| *** dims__ has quit IRC | 00:20 | |
| *** tsekiyam_ has joined #openstack-oslo | 00:29 | |
| *** jaosorior has quit IRC | 00:32 | |
| *** tsekiyama has quit IRC | 00:32 | |
| *** tsekiyam_ has quit IRC | 00:33 | |
| *** dims__ has joined #openstack-oslo | 00:44 | |
| *** pcm_ has joined #openstack-oslo | 00:47 | |
| *** zzzeek has quit IRC | 00:47 | |
| *** pcm_ has quit IRC | 00:48 | |
| *** jaosorior has joined #openstack-oslo | 00:48 | |
| *** pcm_ has joined #openstack-oslo | 00:48 | |
| *** pcm_ has quit IRC | 01:16 | |
| gus | zzzeek, i159: note that OS_TEST_DBAPI_ADMIN_CONNECTION doesn't exist in oslo.db (only oslo-incubator). | 01:30 |
|---|---|---|
| gus | I'm still waiting on reviews for https://review.openstack.org/#/c/104447/ (hint) | 01:30 |
| gus | And it depends on https://review.openstack.org/#/c/104428/ - also awaiting reviews. | 01:31 |
| gus | And viktors|afk: I replied to your comment on https://review.openstack.org/#/c/104425/ - let me know if you need more information before you're willing to approve that change | 01:32 |
| gus | A question for the room: We have some deadlock/concurrency issues with eventlet vs mysqldb driver. I have a test that demonstrates that, but it imports/uses eventlet explicitly. | 01:34 |
| gus | Victor quite reasonably pointed out that requiring eventlet in an olso.db test is frowned upon. This test identifies an important issue, so I'd like to have it somewhere. Any suggestions for how/where this test could exist? | 01:35 |
| gus | ^^ https://review.openstack.org/#/c/104436/ | 01:35 |
| gus | (note that the deadlock could also occur with other n:m-greenthread-style implementations, like asyncio and futures - but reproducing it reliably will probably require explicit asyncio/futures code too) | 01:39 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the core 10k view omnigraffle diagram https://review.openstack.org/105642 | 01:45 |
| *** zhiyan_ is now known as zhiyan | 01:47 | |
| *** arnaud has joined #openstack-oslo | 01:47 | |
| *** zzzeek has joined #openstack-oslo | 01:55 | |
| *** zhiyan is now known as zhiyan_ | 01:56 | |
| *** zhiyan_ is now known as zhiyan | 01:56 | |
| *** arnaud has quit IRC | 02:25 | |
| *** zzzeek has quit IRC | 02:31 | |
| *** oomichi has joined #openstack-oslo | 02:45 | |
| *** SridharG has joined #openstack-oslo | 02:47 | |
| *** harlowja is now known as harlowja_away | 02:50 | |
| *** arnaud has joined #openstack-oslo | 03:01 | |
| *** praneshp has quit IRC | 03:10 | |
| *** SridharG has quit IRC | 03:13 | |
| *** dstanek_zzz is now known as dstanek | 03:20 | |
| *** arnaud has quit IRC | 03:20 | |
| *** arnaud has joined #openstack-oslo | 03:27 | |
| *** pcm_ has joined #openstack-oslo | 03:40 | |
| *** arnaud has quit IRC | 03:42 | |
| *** arnaud has joined #openstack-oslo | 03:43 | |
| *** dims__ has quit IRC | 03:55 | |
| *** arnaud has quit IRC | 03:59 | |
| *** yamahata has joined #openstack-oslo | 04:03 | |
| *** arnaud has joined #openstack-oslo | 04:05 | |
| *** harlowja_at_home has joined #openstack-oslo | 04:23 | |
| harlowja_at_home | haypo, https://wiki.openstack.org/wiki/TaskFlow#Ten_thousand_foot_view (this might help in understand wtf taskflow is, haha) | 04:30 |
| *** praneshp has joined #openstack-oslo | 04:35 | |
| *** SridharG has joined #openstack-oslo | 04:50 | |
| *** dims__ has joined #openstack-oslo | 04:51 | |
| *** dims__ has quit IRC | 04:56 | |
| *** dims__ has joined #openstack-oslo | 04:57 | |
| *** dims__ has quit IRC | 05:02 | |
| *** ildikov has quit IRC | 05:02 | |
| *** praneshp_ has joined #openstack-oslo | 05:11 | |
| *** praneshp has quit IRC | 05:14 | |
| *** praneshp_ is now known as praneshp | 05:14 | |
| *** harlowja_at_home has quit IRC | 05:17 | |
| *** morganfainberg_Z is now known as morganfainberg | 05:18 | |
| *** ildikov has joined #openstack-oslo | 05:46 | |
| *** dims__ has joined #openstack-oslo | 05:58 | |
| *** nacim has quit IRC | 06:01 | |
| *** dims__ has quit IRC | 06:03 | |
| *** dstanek is now known as dstanek_zzz | 06:08 | |
| *** mkoderer has joined #openstack-oslo | 06:09 | |
| *** morganfainberg is now known as morganfainberg_Z | 06:23 | |
| *** AAzza_afk is now known as AAzza | 06:43 | |
| *** dstanek_zzz is now known as dstanek | 06:43 | |
| *** arnaud has quit IRC | 06:53 | |
| *** dstanek is now known as dstanek_zzz | 06:53 | |
| *** dims__ has joined #openstack-oslo | 06:58 | |
| *** flaper87|afk is now known as flaper87 | 06:59 | |
| *** jaosorior has quit IRC | 07:02 | |
| *** dims__ has quit IRC | 07:03 | |
| openstackgerrit | Abhishek Kekane proposed a change to openstack/oslo-incubator: Add default log level for websocket https://review.openstack.org/101908 | 07:12 |
| *** tkelsey has joined #openstack-oslo | 07:26 | |
| *** mrda is now known as mrda-away | 07:28 | |
| *** pcm_ has quit IRC | 07:33 | |
| openstackgerrit | A change was merged to openstack/oslo.config: Hook IPOpt class into the docs https://review.openstack.org/105538 | 07:33 |
| openstackgerrit | Angus Lees proposed a change to openstack/oslo.db: Don't drop pre-existing database before tests https://review.openstack.org/104430 | 07:53 |
| *** ildikov has quit IRC | 07:55 | |
| *** markmc has joined #openstack-oslo | 07:57 | |
| *** ildikov has joined #openstack-oslo | 07:58 | |
| *** dims__ has joined #openstack-oslo | 07:59 | |
| openstackgerrit | Angus Lees proposed a change to openstack/oslo.db: Add DBDuplicateEntry detection for mysqlconnector driver https://review.openstack.org/104425 | 08:04 |
| *** nacim has joined #openstack-oslo | 08:04 | |
| *** dims__ has quit IRC | 08:04 | |
| *** Alexei_9871 has quit IRC | 08:09 | |
| openstackgerrit | Mark McLoughlin proposed a change to openstack/oslo.config: generator: tweak how MultiStrOpt defaults are handled https://review.openstack.org/105671 | 08:11 |
| zhiyan | viktors|afk: hi | 08:13 |
| *** viktors|afk is now known as viktors | 08:14 | |
| viktors | zhiyan: hi! | 08:14 |
| *** AAzza is now known as AAzza_afk | 08:17 | |
| zhiyan | viktors: may i send you a PM, pls check? | 08:17 |
| viktors | zhiyan: sure | 08:18 |
| zhiyan | viktors: thanks! | 08:18 |
| *** i159 has joined #openstack-oslo | 08:19 | |
| *** dims__ has joined #openstack-oslo | 08:28 | |
| *** pblaho has joined #openstack-oslo | 08:32 | |
| *** dims__ has quit IRC | 08:34 | |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo.db: oslo.db.options module documentation https://review.openstack.org/93436 | 08:45 |
| *** ihrachyshka has joined #openstack-oslo | 08:50 | |
| *** AAzza_afk is now known as AAzza | 08:52 | |
| ihrachyshka | gus: hey. still around? | 08:55 |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo.db: oslo.db.exceptions module documentation https://review.openstack.org/93435 | 09:14 |
| openstackgerrit | Romain Soufflet proposed a change to openstack/oslo-incubator: Fix tests.unit.reports.test_base_report with python3.4 https://review.openstack.org/105680 | 09:21 |
| *** dims__ has joined #openstack-oslo | 09:30 | |
| *** dims__ has quit IRC | 09:35 | |
| viktors | dhellmann: hi! Please ping me, when you'll be around | 09:36 |
| *** yamahata has quit IRC | 09:48 | |
| *** AAzza is now known as AAzza_afk | 09:50 | |
| *** SridharG has quit IRC | 09:51 | |
| *** praneshp has quit IRC | 09:56 | |
| *** ildikov has quit IRC | 09:57 | |
| *** viktors is now known as viktors|afk | 10:01 | |
| *** ildikov has joined #openstack-oslo | 10:15 | |
| *** dims__ has joined #openstack-oslo | 10:31 | |
| *** alexpilotti has joined #openstack-oslo | 10:35 | |
| *** dims__ has quit IRC | 10:36 | |
| *** yamahata has joined #openstack-oslo | 10:47 | |
| *** dims__ has joined #openstack-oslo | 11:10 | |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Changes import orders to pass H305 check https://review.openstack.org/99670 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Fixes whitespaces between imports to pass H307 https://review.openstack.org/99671 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Fixes comments to pass E265 check. https://review.openstack.org/99676 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Uses keyword params for i18n string to pass H703 https://review.openstack.org/99674 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Fixes indentations to pass E128 check. https://review.openstack.org/99675 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Updates one line docstring with dot to pass H402 https://review.openstack.org/99672 | 11:14 |
| openstackgerrit | Petr Blaho proposed a change to openstack/oslo.db: Adds empty line to multilines docs to pass H405 https://review.openstack.org/99673 | 11:14 |
| *** dstanek_zzz is now known as dstanek | 11:22 | |
| *** alexpilotti has quit IRC | 11:26 | |
| *** alexpilotti has joined #openstack-oslo | 11:30 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Fix tests.unit.reports.test_base_report with python3.4 https://review.openstack.org/105680 | 11:32 |
| *** tkelsey has quit IRC | 11:34 | |
| *** alexpilotti has quit IRC | 11:35 | |
| openstackgerrit | Miguel Angel Ajo proposed a change to openstack/oslo-specs: Add service-status-interface spec https://review.openstack.org/97748 | 11:36 |
| *** ujjain has quit IRC | 11:37 | |
| openstackgerrit | Miguel Angel Ajo proposed a change to openstack/oslo-specs: Add service-status-interface spec https://review.openstack.org/97748 | 11:41 |
| ajo | dhellmann, https://review.openstack.org/97748 sorry for the delay updating the service status spec | 11:43 |
| *** pcm_ has joined #openstack-oslo | 11:43 | |
| ajo | I decided to turn it into plugin-based, as it makes more sense, and makes everything cleaner & simpler | 11:43 |
| ajo | I have also extended it to include watchdog support (that at least systemd supports) and readyness notification (upstart supports that too in a different way -based in signals AFAIK-) | 11:44 |
| *** viktors|afk is now known as viktors | 11:53 | |
| *** jecarey has joined #openstack-oslo | 11:57 | |
| *** pcm_ has quit IRC | 11:59 | |
| *** pcm_ has joined #openstack-oslo | 12:00 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.config: Changes imports order to pass H305, enables check https://review.openstack.org/99639 | 12:01 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack-dev/pbr: Switch from pypi.python.org to pypi.openstack.org https://review.openstack.org/105344 | 12:05 |
| *** dstanek is now known as dstanek_zzz | 12:11 | |
| pblaho | dims__: ping | 12:12 |
| pblaho | dims__: thanks for update of https://review.openstack.org/#/c/99639/ | 12:13 |
| *** dstanek_zzz is now known as dstanek | 12:18 | |
| *** jecarey has quit IRC | 12:19 | |
| *** dims__ has quit IRC | 12:32 | |
| *** dims__ has joined #openstack-oslo | 12:32 | |
| *** mkoderer has quit IRC | 12:42 | |
| *** jecarey has joined #openstack-oslo | 12:46 | |
| *** pcm_ has quit IRC | 12:48 | |
| *** ndipanov has joined #openstack-oslo | 12:52 | |
| *** jecarey has quit IRC | 12:58 | |
| jd__ | I'm thinking about moving oslo-incubator/o/c/middleware/notifier.py to oslo.messaging | 13:04 |
| jd__ | any opinion? | 13:04 |
| *** Alexei_987 has quit IRC | 13:05 | |
| openstackgerrit | Radoslav Gerganov proposed a change to openstack/oslo.vmware: Refactor the PBM support https://review.openstack.org/102480 | 13:07 |
| *** joesavak has joined #openstack-oslo | 13:08 | |
| *** oomichi has quit IRC | 13:10 | |
| *** Alexei_987 has joined #openstack-oslo | 13:10 | |
| *** ndipanov has quit IRC | 13:12 | |
| *** ildikov has quit IRC | 13:14 | |
| Alexei_987 | markmc: Hi are you available? | 13:15 |
| *** amotoki has quit IRC | 13:16 | |
| *** mriedem has joined #openstack-oslo | 13:18 | |
| *** pcm_ has joined #openstack-oslo | 13:21 | |
| *** tkelsey has joined #openstack-oslo | 13:24 | |
| viktors | dhellmann: ping | 13:30 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo.messaging: Moved main loop to server class https://review.openstack.org/104983 | 13:35 |
| dhellmann | jd__: that's the plan - https://wiki.openstack.org/wiki/Oslo/GraduationStatus#notifier | 13:35 |
| dhellmann | viktors: hi | 13:35 |
| viktors | dhellmann: hello | 13:36 |
| jd__ | dhellmann: thanks! | 13:36 |
| viktors | dhellmann: recenty I found that I can't make `git push gerrit` :( | 13:37 |
| dhellmann | jd__: I don't know if there's a blueprint, yet, since I didn't anticipate doing that during this cycle. | 13:37 |
| dhellmann | viktors: which repo? | 13:37 |
| viktors | dhellmann: oslo.db | 13:37 |
| dhellmann | viktors: we did some cleanup of acls and I might need to add you back to a group, just a sec | 13:37 |
| jd__ | dhellmann: https://blueprints.launchpad.net/oslo/+spec/graduate-notifier-middleware I guess? | 13:37 |
| viktors | it seems to be, that only folks fron grup oslo-db-ptl can do it | 13:37 |
| viktors | dhellmann: ok, sure | 13:38 |
| dhellmann | viktors: yeah, and I'm not in that group yet, either, so let me ping infra | 13:38 |
| viktors | dhellmann: thanks! | 13:38 |
| *** dims_ has joined #openstack-oslo | 13:39 | |
| *** markmcclain has joined #openstack-oslo | 13:39 | |
| *** dstanek is now known as dstanek_zzz | 13:41 | |
| dhellmann | viktors: you should be ready to go | 13:41 |
| *** haypo_ has joined #openstack-oslo | 13:43 | |
| *** GheRiver1 has joined #openstack-oslo | 13:43 | |
| *** jokke__ has joined #openstack-oslo | 13:44 | |
| viktors | dhellmann: thanks, will check | 13:44 |
| viktors | dhellmann: is works, thanks | 13:46 |
| *** krotscheck_ has joined #openstack-oslo | 13:47 | |
| *** zzzeek has joined #openstack-oslo | 13:47 | |
| dims_ | pblaho, welcome | 13:47 |
| *** tkelsey has quit IRC | 13:48 | |
| *** dims__ has quit IRC | 13:48 | |
| *** rpodolyaka has quit IRC | 13:48 | |
| *** GheRivero has quit IRC | 13:48 | |
| *** jogo has quit IRC | 13:48 | |
| *** eezhova has quit IRC | 13:48 | |
| *** haypo has quit IRC | 13:48 | |
| *** jokke_ has quit IRC | 13:48 | |
| *** krotscheck has quit IRC | 13:48 | |
| *** anteaya has quit IRC | 13:48 | |
| *** krotscheck_ is now known as krotscheck | 13:48 | |
| *** jogo has joined #openstack-oslo | 13:49 | |
| *** yamahata has quit IRC | 13:52 | |
| *** yamahata has joined #openstack-oslo | 13:53 | |
| ihrachyshka | zzzeek: hey! I have a question on oslo.db | 13:54 |
| zzzeek | if it can be quick :) i have to leave in 5 minutes | 13:54 |
| *** eezhova has joined #openstack-oslo | 13:54 | |
| ihrachyshka | zzzeek: when I've tried to switch to mysqlconnector, I got some alembic migration failures. And this is because raise_on_warnings is True for the driver. | 13:54 |
| ihrachyshka | zzzeek: I can fix it by setting the option to False passing it thru connection URI | 13:55 |
| zzzeek | how about we fix the warnings | 13:55 |
| *** anteaya has joined #openstack-oslo | 13:55 | |
| ihrachyshka | zzzeek: but I wonder whether we need to set it to False by default somewhere in oslo.db | 13:55 |
| ihrachyshka | zzzeek: we use CREATE TABLE IF NOT EXISTS quite extensively everywhere | 13:55 |
| zzzeek | we could do that in oslo’s create_engine. | 13:55 |
| ihrachyshka | zzzeek: and this issues WARNING (no idea why) | 13:55 |
| zzzeek | that emits a warning ? | 13:55 |
| ihrachyshka | yeah, it's weird, but it does | 13:56 |
| zzzeek | well, SQLAlchemy doesnt emit that so it woudl be nice to move to Table().create() and friends | 13:56 |
| *** markmcclain has quit IRC | 13:56 | |
| *** dstanek_zzz is now known as dstanek | 13:56 | |
| zzzeek | but anyway, oslo.db has its own create_engine where anything can be set up | 13:56 |
| *** tkelsey has joined #openstack-oslo | 13:56 | |
| ihrachyshka | zzzeek: ok, I'll check it, thanks | 13:57 |
| ihrachyshka | zzzeek: you mean, if we use Table().create(), it will be smart enough to detect table existance? | 13:57 |
| ihrachyshka | [I'm dumb in things db related] | 13:57 |
| zzzeek | any of SQLAclhemy or Alembics systems dont rely on “IF NOT EXISTS” | 13:57 |
| ihrachyshka | I see | 13:58 |
| zzzeek | its not supported by all backends | 13:58 |
| ihrachyshka | we have only one rule that uses it actually | 13:58 |
| ihrachyshka | I thought there are more :) | 13:58 |
| *** jokke__ is now known as jokke_ | 13:58 | |
| ihrachyshka | zzzeek: can we use alembic.op.create_table for that? | 14:02 |
| zzzeek | in a migration script, sure | 14:02 |
| ihrachyshka | good, thanks :) | 14:02 |
| zzzeek | gotta run :) | 14:02 |
| *** zzzeek has quit IRC | 14:02 | |
| *** YorikSar has quit IRC | 14:04 | |
| *** markmcclain has joined #openstack-oslo | 14:06 | |
| *** rpodolyaka has joined #openstack-oslo | 14:09 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Update oslo-incubator https://review.openstack.org/105743 | 14:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator/middleware/base https://review.openstack.org/105744 | 14:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator context module https://review.openstack.org/105745 | 14:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import notifier middleware from oslo-incubator https://review.openstack.org/105746 | 14:11 |
| *** lbragstad_ has joined #openstack-oslo | 14:11 | |
| *** alexpilotti has joined #openstack-oslo | 14:14 | |
| *** ihrachyshka has quit IRC | 14:16 | |
| *** ihrachyshka has joined #openstack-oslo | 14:17 | |
| jd__ | dhellmann: is there a plan about memorycache? should we drop it in favor of 'cache'? | 14:19 |
| dhellmann | jd__: morganfainberg_Z has a blueprint for doing something with that | 14:27 |
| dhellmann | jd__: https://review.openstack.org/#/c/97155/ | 14:27 |
| *** pblaho has quit IRC | 14:27 | |
| *** tsekiyama has joined #openstack-oslo | 14:28 | |
| viktors | dhellmann: just for information - how soon the new release comes to http://pypi.openstack.org/ ? | 14:28 |
| jd__ | dhellmann: ah that covers 'cache', not 'memorycache' actually | 14:28 |
| dhellmann | jd__: oops | 14:28 |
| jd__ | I need to read that anyway | 14:28 |
| dhellmann | viktors: you can watch the job run in the release queue on http://status.openstack.org/zuul/ | 14:29 |
| openstackgerrit | Romain Soufflet proposed a change to openstack/oslo-incubator: Fix tests.unit.reports.* with python3.4 https://review.openstack.org/105749 | 14:29 |
| dhellmann | viktors: it looks like it's done (0.3.0, right?) | 14:29 |
| viktors | dhellmann: I see 0.2.0 on http://pypi.openstack.org/openstack/oslo.db/ | 14:29 |
| viktors | but 0.3.0 on tarballs.openstack | 14:30 |
| dhellmann | viktors: hit reload, I'm seeing 0.3.0 | 14:30 |
| viktors | dhellmann: hm... ) | 14:30 |
| dhellmann | viktors: possibly a cache issue | 14:30 |
| dhellmann | viktors: https://pypi.python.org/pypi/oslo.db/0.3.0 | 14:30 |
| viktors | dhellmann: I see 0.3.0 on pypi.openstack.org also | 14:31 |
| viktors | but 0.2.0 on http://pypi.openstack.org/ | 14:31 |
| dhellmann | jd__: do you think we will need a separate memorycache module, since dogpile supports memcache? | 14:32 |
| dhellmann | viktors: ah, I should have read the full url :-/ | 14:32 |
| dhellmann | viktors: the mirror update can take a little while; I don't know how often that job runs | 14:32 |
| viktors | dhellmann: ok, thanks. I just worried a bit ) | 14:33 |
| *** pcm_ has quit IRC | 14:35 | |
| *** pcm_ has joined #openstack-oslo | 14:35 | |
| jd__ | dhellmann: now I'm thinking about dropping memorycache | 14:36 |
| jd__ | s/now/no/ | 14:36 |
| *** mkoderer has joined #openstack-oslo | 14:37 | |
| dhellmann | jd__: I suspect cleaning up the cache library is going to be a theme for the next release | 14:38 |
| *** lbragstad_ has left #openstack-oslo | 14:38 | |
| openstackgerrit | A change was merged to openstack/oslo.config: Hook up config fixture docs https://review.openstack.org/105539 | 14:39 |
| jd__ | dhellmann: ok, what am I supposed to do for this release then? :D | 14:39 |
| dhellmann | jd__: you and morganfainberg_Z are the experts :-) | 14:40 |
| openstackgerrit | A change was merged to openstack/oslo.config: Add release notes for 1.3.0 and 1.4.0.0a1/2 https://review.openstack.org/105540 | 14:41 |
| jd__ | anyway the spec LGTM | 14:41 |
| *** jaosorior has joined #openstack-oslo | 14:44 | |
| dhellmann | jd__: ok, good | 14:45 |
| *** nacim has quit IRC | 14:46 | |
| *** morganfainberg_Z is now known as morganfainberg | 14:46 | |
| *** haypo_ is now known as haypo | 14:48 | |
| *** HenryG has joined #openstack-oslo | 15:02 | |
| *** YorikSar has joined #openstack-oslo | 15:02 | |
| *** AAzza_afk is now known as AAzza | 15:09 | |
| *** zhiyan is now known as zhiyan_ | 15:10 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import notifier middleware from oslo-incubator https://review.openstack.org/105746 | 15:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator context module https://review.openstack.org/105745 | 15:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator/middleware/base https://review.openstack.org/105744 | 15:11 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Update oslo-incubator https://review.openstack.org/105743 | 15:11 |
| *** joesavak has quit IRC | 15:30 | |
| *** ihrachyshka has quit IRC | 15:34 | |
| *** lbragstad_ has joined #openstack-oslo | 15:39 | |
| *** alexpilotti has quit IRC | 15:45 | |
| markmc | jd__, dhellmann, flaper87, need https://review.openstack.org/105671 for the config generator to be usable, if one of you don't mind approving | 15:59 |
| *** viktors is now known as viktors|afk | 15:59 | |
| * flaper87 clicks | 15:59 | |
| * jd__ clicks too | 16:00 | |
| flaper87 | markmc: LGTM | 16:01 |
| jd__ | beaten by flaper87 :( | 16:02 |
| markmc | thanks :) | 16:05 |
| *** markmc has quit IRC | 16:05 | |
| *** pcm__ has joined #openstack-oslo | 16:08 | |
| *** pcm_ has quit IRC | 16:10 | |
| *** arnaud has joined #openstack-oslo | 16:10 | |
| *** SridharG has joined #openstack-oslo | 16:21 | |
| *** arnaud has quit IRC | 16:23 | |
| *** i159 has quit IRC | 16:32 | |
| *** GheRiver1 is now known as GheRivero | 16:43 | |
| *** lbragstad_ is now known as lbragstad | 16:51 | |
| *** jecarey has joined #openstack-oslo | 16:52 | |
| *** tkelsey has quit IRC | 16:54 | |
| *** praneshp has joined #openstack-oslo | 16:59 | |
| *** pcm__ has quit IRC | 17:00 | |
| *** zzzeek has joined #openstack-oslo | 17:03 | |
| *** harlowja_away is now known as harlowja | 17:03 | |
| *** pcm_ has joined #openstack-oslo | 17:04 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo-specs: oslo.messaging main loop refactoring https://review.openstack.org/105796 | 17:05 |
| *** Alexei_987 has quit IRC | 17:07 | |
| *** jecarey has quit IRC | 17:08 | |
| *** AAzza is now known as AAzza_afk | 17:29 | |
| *** arnaud__ has joined #openstack-oslo | 17:34 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement exception interception and filtering layer https://review.openstack.org/105307 | 17:35 |
| *** ildikov has joined #openstack-oslo | 17:59 | |
| *** Alexei_987 has joined #openstack-oslo | 18:02 | |
| boris-42 | zzzeek hi | 18:05 |
| zzzeek | hey | 18:05 |
| boris-42 | zzzeek thank you for advice | 18:05 |
| boris-42 | zzzeek to use before_cursor_execute | 18:05 |
| boris-42 | zzzeek and after_cursor_execute | 18:05 |
| zzzeek | ohhh, right | 18:05 |
| zzzeek | howd that work | 18:05 |
| boris-42 | zzzeek it works super great | 18:06 |
| zzzeek | what wre you trying to do? i forget. log queries right | 18:06 |
| boris-42 | zzzeek cause without it i have to disable osprofiler during migrations | 18:06 |
| boris-42 | zzzeek and now I don't need to do that | 18:06 |
| boris-42 | zzzeek https://github.com/stackforge/osprofiler/blob/master/osprofiler/sqlalchemy.py | 18:06 |
| zzzeek | great | 18:06 |
| boris-42 | zzzeek yep yep it's much better=) | 18:07 |
| zzzeek | did u see my oslo.db refactoring https://review.openstack.org/#/c/105307/ ? | 18:07 |
| zzzeek | who should review this? | 18:07 |
| boris-42 | zzzeek heh I didn't try it | 18:07 |
| boris-42 | zzzeek I'm mostly working now on rally & osprofiler and integration it everywehre=) | 18:07 |
| zzzeek | ok well, here’s the main part, do you like this: https://review.openstack.org/#/c/105307/8/oslo/db/sqlalchemy/exc_filters.py | 18:07 |
| zzzeek | your code is in there, i moved it around | 18:08 |
| boris-42 | zzzeek yep I will take a look at your patch | 18:08 |
| zzzeek | the main impact is that this is how we handle exceptions , using filter functions | 18:08 |
| zzzeek | theres no need for lots of if/then | 18:08 |
| zzzeek | and it works everywhere automatically without any wrapping | 18:08 |
| boris-42 | zzzeek btw | 18:10 |
| boris-42 | zzzeek one thing about OpenStack=) here it's better to split big pathces | 18:10 |
| boris-42 | zzzeek I mean +1k -0.4k is too big change to make a proper and detailed review | 18:10 |
| zzzeek | boris-42: there’s no way to split this up without adding more code that is then thrown away | 18:11 |
| zzzeek | i could add the event in one, and then copy-and-paste @_wrap_db_error onto it | 18:11 |
| zzzeek | but that still means the tests have to change | 18:11 |
| boris-42 | zzzeek okay | 18:12 |
| zzzeek | absoutely, if this could be done in smaller steps id have done it like that but i dont have the impression this patch has a great spot for that. | 18:13 |
| zzzeek | maybe it does but thats more work :) | 18:13 |
| zzzeek | its true though, the handle_error event and then the filtering system are two different things. | 18:14 |
| zzzeek | do i need to split it out ? | 18:14 |
| *** lbragstad has quit IRC | 18:14 | |
| zzzeek | when you have two gerrits for one blueprint, how do you name the branches for each of the reviews? | 18:14 |
| *** pcm_ has quit IRC | 18:15 | |
| zzzeek | the branch is named bp/use-events-for-error-wrapping | 18:15 |
| zzzeek | id need two branches right? | 18:15 |
| zzzeek | bp/use-events-for-error-wrapping-1 bp/use-events-for-error-wrapping-2 ? | 18:15 |
| *** pcm_ has joined #openstack-oslo | 18:17 | |
| *** AAzza_afk is now known as AAzza | 18:18 | |
| boris-42 | zzzeek sorry was not here) | 18:19 |
| boris-42 | zzzeek reading | 18:19 |
| boris-42 | zzzeek btw could you write "boris-42" in msg's so I won't miss your messages | 18:20 |
| boris-42 | zzzeek so about branches | 18:20 |
| boris-42 | zzzeek you have 1 branch | 18:20 |
| boris-42 | zzzeek in this branch you may have multiple patches | 18:20 |
| boris-42 | zzzeek if you run "git review -R" it will push all patches on review and update all dependencies | 18:21 |
| zzzeek | boris-42: ok. hm | 18:21 |
| boris-42 | zzzeek ^ that command is very smart | 18:21 |
| zzzeek | boris-42: how do i do an —amend to just one of them ? | 18:21 |
| boris-42 | zzzeek and it won't update any untouched pathces | 18:21 |
| boris-42 | zzzeek just do it | 18:21 |
| *** AAzza is now known as AAzza_afk | 18:21 | |
| boris-42 | zzzeek via rebase e.g. | 18:21 |
| boris-42 | <hash_of_last> | 18:21 |
| boris-42 | <has_of_last_-1> | 18:22 |
| boris-42 | <has_of_last_-2> | 18:22 |
| boris-42 | you would like to change in <hash_of_last_-1> something | 18:22 |
| boris-42 | you'll need to do git rebase -i <has_of_last_-2> | 18:22 |
| boris-42 | then you should put "e" insetad of "p" on patch that you would like to change | 18:22 |
| openstackgerrit | John Dewey proposed a change to openstack/oslo-incubator: Add healthcheck middleware https://review.openstack.org/105833 | 18:22 |
| boris-42 | and it will stop just after your patch | 18:23 |
| boris-42 | then you do changes | 18:23 |
| boris-42 | commit --amend | 18:23 |
| boris-42 | after that git rebase --continut | 18:23 |
| boris-42 | continue* | 18:23 |
| zzzeek | geez | 18:23 |
| boris-42 | and then git review -R | 18:23 |
| *** dstanek is now known as dstanek_zzz | 18:23 | |
| boris-42 | and it will update hash_of_last_-1 and hash_of_last patches | 18:24 |
| boris-42 | and won't touch hash_of_last | 18:24 |
| boris-42 | zzzeek =) | 18:24 |
| *** dims_ has quit IRC | 18:24 | |
| boris-42 | zzzeek at least it's my work flow=) | 18:24 |
| zzzeek | seems like its required | 18:24 |
| zzzeek | if i need to split this | 18:24 |
| *** morganfainberg is now known as morganfainberg_Z | 18:25 | |
| *** dims_ has joined #openstack-oslo | 18:25 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo-specs: oslo.messaging main loop refactoring https://review.openstack.org/105796 | 18:26 |
| *** SridharG has quit IRC | 18:32 | |
| *** SridharG has joined #openstack-oslo | 18:34 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo.messaging: Moved main loop to server class https://review.openstack.org/104983 | 18:38 |
| Alexei_987 | harlowja: Hi are you available? | 18:39 |
| harlowja | Alexei_987 hey hey, a little ya | 18:39 |
| Alexei_987 | harlowja: I've made a spec for my change and patch seems to be production ready ^^ | 18:39 |
| Alexei_987 | harlowja: sorry for being little late | 18:39 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo.messaging: Moved main loop to server class https://review.openstack.org/104983 | 18:44 |
| *** mriedem has quit IRC | 18:45 | |
| *** mriedem has joined #openstack-oslo | 18:48 | |
| openstackgerrit | John Dewey proposed a change to openstack/oslo-incubator: Add healthcheck middleware https://review.openstack.org/105833 | 18:51 |
| *** Alexei_987 has quit IRC | 18:55 | |
| harlowja | Alexei_987 cool, i'll look it over :) | 18:56 |
| *** Alexei_987 has joined #openstack-oslo | 19:00 | |
| *** praneshp has quit IRC | 19:00 | |
| *** praneshp has joined #openstack-oslo | 19:03 | |
| *** dims_ has quit IRC | 19:03 | |
| *** praneshp has quit IRC | 19:03 | |
| *** praneshp has joined #openstack-oslo | 19:04 | |
| *** ihrachyshka has joined #openstack-oslo | 19:10 | |
| ihrachyshka | zzzeek, gus, do you know when sqlalchemy 1.0 will be released? I'm interested in that fix for unicode + mysqlconnector | 19:12 |
| zzzeek | 1.0? months | 19:12 |
| zzzeek | what fix, the NATIONAL one? i have a workaround in the ticket | 19:12 |
| zzzeek | can use that right now | 19:12 |
| *** AAzza_afk is now known as AAzza | 19:13 | |
| zzzeek | and/or just send encoding=utf8 on the mysqlconnector URL | 19:13 |
| ihrachyshka | zzzeek, yes, that one. you mean, starting mysql with an option? | 19:13 |
| zzzeek | well the best is Unicode().with_variant(mysql.VARCHAR(national=True)) | 19:13 |
| ihrachyshka | zzzeek, can we set it in oslo.db then, in create_engine? | 19:13 |
| zzzeek | sure | 19:13 |
| ihrachyshka | zzzeek, ok. we will still need to set raise_on_warnins to False anyway. alembic.op.create_table won't work for offline migration anyway. | 19:14 |
| zzzeek | whys that | 19:15 |
| ihrachyshka | zzzeek, for offline migration, we just generate sql script, and then run it later. if we put create table with no condition, it will fail. | 19:15 |
| ihrachyshka | zzzeek, in that case, we don't have access to db contents to check whether table is there | 19:16 |
| ihrachyshka | well, I may be wrong, but that's what jlibosva from neutron confirmed | 19:16 |
| zzzeek | ihrachyshka: thats wrong | 19:18 |
| zzzeek | ihrachyshka: menaing, not that waht theyre saying isnt happening | 19:18 |
| zzzeek | ihrachyshka: but, you shoud not be running a .sql script against a DB in an unknown state | 19:18 |
| ihrachyshka | zzzeek, maybe, but still, there is that offline migration thing present, and we support it :) | 19:19 |
| zzzeek | what “offline migration thing” ? | 19:19 |
| *** arnaud__ has quit IRC | 19:19 | |
| zzzeek | you can’t safely run a sql script that has a full set of migrations in it, against a DB that is in an unknown state | 19:19 |
| ihrachyshka | offline migration mode, with sql script run later | 19:19 |
| ihrachyshka | we have online and offline modes | 19:19 |
| zzzeek | if a table has the wrong columns in it, that approach fails | 19:19 |
| ihrachyshka | I think you generate it for some specific revision of models | 19:20 |
| ihrachyshka | but honestly, I don't know details, yet | 19:20 |
| zzzeek | a migration script shoudl never need to check anything whatsoever in the exsiting schema | 19:21 |
| zzzeek | migration script should *only* be run against a schema whos state is exactly determined ahead of time | 19:21 |
| zzzeek | otherwise, it is wrong. it will fail | 19:21 |
| zzzeek | either loudly or silently | 19:21 |
| zzzeek | this is probably why “create if not exists” raises a warning | 19:22 |
| zzzeek | if you are using that, it means you are guessing | 19:22 |
| ihrachyshka | I think those conditions in migration are due to some known errors made previously, everyone knows it's hacks, but still, we have them, and we can't remove them now | 19:22 |
| zzzeek | I am 100% sure it can be fixed :) | 19:22 |
| zzzeek | ihrachyshka - “everyone knows it's hacks, but still, we have them, and we can't remove them now” - anytime someone says that, then I know the problem is fixable :) | 19:23 |
| ihrachyshka | zzzeek, the problem is that earlier, db could have different models created depemding on configuration used, for the same revision. that's why those conditional rules | 19:24 |
| zzzeek | yup, so, you need separate migration scripts for each of those conditions | 19:24 |
| zzzeek | those conditions need to be enumerated | 19:24 |
| zzzeek | “create table if not exists”, is one thing, but there is no “alter column set type = integer if the type is string but is not datetime” | 19:25 |
| zzzeek | so these hacks are of limited scope, they have to be | 19:25 |
| ihrachyshka | there are lots of different configuration setups, it's basically a set of plugins enabled locally that define models | 19:26 |
| ihrachyshka | better, "defined" | 19:26 |
| ihrachyshka | now neutron introduces "healing script" to get back to single sane schema | 19:27 |
| zzzeek | ihrachyshka: OK, so are we referring strictly to the “healing script” part of this? | 19:27 |
| ihrachyshka | hm? no, those conditional create tables are in migration rules defined before people realized it's insane how they handled migration before | 19:28 |
| zzzeek | set the setting in create_engine() for now, that is easiet | 19:30 |
| zzzeek | but if it were me I’d fix all those migrations and for those that definitely need special rules I’d hardwire MySQL directives to suppress the warnings at the SQL level | 19:30 |
| zzzeek | ihrachyshka: http://dev.mysql.com/doc/connector-python/en/connector-python-api-mysqlconnection-raise-on-warnings.html default says “False" | 19:31 |
| ihrachyshka | it's true in sqlalchemy | 19:32 |
| ihrachyshka | for that specific driver | 19:32 |
| zzzeek | well there’s your bug wtf is it like that for | 19:33 |
| ihrachyshka | line 81 | 19:34 |
| zzzeek | i know that | 19:34 |
| zzzeek | this was contributed by the author of mysqlconnector | 19:34 |
| zzzeek | look set it to False on create_engine() | 19:34 |
| zzzeek | this default shouldnt be like that | 19:34 |
| zzzeek | but easy enough to fix | 19:34 |
| ihrachyshka | so probably should be false in sa too? | 19:34 |
| zzzeek | im changing it right now | 19:35 |
| zzzeek | as well as that buffered BS | 19:35 |
| zzzeek | but you need to support back to 0.7.8, so, needs to be set | 19:36 |
| ihrachyshka | yeah, I get it. great, thanks a lot, I'll send a patch tomorrow | 19:36 |
| zzzeek | no patch needed i am doing it here. buffered seems a little more important.... | 19:37 |
| ihrachyshka | i mean oslo.db | 19:37 |
| ihrachyshka | or will you handle it too? | 19:37 |
| ihrachyshka | what's that buffered? | 19:38 |
| zzzeek | wow, it is a PITA | 19:41 |
| zzzeek | no, you do oslo.db | 19:41 |
| zzzeek | buffered is really broken in mysqlconnecvtor. at its default, results arent bufffered. which is great! but then it refuses to let the cursor be closed without fetching all of them | 19:42 |
| zzzeek | that is wrong wrong wrong | 19:42 |
| ihrachyshka | so we better leave buffered to true for now? | 19:43 |
| zzzeek | on your end yes | 19:43 |
| zzzeek | dont name it at all | 19:43 |
| ihrachyshka | ok, got it. doesn't it mean the library is not that mature and we may get other surprises after we switch to it? ;) | 19:44 |
| ihrachyshka | (not that existing one is great) | 19:45 |
| *** dims has joined #openstack-oslo | 19:46 | |
| zzzeek | it passes all tests | 19:46 |
| zzzeek | so it’s pretty strong | 19:47 |
| zzzeek | but this one, ugh. in the early days it had a lot of troublesome things like this | 19:47 |
| *** SridharG has quit IRC | 19:49 | |
| *** AAzza is now known as AAzza_afk | 19:50 | |
| ihrachyshka | btw I think we support sa starting from 0.8.4 now, 0.7.* is gone for us | 19:50 |
| *** mkoderer has quit IRC | 19:52 | |
| *** dims_ has joined #openstack-oslo | 19:55 | |
| *** dstanek_zzz is now known as dstanek | 19:56 | |
| *** dims has quit IRC | 19:57 | |
| *** morganfainberg_Z is now known as morganfainberg | 19:57 | |
| zzzeek | ihrachyshka: wow when the heck did that happen | 19:58 |
| *** ihrachyshka has quit IRC | 19:59 | |
| *** ihrachyshka has joined #openstack-oslo | 19:59 | |
| *** lbragstad_ has joined #openstack-oslo | 19:59 | |
| *** arnaud has joined #openstack-oslo | 20:02 | |
| *** YorikSar has quit IRC | 20:06 | |
| *** YorikSar has joined #openstack-oslo | 20:08 | |
| ihrachyshka | https://review.openstack.org/#/c/96438/ | 20:09 |
| ihrachyshka | zzzeek, ^^ | 20:10 |
| ihrachyshka | it requires sa 0.8 (at least that's why version was bumped) | 20:10 |
| *** ihrachyshka has quit IRC | 20:16 | |
| *** markmcclain has quit IRC | 20:23 | |
| *** YorikSar has quit IRC | 20:38 | |
| *** YorikSar has joined #openstack-oslo | 20:40 | |
| *** bknudson has joined #openstack-oslo | 20:43 | |
| *** arnaud has quit IRC | 20:45 | |
| zzzeek | what feature | 20:47 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the core 10k view omnigraffle diagram https://review.openstack.org/105642 | 21:00 |
| boris-42 | zzzeek one questions about sqla | 21:02 |
| zzzeek | sure | 21:02 |
| boris-42 | zzzeek can I add event listener after engine.connect() | 21:02 |
| *** lbragstad_ is now known as lbragstad | 21:03 | |
| zzzeek | pool events for that | 21:04 |
| zzzeek | the checkout event | 21:04 |
| zzzeek | you should be tracking pool checkouts/checkins, thats where DB connection resources are used | 21:05 |
| boris-42 | zzzeek okay lemme show the code | 21:06 |
| boris-42 | zzzeek this doesn't work https://review.openstack.org/#/c/105635/6/glance/db/sqlalchemy/api.py | 21:06 |
| zzzeek | and why is that | 21:08 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the core 10k view omnigraffle diagram https://review.openstack.org/105642 | 21:11 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the core 10k view omnigraffle diagram https://review.openstack.org/105642 | 21:12 |
| zzzeek | boris-42: what does git review -R do as opposed to git review ? I see it at http://www.mediawiki.org/wiki/Gerrit/Tutorial#Configuring_git-review, says nothing about why, there is no —help | 21:13 |
| boris-42 | zzzeek ? | 21:13 |
| zzzeek | boris-42: ? | 21:13 |
| boris-42 | zzzeek you mean "-R" | 21:13 |
| boris-42 | zzzeek "-R" if you don't specify it, it will update all patches in any case | 21:14 |
| zzzeek | boris-42: here: http://www.mediawiki.org/wiki/Gerrit/Tutorial#Push_your_change_set_to_Gerrit | 21:14 |
| zzzeek | boris-42: where is that documnted? i can’t find any doc for what -R means | 21:14 |
| boris-42 | zzzeek it's secret knowledge | 21:14 |
| boris-42 | zzzeek =) | 21:14 |
| zzzeek | boris-42: so -R gives me some kind of edit screen ? | 21:14 |
| boris-42 | zzzeek nope | 21:14 |
| zzzeek | boris-42: i cant test it b.c. i dont have anything to push up | 21:14 |
| boris-42 | zzzeek -R sends to gerrit patches only that were changed | 21:15 |
| zzzeek | boris-42: oh | 21:15 |
| boris-42 | zzzeek if you have 5 patches in branch | 21:15 |
| boris-42 | zzzeek and you edited only last one | 21:15 |
| boris-42 | zzzeek and run "git review" it will send all 5 | 21:15 |
| boris-42 | zzzeek in case of "git review -R" only last one | 21:16 |
| zzzeek | boris-42: so if i take my changeset at https://review.openstack.org/#/c/105307/, and rework/rebase it into two separate commits, then do “git review -R”, it will update this one, then make a second review for the other changeset? | 21:18 |
| boris-42 | zzzeek it will make to separated change sets | 21:22 |
| boris-42 | zzzeek i mean with 2 unique numbers in url | 21:22 |
| *** markmcclain has joined #openstack-oslo | 21:22 | |
| * zzzeek is giong to do it! | 21:22 | |
| * zzzeek will see if it ruins everything! | 21:22 | |
| boris-42 | zzzeek and btw | 21:22 |
| boris-42 | zzzeek number 105307 is 1-1 to Change-ID number | 21:22 |
| boris-42 | zzzeek it sadness | 21:23 |
| boris-42 | zzzeek but for some reason I am not able to trace sqlalchemy in glance | 21:23 |
| zzzeek | boris-42: i dont know why your api.py doesnt work there’s nothing in that file that uses SQLA APIs directly | 21:23 |
| zzzeek | they use NullPool, but that works: http://paste.openstack.org/show/85833/ | 21:25 |
| *** praneshp has quit IRC | 21:31 | |
| boris-42 | zzzeek I found the reason | 21:31 |
| boris-42 | zzzeek they call another service, and it doesn't have inited profiler | 21:31 |
| *** arnaud has joined #openstack-oslo | 21:40 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement the SQLAlchemy ``handle_error()`` event. https://review.openstack.org/105307 | 21:46 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement new exception interception and filtering layer https://review.openstack.org/105891 | 21:46 |
| openstackgerrit | A change was merged to openstack/oslo.messaging: Synced jsonutils and its dependencies from oslo-incubator https://review.openstack.org/95696 | 21:46 |
| zzzeek | boris-42: wow! https://review.openstack.org/#/c/105307/ and https://review.openstack.org/#/c/105891/ | 21:47 |
| boris-42 | zzzeek yep magic =) | 21:47 |
| zzzeek | magic! | 21:47 |
| boris-42 | zzzeek and don't worry if you change order of patches with git rebase | 21:48 |
| boris-42 | zzzeek it will change in gerrit as well automatically=) | 21:48 |
| *** mrda-away is now known as mrda | 21:51 | |
| *** praneshp has joined #openstack-oslo | 21:53 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the arch/big picture omnigraffle diagram https://review.openstack.org/105642 | 22:02 |
| *** mriedem has quit IRC | 22:09 | |
| *** nacim has joined #openstack-oslo | 22:15 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Add the arch/big picture omnigraffle diagram https://review.openstack.org/105642 | 22:26 |
| *** nacim has quit IRC | 22:28 | |
| *** boris-42 has quit IRC | 22:37 | |
| *** boris-42 has joined #openstack-oslo | 22:39 | |
| *** praneshp has quit IRC | 22:44 | |
| *** praneshp has joined #openstack-oslo | 22:47 | |
| *** dstanek is now known as dstanek_zzz | 22:49 | |
| HenryG | rpodolyaka: around? I have a problem with oslo.db in neutron unit tests | 22:55 |
| *** lbragstad has quit IRC | 23:04 | |
| *** bknudson has quit IRC | 23:10 | |
| *** pcm_ has quit IRC | 23:16 | |
| *** HenryG has quit IRC | 23:16 | |
| *** morganfainberg is now known as morganfainberg_Z | 23:17 | |
| *** markmcclain has quit IRC | 23:21 | |
| gus | ihrachyshka: sorry, we keep missing each other online :/ | 23:28 |
| gus | zzzeek: There's some strangeness in both mysqldb + mysqlconnector wrt unicode enums. | 23:29 |
| gus | mysqldb seems to incorrectly think my connection encoding is latin1, even though in this case it isn't, so that adds an extra round of broken escaping. | 23:30 |
| gus | mysqlconnector doesn't have that issue, but still the contents coming back from the db are interpreted as a byte sequence, not codepoints at some point. I've traced it as far as ResultProxy. | 23:31 |
| gus | (still hunting further backwards) | 23:31 |
| gus | (mysqldb unicode seems broken in several ways without being forced with ?charset=utf8) | 23:34 |
| zzzeek | gus: not sure, i might be running it with charset=utf8 | 23:43 |
| zzzeek | hmmm nope | 23:43 |
| zzzeek | i run mysqldb without it. but the DB encoding might be a factor | 23:44 |
| zzzeek | like the deafulst of mysql | 23:44 |
| gus | zzzeek: huh. yeah, utf8 has to get in there at some point. | 23:44 |
| gus | either your mysqld is compiled with a different default than the usual mysql default (latin1), or your server config, or DB were created to default to utf8 | 23:44 |
| gus | at least, afaiui. | 23:44 |
| zzzeek | ok waht is the SHOW command that will tell me | 23:45 |
| gus | show variables; | 23:45 |
| gus | and there's a bunch of character_set_% settings | 23:46 |
| gus | that shows you server + db defaults. | 23:46 |
| zzzeek | server is utf8 here | 23:47 |
| gus | each table can have a default which should be in show create table foo; | 23:47 |
| gus | zzzeek: yep. | 23:47 |
| zzzeek | tehre are tests that have non-latin1 data, so the connecrtion needs to be utf8 | 23:47 |
| zzzeek | or the server | 23:47 |
| gus | so character_set_server is --character-set-server command line flag, or config file option. | 23:48 |
| gus | then it falls back to the compiled in default, which is supposedly latin1, but can be set to something else at build time. | 23:48 |
| zzzeek | anyway, SQLAlchemy’s job is either to send u’’ as is, or to encode it with ‘encoding’, and it does both correctly. if mysql chokes on it then mysql or client has to be configured | 23:49 |
| gus | zzzeek: I respectfully disagree | 23:49 |
| gus | I can point to several sqlalchemy generated statements that don't specify charset. | 23:49 |
| gus | ... when the intention was clearly to specify unicode. | 23:50 |
| zzzeek | here’s our docs for that http://docs.sqlalchemy.org/en/rel_0_9/dialects/mysql.html#unicode | 23:50 |
| zzzeek | you disagree that SQLA’s job is to send a u’’ string or a correctly encoded ‘’ string ? | 23:51 |
| gus | take for example the one I'm currently looking at: test.dialect.mysql.test_types:EnumSetTest.test_unicode_enum | 23:51 |
| gus | that test tries to create an enum with unicode chars in the names, and then sees if they get reflected correctly. | 23:52 |
| gus | there's nothing there that I can see that tells the DB that these are going to be unicode. | 23:52 |
| gus | Specifically, I think the test cases should be mysql.ENUM((u('...'), ...), charset='utf8') | 23:53 |
| gus | or similar. | 23:53 |
| gus | likewise Unicode and UnicodeText (before my change) didn't do anything at all to specify charset for mysql. | 23:53 |
| zzzeek | gus: so first off, I was talking about the actual data values passed to cursor.execute(stmt, <data goes here>) | 23:54 |
| zzzeek | gus: that’s where we may or may not have to encode depending on DBAPI behavior | 23:54 |
| zzzeek | gus: in this case you’re referring to the DDL passed to the database, and that it should contain directives that guide the DB towards a specific encoding | 23:55 |
| zzzeek | gus: I would argue that because MySQL does the right thing when the server-side or client-side encoding is set correctly, it would be surprising behavior for this encoding to be hardwired into the DDL | 23:55 |
| zzzeek | gus: unless you’re saying, just the test here should have charset=‘utf8’ in it, in which case, sure that would help the test here pass, though im not sure the test suite can run all the way thru without a default encoding of utf-8 | 23:56 |
| zzzeek | gus: or osmethign unicode-capable | 23:56 |
| zzzeek | gus: theoretically someone might want to run the whole test suite using utf-16 | 23:56 |
| zzzeek | and it woudl be nice if that worked too ! | 23:56 |
| gus | right, the results I'm seeing now assume the db is set to default to utf8 | 23:57 |
| zzzeek | gus: does it assume utf-8, or unicode-capable encoding ? | 23:57 |
| gus | in particular, the mysqldb engine seems to get out of sync, and even if the database *is* set to default to utf8, then it assumes latin1 unless ?charset=utf8 | 23:57 |
| gus | zzzeek: erm ... I guess I mean it defaults to unicode-capable. | 23:58 |
| gus | actually no, I mean both. | 23:58 |
| gus | it assumes the db is unicode-capable, and the connection is utf8. | 23:58 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Make greenexecutor not keep greenthreads active https://review.openstack.org/105922 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!