| openstackgerrit | Tim Burke proposed openstack/swift master: Get rid of _delete_db https://review.opendev.org/693273 | 00:22 |
|---|---|---|
| *** timburke_ has quit IRC | 00:23 | |
| *** gyee has quit IRC | 00:27 | |
| *** timburke_ has joined #openstack-swift | 01:00 | |
| *** ChanServ sets mode: +v timburke_ | 01:00 | |
| timburke_ | DHE: if it were me, i'd go with 3.7 -- but that's mainly because i don't want to have to think about upgrading it again for a while :-) | 01:02 |
| DHE | that's good enough for me, as long as 3.7 is supported, reliable, etc. and I assume it is then | 01:09 |
| *** spsurya has joined #openstack-swift | 01:37 | |
| openstackgerrit | Merged openstack/swift stable/rocky: slo: Better handle non-manifest responses when refetching manifest https://review.opendev.org/690740 | 01:54 |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift feature/losf: Erase object header if we fail to register it to the index-server https://review.opendev.org/693127 | 02:08 |
| *** renich has joined #openstack-swift | 02:12 | |
| openstackgerrit | Romain LE DISEZ proposed openstack/swift master: proxy: stop sending chunks to objects with a Queue https://review.opendev.org/693116 | 02:59 |
| *** psachin has joined #openstack-swift | 03:14 | |
| clayg | seongsoocho: those flake8 formatting rules can be tricky, i've been really happy with installing https://github.com/nvie/vim-flake8 | 03:21 |
| clayg | seongsoocho: i install it using "vundle" and it makes the linter run everytime I save the file - so i catch formatting errors very quickly! | 03:22 |
| openstackgerrit | Tim Burke proposed openstack/swift feature/losf: Merge remote-tracking branch 'gerrit/master' into feature/losf https://review.opendev.org/693283 | 03:22 |
| seongsoocho | oh. It is really nice. I was used pure vim (not installed any pulgins) from the beginning... | 03:23 |
| kota_ | timburke: oh, you remember correctly regarding to json.*loads* https://docs.python.org/3/library/json.html#json.loads | 03:29 |
| *** tkajinam has joined #openstack-swift | 03:29 | |
| *** timburke_ has quit IRC | 03:30 | |
| *** timburke_ has joined #openstack-swift | 03:34 | |
| *** ChanServ sets mode: +v timburke_ | 03:34 | |
| rledisez | clayg, https://review.opendev.org/#/c/333331/ | 03:34 |
| patchbot | patch 333331 - swift - Preserve query params in tempurl - 4 patch sets | 03:34 |
| timburke_ | seongsoocho: https://bugs.launchpad.net/swift/+bug/1693149 -- you could probably crib from https://github.com/openstack/swift/commit/c6cebb6 a good bit | 03:38 |
| openstack | Launchpad bug 1693149 in OpenStack Object Storage (swift) "RingBuilder.save isn't atomic" [Wishlist,Confirmed] | 03:38 |
| timburke_ | https://bugs.launchpad.net/swift/+bug/1646410 is good, too | 03:39 |
| openstack | Launchpad bug 1646410 in OpenStack Object Storage (swift) "recon middleware complains too loudly about missing recon files" [Low,Confirmed] | 03:39 |
| timburke_ | https://bugs.launchpad.net/swift/+bug/1794601 - Use stronger hashes in formpost middleware | 03:40 |
| openstack | Launchpad bug 1794601 in OpenStack Object Storage (swift) "Formpost middleware should support stronger hash functions" [Undecided,New] | 03:40 |
| clayg | seongsoocho: if you want to stick in internal_client (since you're familiar with the unit test infrastructure for that module) you could probably bang out https://bugs.launchpad.net/swift/+bug/1680377 | 03:41 |
| openstack | Launchpad bug 1680377 in OpenStack Object Storage (swift) "Delete account with InternalClient" [Wishlist,In progress] - Assigned to HCLTech-SSW (hcltech-ssw) | 03:41 |
| clayg | seongsoocho: or https://bugs.launchpad.net/swift/+bugs?field.tag=low-hanging-fruit has a list of bugs that you could cherry pick from | 03:42 |
| seongsoocho | clayg: oh there are a tags. thanks | 03:43 |
| mattoliverau | clayg: oh nice tip on the vim-flake8 might have to try it out. | 03:46 |
| clayg | hi mattoliverau ! | 03:46 |
| mattoliverau | my vim setup is https://github.com/matthewoliver/vim_config so I can access it anywhere. It's built up over years and probably only useful to me. Though I know csmart uses it :) | 03:48 |
| mattoliverau | clayg: hey, how's the PTG? | 03:48 |
| seongsoocho | How to linking between patch and launchpad bugs? I tried write a "Closes-Bug" in commit message, it doesn't work. | 03:49 |
| mattoliverau | y'all having interesting and usful discussions | 03:49 |
| mattoliverau | seongsoocho: Closes-Bug: #<bugnumber> | 03:50 |
| clayg | mattoliverau: it's decidely LESS fun without you :'( | 03:50 |
| clayg | but still pretty fun :) | 03:50 |
| mattoliverau | seongsoocho: see https://wiki.openstack.org/wiki/GitCommitMessages | 03:50 |
| seongsoocho | https://review.opendev.org/#/c/693128/ <- this is my previous patch. Do I do it right? | 03:50 |
| patchbot | patch 693128 - swift - Fix bug internal client's delete_object logs 499 (MERGED) - 4 patch sets | 03:50 |
| clayg | mattoliverau: the best part by far has been seongsoocho | 03:50 |
| mattoliverau | seongsoocho: yup perfect! if you open it in gerrit, it's a link to the bug | 03:51 |
| seongsoocho | Yey! | 03:51 |
| mattoliverau | clayg: \o/ a new dev | 03:52 |
| mattoliverau | seongsoocho: where are you based? we need more apac devs ;) | 03:52 |
| clayg | mattoliverau: he runs a swift cluster too! | 03:52 |
| kota_ | :D | 03:52 |
| mattoliverau | EVEN BETTER! | 03:52 |
| seongsoocho | mattoliverau: south korea ! . I runs a swift cluster ( public cloud in south korea ) | 03:52 |
| mattoliverau | nice! | 03:53 |
| seongsoocho | This is my first time attending PTG, I am very happy that everyone welcomed me very much. | 03:54 |
| mattoliverau | seongsoocho: the Swift devs are definitely are the most welcoming, fun, and incredible people. It's an awesome community to be involved in! | 03:56 |
| mattoliverau | I may be biased tho :P | 03:56 |
| *** alexlecuyer-ch has joined #openstack-swift | 03:57 | |
| clayg | 🤗 mattoliverau | 03:58 |
| mattoliverau | So I've managed to talk with my new manager, and it seems it's ok for me to spend some of my work time working upstream on an opensource project, I mean that isn't related to the new role. So even though I'm not working on the Suse cloud team anymore, it seems I can still do some swift work. I just need to negoiate how much. | 03:59 |
| mattoliverau | so that's a bit of a win | 03:59 |
| mattoliverau | Not sure how much they'll come to the party on travel though. But baby steps :P | 04:00 |
| timburke_ | \o/ | 04:00 |
| mattoliverau | So we may need to do a hackathon at my beach house sometime ;P | 04:00 |
| clayg | i think that's a GREAT idea! | 04:01 |
| timburke_ | sold! let's do it! | 04:01 |
| timburke_ | https://bugs.launchpad.net/swift/+bug/1740326 | 04:03 |
| openstack | Launchpad bug 1740326 in OpenStack Object Storage (swift) "tempauth: Account ACLs allow users to delete their own accounts" [Undecided,New] | 04:03 |
| openstackgerrit | Merged openstack/swift master: Get rid of _delete_db https://review.opendev.org/693273 | 04:04 |
| kota_ | +100! Let's do it at the beach house! | 04:05 |
| alexlecuyer-ch | mattoliverau: wow good news!! | 04:10 |
| *** alexlecuyer-ch has quit IRC | 04:21 | |
| *** timburke_ has quit IRC | 04:21 | |
| *** timburke_ has joined #openstack-swift | 04:24 | |
| *** ChanServ sets mode: +v timburke_ | 04:24 | |
| *** timburke_ has quit IRC | 04:29 | |
| *** tkajinam has quit IRC | 04:37 | |
| *** timburke_ has joined #openstack-swift | 04:55 | |
| *** ChanServ sets mode: +v timburke_ | 04:55 | |
| *** alexlecuyer-ch has joined #openstack-swift | 04:55 | |
| *** mwheckmann has joined #openstack-swift | 04:58 | |
| *** alexlecuyer-ch has quit IRC | 04:58 | |
| *** alexlecuyer-ch has joined #openstack-swift | 04:58 | |
| mwheckmann | hello. I contacted this channel a couple of days ago in regards to some ghost container listings that I had. I think I may have found a root cause: On another cluster that we have, I have millions of async_pending updates. Largely due to some mis-configured container disks on new nodes that were added in the past couple of months. The object-updater can't keep up. | 05:00 |
| mwheckmann | My question now, is what should I do? speed up the updater (increase concurrency) or delete any async_pending's older than 1 week? | 05:01 |
| mwheckmann | I'm afraid of getting hit by ghost listings again. | 05:02 |
| mwheckmann | The container replicator should have done it's job to get the container DBs in sync, but what if some old pending PUT of a now deleted object comes along? I'm assuming there is something that I'm not getting. | 05:05 |
| zaitcev | I think the pendings are just summary changes. You know, when stat (HEAD) shows how many bytes are used. | 05:06 |
| mwheckmann | The pendings are a lot of DELETEs and PUTs. Many more DELETEs from the object-expirer. roughly 8:1 DELETEs vs PUTs | 05:09 |
| mwheckmann | zaitcev: The pendings will affect container listings. See https://bugs.launchpad.net/swift/+bug/1619408 | 05:10 |
| openstack | Launchpad bug 1619408 in OpenStack Object Storage (swift) "old async pendings result in ghost listings" [Medium,In progress] - Assigned to Matthew Oliver (matt-0) | 05:10 |
| clayg | mwheckmann: I've used something like this to get a picture of what's going in in big async pending piles -> https://gist.github.com/clayg/249c5d3ff580032a0d40751fc3f9f24b | 05:32 |
| clayg | dunno if more information might be helpful | 05:32 |
| clayg | if you have a bunch of async pendings tuning the updater is probably a really good idea - there's concurrency, but workers will probably have the bigger impact | 05:33 |
| clayg | mwheckmann: you also need to remove (or increase) the object-updater "rate limiting" which by default is fairly conservative | 05:33 |
| *** tkajinam has joined #openstack-swift | 05:34 | |
| openstackgerrit | Romain LE DISEZ proposed openstack/swift master: proxy: stop sending chunks to objects with a Queue https://review.opendev.org/693116 | 05:37 |
| *** viks___ has joined #openstack-swift | 05:47 | |
| *** pcaruana has joined #openstack-swift | 05:53 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Stop retrying every deletes in container-sync https://review.opendev.org/691729 | 06:07 |
| *** renich has quit IRC | 06:07 | |
| rledisez | Some though about hashes.pkl: https://etherpad.openstack.org/p/swift-replacing-hashes.pkl | 06:10 |
| rledisez | (I didn't read myself so sorry in advance for every thing :)) | 06:10 |
| *** diablo_rojo has joined #openstack-swift | 06:32 | |
| *** alexlecuyer-ch has quit IRC | 06:41 | |
| *** diablo_rojo__ has joined #openstack-swift | 06:44 | |
| *** diablo_rojoooooo has joined #openstack-swift | 06:45 | |
| *** diablo_rojo has quit IRC | 06:47 | |
| *** diablo_rojo__ has quit IRC | 06:48 | |
| *** diablo_rojo__ has joined #openstack-swift | 06:49 | |
| *** alexlecuyer-ch has joined #openstack-swift | 06:49 | |
| *** alexlecuyer-ch has quit IRC | 06:51 | |
| *** diablo_rojoooooo has quit IRC | 06:52 | |
| *** diablo_rojoooooo has joined #openstack-swift | 06:54 | |
| *** diablo_rojo__ has quit IRC | 06:56 | |
| *** diablo_rojo__ has joined #openstack-swift | 06:58 | |
| *** diablo_rojoooooo has quit IRC | 07:00 | |
| *** rcernin has quit IRC | 07:01 | |
| *** diablo_rojoooooo has joined #openstack-swift | 07:03 | |
| *** diablo_rojo__ has quit IRC | 07:06 | |
| *** diablo_rojoooooo has quit IRC | 07:07 | |
| *** alexlecuyer-ch has joined #openstack-swift | 07:13 | |
| *** timburke_ has quit IRC | 07:13 | |
| *** timburke_ has joined #openstack-swift | 07:14 | |
| *** ChanServ sets mode: +v timburke_ | 07:14 | |
| *** alexlecuyer-ch has quit IRC | 07:15 | |
| *** alexlecuyer-ch has joined #openstack-swift | 07:15 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: WIP: fix DB status replication https://review.opendev.org/693299 | 07:17 |
| *** timburke_ has quit IRC | 07:22 | |
| *** timburke_ has joined #openstack-swift | 07:25 | |
| *** ChanServ sets mode: +v timburke_ | 07:25 | |
| tdasilva | good mornign | 07:33 |
| *** timburke_ has quit IRC | 07:33 | |
| openstackgerrit | SeongSoo Cho proposed openstack/swift master: Add delete_account in InternalClient https://review.opendev.org/693305 | 07:35 |
| *** timburke_ has joined #openstack-swift | 07:38 | |
| *** ChanServ sets mode: +v timburke_ | 07:38 | |
| * tdasilva likes the idea of a hackathon at mattoliverau's beach house :P | 07:41 | |
| *** tesseract has joined #openstack-swift | 07:52 | |
| openstackgerrit | Merged openstack/swift feature/losf: Merge remote-tracking branch 'gerrit/master' into feature/losf https://review.opendev.org/693283 | 07:54 |
| *** alexlecuyer-ch has quit IRC | 08:08 | |
| *** alexlecuyer-ch has joined #openstack-swift | 08:10 | |
| clayg | seongsoocho: my first attempt to use the delete_account gave me a 405 - did you have to do anything special to make the request like a super duper user? | 08:11 |
| seongsoocho | clayg: no, I just add options in internal client config. `allow_account_management = true` | 08:12 |
| *** alexlecuyer-ch has quit IRC | 08:18 | |
| *** rdejoux has joined #openstack-swift | 08:21 | |
| *** ianychoi has joined #openstack-swift | 08:29 | |
| clayg | cschwede: kota_: alecuyer: rledisez: plan is to meet at marriot hotel lobby at 6pm | 08:39 |
| *** timburke_ has quit IRC | 08:40 | |
| *** pcaruana has quit IRC | 08:43 | |
| *** tkajinam has quit IRC | 08:44 | |
| openstackgerrit | Merged openstack/swift stable/rocky: Fix quarantine when object path is not a directory https://review.opendev.org/690738 | 09:09 |
| *** mikecmpbll has joined #openstack-swift | 09:17 | |
| *** pcaruana has joined #openstack-swift | 09:38 | |
| openstackgerrit | Romain LE DISEZ proposed openstack/swift master: proxy: stop sending chunks to objects with a Queue https://review.opendev.org/693116 | 10:02 |
| *** spsurya has quit IRC | 10:06 | |
| clayg | rledisez: hey! We’re here in the lobby. I’m hungry. | 10:16 |
| mwheckmann | clayg: thanks for the tips on the big async_pending piles. So no increased risk of ghost objects in listings? Even if some of the async_pendings are over a month old? I should just acclerate the updaters and let them go at it? | 12:23 |
| *** timburke_ has joined #openstack-swift | 12:25 | |
| *** ChanServ sets mode: +v timburke_ | 12:25 | |
| *** timburke_ has quit IRC | 12:39 | |
| openstackgerrit | Merged openstack/swift feature/losf: Erase object header if we fail to register it to the index-server https://review.opendev.org/693127 | 13:10 |
| *** timburke_ has joined #openstack-swift | 13:16 | |
| *** ChanServ sets mode: +v timburke_ | 13:16 | |
| *** timburke_ has quit IRC | 13:21 | |
| *** mwheckmann has quit IRC | 13:41 | |
| *** spsurya has joined #openstack-swift | 14:11 | |
| *** onovy has quit IRC | 14:15 | |
| *** timburke_ has joined #openstack-swift | 14:18 | |
| *** ChanServ sets mode: +v timburke_ | 14:18 | |
| *** timburke_ has quit IRC | 14:22 | |
| *** mwheckmann has joined #openstack-swift | 14:23 | |
| *** onovy has joined #openstack-swift | 14:26 | |
| *** nanzha has joined #openstack-swift | 14:46 | |
| *** nanzha has quit IRC | 14:50 | |
| *** mwheckmann has quit IRC | 14:52 | |
| *** mikecmpbll has quit IRC | 15:09 | |
| *** mikecmpbll has joined #openstack-swift | 15:10 | |
| *** mikecmpbll has quit IRC | 16:02 | |
| *** mikecmpbll has joined #openstack-swift | 16:05 | |
| *** timburke_ has joined #openstack-swift | 16:40 | |
| *** ChanServ sets mode: +v timburke_ | 16:40 | |
| *** rdejoux has quit IRC | 16:51 | |
| *** tesseract has quit IRC | 17:03 | |
| *** timburke_ has quit IRC | 17:04 | |
| *** psachin has quit IRC | 17:05 | |
| *** henriqueof has joined #openstack-swift | 17:14 | |
| *** timburke_ has joined #openstack-swift | 17:17 | |
| *** ChanServ sets mode: +v timburke_ | 17:17 | |
| *** timburke_ has quit IRC | 17:21 | |
| *** gyee has joined #openstack-swift | 17:26 | |
| *** mikecmpbll has quit IRC | 17:35 | |
| *** henriqueof1 has joined #openstack-swift | 17:42 | |
| *** henriqueof has quit IRC | 17:42 | |
| openstackgerrit | Darrell Bishop proposed openstack/swift master: Seamlessly reload servers with SIGUSR1 https://review.opendev.org/691423 | 18:16 |
| *** timburke_ has joined #openstack-swift | 18:52 | |
| *** ChanServ sets mode: +v timburke_ | 18:52 | |
| *** pcaruana has quit IRC | 18:54 | |
| *** timburke_ has quit IRC | 18:57 | |
| *** spsurya has quit IRC | 20:02 | |
| *** threestrands has quit IRC | 20:17 | |
| *** henriqueof1 has quit IRC | 20:19 | |
| *** mwheckmann has joined #openstack-swift | 21:14 | |
| *** ianychoi has quit IRC | 21:35 | |
| *** ianychoi has joined #openstack-swift | 21:36 | |
| *** ccamel has quit IRC | 21:44 | |
| *** camelCaser has joined #openstack-swift | 21:44 | |
| mattoliverau | morning | 22:00 |
| *** mwheckmann has quit IRC | 22:15 | |
| *** rcernin has joined #openstack-swift | 22:23 | |
| *** onovy has quit IRC | 23:01 | |
| *** timburke_ has joined #openstack-swift | 23:18 | |
| *** ChanServ sets mode: +v timburke_ | 23:18 | |
| seongsoocho | good morning. | 23:19 |
| seongsoocho | https://review.opendev.org/#/c/693305/ The zuul gate was failed. I saw a log for failed test case, but I don't know why the test fail. | 23:20 |
| patchbot | patch 693305 - swift - Add delete_account in InternalClient - 1 patch set | 23:20 |
| timburke_ | seongsoocho: so the gate tests ten to be a little flakey -- the resources available to us via zuul are a lot more constrained than what we'd typically have in a dev environment, where we'd originally written the test | 23:24 |
| timburke_ | probe tests in particular tend to have cascading failures, since it ties to leave the environment more-or-less intact following a failure. so in this particular case, test_post_ctype_replicated_when_previous_incomplete_puts failed, and then there must've still been a process somewhere with a file open on the loopback that caused resetswift to fail fo a bunch of subsequent tests | 23:27 |
| timburke_ | all of which is to say, i wouldn't worry about it -- i'll just recheck it, and (hopefully) it'll merge this time around | 23:28 |
| seongsoocho | oh. ok. I was worried that the failure was the cause of my patch. | 23:30 |
| timburke_ | nope, it looks great :-) | 23:31 |
| seongsoocho | \0/ yey | 23:35 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!