| mattoliverau | timburke: done, shall I +a or leave that to clayg? | 00:11 |
|---|---|---|
| timburke | either way, i suppose -- pretty sure he or i will be packaging a swift that includes that next week | 00:12 |
| mattoliverau | well in the that case.. let's get it in ;) | 00:47 |
| mattoliverau | +A'ed based on his previous +2 anyway | 00:47 |
| *** jv|afk has quit IRC | 01:17 | |
| *** rcernin has joined #openstack-swift | 02:34 | |
| *** rcernin has quit IRC | 02:41 | |
| *** viks____ has joined #openstack-swift | 02:47 | |
| *** tkajinam is now known as tkajinam|lunch | 03:19 | |
| *** tkajinam|lunch is now known as tkajinam | 03:19 | |
| *** manuvakery has joined #openstack-swift | 03:31 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-swift | 04:33 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Extend probe sharder tests https://review.opendev.org/707521 | 05:11 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Add a new URL parameter to allow for async cleanup of SLO segments https://review.opendev.org/733026 | 05:54 |
| openstackgerrit | Merged openstack/swift master: relinker: Improve performance by limiting I/O https://review.opendev.org/695344 | 06:09 |
| *** viks____ has quit IRC | 06:10 | |
| *** viks____ has joined #openstack-swift | 06:52 | |
| *** rpittau|afk is now known as rpittau | 07:04 | |
| viks____ | s | 07:09 |
| *** ccamacho has joined #openstack-swift | 07:40 | |
| *** zigo_ is now known as zigo | 07:58 | |
| *** ravsingh has joined #openstack-swift | 08:19 | |
| *** tkajinam has quit IRC | 08:31 | |
| *** dtantsur|afk is now known as dtantsur | 08:34 | |
| *** rpittau is now known as rpittau|bbl | 10:08 | |
| *** dtantsur has left #openstack-swift | 10:37 | |
| *** jv|afk has joined #openstack-swift | 11:46 | |
| *** rpittau|bbl is now known as rpittau | 11:59 | |
| clayg | timburke: so many recheck 😥 | 14:20 |
| timburke | clayg, yeah, and i feel like i'm seeing a lot more timeouts than we've been seeing lately. might be time to increase them more :-/ | 14:28 |
| timburke | it's weird though -- we'll often see some jobs pass in half the time limit *on the same build* as the ones that timeout while running a similar set of tests | 14:30 |
| timburke | (seems to have certain parallels with our own concerns for latency variability) | 14:31 |
| *** jv|afk has quit IRC | 14:45 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Quiet eventlet exceptions in test https://review.opendev.org/733863 | 14:50 |
| *** jv|afk has joined #openstack-swift | 15:03 | |
| *** jv|afk has quit IRC | 15:10 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Add a new URL parameter to allow for async cleanup of SLO segments https://review.opendev.org/733026 | 15:20 |
| timburke | i hadn't thought about it much before, but the probe test i added for ^^^ revealed that it'd be handy to have sub-second precision for the expirer deletes | 15:21 |
| *** gyee has joined #openstack-swift | 15:44 | |
| *** tkajinam has joined #openstack-swift | 16:08 | |
| clayg | timburke: I've missed some of the PTG discussions this week - is there a strong consensus on p 733026 | 16:10 |
| patchbot | https://review.opendev.org/#/c/733026/ - swift - Add a new URL parameter to allow for async cleanup... - 3 patch sets | 16:10 |
| *** rpittau is now known as rpittau|afk | 16:10 | |
| timburke | the impression i got was that people found it reasonable, particularly for s3api. it's probably still going to trip people up when it comes to quotas, though | 16:29 |
| timburke | i'm not sure that's avoidable though :-( | 16:30 |
| timburke | unless... we get the segments out of the user's account entirely... 🤔 | 16:30 |
| timburke | use container-update overrides to get the total size into the user-namespace container, and have a hack like we've got for sharding to zero-out the segment container's account update so we don't double count... that all feels pretty round about and hacky though | 16:32 |
| timburke | anyway, those seem like ideas for ALOs. p 733026 helps *now* (and covers yesterday's data) | 16:40 |
| patchbot | https://review.opendev.org/#/c/733026/ - swift - Add a new URL parameter to allow for async cleanup... - 3 patch sets | 16:40 |
| timburke | still need to beef up unit tests, but i'm glad i got the probe test written (especially since it revealed issues where the DELETE might 409) | 16:41 |
| *** tkajinam has quit IRC | 16:49 | |
| *** manuvakery has quit IRC | 17:30 | |
| openstackgerrit | Merged openstack/swift-specs master: Retire swift-specs https://review.opendev.org/733001 | 17:40 |
| *** AJaeger has joined #openstack-swift | 17:43 | |
| *** AJaeger has left #openstack-swift | 17:45 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: wip: asyc concurrent ecfragfetcher https://review.opendev.org/711342 | 18:56 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Start to decouple Object GET path https://review.opendev.org/733911 | 18:56 |
| clayg | ^ there's still a ton to do there; i'm just shelving it to look at p 722552 | 18:56 |
| patchbot | https://review.opendev.org/#/c/722552/ - swift - s3api: Check whether versioning is enabled more - 3 patch sets | 18:56 |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: Add config option to include uploadId in GET/HEAD responses https://review.opendev.org/728589 | 19:06 |
| *** ravsingh has quit IRC | 19:11 | |
| openstackgerrit | Merged openstack/swift master: func tests: work with etag-quoter on by default https://review.opendev.org/695131 | 20:52 |
| tdasilva | heh, only took 24 rechecks | 21:00 |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: Allow CompleteMultipartUpload requests to be retried https://review.opendev.org/733919 | 21:09 |
| timburke | clayg, ^^^ is an alternate approach for the problem behind p 728589 -- needs a bunch of tests, though. and i think we should be able to use the etag that's returned to decide whether we can just skip the SLO PUT entirely | 21:10 |
| patchbot | https://review.opendev.org/#/c/728589/ - swift - s3api: Add config option to include uploadId in GE... - 3 patch sets | 21:10 |
| timburke | (which should be handy if the reason the client felt the need to retry in the first place was a timeout waiting for everything to verify) | 21:11 |
| clayg | interesting! | 21:12 |
| openstackgerrit | Tim Burke proposed openstack/swift master: tests: Bump up timeout for unit and in-process func tests https://review.opendev.org/733922 | 21:21 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: s3api: pre-flight versions-enabled https://review.opendev.org/733930 | 22:11 |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: Allow CompleteMultipartUpload requests to be retried https://review.opendev.org/733919 | 22:39 |
| *** rcernin has joined #openstack-swift | 23:02 | |
| timburke | ah, crap -- i forgot to have clayg and zaitcev talk through https://review.opendev.org/#/c/724141/ this week :-( | 23:05 |
| patchbot | patch 724141 - swift - swift-get-nodes: Unquote paths before working with... - 2 patch sets | 23:05 |
| zaitcev | bah | 23:06 |
| zaitcev | Just have it your way. | 23:06 |
| zaitcev | I would prefer if the new quoting was enabled by an explicit flag, but whatever. | 23:06 |
| clayg | zaitcev: EVERYTIME I lookup an object with swift-get-nodes I'm pulling the name from a log message | 23:08 |
| clayg | and cussing that it doesn't already expect quoted | 23:08 |
| zaitcev | Hmm. | 23:08 |
| clayg | it's *worse* that we compounded the problem with null namespaces | 23:08 |
| clayg | I don't think anyone's scripted it - it MIGHT surprise some people | 23:08 |
| zaitcev | Well, you know... I always type df -k even though it's the default on Linux. | 23:08 |
| clayg | we can do the --do-what-i-already-want flag - honestly I'd prefer to type that then trying to ^H^H^H %20 everytime I have an even slightly interesting name | 23:09 |
| clayg | zaitcev: +1 | 23:09 |
| zaitcev | well | 23:09 |
| zaitcev | I'd prefer -n or something | 23:10 |
| zaitcev | The --do-what-i-already-want is seriously asking for an alias, like git ffpull or git pullff (I wish people agreed on that one too). | 23:11 |
| clayg | yeah -q sounds perfect! I can type swift-get-nodes -q all day! | 23:11 |
| zaitcev | My first thought was -q but it's used as "quiet" so often that I asked for -n. But either way works by me. I agree that it's unlikely that anyone scripts it among my customers. | 23:12 |
| timburke | the prevalence of -q being a shorthand for --quiet was why i went with -Q originally | 23:19 |
| *** ab-a has quit IRC | 23:20 | |
| *** ab-a has joined #openstack-swift | 23:21 | |
| timburke | so i guess my remaining question is, should we try to move toward expecting quoted by default? | 23:21 |
| *** rcernin has quit IRC | 23:27 | |
| clayg | -Q is a bridge to far. Shift AND a q!? What’s the pneumonic for -n? —no-quote!? 😅 | 23:31 |
| clayg | I’m already off work today. Don’t listen to me. (Unless you’re doing virtual PTG after hours) | 23:32 |
| clayg | This was a good week! Thanks y’all! 💪 | 23:33 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Remove etag-quoter from 2.25.0 release notes https://review.opendev.org/733934 | 23:36 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!