| kota_ | good morning | 00:00 |
|---|---|---|
| openstackgerrit | Matthew Oliver proposed openstack/swift master: fixup! Add container sharding to Swift containers https://review.openstack.org/450491 | 00:12 |
| mattoliverau | kota_: morning | 00:14 |
| kota_ | mattoliverau: o/ | 00:14 |
| kota_ | yey, container sharding | 00:15 |
| mattoliverau | well a sharding fixup thanks to timburke :) | 00:18 |
| *** chsc has quit IRC | 00:19 | |
| timburke | mattoliverau: good call on counting the number of calls to get_items_since -- i wonder if we could inspect the args to verify which sub-brokers we hit? it makes me a little nervous that the first and last of the added checks look so similar :-/ | 00:20 |
| mattoliverau | yeah I know, but getting the name of the broker called, is what we'd have to figure out | 00:23 |
| mattoliverau | hmm, maybe, we do have access to the self call. | 00:25 |
| timburke | yeah, i was just about to play with trying to track args[0]._db_file or something... | 00:25 |
| mattoliverau | timburke: yup, figured it out | 00:28 |
| mattoliverau | args[0].db_file. seems to be what we want | 00:28 |
| mattoliverau | or maybe via the conn | 00:28 |
| mattoliverau | let me add it to the test and see how it goes | 00:29 |
| mattoliverau | ok args[0].conn.db_file is what we want | 00:29 |
| mattoliverau | just args[0].db_file will always point to pivot (obviously) | 00:30 |
| timburke | ah, yeah, makes sense | 00:30 |
| *** hoonetorg has quit IRC | 00:32 | |
| *** catintheroof has quit IRC | 00:34 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift master: fixup! Add container sharding to Swift containers https://review.openstack.org/450491 | 00:36 |
| mattoliverau | timburke: ^ how about that? | 00:37 |
| *** sams-gleb has joined #openstack-swift | 00:39 | |
| *** sams-gleb has quit IRC | 00:43 | |
| *** hoonetorg has joined #openstack-swift | 00:45 | |
| timburke | mattoliverau: love it | 00:48 |
| *** jamielennox is now known as jamielennox|away | 01:02 | |
| openstackgerrit | XieYingYun proposed openstack/swift master: Fix some reST field lists in docstrings https://review.openstack.org/449890 | 01:15 |
| *** jamielennox|away is now known as jamielennox | 01:16 | |
| *** NM has quit IRC | 01:21 | |
| *** furlongm has quit IRC | 01:28 | |
| openstackgerrit | liuyamin proposed openstack/python-swiftclient master: Fix some reST field lists in docstrings https://review.openstack.org/450656 | 01:30 |
| *** bkopilov has quit IRC | 01:38 | |
| *** bkopilov_ has quit IRC | 01:38 | |
| *** hoonetorg has quit IRC | 01:42 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift master: Add Composite Ring Functionality https://review.openstack.org/441921 | 01:46 |
| *** hoonetorg has joined #openstack-swift | 01:54 | |
| *** tanee is now known as tanee_away | 01:57 | |
| *** tanee_away is now known as tanee | 01:58 | |
| *** JimCheung has quit IRC | 02:02 | |
| *** JimCheung has joined #openstack-swift | 02:03 | |
| *** zhurong has joined #openstack-swift | 02:04 | |
| *** JimCheung has quit IRC | 02:07 | |
| openstackgerrit | liuyamin proposed openstack/swift master: Fix some reST field lists in docstrings https://review.openstack.org/451143 | 02:17 |
| *** chosafine has joined #openstack-swift | 02:36 | |
| *** sams-gleb has joined #openstack-swift | 02:41 | |
| *** tone_zrt has joined #openstack-swift | 02:41 | |
| *** tone_zrt has quit IRC | 02:42 | |
| *** sams-gleb has quit IRC | 02:45 | |
| *** bkopilov has joined #openstack-swift | 02:57 | |
| *** chosafine has quit IRC | 03:08 | |
| *** links has joined #openstack-swift | 03:12 | |
| *** raginbajin has quit IRC | 03:14 | |
| *** pdardeau_ has quit IRC | 03:15 | |
| *** pdardeau_ has joined #openstack-swift | 03:17 | |
| *** raginbajin has joined #openstack-swift | 03:21 | |
| *** jeblair has quit IRC | 03:24 | |
| *** jeblair has joined #openstack-swift | 03:25 | |
| *** kei_yama has joined #openstack-swift | 03:38 | |
| *** SkyRocknRoll has quit IRC | 03:50 | |
| *** Dinesh_Bhor has joined #openstack-swift | 04:00 | |
| *** psachin has joined #openstack-swift | 04:23 | |
| *** gkadam has joined #openstack-swift | 04:42 | |
| *** sams-gleb has joined #openstack-swift | 04:43 | |
| *** jordanP has joined #openstack-swift | 04:47 | |
| *** klrmn has quit IRC | 04:48 | |
| *** sams-gleb has quit IRC | 04:49 | |
| *** jordanP has quit IRC | 04:51 | |
| *** rcernin has joined #openstack-swift | 05:39 | |
| *** zaitcev has quit IRC | 05:50 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift master: sharding - Fix container server put/delete piv https://review.openstack.org/451197 | 05:51 |
| mattoliverau | timburke: ^ theres one clean up from the many you spotted. | 05:53 |
| *** zhurong has quit IRC | 05:54 | |
| *** jamielennox is now known as jamielennox|away | 06:10 | |
| *** jaosorior has joined #openstack-swift | 06:14 | |
| *** pcaruana has joined #openstack-swift | 06:24 | |
| *** jamielennox|away is now known as jamielennox | 06:25 | |
| *** hseipp has joined #openstack-swift | 06:40 | |
| *** sams-gleb has joined #openstack-swift | 06:46 | |
| openstackgerrit | Merged openstack/liberasurecode master: ISA-L: Only calculate gf tables on init, not every encode https://review.openstack.org/431850 | 06:49 |
| *** jamielennox is now known as jamielennox|away | 06:50 | |
| *** sams-gleb has quit IRC | 06:50 | |
| *** sams-gleb has joined #openstack-swift | 06:51 | |
| *** McMurlock1 has joined #openstack-swift | 06:54 | |
| *** oshritf has joined #openstack-swift | 06:56 | |
| *** jamielennox|away is now known as jamielennox | 06:57 | |
| *** tesseract has joined #openstack-swift | 07:14 | |
| acoles | good morning | 07:21 |
| acoles | mattoliverau: do I almost overlap with you now? (in time I mean!) | 07:22 |
| mattoliverau | acoles: morning.. a little, and you will more so next week when my time changes. My clocks go back so I'll be around for another hour each day :) | 07:23 |
| acoles | mattoliverau: oh so you haven't changed yet. even better | 07:24 |
| *** amoralej|off is now known as amoralej | 07:24 | |
| admin6 | acoles: sorry for the noise yesterday, it probably was an error on my side. Did you receive my email? | 07:25 |
| acoles | admin6: ok no problem, glad to hear it. | 07:26 |
| admin6 | acoles: me too, i do prefer that ;-) | 07:26 |
| acoles | admin6: I didn't receive email | 07:27 |
| acoles | admin6: btw, there have been some other bug fixes in recent swift releases, so upgrade when you are able to :) | 07:28 |
| kota_ | acoles, admin6: good morning | 07:29 |
| acoles | kota_: o/ | 07:29 |
| kota_ | and glad that it sounds no bug in swift | 07:29 |
| *** sams-gleb has quit IRC | 07:29 | |
| acoles | hehe. I'm sure there may be *a* bug in swift - but i'm glad if it's not *that* bug :) | 07:30 |
| admin6 | acoles: ok, basically, another sys amdin upgraded the ubuntu system (including some swift packages) one month ago without telling me, and the patched version of ssync script was overwritten. So new fragment with error was created and not detected… again. | 07:30 |
| acoles | admin6: :/ so the updated auditor should find that bad frag and remove it | 07:31 |
| *** geaaru has joined #openstack-swift | 07:32 | |
| admin6 | acoles: in fact, at least ssync_sender.py, ssync_receiver.py and diskfile.py was overwritten | 07:33 |
| *** zhurong has joined #openstack-swift | 07:34 | |
| *** sams-gleb has joined #openstack-swift | 07:34 | |
| kota_ | acoles: oh, that's my English fault :/ | 07:35 |
| *** sams-gleb has quit IRC | 07:40 | |
| *** sams-gleb has joined #openstack-swift | 07:40 | |
| acoles | admin6: makes sense...the check for corrupt fragments is actually in diskfile.py , the auditor just happens to trigger the check. so if diskfile regressed then the bad fragment would not be removed. | 07:41 |
| acoles | kota_: thanks for the update to patch 441921 | 07:44 |
| patchbot | https://review.openstack.org/#/c/441921/ - swift - Add Composite Ring Functionality | 07:44 |
| *** sams-gleb has quit IRC | 07:45 | |
| *** sileht has quit IRC | 07:45 | |
| acoles | timburke: oops https://review.openstack.org/#/c/446142/2//COMMIT_MSG@24 I guess I need to learn to scroll down in the commit message frame :) | 07:55 |
| *** tanee is now known as tanee_away | 08:03 | |
| *** tanee_away is now known as tanee | 08:07 | |
| *** jordanP has joined #openstack-swift | 08:14 | |
| *** cbartz has joined #openstack-swift | 08:15 | |
| *** jamielennox is now known as jamielennox|away | 08:17 | |
| kota_ | acoles: sure | 08:24 |
| * kota_ is leaving his office earlier than usual, today, for heading dinner with his collegues. | 08:25 | |
| openstackgerrit | Merged openstack/swift master: Test that Manager.reload does stop/start in that order https://review.openstack.org/449567 | 08:26 |
| *** gkadam is now known as gkadam-mtg | 08:30 | |
| *** gkadam_ has joined #openstack-swift | 08:32 | |
| *** gkadam-mtg has quit IRC | 08:35 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift master: Modify _get_hashes() arguments to be more generic https://review.openstack.org/449709 | 08:39 |
| openstackgerrit | Merged openstack/swift master: Fix some reST field lists in docstrings https://review.openstack.org/449890 | 08:56 |
| *** JimCheung has joined #openstack-swift | 09:04 | |
| *** JimCheung has quit IRC | 09:08 | |
| *** gkadam_ has quit IRC | 09:20 | |
| *** jordanP has quit IRC | 09:24 | |
| *** SkyRocknRoll has joined #openstack-swift | 09:27 | |
| *** gkadam_ has joined #openstack-swift | 09:31 | |
| *** gkadam_ is now known as gkadam | 09:35 | |
| openstackgerrit | Merged openstack/swift master: Use swift tempurl instaed of swift-temp-url https://review.openstack.org/450494 | 09:53 |
| *** McMurlock1 has quit IRC | 09:54 | |
| *** mvk has quit IRC | 10:00 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift master: Modify _get_hashes() arguments to be more generic https://review.openstack.org/449709 | 10:02 |
| *** silor has joined #openstack-swift | 10:04 | |
| acoles | alecuyer: in answer to your question here https://review.openstack.org/#/c/449709/1/swift/obj/diskfile.py@1062 I think it is ok for the function to be in diskfile. If you can modify the other potential uses for it in diskfile then that would be even better. | 10:12 |
| *** jordanP has joined #openstack-swift | 10:30 | |
| *** mvk has joined #openstack-swift | 10:30 | |
| *** McMurlock1 has joined #openstack-swift | 10:30 | |
| *** klrmn has joined #openstack-swift | 10:36 | |
| *** klrmn has quit IRC | 10:44 | |
| cbartz | Anyone here for reviewing patch 414232 ? | 11:10 |
| patchbot | https://review.openstack.org/#/c/414232/ - swift - Exclude containers for account quota | 11:10 |
| *** SkyRocknRoll has quit IRC | 11:11 | |
| *** bkopilov has quit IRC | 11:21 | |
| *** zhurong has quit IRC | 11:34 | |
| *** spotz is now known as spotz_zzz | 11:40 | |
| *** NM has joined #openstack-swift | 11:43 | |
| *** gkadam is now known as gkadam-afk | 11:47 | |
| *** bikmak has joined #openstack-swift | 11:49 | |
| *** NM has quit IRC | 11:49 | |
| *** amoralej is now known as amoralej|lunch | 11:56 | |
| *** spotz_zzz is now known as spotz | 11:58 | |
| *** chlong has quit IRC | 12:05 | |
| alecuyer | acoles: noted, thanks a lot for your review. I will do this and fix the docstring | 12:10 |
| *** sams-gleb has joined #openstack-swift | 12:10 | |
| *** NM has joined #openstack-swift | 12:11 | |
| *** gkadam-afk is now known as gkadam | 12:14 | |
| *** mvk has quit IRC | 12:25 | |
| *** mvk has joined #openstack-swift | 12:38 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements https://review.openstack.org/88736 | 12:41 |
| *** gkadam has quit IRC | 12:53 | |
| *** bkopilov has joined #openstack-swift | 12:59 | |
| *** PavelK has joined #openstack-swift | 13:13 | |
| *** klamath has joined #openstack-swift | 13:14 | |
| *** klamath has quit IRC | 13:14 | |
| *** klamath has joined #openstack-swift | 13:15 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift master: Modify _get_hashes() arguments to be more generic https://review.openstack.org/449709 | 13:17 |
| *** amoralej|lunch is now known as amoralej | 13:17 | |
| *** sileht has joined #openstack-swift | 13:19 | |
| *** links has quit IRC | 13:20 | |
| *** zhurong has joined #openstack-swift | 13:22 | |
| *** catintheroof has joined #openstack-swift | 13:30 | |
| *** jaosorior has quit IRC | 13:38 | |
| *** chlong has joined #openstack-swift | 13:40 | |
| *** catintheroof has quit IRC | 13:41 | |
| openstackgerrit | Pavel Kvasnička proposed openstack/swift master: Container/account disk drive error results double space usage on rest drives https://review.openstack.org/448480 | 13:43 |
| *** jaosorior has joined #openstack-swift | 13:43 | |
| *** vint_bra has joined #openstack-swift | 13:46 | |
| *** tongli has joined #openstack-swift | 13:51 | |
| PavelK | Hi guys. I'd like a review of patch 448480 - is someone able to look at it? | 13:52 |
| patchbot | https://review.openstack.org/#/c/448480/ - swift - Container/account disk drive error results double ... | 13:52 |
| *** vinsh has quit IRC | 13:52 | |
| *** vinsh has joined #openstack-swift | 13:53 | |
| acoles | PavelK: I have it on my list, but I am working a short week so may not get there until next week | 13:54 |
| PavelK | acoles: thanks that you notice it. we would like to apply the patch in production because some drives looks unmounted sometimes and we expect that excessive replication helps others to fail :-) | 13:57 |
| *** JimCheung has joined #openstack-swift | 14:04 | |
| *** zhurong_ has joined #openstack-swift | 14:06 | |
| *** zhurong has quit IRC | 14:07 | |
| *** JimCheung has quit IRC | 14:08 | |
| *** zhurong_ has quit IRC | 14:18 | |
| *** vinsh has quit IRC | 14:29 | |
| *** vinsh has joined #openstack-swift | 14:43 | |
| *** ma9_ has joined #openstack-swift | 14:45 | |
| *** psachin has quit IRC | 14:53 | |
| openstackgerrit | liuyamin proposed openstack/swift master: Fix the reST field raises in docstrings https://review.openstack.org/451143 | 14:54 |
| *** jistr is now known as jistr|mtg | 15:01 | |
| *** rcernin has quit IRC | 15:05 | |
| *** klrmn has joined #openstack-swift | 15:13 | |
| *** oshritf has quit IRC | 15:18 | |
| *** tongli has quit IRC | 15:26 | |
| acoles | notmyname: when adding docs (in this case for composite rings) is there any preference between including by reference text from a py module into the rst vs. writing the doc directly in the .rst (overview_ring.rst in this case)? | 15:27 |
| *** ma9_ has quit IRC | 15:29 | |
| openstackgerrit | Merged openstack/python-swiftclient master: Fix some reST field lists in docstrings https://review.openstack.org/450656 | 15:39 |
| *** ma9_ has joined #openstack-swift | 15:40 | |
| *** ma9_ has left #openstack-swift | 15:44 | |
| *** tesseract has quit IRC | 15:58 | |
| *** klrmn has quit IRC | 16:14 | |
| notmyname | good morning | 16:16 |
| notmyname | acoles: I'm not sure I completely understand | 16:17 |
| *** d0ugal has quit IRC | 16:17 | |
| *** oshritf has joined #openstack-swift | 16:18 | |
| *** JimCheung has joined #openstack-swift | 16:20 | |
| acoles | notmyname: hmm, I would link to an example but unfortunately github renders the rst so can't link to source lines | 16:20 |
| acoles | notmyname: but in overview_ring.rst search for "Ring Builder Analyzer" where the test is included by ref from the py module docstring, vs there being text written in the rst file | 16:21 |
| acoles | s/test/text/ | 16:21 |
| notmyname | ah | 16:21 |
| acoles | I like having the doc in the py module where it can be read along with source, but it does mean you can't just go read the rst file | 16:22 |
| notmyname | I think the clear wrong answer is to have it in both places | 16:23 |
| acoles | notmyname: noooo! | 16:23 |
| acoles | three places at least ;) | 16:23 |
| notmyname | but I don't have a strong preference for one over the other (I'd probably have a weak preference for in the rst, but I wouldn't -1 anything for that) | 16:23 |
| *** gpocentek has joined #openstack-swift | 16:24 | |
| acoles | notmyname: ok, thanks | 16:24 |
| acoles | notmyname: of course, the easy part is writing the text :P | 16:25 |
| *** saltsa has quit IRC | 16:30 | |
| *** saltsa has joined #openstack-swift | 16:32 | |
| *** d0ugal has joined #openstack-swift | 16:39 | |
| *** chsc has joined #openstack-swift | 16:39 | |
| *** chsc has joined #openstack-swift | 16:39 | |
| *** oshritf has quit IRC | 16:40 | |
| notmyname | acoles: I'll never believe that the docs are "the easy part". (unless someone else is doing it. then it's trivial ;-) | 16:41 |
| notmyname | it's just typing! | 16:41 |
| acoles | heh | 16:42 |
| *** ChubYann has joined #openstack-swift | 16:42 | |
| *** d0ugal has quit IRC | 16:43 | |
| *** cbartz has left #openstack-swift | 16:45 | |
| notmyname | has anyone else done the repro steps from PavelK in https://bugs.launchpad.net/swift/+bug/1675500 yet? | 16:45 |
| openstack | Launchpad bug 1675500 in OpenStack Object Storage (swift) "Container/account disk drive fault results replication on all rest drives" [Undecided,In progress] - Assigned to Pavel Kvasnička (pavel-kvasnicka) | 16:45 |
| timburke | good morning | 16:47 |
| *** jistr|mtg is now known as jistr | 16:47 | |
| *** SkyRocknRoll has joined #openstack-swift | 16:51 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Add Composite Ring Functionality https://review.openstack.org/441921 | 16:54 |
| *** Renich has quit IRC | 16:54 | |
| *** McMurlock1 has quit IRC | 16:55 | |
| *** silor has quit IRC | 16:58 | |
| *** vinsh has quit IRC | 16:58 | |
| *** sams-gleb has quit IRC | 17:00 | |
| *** chlong has quit IRC | 17:03 | |
| *** klrmn has joined #openstack-swift | 17:05 | |
| *** links has joined #openstack-swift | 17:07 | |
| *** sams-gleb has joined #openstack-swift | 17:19 | |
| *** links has quit IRC | 17:19 | |
| *** hseipp has quit IRC | 17:20 | |
| *** ouchkernel has joined #openstack-swift | 17:29 | |
| *** mathiasb has quit IRC | 17:35 | |
| *** mathiasb has joined #openstack-swift | 17:39 | |
| *** mvk has quit IRC | 17:41 | |
| *** tonanhngo has joined #openstack-swift | 17:44 | |
| *** tonanhngo_ has joined #openstack-swift | 17:48 | |
| *** tonanhngo has quit IRC | 17:48 | |
| *** eckesicle has joined #openstack-swift | 17:50 | |
| *** pcaruana has quit IRC | 17:52 | |
| *** tonanhngo_ has quit IRC | 17:52 | |
| *** geaaru has quit IRC | 17:53 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Add composite ring doc https://review.openstack.org/451500 | 17:53 |
| *** tonanhngo has joined #openstack-swift | 17:54 | |
| *** chlong has joined #openstack-swift | 17:58 | |
| *** amoralej is now known as amoralej|off | 18:07 | |
| *** chlong has quit IRC | 18:07 | |
| *** oshritf has joined #openstack-swift | 18:07 | |
| *** mvk has joined #openstack-swift | 18:13 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Experimental swift-ring-composer CLI to build composite rings https://review.openstack.org/451507 | 18:13 |
| acoles | kota_: ^^ found a bug, plus wrote some docs and revamped a very simple CLI | 18:14 |
| *** zaitcev has joined #openstack-swift | 18:17 | |
| *** ChanServ sets mode: +v zaitcev | 18:17 | |
| *** chlong has joined #openstack-swift | 18:21 | |
| openstackgerrit | Thiago da Silva proposed openstack/python-swiftclient master: ISO 8601 timestamps for tempurl https://review.openstack.org/423377 | 18:27 |
| *** hseipp has joined #openstack-swift | 18:29 | |
| *** chlong has quit IRC | 18:34 | |
| *** openstackstatus has joined #openstack-swift | 18:45 | |
| *** ChanServ sets mode: +v openstackstatus | 18:45 | |
| MooingLemur | I forget, can you create containers referring to storage policy by number or only by name? | 18:47 |
| notmyname | MooingLemur: name | 18:47 |
| MooingLemur | gotcha | 18:48 |
| MooingLemur | thanks | 18:48 |
| *** chlong has joined #openstack-swift | 18:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-swiftclient master: Updated from global requirements https://review.openstack.org/89250 | 18:51 |
| *** oshritf has quit IRC | 19:13 | |
| *** SkyRocknRoll has quit IRC | 19:14 | |
| notmyname | golang-for-swift tc document has been proposed at https://review.openstack.org/#/c/451524/ | 19:14 |
| notmyname | we'll talk about it in the meeting today too | 19:15 |
| notmyname | thank you tdasilva for you hard work on it | 19:15 |
| *** gyee has joined #openstack-swift | 19:15 | |
| *** hseipp has quit IRC | 19:19 | |
| tdasilva | notmyname: it was a real team effort again! | 19:22 |
| *** openstackgerrit has quit IRC | 19:33 | |
| *** silor has joined #openstack-swift | 19:33 | |
| *** jamielennox|away is now known as jamielennox | 19:33 | |
| *** chlong has quit IRC | 19:34 | |
| *** d0ugal has joined #openstack-swift | 19:49 | |
| *** vinsh has joined #openstack-swift | 19:55 | |
| jrichli | I just came across a good explanation of why some minor code changes can be problematic: https://docs.openstack.org/developer/glance/contributing/minor-code-changes.html | 19:59 |
| *** d0ugal has quit IRC | 20:00 | |
| *** silor has quit IRC | 20:04 | |
| notmyname | jrichli: oh interesting | 20:15 |
| notmyname | jrichli: similar thoughts in https://youtu.be/voXVTjwnn-U?t=31m (which I first heard of via clayg) | 20:23 |
| timburke | notmyname: i was thinking of that very talk while looking at the tests for PivotRange (which defines __lt__ and __gt__)... | 20:26 |
| notmyname | there it is! starting close to minute 39 in that video | 20:27 |
| notmyname | https://review.openstack.org/#/c/342830/ | 20:28 |
| patchbot | patch 342830 - swift - Use more specific asserts in test/unit/obj tests | 20:28 |
| *** Guest6666 has quit IRC | 20:28 | |
| *** Guest6666 has joined #openstack-swift | 20:29 | |
| jrichli | notmyname: good talk. i hadn't seen that before | 20:30 |
| notmyname | jrichli: oh yeah. definitely worth watching the whole thing | 20:30 |
| mattoliverau | Morning | 20:33 |
| notmyname | hello mattoliverau | 20:34 |
| jrichli | mattoliverau: good morning. I was wondering, about how much time / effort does it require from you to support the swift driver in glance-store? | 20:36 |
| jrichli | I have been asked to take a wag on what type of time / effort it would take for somebody to volunteer to bring back and maintain the S3 driver | 20:37 |
| timburke | jrichli: fwiw, it's not real high-traffic. i've been watching new patchsets/new changes for it via https://review.openstack.org/#/settings/projects with a filter for file:^glance_store/_drivers/swift/.* | 20:39 |
| jrichli | thanks, timburke | 20:39 |
| *** m_kazuhiro has joined #openstack-swift | 20:45 | |
| mattoliverau | Yeah, it hasn't been that crazy, there was a week or so last year when I was dragged into a embargo glace bug, which was a little time consuming, but usually I get poked to check from a swift pov. | 20:47 |
| mattoliverau | Fairly rarely in fact | 20:47 |
| mattoliverau | Tho maybe I could personally do better | 20:47 |
| timburke | https://review.openstack.org/#/c/120866/ still seems like a pretty nice idea... | 20:48 |
| patchbot | patch 120866 - glance_store - Buffered reader: Upload recovery for swift store | 20:48 |
| jrichli | mattoliverau: kudos to you for volunteering to be the maintainer! I am sure you are doing fine. thanks for the input. | 20:49 |
| jrichli | timburke: oh, thanks for sharing that. In my research today, I came across the spec and wondered about it. I see the patch points to a blueprint. so they have blueprints AND specs ... wow. | 20:53 |
| kota_ | morning | 20:54 |
| mattoliverau | Its not maintainer it's an liason, so someone for them to blame :p | 20:54 |
| jrichli | mattoliverau: lol, i see :-) | 20:55 |
| kota_ | acoles: nice, thanks for updating! | 20:55 |
| jrichli | kota_ o/ | 20:55 |
| kota_ | jrichli: o/ | 20:56 |
| *** dmorita has joined #openstack-swift | 20:57 | |
| *** JimCheung has quit IRC | 20:57 | |
| notmyname | swift team meeting in 2 minutes in #openstack-meeting | 20:58 |
| *** JimCheung has joined #openstack-swift | 20:58 | |
| *** openstackgerrit has joined #openstack-swift | 21:00 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Clean up some doc formatting https://review.openstack.org/451567 | 21:00 |
| *** NM has quit IRC | 21:29 | |
| *** dmorita has quit IRC | 21:43 | |
| acoles | tdasilva: I'm interested in your DLO versions thoughts | 21:43 |
| tdasilva | timburke: I was just remembering that updates to metadata does not cause a new version of an object | 21:43 |
| tdasilva | acoles, timburke ^ is that right? | 21:44 |
| mattoliverau | \o/ earlier breakfast. bbs | 21:44 |
| *** timss has quit IRC | 21:44 | |
| acoles | hmm, can x-object-manifest be updated by a POST? | 21:44 |
| timburke | tdasilva: correct. including with post-as-copy | 21:44 |
| timburke | acoles: yup | 21:45 |
| *** timss has joined #openstack-swift | 21:45 | |
| timburke | still not sure we came down on the right side with that :-/ but i also couldn't figure out how to behave sensibly with pre-existing data | 21:45 |
| tdasilva | timburke: yeah, I think even that could be considered a "bug" on its own | 21:46 |
| tdasilva | what do we treat metadata differently then content data | 21:46 |
| tdasilva | but just wanted to point that out... | 21:46 |
| tdasilva | for some reason I remember that point being made a while back during "pull out versioning to middleware" work | 21:47 |
| tdasilva | s/what do we treat/why do we treat/ | 21:47 |
| timburke | tdasilva: yeah... but if we treat it the same as object data, then a POST to a versioned object behaves like post-as-copy even under fast-post :-( | 21:48 |
| timburke | in addition to increasing your storage usage | 21:48 |
| tdasilva | timburke: well...i'm thinking just from a user PV, getting out of the engineering mindset... | 21:49 |
| tdasilva | If I'm a user and a write new content, I expect the old copy to be versioned | 21:49 |
| tdasilva | I guess one could argue that the expectation for metadata should be the same | 21:50 |
| tdasilva | if for whatever reason my use case made a lot of use of custom metadata | 21:50 |
| timburke | fair enough. and maybe you're content to accept the worsened performance/usage as a result... | 21:50 |
| rledisez | timburke: I tested https://bugs.launchpad.net/swift/+bug/1675500 on a 3 replicas / 6 devices cluster. can’t reproduce. after first run of container-replicator new replicas of impacted partitions were created on their first handoff devices. following runs of container-replicator didn’t create new replicas | 21:50 |
| openstack | Launchpad bug 1675500 in OpenStack Object Storage (swift) "Container/account disk drive fault results replication on all rest drives" [High,In progress] - Assigned to Pavel Kvasnička (pavel-kvasnicka) | 21:50 |
| rledisez | i’m gonna put that in the bugreport also | 21:51 |
| timburke | rledisez: interesting... good data point. now we just need someone to try the 4 replica case :-) | 21:51 |
| timburke | tdasilva: i think there may be another bug (or at any rate, unexpected behavior) with expiring objects in versioned containers... | 21:53 |
| tdasilva | timburke: oh, is that not being copied over? | 21:53 |
| timburke | tdasilva: i don't think so... maybe it depends on how your object-expirer is configured? | 21:54 |
| tdasilva | timburke: ok, i will try to play with that tomorrow | 21:54 |
| timburke | thanks | 21:54 |
| tdasilva | timburke: I'm going to run too, have a great vacation! | 21:55 |
| acoles | tdasilva: I see your logic wrt the use case i.e. why is metadata not versioned like data? But metadata updates can't be undone/deleted, so there's no way to unwind only a metadata update. | 21:55 |
| acoles | good night | 21:56 |
| rledisez | timburke: if you’re thinking of more tests just tell me. i’m going offline now but i will try tomorrow. i’ll try the 4-replicas tomorrow and also the 2-replicas to see if I can at least reproduce the original bug | 21:56 |
| timburke | rledisez: thanks! | 21:57 |
| rledisez | good night | 21:58 |
| timburke | i think that should establish whether it's a quorum-related bug as notmyname suggested -- in which case we've got a better handle on what configurations are affected | 21:59 |
| *** m_kazuhiro has quit IRC | 22:04 | |
| *** jamielennox is now known as jamielennox|away | 22:09 | |
| *** jamielennox|away is now known as jamielennox | 22:12 | |
| timburke | hrm... i'm seeing intermittent (~0.5-1.5% chance) failures on test/unit/common/test_utils.py:TestUnlinkOlder.test_unlink_older_than_now | 22:14 |
| *** sams-gleb has quit IRC | 22:17 | |
| jrichli | timburke: I have seen that a long time. i am pretty sure i wrote a bug on it at some point - ill poke around for that | 22:19 |
| jrichli | https://bugs.launchpad.net/swift/+bug/1579578 | 22:21 |
| openstack | Launchpad bug 1579578 in OpenStack Object Storage (swift) "test_unlink_* unittests fail intermittently" [Undecided,New] | 22:21 |
| *** mahatic has quit IRC | 22:22 | |
| *** JimCheung has quit IRC | 22:23 | |
| *** mahatic has joined #openstack-swift | 22:23 | |
| jrichli | there was one time recently I hit one in test_meta_file_sync (test.unit.obj.test_ssync.TestSsyncReplication), but haven't seen it since and have not researched it. | 22:24 |
| timburke | jrichli: thanks! and the referenced patch fixes it for me... did we (clayg?) ever submit that? | 22:24 |
| jrichli | i dont think so. its been on my list ... but never go there :/ | 22:25 |
| jrichli | s/go/got/ | 22:26 |
| jrichli | timburke: if you are busy with other things, I can do that in the next couple days | 22:27 |
| *** JimCheung has joined #openstack-swift | 22:29 | |
| jrichli | well, let me know. hubby is staring me down cause we gotta go eat now :-) | 22:30 |
| timburke | jrichli: go eat! and don't worry about it :-) | 22:30 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Fix intermittent test_unlink_* failures https://review.openstack.org/451593 | 22:31 |
| *** JimCheung has quit IRC | 22:33 | |
| *** JimCheung has joined #openstack-swift | 22:34 | |
| *** vint_bra has quit IRC | 23:11 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: fixup! Add container sharding to Swift containers https://review.openstack.org/450491 | 23:16 |
| *** klamath has quit IRC | 23:31 | |
| mattoliverau | timburke: dude, bisect simplified that function nicely! great job! | 23:39 |
| *** chsc has quit IRC | 23:39 | |
| timburke | mattoliverau: i think it still behaves properly? probably want some more tests. oooh! be sure to include some holes! like, [a..f, l..r, r..z] and go looking for a spot for h | 23:41 |
| mattoliverau | timburke: great idea | 23:41 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!