Wednesday, 2022-03-09

opendevreviewTim Burke proposed openstack/swift master: diskfile: Quarantine hashdirs on ENODATA
opendevreviewTim Burke proposed openstack/swift master: diskfile: Quarantine hashdirs on ENODATA
opendevreviewMatthew Oliver proposed openstack/swift master: Refactor memcache config and MemcacheRing loading
opendevreviewMatthew Oliver proposed openstack/swift master: Refactored errorlimited out of proxy + global cache
opendevreviewMatthew Oliver proposed openstack/swift master: Add global error_limiting to object-updater
opendevreviewAndre Aranha proposed openstack/swift master: tempurl: Deprecate sha1 signatures
afaranhatimburke__,  thanks, I'm updating your commit, I sent a review to tempest to update the tests as well
opendevreviewTim Burke proposed openstack/swift master: tempurl: Deprecate sha1 signatures
timburke__PTG etherpad:
opendevreviewTim Burke proposed openstack/swift master: s3api: Delete all parts when aborting MPU with non-ASCII characters
opendevreviewMerged openstack/swift master: s3api: Properly decode MPU request parameters before using them
mattoliverMorning, and morning kota! 20:58
kotamattoliver: o/20:59
timburke_#startmeeting swift21:01
opendevmeetMeeting started Wed Mar  9 21:01:31 2022 UTC and is due to finish in 60 minutes.  The chair is timburke_. Information about MeetBot at
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.21:01
opendevmeetThe meeting name has been set to 'swift'21:01
timburke_who's here for the swift meeting?21:01
timburke_as usual, the agenda's at
timburke_#topic PTG21:03
timburke_i finally got the etherpad started!21:03
timburke_thanks everybody, for responding to the doodle poll -- i'll work on getting time slots booked later today21:04
timburke_are there any questions or comments on the PTG? it's less than a month away, now!21:04
timburke_all right.21:06
timburke_#topic sha1 signatures21:06
timburke_afaranha has been trying to get a fips-mode tempest job passing, and some swift tests are currently failing21:07
timburke_because the tempest swift tests cover tempurls, and only use sha1 signatures21:08
timburke_which aren't secure enough to be used in fips-mode21:08
timburke_there's a tempest change to address that:
timburke_but i'd also like to deprecate sha1 signatures in tempurl:
timburke_now that it's got a dependency on the tempest change, that mostly ready to go... except for some odd test failures in the dsvm jobs. i'll look into those more21:11
timburke_i know we *also* use sha1 signatures in formpost:
timburke_anybody know of any other places we ought to update to use (or at least allow) sha256/sha512 sigs?21:12
mattolivernothing straight off the top of my head21:13
timburke_oh! admin /info requests21:14
mattolivercontainer sync realms21:15
timburke_good catch21:16
timburke_anyone available to champion updating formpost, the info controller, or container sync realms?21:17
timburke_or at least writing up the bugs for the latter two?21:17
mattoliverI'll have a go21:18
timburke_thanks mattoliver!21:18
timburke_that's all i've got21:19
timburke_#topic open discussion21:19
timburke_what else should we talk about this week?21:20
acolesdoes any one know how these s3api doc files are used ?21:21
acoleskota: ^^ do you remember?21:21
acolesthey seem to define schema for s3api requests but I can't see how they are built into docs (if at all)21:22
kotaat least rnc is used for generating assertion files for response xml21:22
timburke_and the conf files in are used when we run the swift-tox-func-s3api-ceph-s3tests-tempauth jobs21:22
kotaon the conf files, I'm not sure on my memory but it looks like to pass ceph-s3-tests21:23
timburke_kota, can you refresh my memory on the difference between those and ?21:23
kotaoh, timburke_ is sure on that.21:23
acolesbut we have rng files here for the request validation I think
kotai think the rnc is human readable rather than schema xml so at that time, I wanted to keep the original source of the xml21:24
kotaif we decided to maintain xml directly, we don't have to keep the rnc files21:24
kotaah... maybe rng, i meant21:25
timburke_do you remember what tool converts between the two? it'd probably be good to confirm that they're actually the definitions...21:25
timburke_actually the *same* definitions, i mean21:26
kotalet me take it as my homework, probably I discussed it on the review (the patch is proposed another one, IIRC). So i can look for that21:27
acolesthanks kota !21:27
timburke_thanks kota21:27
mattoliverI have a few chains lingering (v2 serialisation, global error limitting, ACTIVATING and CLEAVING states to the shading state machine, etc.) but maybe rather then disccuss here and now I'll add them to the PTG etherpad :) 21:29
timburke_👍 i keep meaning to keep pushing on the ring v2 stuff21:30
mattoliverI know, nps. busy times :) 21:30
acolesmattoliver: I am (slowly) reloading the ACTIVATING and CLEAVING states stuff21:30
acolesinto my brain21:30
mattoliverlol, great! yeah it's a long reload21:30
mattolivertook me a while as well :)21:31
mattoliverhaving full confidence and getting shrinking "done" would be awesome 21:31
timburke_i think i'll try to cut a 2.29.1 as the final yoga release -- and be sure to call out how it'll be the final py2-supporting release21:32
timburke_or, final py2-supporting stable branch, anyway21:33
timburke_all right, i think i'll call it21:36
timburke_thank you all for coming, and thank you for working on swift!21:36
opendevmeetMeeting ended Wed Mar  9 21:36:16 2022 UTC.  Information about MeetBot at . (v 0.1.4)21:36
opendevmeetMinutes (text):
kotaOK, i got it on schema21:55
kotaacoles: the commit that merges the schema file is at
kotaand the commit message says "trang" command for converts21:56
kotaI tested the trang command locally (on Ubuntu Bionic), a rnc file can be converted to the rng xml file with same content21:57
kotanote that I'm not a trang command master and hit some failures for file name... in my local trang only find "common.rnc" for source. No other file name can not be fed...21:59
kotafyi, the discussion for the commit is still found on gerrit at
kota^^^ also for timburke_22:02
opendevreviewMatthew Oliver proposed openstack/swift master: Prepare tracing by adding a WSGI mixin for middlewares
opendevreviewMatthew Oliver proposed openstack/swift master: trace: Add RequestTraceMiddleware
mattoliver^ just a rebase22:38
opendevreviewMerged openstack/swift master: s3api: Delete all parts when aborting MPU with non-ASCII characters

Generated by 2.17.3 by Marius Gedminas - find it at!