| harlowja | ok, lets see where that goes | 00:01 |
|---|---|---|
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 00:07 |
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 00:10 |
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 00:11 |
| *** dims__ has quit IRC | 00:15 | |
| *** dims__ has joined #openstack-oslo | 00:15 | |
| *** dims__ has quit IRC | 00:18 | |
| *** dims__ has joined #openstack-oslo | 00:18 | |
| *** exploreshaifali has quit IRC | 00:21 | |
| *** david-lyle is now known as david-lyle_afk | 00:23 | |
| *** dims__ has quit IRC | 00:45 | |
| *** dims__ has joined #openstack-oslo | 00:46 | |
| *** andreykurilin__ has quit IRC | 00:47 | |
| *** dims__ has quit IRC | 00:50 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 01:07 |
| *** tsekiyam_ has joined #openstack-oslo | 01:15 | |
| *** tsekiyama has quit IRC | 01:18 | |
| *** tsekiyam_ has quit IRC | 01:19 | |
| *** mtanino has quit IRC | 01:26 | |
| *** dims__ has joined #openstack-oslo | 01:27 | |
| *** _amrith_ is now known as amrith | 01:38 | |
| *** boris-42 has quit IRC | 02:03 | |
| *** zzzeek has quit IRC | 02:08 | |
| *** oomichi has joined #openstack-oslo | 02:46 | |
| *** dims__ has quit IRC | 02:49 | |
| *** dims__ has joined #openstack-oslo | 02:50 | |
| *** dims__ has quit IRC | 02:54 | |
| *** sabari is now known as zz_sabari | 03:10 | |
| *** jeckersb is now known as jeckersb_gone | 03:29 | |
| *** harlowja is now known as harlowja_away | 03:36 | |
| *** zz_sabari is now known as sabari | 03:39 | |
| *** sabari is now known as zz_sabari | 03:47 | |
| *** zz_sabari is now known as sabari | 03:49 | |
| *** oomichi has quit IRC | 04:12 | |
| *** sabari is now known as zz_sabari | 04:16 | |
| *** zz_sabari is now known as sabari | 04:17 | |
| *** stevemar has joined #openstack-oslo | 04:30 | |
| *** lbragstad has quit IRC | 04:32 | |
| *** lbragstad has joined #openstack-oslo | 04:33 | |
| *** zzzeek has joined #openstack-oslo | 04:33 | |
| *** zzzeek has quit IRC | 04:33 | |
| *** ujjain has quit IRC | 04:34 | |
| *** ujjain has joined #openstack-oslo | 04:35 | |
| *** sabari is now known as zz_sabari | 04:44 | |
| *** zz_sabari is now known as sabari | 05:00 | |
| *** sabari is now known as zz_sabari | 05:32 | |
| *** zz_sabari is now known as sabari | 05:37 | |
| *** rpodolyaka2 has joined #openstack-oslo | 05:38 | |
| *** stevemar has quit IRC | 05:43 | |
| *** sgotliv has joined #openstack-oslo | 06:08 | |
| *** harlowja_at_home has joined #openstack-oslo | 06:12 | |
| *** sgotliv has quit IRC | 06:13 | |
| *** vigneshvar has joined #openstack-oslo | 06:39 | |
| *** harlowja_at_home has quit IRC | 06:46 | |
| openstackgerrit | Merged openstack/taskflow: Avoid holding the lock while scanning for existing jobs https://review.openstack.org/141169 | 06:53 |
| *** sgotliv has joined #openstack-oslo | 06:57 | |
| *** sgotliv has quit IRC | 07:04 | |
| *** oomichi has joined #openstack-oslo | 07:11 | |
| *** oomichi has quit IRC | 07:21 | |
| *** SridharGaddam has joined #openstack-oslo | 07:22 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: safe_log Sanitize Passwords in List of Dicts https://review.openstack.org/123759 | 07:31 |
| *** lifeless has quit IRC | 07:36 | |
| *** lifeless has joined #openstack-oslo | 07:43 | |
| *** noelbk has joined #openstack-oslo | 08:00 | |
| *** jamespage_ has joined #openstack-oslo | 08:02 | |
| *** jamespage_ has quit IRC | 08:04 | |
| *** dtantsur|afk is now known as dtantsur | 08:05 | |
| *** sabari is now known as zz_sabari | 08:06 | |
| *** xek has joined #openstack-oslo | 08:08 | |
| *** jaypipes_ has joined #openstack-oslo | 08:10 | |
| *** zz_sabari is now known as sabari | 08:11 | |
| *** jaypipes has quit IRC | 08:14 | |
| *** rpodolyaka2 has quit IRC | 08:16 | |
| *** e0ne has joined #openstack-oslo | 08:28 | |
| *** ajo has joined #openstack-oslo | 08:32 | |
| *** stannie has joined #openstack-oslo | 08:42 | |
| *** ajo has quit IRC | 08:45 | |
| *** ajo has joined #openstack-oslo | 08:47 | |
| *** dtantsur is now known as dtantsur|bbl | 08:59 | |
| *** stannie has quit IRC | 09:00 | |
| *** sabari is now known as zz_sabari | 09:05 | |
| *** raildo_ has joined #openstack-oslo | 09:14 | |
| *** e0ne has quit IRC | 09:28 | |
| *** raildo_ has quit IRC | 09:32 | |
| *** ihrachyshka has joined #openstack-oslo | 09:32 | |
| *** ihrachyshka_ has joined #openstack-oslo | 09:34 | |
| *** pblaho has joined #openstack-oslo | 09:36 | |
| *** ihrachyshka has quit IRC | 09:37 | |
| openstackgerrit | Julien Danjou proposed openstack/tooz: Add create/join/leave group support in IPC driver https://review.openstack.org/141314 | 09:56 |
| *** alexpilotti has joined #openstack-oslo | 10:08 | |
| *** sgotliv has joined #openstack-oslo | 10:08 | |
| *** _gryf has quit IRC | 10:15 | |
| *** _gryf has joined #openstack-oslo | 10:15 | |
| *** e0ne has joined #openstack-oslo | 10:25 | |
| *** andreykurilin_ has joined #openstack-oslo | 10:36 | |
| *** sgotliv has quit IRC | 10:40 | |
| *** ihrachyshka_ has quit IRC | 10:43 | |
| openstackgerrit | Roman Podoliaka proposed openstack/oslo.db: Ensure DBConnectionError is raised on failed revalidate https://review.openstack.org/139738 | 10:50 |
| openstackgerrit | Roman Podoliaka proposed openstack/oslo.db: Repair string-based disconnect filters for MySQL, DB2 https://review.openstack.org/139733 | 10:50 |
| openstackgerrit | Roman Podoliaka proposed openstack/oslo.db: Ensure PostgreSQL connection errors are wrapped https://review.openstack.org/141323 | 10:50 |
| *** dims__ has joined #openstack-oslo | 10:56 | |
| *** dims__ has quit IRC | 11:00 | |
| *** ihrachyshka_ has joined #openstack-oslo | 11:35 | |
| *** ihrachyshka__ has joined #openstack-oslo | 11:38 | |
| *** ihrachyshka_ has quit IRC | 11:41 | |
| *** dims__ has joined #openstack-oslo | 11:45 | |
| *** dims__ has quit IRC | 11:49 | |
| *** dims__ has joined #openstack-oslo | 11:49 | |
| *** dims__ has quit IRC | 11:54 | |
| *** boris-42 has joined #openstack-oslo | 11:55 | |
| *** dims__ has joined #openstack-oslo | 11:56 | |
| *** dtantsur|bbl is now known as dtantsur | 12:10 | |
| openstackgerrit | Merged openstack/oslotest: Activate pep8 check that _ is imported https://review.openstack.org/141205 | 12:18 |
| *** sgotliv has joined #openstack-oslo | 12:18 | |
| *** ihrachyshka__ is now known as ihrachyshka | 12:23 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/oslo-incubator: Introduce fileutils ensure_dir creation mode parameter https://review.openstack.org/141348 | 12:28 |
| dims__ | jd__: when you get a chance, please bless this oslo.vmware bump | 12:29 |
| dims__ | jd__: https://review.openstack.org/#/c/141099/ | 12:29 |
| ajo | dhellmann ^ | 12:30 |
| ajo | http://review.openstack.org/141348 | 12:30 |
| ajo | I hope it's ok, not sure if I need to fill a bug or blueprint for such a tiny thing, but if I need, I will | 12:30 |
| *** e0ne_ has joined #openstack-oslo | 12:40 | |
| openstackgerrit | Merged openstack/oslo.rootwrap: Activate pep8 check that _ is imported https://review.openstack.org/141202 | 12:41 |
| *** tedross has joined #openstack-oslo | 12:41 | |
| *** e0ne has quit IRC | 12:42 | |
| openstackgerrit | Merged openstack/oslo.messaging: rabbit: add some tests when rpc_backend is set https://review.openstack.org/138974 | 12:43 |
| openstackgerrit | Merged openstack/oslo.serialization: Activate pep8 check that _ is imported https://review.openstack.org/141203 | 12:44 |
| openstackgerrit | Merged openstack/oslo.db: Upgrade exc_filters for 'engine' argument and connect behavior https://review.openstack.org/139725 | 12:46 |
| *** vigneshvar has quit IRC | 12:47 | |
| *** dims__ has quit IRC | 12:48 | |
| openstackgerrit | Merged openstack/oslo.db: Repair string-based disconnect filters for MySQL, DB2 https://review.openstack.org/139733 | 12:48 |
| *** dims__ has joined #openstack-oslo | 12:48 | |
| ajo | ihrachyshka, there is no constant for 0x777 | 12:49 |
| ajo | I can construct it btw... | 12:49 |
| ajo | but python itself uses 0o777 | 12:50 |
| ajo | (octal, sorry) | 12:50 |
| *** dims__ has quit IRC | 12:52 | |
| *** dims__ has joined #openstack-oslo | 13:01 | |
| openstackgerrit | Merged openstack/oslo.concurrency: Port processutils to Python 3 https://review.openstack.org/141206 | 13:16 |
| openstackgerrit | Merged openstack/oslo.concurrency: Drop requirements-py3.txt https://review.openstack.org/140980 | 13:17 |
| openstackgerrit | Julien Danjou proposed openstack/tooz: Add create/join/leave group support in IPC driver https://review.openstack.org/141314 | 13:22 |
| *** dims__ has quit IRC | 13:26 | |
| *** jecarey has joined #openstack-oslo | 13:26 | |
| *** vigneshvar has joined #openstack-oslo | 13:27 | |
| *** dims__ has joined #openstack-oslo | 13:27 | |
| *** dims__ has quit IRC | 13:31 | |
| ihrachyshka | ajo: well, yes, you can construct it | 13:42 |
| *** jecarey has quit IRC | 13:46 | |
| *** amrith is now known as _amrith_ | 13:50 | |
| *** jeckersb_gone is now known as jeckersb | 14:00 | |
| *** dims__ has joined #openstack-oslo | 14:02 | |
| *** k4n0 has quit IRC | 14:17 | |
| dims__ | dhellmann: do we have a process for moving attic? | 14:26 |
| dims__ | dhellmann: (oslo.version) | 14:26 |
| *** noelbk has quit IRC | 14:31 | |
| *** jecarey has joined #openstack-oslo | 14:39 | |
| *** jeckersb is now known as jeckersb_gone | 14:45 | |
| openstackgerrit | Merged openstack/oslo.log: Add pbr to installation requirements https://review.openstack.org/130785 | 14:48 |
| openstackgerrit | Merged openstack/oslo.messaging: Updated from global requirements https://review.openstack.org/140950 | 14:48 |
| *** _amrith_ is now known as amrith | 14:49 | |
| *** xek has quit IRC | 14:51 | |
| *** jeckersb_gone is now known as jeckersb | 14:53 | |
| *** vigneshvar has quit IRC | 14:53 | |
| *** andymaier has joined #openstack-oslo | 15:01 | |
| andymaier | Hi. We did not find a tag in the oslo.log git repo that is specific to the Juno release. | 15:03 |
| andymaier | oslo.log is one of five repos that are installed by devstack. | 15:03 |
| andymaier | Question: Should such a tag be added to the oslo.log repo? | 15:03 |
| andymaier | correction: oslo.log is one of five repos that are installed by devstack that does not have such a tag. | 15:04 |
| dhellmann | andymaier: we haven't made an official release of oslo.log yet, so nothing is using it | 15:04 |
| andymaier | devstack offers a variable to set the branch for oslo.log. Does that mean that devstack went ahead of things there? | 15:05 |
| dhellmann | andymaier: our plan was to release it last cycle, but we hit a snag, so it was unfinished | 15:05 |
| dhellmann | andymaier: if you change devstack w.r.t. oslo.log, please limit the changes to the stable/juno branch and keep master as it is because we're working on it again this cycle | 15:06 |
| andymaier | ok. thanks for the info, doug. | 15:07 |
| *** mtanino has joined #openstack-oslo | 15:18 | |
| *** eezhova has quit IRC | 15:21 | |
| *** andymaier has left #openstack-oslo | 15:25 | |
| *** eezhova has joined #openstack-oslo | 15:26 | |
| openstackgerrit | Merged openstack/oslo.log: Updated from global requirements https://review.openstack.org/140949 | 15:29 |
| *** kgiusti has joined #openstack-oslo | 15:33 | |
| openstackgerrit | Merged openstack/oslo.concurrency: Activate pep8 check that _ is imported https://review.openstack.org/141193 | 15:39 |
| *** mriedem has joined #openstack-oslo | 15:39 | |
| dhellmann | dims__: there were some notes on the attic stuff in one of the etherpads from the summit, just a sec | 15:39 |
| dhellmann | dims__: bottom of https://etherpad.openstack.org/p/kilo-oslo-library-proposals | 15:40 |
| *** e0ne_ is now known as e0ne | 15:41 | |
| dims__ | dhellmann: those make sense, we probably need to request infra for a rename before those | 15:43 |
| *** jaypipes_ is now known as jaypipes | 15:44 | |
| dhellmann | dims__: yeah, I think they want the patch up for review first but you'd have to ask fungi | 15:44 |
| dims__ | dhellmann: ack thanks! | 15:46 |
| *** andymaier has joined #openstack-oslo | 15:53 | |
| *** andymaier has quit IRC | 15:55 | |
| *** andymaier has joined #openstack-oslo | 15:56 | |
| *** gordc has joined #openstack-oslo | 15:57 | |
| *** SridharGaddam has quit IRC | 16:08 | |
| *** prad has joined #openstack-oslo | 16:08 | |
| *** andymaier has quit IRC | 16:14 | |
| *** tsekiyama has joined #openstack-oslo | 16:16 | |
| *** noelbk has joined #openstack-oslo | 16:18 | |
| *** e0ne_ has joined #openstack-oslo | 16:21 | |
| *** e0ne has quit IRC | 16:24 | |
| *** boris-42 has quit IRC | 16:32 | |
| *** dims__ has quit IRC | 16:33 | |
| *** dims__ has joined #openstack-oslo | 16:34 | |
| *** dims__ has quit IRC | 16:38 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.rootwrap: Move files out of the namespace package https://review.openstack.org/141147 | 16:43 |
| *** kbyrne has quit IRC | 16:44 | |
| *** dims__ has joined #openstack-oslo | 16:44 | |
| *** kbyrne has joined #openstack-oslo | 16:46 | |
| *** david-lyle_afk is now known as david-lyle | 16:49 | |
| *** jaosorior has joined #openstack-oslo | 16:53 | |
| *** kbyrne has quit IRC | 16:53 | |
| *** kbyrne has joined #openstack-oslo | 16:57 | |
| *** haypo has left #openstack-oslo | 16:57 | |
| *** andreykurilin_ has quit IRC | 16:58 | |
| *** andreykurilin_ has joined #openstack-oslo | 16:58 | |
| *** miqui_ has quit IRC | 17:04 | |
| *** miqui_ has joined #openstack-oslo | 17:04 | |
| *** mriedem has quit IRC | 17:11 | |
| *** dtantsur is now known as dtantsur|afk | 17:15 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.rootwrap: Move files out of the namespace package https://review.openstack.org/141147 | 17:21 |
| *** jogo is now known as flashgordon | 17:24 | |
| *** jaypipes is now known as leakypipes | 17:25 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.rootwrap: Move files out of the namespace package https://review.openstack.org/141147 | 17:31 |
| *** openstackgerrit has quit IRC | 17:34 | |
| *** openstackgerrit has joined #openstack-oslo | 17:34 | |
| openstackgerrit | Ihar Hrachyshka proposed openstack/oslo.concurrency: Don't use ConfigFilter for lockutils https://review.openstack.org/141436 | 17:35 |
| *** andreykurilin_ has quit IRC | 17:38 | |
| *** andreykurilin_ has joined #openstack-oslo | 17:39 | |
| *** zzzeek has joined #openstack-oslo | 17:41 | |
| *** e0ne_ has quit IRC | 17:48 | |
| *** openstackgerrit has quit IRC | 17:49 | |
| *** openstackgerrit has joined #openstack-oslo | 17:49 | |
| *** ihrachyshka has quit IRC | 17:49 | |
| *** andreykurilin_ has quit IRC | 17:54 | |
| *** dims__ is now known as dimsum__ | 17:57 | |
| *** pblaho has quit IRC | 18:16 | |
| *** harlowja_away is now known as harlowja | 18:17 | |
| *** noelbk has quit IRC | 18:21 | |
| *** leakypipes is now known as leaky-tennis | 18:23 | |
| *** russellb is now known as rustlebee | 18:29 | |
| *** zz_sabari is now known as sabari | 18:29 | |
| *** amrith is now known as _amrith_ | 18:41 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 18:53 |
| *** sgotliv has quit IRC | 18:57 | |
| *** boris-42 has joined #openstack-oslo | 19:03 | |
| *** achanda has joined #openstack-oslo | 19:04 | |
| achanda | I've been trying to 'pip install oslo.log' and it says there is no distribution for oslo.log | 19:05 |
| dhellmann | achanda: that is correct, we have not released oslo.log yet | 19:05 |
| achanda | dhellman: thanks! | 19:05 |
| *** ruhe has joined #openstack-oslo | 19:05 | |
| achanda | I guess I rushed into it :) | 19:05 |
| * dhellmann wonders why so many people have been asking that question lately :-) | 19:05 | |
| achanda | maybe because it shows up in pip search! :) | 19:06 |
| achanda | in pypi | 19:06 |
| dhellmann | yes, we registered the project to reserve the name | 19:07 |
| achanda | ok, cool | 19:07 |
| *** vigneshvar has joined #openstack-oslo | 19:10 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.concurrency: Report import warnings where the import occurs https://review.openstack.org/141125 | 19:11 |
| dhellmann | viktors|afk, zzzeek, rpodolyaka1: I'm going to have a lot of trouble rebasing https://review.openstack.org/138787 if we keep landing changes in oslo.db. Can we prioritize reviewing that one when I have a new patch ready? | 19:13 |
| zzzeek | dhellmann: aboslutely say the word | 19:20 |
| dhellmann | zzzeek: I hate to block everyone else, but I have to recreate that patch by hand every time someone changes a file so I don't lose their change :-/ | 19:20 |
| zzzeek | dhellmann: im sure , so, when is that finally going to go in ? | 19:20 |
| dhellmann | git doesn't seem to be smart enough to figure this out | 19:20 |
| dhellmann | well, I'm going to start working on fixing the current patch now, so then it's up to the reviewers as to when it lands | 19:21 |
| zzzeek | dhellmann: oh wow there was a pymysql gerrit, heh | 19:23 |
| zzzeek | dhellmann: IMO we shoudl try to get mysql-connector to work for us | 19:23 |
| dhellmann | yeah, we have a spec for that, too, and I think we approved the spec | 19:23 |
| dhellmann | I'll let you and ihar work that out | 19:24 |
| harlowja | zzzeek sooo who wants to fork mysql-connector ;) | 19:24 |
| harlowja | i don't have larry's phone nmber | 19:24 |
| harlowja | maybe if i got that | 19:24 |
| harlowja | i could run over to the oracle buildings and yell like a crazy person in them | 19:24 |
| harlowja | that might help ? | 19:24 |
| dhellmann | zzzeek: oh, here's the spec review: https://review.openstack.org/#/c/125044/ | 19:24 |
| zzzeek | dhellmann: wasnt sure I’d be the only one uncomforatble with “many bug fixes and improvements” | 19:24 |
| dhellmann | yeah, that's not great -- I mean, maybe it's OK, but I don't want to assume it is and I don't have time to do a real review myself | 19:25 |
| harlowja | better than no bug fixes, lol | 19:25 |
| dhellmann | heh | 19:25 |
| harlowja | i know where the oracle tin can buildings are, maybe if i scope out the buildings and do a raid, i can get that published to pypi | 19:26 |
| harlowja | * http://www.bloomberg.com/image/iBvkPglvU38M.jpg | 19:27 |
| harlowja | could be like a stealth mission | 19:27 |
| harlowja | * http://cdn.cultofmac.com/wp-content/uploads/2010/03/mission-impossible-SPLASH.jpg (would be me instead of tom cruise) | 19:29 |
| harlowja | and i'd be uploading the package to pypi instead of doing whatever he's doing there | 19:30 |
| harlowja | lol | 19:30 |
| *** sabari is now known as zz_sabari | 19:35 | |
| *** zz_sabari is now known as sabari | 19:38 | |
| *** _amrith_ is now known as amrith | 19:43 | |
| dimsum__ | harlowja: haha | 19:49 |
| harlowja | :) | 19:50 |
| openstackgerrit | Joshua Harlow proposed openstack/oslo-specs: Add debtcollector adoption https://review.openstack.org/141220 | 19:56 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.db: Move files out of the namespace package https://review.openstack.org/138787 | 19:58 |
| *** e0ne has joined #openstack-oslo | 20:02 | |
| *** alexpilotti has quit IRC | 20:10 | |
| *** dtantsur|afk has quit IRC | 20:11 | |
| *** alexpilotti has joined #openstack-oslo | 20:12 | |
| *** noelbk has joined #openstack-oslo | 20:15 | |
| *** e0ne has quit IRC | 20:15 | |
| *** sabari is now known as zz_sabari | 20:16 | |
| *** alexpilotti has quit IRC | 20:22 | |
| *** jaosorior has quit IRC | 20:23 | |
| *** alexpilotti has joined #openstack-oslo | 20:23 | |
| *** vigneshvar has quit IRC | 20:23 | |
| *** dtantsur has joined #openstack-oslo | 20:24 | |
| *** alexpilotti has quit IRC | 20:27 | |
| *** dteselkin has left #openstack-oslo | 20:33 | |
| *** vigneshvar has joined #openstack-oslo | 20:37 | |
| *** alexpilotti has joined #openstack-oslo | 20:37 | |
| *** achanda has quit IRC | 20:37 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a moved_inheritable_class deprecation helper https://review.openstack.org/140132 | 20:51 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Remove usage of listener base postfix https://review.openstack.org/141490 | 20:51 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Remove usage of listener base postfix https://review.openstack.org/141490 | 20:53 |
| *** zz_sabari is now known as sabari | 20:54 | |
| *** ajo has quit IRC | 20:56 | |
| *** alexpilotti has quit IRC | 20:57 | |
| *** alexpilotti has joined #openstack-oslo | 20:58 | |
| *** alexpilotti has quit IRC | 20:59 | |
| *** alexpilotti has joined #openstack-oslo | 21:01 | |
| *** alexpilotti has quit IRC | 21:01 | |
| *** alexpilotti has joined #openstack-oslo | 21:02 | |
| *** alexpilotti has quit IRC | 21:02 | |
| *** alexpilotti has joined #openstack-oslo | 21:03 | |
| *** prad has quit IRC | 21:04 | |
| *** alexpilotti has quit IRC | 21:05 | |
| *** ajo has joined #openstack-oslo | 21:06 | |
| *** jecarey has quit IRC | 21:07 | |
| *** alexpilotti has joined #openstack-oslo | 21:07 | |
| *** alexpilotti has quit IRC | 21:09 | |
| *** ajo has quit IRC | 21:12 | |
| *** alexpilotti has joined #openstack-oslo | 21:14 | |
| *** ajo has joined #openstack-oslo | 21:15 | |
| *** alexpilotti has quit IRC | 21:15 | |
| *** leaky-tennis is now known as leakypipes | 21:17 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.config: Move files out of the namespace package https://review.openstack.org/139735 | 21:23 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.i18n: Move out of the oslo namespace package https://review.openstack.org/127323 | 21:27 |
| *** boris-42 has quit IRC | 21:33 | |
| *** e0ne has joined #openstack-oslo | 21:33 | |
| dhellmann | sileht, jd__ : same question for you as for the oslo.db guys, can we prioritize my patch to remove the namespace package so I don't have to keep rebasing it by hand? I need to submit a fixed up patch | 21:37 |
| flashgordon | zzzeek: ping | 21:40 |
| zzzeek | flashgordon: hi | 21:40 |
| flashgordon | zzzeek: https://review.openstack.org/#/c/135296/2/specs/kilo/approved/lock-free-quota-management.rst,cm thought you may find that spec interesting | 21:41 |
| flashgordon | note I have only just started reading it, but thought you may have a thought or two | 21:41 |
| zzzeek | flashgordon: have seen this already being implemented in lots of places | 21:42 |
| zzzeek | flashgordon: its the basic optimistic approach and is what we shoudl be doing | 21:42 |
| flashgordon | zzzeek: glad to hear it, a +1 from you would be nice | 21:44 |
| flashgordon | it would help me feel comfortable +2ing it | 21:44 |
| zzzeek | i approve everything jay pipes does :) | 21:45 |
| leakypipes | lol | 21:45 |
| flashgordon | leakypipes: oh no more tennis? | 21:45 |
| *** tedross has quit IRC | 21:45 | |
| leakypipes | flashgordon: no, we got beaten badly by two people 30 years older than us. It was mucho embarrassing. | 21:46 |
| leakypipes | but hey, whatever makes the wife happy, right? :) | 21:46 |
| flashgordon | leakypipes: haha yup | 21:47 |
| *** jeckersb is now known as jeckersb_gone | 21:47 | |
| flashgordon | leakypipes: naive question | 21:47 |
| flashgordon | the write-intent lock, is that on the whole table? And will the compare and swap method hit the backoff less often then the deadlock is raised now? | 21:49 |
| leakypipes | flashgordon: write-intent locks are on single rows and what are called "gap locks" in InnoDB | 21:50 |
| leakypipes | flashgordon: they are definitely not on the whole table, no. | 21:51 |
| flashgordon | leakypipes: so the part I am a little confused on, is how will this make for better concurrency. I get the fewer db deadlock based retries | 21:53 |
| flashgordon | leakypipes: I am asking to learn, not because I doubt it does that | 21:54 |
| *** kgiusti has quit IRC | 21:54 | |
| *** boris-42 has joined #openstack-oslo | 21:55 | |
| flashgordon | oh and there is a section with TODO in it so | 21:57 |
| sdague | flashgordon: so the issue is select for update gets a write lock | 21:57 |
| sdague | because that's what you were intending to do | 21:57 |
| leakypipes | flashgordon: no worries. so right now, there is a set of write-intent locks held for all quota usage records for a project during the reserve() and commit() methods of the QuotaDbDriver. These methods take a lock on all the rows in the table that match the project_id, loop through each record, adjust the usage amount, and then calls an UPDATE statement via the ORM in SQLalchemy. Again, all those loops are run while th | 21:57 |
| leakypipes | e locks on the rows are held. This means that users in the same tenant cannot concurrently take actions that impact quotas. | 21:57 |
| sdague | vs. just a select which could run in parallel | 21:58 |
| leakypipes | flashgordon: because user B will block waiting for user A's thread to release its locks on the usage records. | 21:58 |
| leakypipes | flashgordon: with the compare and swap technique, both threads can interleave their requests concurrently, and the window of time that a lock is held is reduced to its smallest possible period. | 21:59 |
| flashgordon | leakypipes: ahhh makes sense | 21:59 |
| flashgordon | leakypipes: so in theory randomizing the order of the resources being updated would help? | 22:00 |
| leakypipes | flashgordon: now, the compare and swap technique also has the benefit of "fixing" the SELECT FOR UPDATE problem that Galera has (Galera throws a certification timeout error when it detects two threads on two writer nodes have updated the same rows, and a SELECT FOR UPDATE in Galera only locks rows on the node that received the SELECT query. | 22:00 |
| flashgordon | at least in a pathological case | 22:00 |
| sdague | leakypipes: any major concerns on races on update? or is that going to be healed in some offline manner? | 22:00 |
| leakypipes | sdague: not really, no. say thread A and thread B are simultaneously launching an instance. | 22:01 |
| flashgordon | sdague leakypipes: now that we are consuming resources one at a time, I assume roll back will take care of things. | 22:01 |
| flashgordon | say two instances being spawned at once, and there is enough memory quota for both but not enough CPU quota | 22:02 |
| leakypipes | sdague: thread A gets to the quota reserve() part and reads the usage records. thread B does the same. thread A then attempts to update the usage record for a particular resource and project ID. it does so by issuing a SQL statement like so: | 22:02 |
| flashgordon | both consume memory quota, one gets CPU, and the other undoes the memory quota usage? | 22:02 |
| flashgordon | and fails | 22:02 |
| leakypipes | UPDATE usages SET amount = <NEW_AMOUNT> WHERE project_id = <PROJECT> and resource_id = <REOURCE> and amount = <OLD_AMOUNT> | 22:03 |
| sdague | leakypipes: right | 22:03 |
| sdague | but if we have two attempts there, the 2nd one won't increment the allocation | 22:03 |
| leakypipes | sdague: if thread B happens to have been able to update that row before thread A, thread A will simply re-read the usage row and try again (after checking to see if the quota has been exceeded) | 22:03 |
| leakypipes | sdague: the only time a lock is held is for the very brief time that the UPDATE statement is being executed on the MySQL server. | 22:04 |
| sdague | oh, you'll check the row count on update, and if it's 0 retry? | 22:04 |
| sdague | yep... that's even there in the spec | 22:04 |
| sdague | I jumped past it | 22:05 |
| sdague | very good | 22:05 |
| sdague | flashgordon: you should +2 that | 22:05 |
| flashgordon | leakypipes: what about the case where we hit quota limit, as I attempted to describe above | 22:05 |
| flashgordon | sdague: I want to, waiting for leakypipes to remove the TODO from the spec | 22:05 |
| leakypipes | sdague: yes, exactly correct. | 22:05 |
| flashgordon | http://docs-draft.openstack.org/96/135296/2/check/gate-nova-specs-docs/f4b24fe/doc/build/html/specs/kilo/approved/lock-free-quota-management.html#work-items | 22:06 |
| leakypipes | flashgordon: then we raise Overlimit as per usual. | 22:06 |
| leakypipes | flashgordon: oh, crap... what TODO is left on that spec? | 22:06 |
| flashgordon | leakypipes: and for the request that failed, we undo the quota usage? | 22:06 |
| flashgordon | even though it is partia | 22:06 |
| flashgordon | partial | 22:06 |
| sdague | flashgordon: it's like any Overlimit fail, you have to back track it all | 22:07 |
| sdague | which we do terribly today | 22:07 |
| sdague | which is a whole other can of worms | 22:07 |
| leakypipes | flashgordon: correct. | 22:07 |
| sdague | but this at least solves one problem | 22:07 |
| flashgordon | sdague: right, just making sure this doesn't make that worse. | 22:07 |
| sdague | no | 22:07 |
| flashgordon | leakypipes: once you make that TODO go away, +2 from me | 22:08 |
| flashgordon | wow the quota logic is funny | 22:11 |
| flashgordon | it looks like we check for overquota in python not in sql | 22:11 |
| flashgordon | anyway this is a massive step in the right direction so huzzah | 22:13 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo.messaging: Updated from global requirements https://review.openstack.org/141509 | 22:20 |
| *** e0ne is now known as e0ne_ | 22:25 | |
| *** e0ne_ has quit IRC | 22:31 | |
| *** stevemar has joined #openstack-oslo | 22:36 | |
| *** gordc has quit IRC | 22:43 | |
| *** dimsum__ has quit IRC | 22:51 | |
| *** dimsum__ has joined #openstack-oslo | 22:52 | |
| *** dimsum__ has quit IRC | 22:57 | |
| *** vigneshvar has quit IRC | 23:00 | |
| *** tsekiyam_ has joined #openstack-oslo | 23:04 | |
| *** andreykurilin_ has joined #openstack-oslo | 23:07 | |
| *** tsekiyama has quit IRC | 23:08 | |
| *** tsekiyam_ has quit IRC | 23:09 | |
| *** ftcjeff has joined #openstack-oslo | 23:12 | |
| *** ftcjeff has quit IRC | 23:13 | |
| *** ajo_ has joined #openstack-oslo | 23:15 | |
| *** ajo has quit IRC | 23:15 | |
| openstackgerrit | Miguel Angel Ajo proposed openstack/oslo-incubator: Introduce fileutils ensure_dir creation mode parameter https://review.openstack.org/141348 | 23:22 |
| ajo_ | repeating myself in a second... | 23:23 |
| openstackgerrit | Miguel Angel Ajo proposed openstack/oslo-incubator: Introduce fileutils ensure_dir creation mode parameter https://review.openstack.org/141348 | 23:23 |
| *** andreykurilin_ has quit IRC | 23:24 | |
| *** andreykurilin_ has joined #openstack-oslo | 23:25 | |
| *** ajo_ has quit IRC | 23:44 | |
| *** sabari is now known as zz_sabari | 23:46 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
