21:01:47 <timburke> #startmeeting swift
21:01:47 <openstack> Meeting started Wed Apr 17 21:01:47 2019 UTC and is due to finish in 60 minutes.  The chair is timburke. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:01:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:01:51 <openstack> The meeting name has been set to 'swift'
21:01:57 <timburke> who's here for the swift meeting?
21:01:58 <clayg> o/
21:02:04 <alecuyer> hi o/
21:03:09 <rledisez> o/
21:03:57 <timburke> all right, i'm gonna try to make this quick, since (1) there isn't much new on the agenda and (2) i've got an almost-one-year-old who's still a bit angry that his mattress had to get up and host a meeting
21:04:20 <alecuyer> hehe
21:04:21 <timburke> #topic PTG
21:04:46 <timburke> #link https://etherpad.openstack.org/p/swift-ptg-train
21:05:09 <timburke> everyone's been doing a great job of filling out the agenda, as well as when they'll be in town
21:05:27 <timburke> it looks like most of the active contributors will be there in Denver, which is great!
21:05:59 <clayg> noice
21:06:55 <timburke> i don't have much more to say there, but if anyone has anything else to bring up, feel free to add it to the etherpad (or mention it here, or in -swift, or... just bring it up)
21:07:39 <timburke> moving on to updates!
21:07:48 <timburke> #topic updates - py3
21:08:14 <timburke> zaitcev's dlo patch is working its way through the gate
21:08:27 <timburke> i'm trying to get a py3 dsvm job going
21:08:43 <timburke> mattoliverau took a stab at tempurl
21:09:27 <timburke> i'm feeling like getting some gating func tests will be incredibly valuable, so if anyone has time to look at the chain starting at https://review.openstack.org/#/c/645284/ i'd greatly appreciate it
21:10:26 <timburke> oh yeah, and i did s3api! which covers a *bunch* of unit tests -- once we get that landed, something like 3/4 of tests will be running under py3
21:11:31 <zaitcev> s3api is on my plate to review. That chain you mentioned looks very familiar, but I forgot about it.
21:11:50 <timburke> thanks zaitcev!
21:12:16 <timburke> i know you looked at the next one down, that makes func tests send non-ascii request paths
21:12:45 <timburke> #topic updates - losf
21:12:58 <timburke> alecuyer, rledisez, how's it going?
21:13:27 <alecuyer> We have more machines using LOSF now, I did ops work this week, we found some issues with LOSF and xfs_repair,
21:13:46 <alecuyer> xfs_repair may remove inodes (volumes or leveldb files) and we do not handle this automatically currently
21:13:53 <timburke> i *love* finding issues before they hit master! ;-)
21:14:00 <alecuyer> so, i've put some details here https://docs.google.com/document/d/1YyWyMxNKAYek-wUAps1cv_Bu4guojc2Yu31RVVNxMHw/edit#
21:14:33 <timburke> #link https://docs.google.com/document/d/1YyWyMxNKAYek-wUAps1cv_Bu4guojc2Yu31RVVNxMHw/edit
21:14:40 <timburke> (for the logs)
21:15:05 <alecuyer> So, no progress to report otherwise :/
21:15:09 <rledisez> question about that: how far should the automatic repair go? is trashing the KV and starting de rebuild ok?
21:15:25 <rledisez> or should it be an operator action?
21:15:33 <rledisez> like running xfs_repair actually
21:16:15 <rledisez> i'm in favor of automation as far as we know there is no options and there will be no data loss. otherwise, it should be an operator action
21:16:44 <timburke> maybe we could quarantine? starting a rebuild as soon as possible seems like a good remediation, at first glance...
21:17:48 <timburke> do we have anything to clear out quarantined data after some amount of time, or do we just let it pile up until the operator removes it manually?
21:18:09 <alecuyer> I don't think there is anything automatic to clear it at the moment
21:18:12 <rledisez> right now it piles up, as the current quarantine
21:18:31 <rledisez> that's an interesting topic btw, a tool to clean quarantine would be nice
21:19:01 <rledisez> like, thi object was corrupted, but I know there is a correct version in the cluster => clean
21:20:27 <timburke> rledisez, do you feel like you have a good enough idea of what that tool would look like to write up a "bug" (feature request) describing it?
21:21:18 <rledisez> timburke: I might start something. maybe on the etherpad first. when we come up to an aggrement on the feature i could move it to a bug
21:21:28 <timburke> sounds good
21:22:15 <timburke> anyone have any other topics that should have an update?
21:24:03 <timburke> #open discussion
21:24:11 <timburke> any other topics in general?
21:24:35 <zaitcev> Not me... PTG is still 2 weeks away, so normal work.
21:26:19 <timburke> oh! i wrote a client patch for the first time in forever: https://review.openstack.org/#/c/651666/
21:26:56 <timburke> so there can be an actual bug fix for its eventual train release ;-)
21:27:05 <zaitcev> Reviewers: Darrel Bishop
21:27:10 <zaitcev> get real
21:27:31 <timburke> he may have been the one to report the issue
21:27:53 <zaitcev> Oh, wow.
21:29:00 <zaitcev> Waaaaait a momen. I thought that he occupied the DofEng position that John aims for. I was sure he'd go up, to a VP or COO.
21:29:37 <timburke> *shrug* startup titles are mostly made up anyway
21:29:58 <timburke> i believe my business cards say "Squasher of Bugs" :P
21:30:06 <alecuyer> sounds good :-)
21:30:26 <timburke> all right, i'm'a call it and get this kid some meds to get his fever back down
21:30:37 <timburke> thank you all for coming, and thank you for working on swift!
21:30:46 <timburke> #endmeeting