21:16:52 <notmyname> #startmeeting swift
21:16:52 <openstack> Meeting started Wed Dec  6 21:16:52 2017 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:16:53 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:16:55 <timburke> whoooo!
21:16:56 <openstack> The meeting name has been set to 'swift'
21:17:14 <notmyname> who's here for the swift team meeting?
21:17:17 <mattoliverau> o/
21:17:18 <m_kazuhiro> o/
21:17:19 <clayg> o/
21:17:24 <clayg> m_kazuhiro: !!!
21:17:25 <patchbot> I am!
21:17:27 <torgomatic> o/
21:17:41 <acoles> I am
21:17:55 <notmyname> is kota_ here?
21:18:01 <notmyname> tdasilva: cschwede: rledisez: ?
21:18:12 <tdasilva> yep! sorry i'm here
21:18:29 <tdasilva> currently discussion swift+barbican integration on #tripleo
21:18:33 <notmyname> fun
21:18:37 <tdasilva> super
21:18:46 <tdasilva> ;)
21:19:05 <notmyname> ok,let's get started
21:19:43 <notmyname> the swiftstack hiatus is coming to a close. we'll be jumping back in to container sharding and other things
21:19:57 <notmyname> so, sorry about that. the last few weeks of not doing anything
21:20:07 <mattoliverau> \o/
21:20:16 <notmyname> I went through the backports earlier to land those
21:20:26 <notmyname> please let me know if there are other things that should be backported
21:20:57 <timburke> one upside to it though: i'm currently working on some documentation about bimodal support in proxyfs. i figure kota_ at least is interested :-)
21:21:09 <notmyname> cool
21:21:28 <notmyname> also, pay attention to the PTG registration. I hear the hotel is tightly booked already, and there isn't a lot of stuff in the area
21:21:35 <notmyname> so don't delay on those logistics
21:21:56 <mattoliverau> Hopefully I won't here last minute again then :(
21:22:04 <notmyname> ok, before we talk about symlinks with m_kazuhiro, is there any other announcement/info things to bring up?
21:22:30 <notmyname> ok
21:22:33 <notmyname> #topic symlinks
21:22:34 <tdasilva> for anyone that won't be staying at the ptg hotel and happens to be looking for another hotel in the area, cschwede and I have booked elsewhere, ping me if you want details...
21:22:43 <notmyname> tdasilva: great! thanks
21:22:52 <mattoliverau> tdasilva: cool, ta
21:22:52 <notmyname> m_kazuhiro has been hard at work on https://review.openstack.org/#/c/232162/
21:22:53 <patchbot> patch 232162 - swift - Symlink implementation.
21:23:01 <notmyname> and it has a +2 from kota_!
21:23:12 <tdasilva> sweeet
21:23:14 <notmyname> m_kazuhiro: what do we need to discuss on it this week?
21:23:28 <mattoliverau> I plan on going over it today, now that kota_s follow up has been squashed in
21:23:54 <mattoliverau> It looks good, I'm keen to get it in
21:23:56 <notmyname> mattoliverau: thanks
21:24:04 <notmyname> yeah, it's a great feature to have
21:24:16 <joeljwright> damn these timezones!
21:24:31 <notmyname> joeljwright: no worries. we started late anyway
21:24:43 <mattoliverau> It's a joeljwright
21:24:59 <notmyname> m_kazuhiro: for those of us who haven't been keeping close track on it, where's the best starting place?
21:25:14 <joeljwright> :)
21:25:27 <m_kazuhiro> please one minute...
21:25:40 <notmyname> also, ( mattoliverau, timburke, acoles, tdasilva, clayg) do we want an extra +2 on this (ie more than 2)?
21:26:17 <clayg> i see api docs and functional tests - so I'm guessing it's pretty done at this point
21:27:04 <tdasilva> notmyname: i like that idea...honestly...the concept of the middlware is really not that complicated, the code is somewhat short for the middleware itself..of the ~500 lines, i think over half of it is documentation
21:27:09 <m_kazuhiro> Yes. API-docs and comments of head of symlink.py will be good start point.
21:27:30 <notmyname> m_kazuhiro: ok, thanks. good to know
21:27:36 <mattoliverau> I think people should go have a play, see if they like the API. Now's the chance if they want something changed.
21:28:10 <clayg> is kota around?
21:28:17 <clayg> kota_: what would YOU like to see as far as additional review
21:28:23 <notmyname> haven't seen kota_ here yet
21:30:00 <m_kazuhiro> I want merge symlink before christmas weeks. Please ask me questions which you have about symlink!
21:30:20 <clayg> So my read is kota said he thinks it's ready to be on master - he thinks it's better to have it running on his production clusters than not - that carries a lot of weight in my book.  I'd be hard pressed to ask for a significant design change at this point
21:30:22 <notmyname> m_kazuhiro: yes. i know you've really wanted to get it in soon :-)
21:30:42 <mattoliverau> As tdasilva says, the middleware is pretty straight forward. It's easy to grok and is good.
21:31:23 <mattoliverau> clayg: I agree, but we're at a unique point where we're about to land and commit to an API, so bows the last chance
21:31:32 <clayg> I think we should focus on documentation and maintainability - it's implemented at this point - getting used to it and building familiarity is the most important thing.
21:31:34 <notmyname> ok, so mattoliverau is going to be looking at it. who else? tdasilva? clayg?
21:32:02 <tdasilva> o/
21:32:09 <mattoliverau> Me personally like it.. and after looking at the functional test changes (which I havent seen yet) I'm very close to +2
21:32:17 <notmyname> thanks tdasilva
21:32:47 <m_kazuhiro> tdasilva: Thank you!
21:33:13 <notmyname> m_kazuhiro: is there anything else specifically you wanted to bring up on this topic in this meeting today?
21:33:27 <timburke> i'll probably try to take a look at it again. i had most of it loaded in my head at one point...
21:33:34 <notmyname> timburke: thank you
21:33:37 <m_kazuhiro> notmyname: Nothing
21:33:48 <acoles> me too, I'll try to play with it
21:33:53 <clayg> ^ same
21:34:10 <notmyname> everyone will review it! :-)
21:34:10 <clayg> i checked it out - i want to build the docs and read up on the api - i'm sure I'll spin it up and play with it
21:34:31 <m_kazuhiro> great!
21:34:33 <notmyname> I'll add an agenda item to next week's meeting to see where we are. hopefully, it will be on master by then!!
21:34:34 <clayg> i'll probably look at the probe test and try to break that
21:34:40 <clayg> notmyname: +1
21:35:10 <notmyname> #topic open discussion
21:35:11 <m_kazuhiro> notmyname: +1
21:35:24 <tdasilva> clayg: container_sync is the part i've looked at the least, so any extra coverage there is a plus
21:35:28 <notmyname> any other topics to bring up this week? there's nothing else as an agenta topic
21:35:44 <tdasilva> clayg: if you have the chance...
21:36:06 <joeljwright> just wanted to say thanks to timburke for following up on the data-in-slo patch
21:36:23 <notmyname> joeljwright: that's another one I'd love to see landed soon :-)
21:36:30 <timburke> oh yeah! that thing's looking pretty good now, as i recall...
21:36:42 <joeljwright> I'm really happy with the change to data/object segments
21:36:54 <notmyname> last week at AWS, the BBC was talking about "data formats where you can concatenate objects to get a continuous stream"
21:37:06 <notmyname> sounded like a great fit for the data segments :-)
21:37:10 <joeljwright> cool!
21:37:19 <timburke> #link https://review.openstack.org/#/c/365371/
21:37:20 <patchbot> patch 365371 - swift - Add support for data segments to SLO and Segmented...
21:37:57 <notmyname> I'll try to take a look at that one, since I've looked before
21:38:09 <notmyname> anything else? other patches or info to share?
21:38:19 <timburke> note that there's a dep on a little refactor patch: https://review.openstack.org/#/c/512804/
21:38:20 <patchbot> patch 512804 - swift - Add base64decode function to common/utils
21:39:46 <notmyname> doesn't seem like there's other stuff to bring up
21:40:02 <notmyname> thanks for coming today. and as always thanks for working on swift!
21:40:04 <mattoliverau> Too early for brain to think :p
21:40:05 <notmyname> #endmeeting