21:03:28 #startmeeting swift 21:03:29 Meeting started Wed Oct 17 21:03:28 2018 UTC and is due to finish in 60 minutes. The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:03:30 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:03:33 The meeting name has been set to 'swift' 21:03:38 who's here for the swift team meeting? 21:03:41 o/ 21:03:46 good morning 21:03:56 o/ 21:04:02 hi 21:04:03 hi o/ 21:04:12 welcome 21:04:33 agenda is the same as last week (and it looks like *someone* forgot to update the date...) 21:04:35 #link https://wiki.openstack.org/wiki/Meetings/Swift 21:05:17 there have been quite a few reviews on py3 and s3api patches and several have landed 21:05:34 whoooo! 21:05:40 timburke: we're getting closer! 21:05:44 nice 21:06:07 also, multiple keymaster support. thanks mattoliverau 21:06:15 I should probably look at some of those py3 ones seeing as SUSE want a py3 release soonish. 21:06:22 no thank timburke for the patch :) 21:06:27 that would be great, thanks 21:06:55 kota_: rledisez: anything to update the rest of us on this week? 21:07:20 lots of patches from OVH this week, too! i'm excited about them 21:07:20 rledisez: looks like there's some work from your team (eg patch 611325) 21:07:21 https://review.openstack.org/#/c/611325/ - swift - Remove empty directories after a revert job - 1 patch set 21:07:23 yeah 21:07:38 rledisez: I still haven't had a chance to really look and play with your ssync patches, but sound kinda exciting. 21:07:58 while we were looking at SSYNC, we optimized some part. there is still some works on other part, but it gives good result for now 21:08:31 not so much updates to me, I randomly walk some patches related to s3api's... including keystone secret caching. 21:08:35 thx timburke for the reviews. I think your concerns atomicity must be addressed now 21:08:45 rledisez: cool 21:10:21 timburke or tdasilva: do you have anything to update on? 21:11:08 nothing from me 21:11:40 there's a patch at https://review.openstack.org/#/c/610747/ that addresses a long-standing bug from swift3. i saw kota_ and zaitcev already took a glance at it 21:11:41 patch 610747 - swift - s3api: Stop relying on container listings during m... - 2 patch sets 21:11:47 FWIW, at swiftstack we've been working on some s3 migration functionality for 1space. the s3api patches are getting more important to us (blocking future s3 api compat work). 21:12:01 it requires a new feature in slo, found in https://review.openstack.org/#/c/609860/ 21:12:02 patch 609860 - swift - Add slo_validator callback - 2 patch sets 21:12:56 well better s3api is better for us all :) 21:13:02 thanks kota_ and zaitcev for looking at those 21:13:32 :) 21:13:47 oh, and the multi-keymaster patch isn't *quite* landed yet; i wound up putting it on a chain with a couple refactorings: https://review.openstack.org/#/c/611172/ and https://review.openstack.org/#/c/611173/ 21:13:48 patch 611172 - swift - Unify handle_get/handle_head in decrypter - 1 patch set 21:13:49 patch 611173 - swift - Simplify the decryption of container listings - 1 patch set 21:13:56 I'm building a vitual multinode swift env to attempt to recreate the sharding bugs.. and maybe use it to test ssync while I'm at it. 21:14:01 cool 21:14:17 personally, I'm getting a little nervous about summit prep, so I'll be working on that Real Soon Now(tm) :-) 21:14:17 * mattoliverau misses the days of unlimitted public cloud, or hardware for that matter. 21:14:24 mattoliverau: lol I hear ya 21:14:26 but one's gotta do what ones gotta do :P 21:15:15 does anyone want to talk about any of these things in more detail right now? highlight any current work you're doing? 21:15:35 oh yeah! and https://review.openstack.org/#/c/609843/ was a fun thing to shake out of some more s3api compat testing 21:15:36 patch 609843 - swift - Allow arbitrary UTF-8 strings as delimiters in con... - 2 patch sets 21:16:39 oh right. timburke that came from a third-party client we were testing, right? 21:17:00 yep. i forget which one offhand 21:17:06 IIRC splunk 21:17:59 it seemed like a bit of a weird client -- the delimiter was literally 'Guid'; it *really* seemed like they forgot to interpolate a string or something 21:18:16 lol 21:18:33 I'm wondering if we should do an upstream swift release when we get the s3api patches landed 21:19:11 not sure yet, but maybe just a 2.19.1? 21:19:28 sounds good to me, release often I say :) 21:20:11 oh yeah, i should look at timur's https://review.openstack.org/#/c/605568/ 21:20:12 patch 605568 - swift - Handle non-ASCII characters in S3 v2 listings. - 2 patch sets 21:21:12 the last few weeks in the community (and here in these meetings) it's felt a little quieter, but I think that's ok for now. we're reviewing patches, making progress on s3api and py3, making (slower) progress on losf. thank you for the "slow and steady" work you're doing :-) 21:21:50 does anyone have anythign else to bring up this week? 21:22:20 I'd like to ask to tdasilva for FOSDEM 2019 21:22:35 I have to play with ipv6 stuff in a new sqaud I'm apart of. I assume Swift is fine at that.. I remember some patches re ipv6 landed ages ago. 21:22:36 kota_: hey! what's up? 21:22:48 in any case, I'll soon find out :) 21:22:50 mattoliverau: yeah, swift is good with ipv6 everywhere 21:22:59 (if it's not, it's a bug) 21:23:04 good, cause that's what I've been telling people ;) 21:23:15 it sounds like currently it's in the term to gather topics so... I'm wondering someone is planning to do it 21:23:55 we've even got an IPv6 gate job :-) 21:24:11 kota_: are you planning on going to fosdem? 21:24:28 kota_: yeah, last year acoles cschwede and I kinda of organized ourselves to propose multiple talks, maybe we could do the same this year? 21:24:45 rledisez: any chance you might be able to go? 21:25:17 notmyname: i hope but I'd like to find the reason to negotiate to my boss :-) 21:25:51 tdasilva: i can't say. for now it's not part of my plan. i'll check and let you know 21:25:54 timburke: oh yeah of course :) 21:25:58 cool 21:26:12 rledisez: ack 21:26:48 so for example, if we could have some talks on losf and can do some hacking work with rledisez (and any attendees) it's obviously reasonable to go there. 21:27:24 rledisez: now it sounds like you must go! :-) 21:27:28 lol 21:27:41 yeah, i'll told my boss I have no choice but to go ;) 21:27:41 lol 21:27:58 good. I'm glad we got all that settled :-) 21:28:11 kota_: rledisez: I'll start a mail thread and we can try to get organized... 21:28:23 tdasilva: it would be great if you could help coordinate .... awesome. way ahead of me as always :-) 21:28:28 tdasilva: much appreciated 21:29:00 anything other topics this week? 21:29:10 rledisez: mattoliverau: do you have anything else? 21:29:25 notmyname: nope 21:29:32 not that I can think of, at least until after the meeting closes :P 21:29:39 heh 21:30:02 ok then. let's close it (and so mattoliverau can get his morning coffee and remember what he really wanted to talk about) 21:30:08 thanks for coming 21:30:11 #endmeeting