| *** vintbra has joined #openstack-swift | 00:16 | |
| *** vintbra has quit IRC | 00:16 | |
| *** itlinux has joined #openstack-swift | 00:26 | |
| *** itlinux has quit IRC | 00:33 | |
| *** itlinux has joined #openstack-swift | 00:34 | |
| *** itlinux has quit IRC | 00:38 | |
| *** itlinux has joined #openstack-swift | 00:39 | |
| *** itlinux has quit IRC | 00:44 | |
| *** itlinux has joined #openstack-swift | 00:46 | |
| *** itlinux has quit IRC | 00:48 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: encryption: Stop being cutesy with os.path.join() https://review.openstack.org/635380 | 00:51 |
|---|---|---|
| timburke | zigo: i'm so incredibly sorry. ^^^ may be a pre-req for p 633671, and it *definitely* complicates the resolution of https://bugs.launchpad.net/swift/+bug/1813725 :-( | 00:53 |
| openstack | Launchpad bug 1813725 in OpenStack Object Storage (swift) "Object encryption key truncated sometimes when used with Barbican" [Undecided,New] | 00:53 |
| patchbot | https://review.openstack.org/#/c/633671/ - swift - Fix decryption for broken objects - 6 patch sets | 00:53 |
| *** itlinux has joined #openstack-swift | 01:06 | |
| *** itlinux has quit IRC | 01:10 | |
| *** gyee has quit IRC | 01:22 | |
| *** itlinux has joined #openstack-swift | 01:45 | |
| *** itlinux has quit IRC | 01:47 | |
| *** itlinux has joined #openstack-swift | 01:49 | |
| *** itlinux has quit IRC | 01:52 | |
| *** itlinux has joined #openstack-swift | 01:56 | |
| *** itlinux has quit IRC | 02:01 | |
| *** mahatic has quit IRC | 02:20 | |
| *** psachin has joined #openstack-swift | 02:56 | |
| *** rcernin has joined #openstack-swift | 03:07 | |
| *** rchurch_ has joined #openstack-swift | 03:24 | |
| *** rchurch has quit IRC | 03:25 | |
| *** itlinux has joined #openstack-swift | 03:41 | |
| *** gkadam has joined #openstack-swift | 04:06 | |
| *** rcernin has quit IRC | 04:35 | |
| *** itlinux has quit IRC | 04:44 | |
| *** itlinux_ has joined #openstack-swift | 04:45 | |
| *** itlinux_ has quit IRC | 04:45 | |
| *** itlinux has joined #openstack-swift | 04:59 | |
| *** itlinux has quit IRC | 05:03 | |
| *** itlinux has joined #openstack-swift | 05:04 | |
| *** itlinux has quit IRC | 05:26 | |
| *** itlinux has joined #openstack-swift | 05:40 | |
| *** e0ne has joined #openstack-swift | 06:10 | |
| *** pcaruana has joined #openstack-swift | 07:02 | |
| *** e0ne has quit IRC | 07:10 | |
| *** e0ne has joined #openstack-swift | 07:20 | |
| *** e0ne has quit IRC | 07:31 | |
| *** itlinux has quit IRC | 07:50 | |
| *** itlinux has joined #openstack-swift | 07:58 | |
| *** itlinux_ has joined #openstack-swift | 08:06 | |
| *** itlinux has quit IRC | 08:09 | |
| *** hseipp has joined #openstack-swift | 08:12 | |
| *** tkajinam has quit IRC | 08:15 | |
| *** ccamacho has joined #openstack-swift | 08:20 | |
| *** itlinux_ has quit IRC | 08:22 | |
| *** hseipp has quit IRC | 08:33 | |
| *** hseipp has joined #openstack-swift | 08:34 | |
| *** hseipp has quit IRC | 09:09 | |
| *** hseipp has joined #openstack-swift | 09:10 | |
| *** ccamacho has quit IRC | 09:23 | |
| *** e0ne has joined #openstack-swift | 10:03 | |
| *** ccamacho has joined #openstack-swift | 11:49 | |
| *** e0ne has quit IRC | 11:56 | |
| *** e0ne has joined #openstack-swift | 11:57 | |
| *** gkadam has quit IRC | 13:02 | |
| *** waverider has joined #openstack-swift | 13:56 | |
| *** psachin has quit IRC | 14:26 | |
| *** irclogbot_2 has joined #openstack-swift | 14:31 | |
| *** ccamacho has quit IRC | 15:24 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Rebuild frags for unmounted disks https://review.openstack.org/629056 | 15:33 |
| *** gyee has joined #openstack-swift | 16:13 | |
| *** pcaruana has quit IRC | 16:21 | |
| zigo | timburke: For our patch, it looks like I overwrote stuff you commited, what you proposed was much nicer. | 16:22 |
| *** ccamacho has joined #openstack-swift | 16:24 | |
| *** e0ne has quit IRC | 17:08 | |
| *** vr42 has joined #openstack-swift | 17:12 | |
| *** itlinux has joined #openstack-swift | 17:18 | |
| *** hseipp has quit IRC | 17:22 | |
| notmyname | good morning | 17:22 |
| *** admin6 has left #openstack-swift | 17:23 | |
| *** itlinux_ has joined #openstack-swift | 17:26 | |
| *** itlinux has quit IRC | 17:27 | |
| openstackgerrit | Merged openstack/swift master: Fix flakey func test teardown https://review.openstack.org/634449 | 17:38 |
| openstackgerrit | Merged openstack/swift master: Fix flakey func test setup https://review.openstack.org/634362 | 17:38 |
| *** waverider has quit IRC | 17:47 | |
| *** itlinux_ has quit IRC | 17:58 | |
| *** vr42 has quit IRC | 18:04 | |
| timburke | zigo: i'll see what i can do to get those stacked into something reasonable... but as a heads up, if you have objects (1) that were uploaded while you had encryption enabled, (2) whose name started with a slash on that initial PUT, and (3) got copied to another container (which would include a container in a different account)... it's going to be really difficult to find the correct path to use to decrypt container listings :-( | 18:05 |
| timburke | i *think* decrypting the actual object data may still be fairly straight-forward, though? | 18:05 |
| timburke | we might want to prioritize getting https://review.openstack.org/#/c/634507/ to address https://bugs.launchpad.net/swift/+bug/1729921 | 18:06 |
| openstack | Launchpad bug 1729921 in OpenStack Object Storage (swift) "Changing encryption root secret causes container listings to 500" [Low,Confirmed] | 18:06 |
| patchbot | patch 634507 - swift - encryption: Be more tolerant of errors while decry... - 2 patch sets | 18:06 |
| timburke | (because it'd present in the same way) | 18:06 |
| *** mvkr has quit IRC | 18:12 | |
| *** e0ne has joined #openstack-swift | 18:15 | |
| *** e0ne has quit IRC | 18:22 | |
| clayg | timburke: what do you think about https://review.openstack.org/#/c/629056/7/swift/obj/reconstructor.py | 18:30 |
| patchbot | patch 629056 - swift - Rebuild frags for unmounted disks - 7 patch sets | 18:30 |
| clayg | the more I think about it, the more i'm leaning towards ... maybe call it... "handoff_rebuild_node_count" ??? | 18:31 |
| timburke | clayg: with the replicator, do we have a cap on how deep we'll go? | 18:32 |
| clayg | that's a good thought | 18:35 |
| clayg | doesn't look like it | 18:35 |
| timburke | i kinda liked the idea of grabbing the next ec_n_unique_fragments nodes and rotating them according to the frag we're trying to place... then maybe we could repeat that until we run out of disks? | 18:39 |
| timburke | like, the handoff serves two purposes -- first, ensure durability; we'll get that wherever we manage to rebuild. but second, to increase availability -- and we can only get that if it's early enough in the handoff list | 18:41 |
| *** vr42 has joined #openstack-swift | 18:41 | |
| *** e0ne has joined #openstack-swift | 18:42 | |
| timburke | i *do* think that the doubled-up-frag behavior may need to be different when we're sending to a handoff instead of a primary, though. primaries may need to double-up so the frag gets where it's going faster. but it didn't belong on the handoff in the first place, so a 409 (or something) when we've already got a frag we're responsible for seems reasonable | 18:43 |
| *** cschwede has quit IRC | 18:43 | |
| timburke | i'll be back in a bit -- got some other obligations to take care of -- but i'll keep thinking on it | 18:44 |
| clayg | If we do the rotate thing I think there’s a higher chance of doubling... if we do the skip thing availability doesn’t go up after a couple of replica counts because the proxy won’t look out there | 19:01 |
| *** remi_ness has joined #openstack-swift | 19:04 | |
| openstackgerrit | vianney rancurel proposed openstack/liberasurecode master: fix: data access when having non-zero metadata size https://review.openstack.org/635603 | 19:12 |
| openstackgerrit | vianney rancurel proposed openstack/liberasurecode master: fix: proper support of more than 32 fragments https://review.openstack.org/635604 | 19:12 |
| openstackgerrit | vianney rancurel proposed openstack/liberasurecode master: feature: quadiron backend and tests https://review.openstack.org/635605 | 19:12 |
| *** rchurch has joined #openstack-swift | 19:26 | |
| *** rchurch_ has quit IRC | 19:28 | |
| *** vr42 has quit IRC | 19:42 | |
| *** remi_ness has quit IRC | 20:07 | |
| timburke | tdasilva: how crazy would it be to include encryption in the pipeline for https://review.openstack.org/#/c/626663/ ? thinking about ways to ensure we can still read existing data following https://review.openstack.org/#/c/635380/ | 20:14 |
| patchbot | patch 626663 - swift - Add rolling upgrade tests - 4 patch sets | 20:14 |
| patchbot | patch 635380 - swift - encryption: Stop being cutesy with os.path.join() - 1 patch set | 20:14 |
| tdasilva | timburke: mmm...it would be great to test encryption, but a drawback would be the inability to test earlier releases | 20:16 |
| timburke | yeah, i was thinking about that too :-/ | 20:16 |
| tdasilva | but maybe covering encryption might outweigh the benefit of testing earlier releases ?? | 20:16 |
| *** ccamacho has quit IRC | 20:28 | |
| timburke | maybe we could make it conditional on which version we start from? | 20:40 |
| tdasilva | adding encryption you mean? | 20:56 |
| tdasilva | that might work | 20:56 |
| *** rdejoux has joined #openstack-swift | 21:03 | |
| *** rdejoux has quit IRC | 21:12 | |
| *** e0ne has quit IRC | 21:19 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Rebuild frags for unmounted disks https://review.openstack.org/629056 | 21:35 |
| clayg | timburke: ^ well I talked myself into rebuild_handoff_node_count 🤷♂️ | 21:35 |
| timburke | might be time fo me to cherry-pick https://review.openstack.org/#/c/586903/ again... | 21:41 |
| patchbot | patch 586903 - swift - Display crypto data/metadata details in swift-obje... - 4 patch sets | 21:41 |
| openstackgerrit | Tim Burke proposed openstack/swift master: encryption: Stop being cutesy with os.path.join() https://review.openstack.org/635380 | 21:50 |
| openstackgerrit | Pete Zaitcev proposed openstack/swift master: py3: port the container sharder https://review.openstack.org/570320 | 22:04 |
| tdasilva | why don't we include middlewares like slo in the pipeline of internal_client? | 22:09 |
| timburke | i'm guessing just because we've mainly thought about it in terms of the expirer (and later, reconciler), neither of which need it, and nobody's thought too hard about it otherwise | 22:16 |
| timburke | does the reconciler include ?multipart-manifest=get when re-PUTting data? with SLO enabled, will it be able to reconcile a busted SLO? | 22:18 |
| zaitcev | timburke, mattoliverau: I dropped the _path out from sharder, this should be ready. | 22:22 |
| tdasilva | timburke: not sure, will check out | 22:23 |
| openstackgerrit | Pete Zaitcev proposed openstack/swift master: Rename ContainerBroker.path to cpath https://review.openstack.org/635677 | 22:35 |
| *** tkajinam has joined #openstack-swift | 22:55 | |
| *** mvkr has joined #openstack-swift | 23:03 | |
| *** takamatsu has quit IRC | 23:19 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!