rebase his reconstructor on top of it
19:08:00 <notmyname> https://review.openstack.org/164561 - refactor proxy PUT
19:08:13 <notmyname> this is by tdasilva and helps clean stuff up to ease merging
19:08:18 <notmyname> https://review.openstack.org/163620 - EC ranged GETs
19:08:34 <notmyname> this is torgomatic's focus right now and is required for getting functional tests to pass against EC
19:08:40 <notmyname> https://review.openstack.org/159205 - functests
19:08:50 <notmyname> this is from acoles and is the start of functional tests
19:09:09 <notmyname> jrichli_ is also looking at some functional test updates to test EC, but the in-process ones are a good start
19:09:24 <notmyname> those 4 patch chains are what to look at
19:09:35 <notmyname> they need to land asap (with reviews, of course)
19:10:03 <notmyname> also, https://trello.com/board/swift-erasure-codes/51e0814d4ee9022d2b002a2c is up to date as of about 20 minutes ago
19:10:41 <notmyname> things tagged red are marked for the ec beta in kilo
19:10:51 <notmyname> so....all that being said, what questions do you have :-)
19:12:11 <notmyname> ok then :-)
19:12:35 <mattoliverau> i think we're all reading trello (well i am)
19:13:21 <notmyname> in my opinion, there's a lot to do, and it's going to be tight to get all 4 of those landed on feature/ec by the end of next week. but I think it might happen
19:13:46 <torgomatic> I want to know why my brain can't hold all this code and I keep writing stupid bugs, but that's not specific to EC
19:14:14 <notmyname> after feature/ec has all the stuff on it that we want in the beta, then we'll refactor that feature branch and propose it for merge to master (like we did with storage policies last june)
19:14:15 <mattoliverau> depends on how long each chain is. We definittely need to get the reconstructer in, EC kinda needs it for the beta
19:14:19 <tdasilva> notmyname: from following converstations earlier in the day on irc channel, it sounded like there were two "competing" chains on feature/ec , is that still the case?
19:14:22 <notmyname> mattoliverau: of course
19:14:50 <notmyname> tdasilva: not as much, but I haven't talked to clayg yet today. while I need to confirm with him, I think the "winner" is starred and linked above
19:15:17 <notmyname> right now, peluse, torgomatic, acoles (via phone), and cschwede (via phone) are together talking about all of this
19:15:25 <notmyname> and that's the conclusion we had
19:15:35 <tdasilva> notmyname: ok, so you are already taking that into account.
19:15:39 <notmyname> yes
19:15:52 <notmyname> with the caveat of not having talked to clayg today :-)
19:15:57 <tdasilva> ok
19:16:20 <notmyname> during the merge to master, we'll have a freeze on other stuff landing on master (except on a case-by-case basis)
19:16:25 <notmyname> I want to talk about that more next week
19:16:29 <notmyname> not this week
19:17:00 <mattoliverau> tdasilva: so I guess I need to move my review comments to your EC based putt refactor :)
19:17:01 <notmyname> ok, so everyone ok with what's going on in EC? everyone know what needs done and what to work on?
19:17:11 <mattoliverau> *put
19:17:19 * tdasilva is just wondering if it makes sense to have some sort of plan on how to "attack" those 4 chains so we make the best progress and don't duplicate/waste work
19:17:33 <notmyname> tdasilva: as in overlapping features?
19:17:45 <notmyname> I think they're pretty isolated right now. at least as much as possible
19:18:00 <tdasilva> ok
19:18:56 <mattoliverau> ok, I'll review the changes and then take a look at trello and see if there is an item that needs to be picked up.. let get this done!
19:19:04 <notmyname> thanks
19:19:16 <notmyname> ok, then moving on to other topics
19:19:29 <notmyname> #topic encryption
19:19:40 <notmyname> acoles: did you put this on the agenda? I don't see a nick next to it
19:19:54 <acoles> notmyname: yes i did, sorry
19:19:59 <notmyname> no worries
19:20:04 <notmyname> acoles: the floor is yours
19:20:11 <acoles> and it is just for information, to let folks know that work has started
19:20:19 <acoles> and we are tracking on trello https://trello.com/b/63l5zQhq/swift-encryption
19:20:45 <acoles> and if anyone is bored of EC and wants an alternative review then come on over :)
19:21:11 <tdasilva> haha
19:21:13 <ho> :-)
19:21:21 <notmyname> acoles: no!
19:21:28 <notmyname> after EC!
19:21:30 <acoles> i have cookies...
19:21:31 <notmyname> ;-)
19:21:37 <mattoliverau> lol
19:21:40 <kota_> lol
19:21:52 <cutforth> mmm... cookies
19:21:54 * acoles is being kicked by notmyname
19:22:12 <notmyname> ther links are
19:22:17 <notmyname> #linkhttps://review.openstack.org/#/c/154318/
19:22:22 <notmyname> #link https://review.openstack.org/#/c/154318/
19:22:29 <notmyname> #link https://review.openstack.org/#/q/branch:feature/crypto+status:open,n,z
19:22:42 <notmyname> those are the links to see what' going on with encryption
19:22:52 <notmyname> any questions from anyone on encryption?
19:23:10 <acoles> thats all, thank
19:23:11 <notmyname> acoles: thanks for the info update
19:23:19 <notmyname> #topic open discussion
19:23:31 <notmyname> OpenStack summit registration prices go up soon
19:23:39 <notmyname> so if you haven't registered, do so quickly
19:24:14 <mattoliverau> Got my ticket and travel all booked, so your going to have to put up with the noisy Australian again :P
19:24:15 <notmyname> if you are an ATC (active technical contributor) then you should have a free pass
19:24:39 <notmyname> if you haven't gotten one, please let me know. either pm on IRC or email me@not.mn and I'll look into it for you
19:25:39 <notmyname> anything else from anyone else?
19:25:40 <tdasilva> trying not add confusion to patches to review but also answering matts questions about refactor patch. If I understand clayg's plan correctly, we would still like to see patch 156825 land on master
19:26:02 <tdasilva> where is patchbot?
19:26:04 <tdasilva> https://review.openstack.org/#/c/156825/
19:26:07 <notmyname> thanks
19:26:15 <notmyname> patchbot isn't in here
19:26:25 <tdasilva> hehe
19:26:37 <gvernik> we uploaded spec related storlets...would be nice if people start review it
19:26:42 <tdasilva> I think the plan is that if this lands on master, it would be easier to merge later ????
19:26:44 <notmyname> gvernik: thanks!
19:27:06 <notmyname> tdasilva: ya, but what about https://review.openstack.org/#/c/164561/
19:27:10 <notmyname> tdasilva: same refactor?
19:27:55 <tdasilva> yeah, but this one is just simple break up into methods without any of the EC code
19:28:15 <tdasilva> so that in the ec branch, the new EC code will just go into ECObjectController
19:28:38 <mattoliverau> gvernik: I saw that, EC reviews first, I'll take a look at the SPEC next (and try get to it today)
19:28:40 <tdasilva> and the replica put path would not be disrupted (is that a word? )
19:28:49 <notmyname> tdasilva: ya, looking to the merge, whatever makes the merge easier is good
19:29:50 <gvernik> thanks
19:30:44 <notmyname> gvernik: ya, I'm looking forward to seeing more in vancouver about it
19:30:56 <mattoliverau> +1
19:31:15 <notmyname> anything else?
19:31:23 <peluse> lunch
19:32:10 <notmyname> ok, thanks everyone for coming
19:32:16 <mattoliverau> peluse: lol, your here.. and hungry obvousily :P
19:32:18 <notmyname> please help out with code and reviews for EC
19:32:27 <notmyname> thanks for working on swift
19:32:30 <mattoliverau> sir, yes sir!
19:32:31 <notmyname> #endmeeting