| openstackgerrit | Merged openstack/swift master: Make ceph tests more portable https://review.opendev.org/658444 | 00:31 |
|---|---|---|
| *** diablo_rojo has joined #openstack-swift | 00:37 | |
| *** gyee has quit IRC | 00:52 | |
| *** tkajinam has quit IRC | 01:24 | |
| *** tkajinam has joined #openstack-swift | 01:24 | |
| *** diablo_rojo has quit IRC | 01:38 | |
| *** BjoernT_ has quit IRC | 02:42 | |
| *** BjoernT has joined #openstack-swift | 02:44 | |
| *** BjoernT has quit IRC | 02:47 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: func tests: Add more UTF8 tests for versioning https://review.opendev.org/684146 | 02:47 |
| *** psachin has joined #openstack-swift | 03:32 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: WIP: New Object Versioning mode https://review.opendev.org/682382 | 04:17 |
| openstackgerrit | Merged openstack/swift stable/ocata: Fix stable gate https://review.opendev.org/683757 | 04:17 |
| *** seongsoocho has joined #openstack-swift | 05:34 | |
| *** zaitcev_ has quit IRC | 05:40 | |
| *** e0ne has joined #openstack-swift | 06:16 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: tests: Better handle stale account listings https://review.opendev.org/684519 | 06:18 |
| *** new_student1411 has joined #openstack-swift | 06:28 | |
| *** pcaruana has joined #openstack-swift | 06:42 | |
| new_student1411 | DHE, I got disconnected yesterday. To make it more clear, the scenario is like: | 06:45 |
| new_student1411 | make a container public. | 06:45 |
| new_student1411 | I have a keystone admin domain 'xyz_admin' and I have created user and project named 'new_user' inside the 'xyz_admin' domain. Now I have assigned a role 'swift_operator' to 'new_user' user on 'new_user' project. Now I do not want this 'new_user' user to be able to make a container public. | 06:48 |
| *** e0ne has quit IRC | 06:56 | |
| *** tesseract has joined #openstack-swift | 07:01 | |
| *** e0ne has joined #openstack-swift | 07:39 | |
| *** ccamacho has joined #openstack-swift | 07:47 | |
| openstackgerrit | Merged openstack/swift master: func tests: Add more UTF8 tests for versioning https://review.opendev.org/684146 | 08:06 |
| *** e0ne has quit IRC | 08:28 | |
| *** tkajinam has quit IRC | 08:52 | |
| *** ccamacho has quit IRC | 09:03 | |
| *** e0ne has joined #openstack-swift | 09:31 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift master: tests: Better handle stale account listings https://review.opendev.org/684519 | 10:09 |
| *** new_student1411 has quit IRC | 10:22 | |
| *** BjoernT has joined #openstack-swift | 10:22 | |
| *** pcaruana has quit IRC | 10:23 | |
| *** rcernin has quit IRC | 10:23 | |
| *** BjoernT_ has joined #openstack-swift | 10:28 | |
| *** BjoernT has quit IRC | 10:31 | |
| *** pcaruana has joined #openstack-swift | 10:36 | |
| *** BjoernT_ has quit IRC | 10:59 | |
| *** ccamacho has joined #openstack-swift | 11:25 | |
| *** ccamacho has quit IRC | 11:25 | |
| *** ccamacho has joined #openstack-swift | 11:25 | |
| *** NM has joined #openstack-swift | 12:39 | |
| *** BjoernT has joined #openstack-swift | 12:54 | |
| *** BjoernT_ has joined #openstack-swift | 12:56 | |
| *** BjoernT has quit IRC | 12:59 | |
| *** BjoernT_ has quit IRC | 13:00 | |
| *** BjoernT has joined #openstack-swift | 13:01 | |
| *** NM has quit IRC | 13:01 | |
| *** new_student1411 has joined #openstack-swift | 13:02 | |
| *** NM has joined #openstack-swift | 13:03 | |
| *** psachin has quit IRC | 13:04 | |
| *** NM has quit IRC | 13:21 | |
| *** NM has joined #openstack-swift | 13:23 | |
| *** diablo_rojo has joined #openstack-swift | 14:34 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift feature/losf: Add tests for vfile.py https://review.opendev.org/666378 | 14:51 |
| *** e0ne has quit IRC | 14:59 | |
| *** e0ne has joined #openstack-swift | 14:59 | |
| *** BjoernT has quit IRC | 15:07 | |
| *** BjoernT has joined #openstack-swift | 15:08 | |
| *** ccamacho has quit IRC | 15:20 | |
| *** openstackgerrit has quit IRC | 15:21 | |
| *** dsariel has joined #openstack-swift | 15:29 | |
| *** pcaruana has quit IRC | 15:30 | |
| *** dsariel has quit IRC | 15:34 | |
| *** gyee has joined #openstack-swift | 15:35 | |
| *** itlinux has joined #openstack-swift | 15:57 | |
| timburke | good morning | 16:26 |
| ormandj | morning! | 16:39 |
| ormandj | question for folks, we've been seeing a good bit of: Sep 25 11:27:51 sw-proxy01.happyplace.com proxy-server: STDERR: WARNING:urllib3.connectionpool:Connection pool is full, discarding connection: identity.server.address.com (txn: LDFKJDSLFKJDSLKFJLKDSJF-LKDJLKDJLFKDJ) (client_ip: x.x.x.x) <-- that lately | 16:40 |
| ormandj | i don't see any connection pooling config in the proxy server ocnfig re: keystone/identity | 16:40 |
| timburke | ormandj, interesting... are you mostly using the Swift api, or S3? just trying to figure out which middleware i ought to be investigating... | 16:42 |
| ormandj | s3 | 16:42 |
| ormandj | this is stein btw, on ubuntu 18.04 | 16:43 |
| ormandj | i do have secret cache enabled w/ duration of 300s | 16:43 |
| ormandj | delay auth decision is also on | 16:43 |
| ormandj | (for s3token) | 16:43 |
| timburke | something tells me s3token ought to be closing the response it gets back around https://github.com/openstack/swift/blob/2.22.0/swift/common/middleware/s3api/s3token.py#L228-L231 ... | 16:49 |
| timburke | pretty sure it's benign (we should be reading the response fairly quickly and i'd bet that urllib3 does the sane thing of preferring to discard older connections) but the extra logging's sure to be annoying, along with the resources being needlessly tied up | 16:50 |
| ormandj | yeah, everything seems to be working, but we're getting quite a lot of these errors, we've had to filter them out as noise for now | 16:52 |
| *** henriqueof has joined #openstack-swift | 16:54 | |
| *** henriqueof1 has quit IRC | 16:55 | |
| *** zaitcev has joined #openstack-swift | 16:57 | |
| *** ChanServ sets mode: +v zaitcev | 16:57 | |
| *** e0ne has quit IRC | 16:59 | |
| *** openstackgerrit has joined #openstack-swift | 17:31 | |
| openstackgerrit | Tim Burke proposed openstack/swift stable/pike: Fix stable gate https://review.opendev.org/683284 | 17:31 |
| *** henriqueof1 has joined #openstack-swift | 17:53 | |
| *** henriqueof has quit IRC | 17:53 | |
| *** idlemind has joined #openstack-swift | 18:00 | |
| *** tesseract has quit IRC | 18:08 | |
| *** e0ne has joined #openstack-swift | 18:25 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: tests: Better handle stale account listings https://review.opendev.org/684519 | 18:53 |
| *** e0ne has quit IRC | 19:23 | |
| openstackgerrit | Merged openstack/swift-bench master: Allow users to specify object sizes with k/M/G suffixes https://review.opendev.org/679329 | 19:48 |
| clayg | ok, i think i'm caught up on https://etherpad.openstack.org/p/swift-object-versioning - awesome | 20:12 |
| clayg | I put together https://etherpad.openstack.org/p/swift-null-namespace in-case we want to talk about the null-namespace today | 20:13 |
| tdasilva | clayg: I like the distinction between 2.1 and 2.2, had not thought of it that way | 20:18 |
| clayg | I'm not sure if in the future we'll advocate for creating system accounts in the null-namespace - I can sort of imagine it might be possible to configure the auto_create_account_prefix to '%00' or something 🤣 | 20:23 |
| *** diablo_rojo has quit IRC | 20:24 | |
| timburke | clayg, put links for both on https://wiki.openstack.org/wiki/Meetings/Swift | 20:33 |
| clayg | fab | 20:33 |
| openstackgerrit | Thiago da Silva proposed openstack/swift master: WIP: New Object Versioning mode https://review.opendev.org/682382 | 20:36 |
| *** new_student1411 has quit IRC | 20:37 | |
| *** NM has quit IRC | 20:56 | |
| kota_ | morning | 20:57 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Make sharding methods with only one job https://review.opendev.org/684844 | 20:59 |
| timburke | meeting time! | 21:02 |
| timburke | tdasilva, mattoliverau, zaitcev, rledisez, alecuyer, clayg ^^^ | 21:03 |
| *** diablo_rojo has joined #openstack-swift | 21:06 | |
| *** NM has joined #openstack-swift | 21:10 | |
| *** BjoernT has quit IRC | 21:17 | |
| *** BjoernT has joined #openstack-swift | 21:54 | |
| mattoliverau | ok, time for breakfast | 21:58 |
| *** rcernin has joined #openstack-swift | 22:15 | |
| *** diablo_rojo has quit IRC | 22:16 | |
| *** diablo_rojo has joined #openstack-swift | 22:16 | |
| *** NM has quit IRC | 22:34 | |
| *** tkajinam has joined #openstack-swift | 22:59 | |
| timburke | zaitcev, fwiw, i could repro the problem demonstrated in http://paste.openstack.org/show/779200/ | 23:05 |
| zaitcev | timburke: wait, I don't follow. So their proposed workaround #2 does not work either? | 23:05 |
| zaitcev | They wanted to switch to inject | 23:06 |
| zaitcev | IIRC | 23:06 |
| timburke | looks like the SymbolicReference used in https://github.com/gitpython-developers/GitPython/blob/master/git/refs/symbolic.py isn't the same one in the Head object's __mro__ | 23:06 |
| timburke | i *think* this stems from some combination of https://github.com/eventlet/eventlet/commit/05d613d8ed123b87f64f8d1987968c5bf825b07a nuking all submodules (to make sure they get greened, too) and the lazy importing going on at https://github.com/gitpython-developers/GitPython/blob/master/git/objects/commit.py#L391 | 23:08 |
| timburke | ...but that's about where i lose the thread :-( not sure how to actually *fix* the situation... | 23:08 |
| openstackgerrit | Merged openstack/swift master: Sharding: Use the metadata timestamp as last_modified https://review.opendev.org/683760 | 23:11 |
| openstackgerrit | Merged openstack/swift master: sharding: Update probe test to verify CleavingContext cleanup https://review.opendev.org/683257 | 23:11 |
| zaitcev | If I were a TripleO developer, at this point I would start looking at a git library. Surely git has some programmatic access that does not require forking and subprocess. | 23:11 |
| zaitcev | Thanks for looking at it. | 23:13 |
| timburke | so: patchset 1 makes sure everything gets patched, but has things in this funky split-brain state with regard to what a particular class name ought to mean | 23:14 |
| timburke | patchsets 2 and 3 prevent the split brain, but only patch submodules under git.refs afaict -- anywhere else that GitPython might be using subprocesses will still block :-/ | 23:15 |
| timburke | is there a reason they can't monkey-patch globally? | 23:16 |
| zaitcev | Dunno, sorry. | 23:17 |
| zaitcev | TripleO is on my list... In fact, this issue floated up because I made an effort to learn it yet again. | 23:18 |
| *** BjoernT has quit IRC | 23:21 | |
| timburke | oh yeah, the feature counts i promised: http://paste.openstack.org/show/779410/ | 23:40 |
| timburke | (i'm kinda surprised container sync is as high as it is) | 23:40 |
| *** gyee has quit IRC | 23:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!