21:01:28 <timburke> #startmeeting swift
21:01:28 <opendevmeet> Meeting started Wed Feb 19 21:01:28 2025 UTC and is due to finish in 60 minutes.  The chair is timburke. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:01:28 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:01:28 <opendevmeet> The meeting name has been set to 'swift'
21:01:34 <timburke> who's here for the swift meeting?
21:01:46 <mattoliver> o/
21:02:25 <timburke> i updated the agenda, but i've only really got one topic
21:02:31 <kota> o/
21:02:32 <timburke> #topic PTG
21:02:42 <timburke> it's PTG sign-up time again!
21:03:01 <timburke> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/GPE672CBWJXGVDTJM2YY4GIKOZRRV7RD/
21:03:05 <mattoliver> oh cool!
21:03:19 <timburke> it's virtual again, Apr 7-11
21:03:48 <timburke> register at
21:03:49 <timburke> #link https://ptg.openinfra.dev/
21:04:43 <mattoliver> ok this time we need to remember to fill out the survey otherwise we'll get an email reminder from cschwede again :D
21:05:11 <timburke> just did it ;-)
21:05:28 <mattoliver> \o/ nice one!
21:05:42 <timburke> though we'll still need to pick time slots -- i'll try to get a poll up by next week
21:06:00 <kota> sounds good
21:07:11 <timburke> (at least, assuming the same time slots are available as last time -- i suppose they didn't actually specify them yet)
21:07:56 <timburke> i'll also work on getting an etherpad together -- sounds like some of our sre already have ideas they'd like discussed :-)
21:08:21 <mattoliver> yeah, and because we use meetpad now, it probably doesn't matter as much, probably more for us and advertising.
21:08:38 <timburke> that's all i've got this week
21:08:42 <timburke> #topic open discussion
21:08:50 <timburke> anything else we ought to bring up?
21:08:58 <mattoliver> oh a nice quick one :)
21:09:38 <mattoliver> Nah, still the same same for me. Got distracted trying to get ringv2 tested and working properly downstream with out controller etc. So been distracted.
21:09:58 <mattoliver> Are we getting closer with the chunked s3 patch to fix boto?
21:11:01 <timburke> yeah -- one of the patches landed (the one to no longer require content-md5, as it was breaking some test clean-up), and acoles and i have continued to have good back and forth on the main aws-chunked patch
21:11:15 <timburke> #link https://review.opendev.org/c/openstack/swift/+/940172
21:11:19 <timburke> (merged)
21:11:35 <mattoliver> kk cool
21:11:46 <timburke> #link https://review.opendev.org/c/openstack/swift/+/836755
21:11:49 <timburke> (aws-chunked)
21:12:30 <timburke> i also just today finally got the start of the checksum patches stacked on top of that (previously, the chain was flipped, with the checksum stuff coming first)
21:12:47 <timburke> #link https://review.opendev.org/c/openstack/swift/+/909801
21:13:26 <mattoliver> cool
21:13:48 <mattoliver> I'll try and take a look today or tomorrow
21:13:53 <timburke> that's still a little rough -- i need to look at it from gerrit with my reviewer hat on; i'm sure there are some rough edges that fell out of the re-order
21:14:08 <mattoliver> The dickinson college students and I have been playing with adding atomic write to our db layer to fix a bug. It's been interesting.
21:14:14 <timburke> nice!
21:14:21 <boosungkim[m]> Hi!
21:14:32 <mattoliver> hey boosungkim !
21:14:33 <timburke> #link https://review.opendev.org/c/openstack/swift/+/941600
21:15:29 <timburke> sorry to have missed this week's check-in; with the holiday, i quite forgot
21:15:49 <boosungkim[m]> No worries, we forgot about the holiday since we don't get the day off
21:16:28 <mattoliver> if someone wants to take a look. Had to make some changes to GreenDBConnection so we can have say when we're in a transaction and not call rollback (which would be bad) or commit, like we do in some functions.
21:17:37 <mattoliver> Also need to grab a connection through the get contextmanager, but then put it back for people in the transaction context manager, which is a little different. So not 100% sure about it
21:18:38 <mattoliver> Other option is to dig in and refactor things so nothing can call rollback and commits and leave that all to transaction and get context managers.. but that would be alot more DB interface changes and lot of code churn.
21:20:09 <timburke> hmm... it'd be interesting to do an inventory of container-server DB calls per API call...
21:20:16 <mattoliver> anyway. Atomic write kinda fixes a bug nicely and I feel it would be awesome to have in swift's db layer. No doubt there are plenty of those race condition bugs elsewhere due to the distributed nature and even eventlet.
21:20:32 <timburke> but yeah, avoiding the churn for now seems best, if we can
21:20:35 <mattoliver> timburke: true.
21:20:59 <timburke> i'll try to get a review in this week
21:21:12 <mattoliver> thanks timburke
21:21:29 <timburke> both on that and the heartbeat-on-COPY one
21:21:31 <timburke> #link https://review.opendev.org/c/openstack/swift/+/940762
21:21:50 <boosungkim[m]> Thanks! I think James has a working unit test on his local, and we're hoping to push that patch later this week as well.
21:22:09 <mattoliver> nice!
21:22:28 <timburke> 👍
21:23:22 <mattoliver> Thats all I got for now :)
21:23:58 <timburke> all right, we'll call it early then
21:24:11 <timburke> thanks for coming, and thanks for working on swift!
21:24:15 <timburke> #endmeeting