| openstackgerrit | A change was merged to openstack/pycadf: Updated from global requirements https://review.openstack.org/106572 | 00:00 |
|---|---|---|
| *** dims_ has quit IRC | 00:03 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement a dialect-level function dispatch system https://review.openstack.org/109436 | 00:16 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 00:16 |
| zzzeek | that’s a big one | 00:21 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 00:24 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 00:31 |
| *** zzzeek has quit IRC | 00:41 | |
| *** nacim has quit IRC | 01:17 | |
| *** jecarey has joined #openstack-oslo | 01:25 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Allow endpoints to be fetched from factories https://review.openstack.org/110181 | 01:52 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Allow endpoints to be fetched from factories https://review.openstack.org/110181 | 02:00 |
| *** yamahata has quit IRC | 02:16 | |
| *** penguinRaider__ has quit IRC | 02:49 | |
| *** praneshp has joined #openstack-oslo | 02:51 | |
| *** penguinRaider__ has joined #openstack-oslo | 02:52 | |
| *** praneshp_ has joined #openstack-oslo | 02:54 | |
| *** praneshp has quit IRC | 02:56 | |
| *** praneshp_ is now known as praneshp | 02:56 | |
| *** harlowja is now known as harlowja_away | 03:07 | |
| *** harlowja_away is now known as harlowja | 03:08 | |
| *** jecarey has quit IRC | 03:32 | |
| *** amotoki has joined #openstack-oslo | 03:55 | |
| *** SridharG has joined #openstack-oslo | 04:15 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Allow endpoints to be fetched from factories https://review.openstack.org/110181 | 04:25 |
| *** HenryG is now known as HenryG_afk | 04:45 | |
| *** morganfainberg is now known as morganfainberg_Z | 04:59 | |
| *** harlowja is now known as harlowja_away | 05:26 | |
| *** ildikov has quit IRC | 05:34 | |
| *** jaosorior has joined #openstack-oslo | 05:42 | |
| *** jraim has quit IRC | 06:00 | |
| *** zhiyan has quit IRC | 06:00 | |
| *** harlowja_at_home has joined #openstack-oslo | 06:00 | |
| *** jraim__ has joined #openstack-oslo | 06:05 | |
| *** zhiyan has joined #openstack-oslo | 06:05 | |
| *** zhiyan has quit IRC | 06:23 | |
| *** jraim__ has quit IRC | 06:24 | |
| *** noelbk has quit IRC | 06:25 | |
| *** zhiyan has joined #openstack-oslo | 06:28 | |
| *** jraim__ has joined #openstack-oslo | 06:28 | |
| *** ildikov has joined #openstack-oslo | 06:33 | |
| *** viktors|afk has quit IRC | 06:36 | |
| *** harlowja_at_home has quit IRC | 06:38 | |
| *** pblaho has joined #openstack-oslo | 06:42 | |
| *** SridharG has quit IRC | 06:56 | |
| *** SridharG has joined #openstack-oslo | 07:00 | |
| *** SridharG has quit IRC | 07:06 | |
| *** flaper87|afk is now known as flaper87 | 07:26 | |
| *** ihrachyshka has joined #openstack-oslo | 07:46 | |
| gus | urgh. mysqlconnector doesn't accept execute("sql stmt1; sql stmt 2"), and the sqlalchemy docs are unclear on whether that should be allowed or not :/ | 07:49 |
| *** ihrachyshka has quit IRC | 07:50 | |
| *** ihrachyshka has joined #openstack-oslo | 07:50 | |
| gus | zzzeek: ^ is this a user error, or should dialects/mysqlconnector.py be fixed to use cursor.execute(multi=True)? | 07:51 |
| *** nacim has joined #openstack-oslo | 08:02 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/oslo-incubator: Remove usage of readlines() https://review.openstack.org/110228 | 08:02 |
| *** i159 has joined #openstack-oslo | 08:03 | |
| *** praneshp has quit IRC | 08:22 | |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo.db: oslo.db.exceptions module documentation https://review.openstack.org/93435 | 08:27 |
| haypo | win 3 | 08:32 |
| openstackgerrit | Roman Podoliaka proposed a change to openstack/oslo.db: Implement a dialect-level function dispatch system https://review.openstack.org/109436 | 08:42 |
| *** Alexei_987 has joined #openstack-oslo | 08:50 | |
| openstackgerrit | Tim Kelsey proposed a change to openstack/oslo.messaging: Adds message security plugins https://review.openstack.org/109806 | 08:56 |
| *** ajo has joined #openstack-oslo | 08:58 | |
| openstackgerrit | A change was merged to openstack/oslo.db: oslo.db.api module documentation https://review.openstack.org/93434 | 09:16 |
| openstackgerrit | A change was merged to openstack/oslo.db: Updated from global requirements https://review.openstack.org/110105 | 09:20 |
| openstackgerrit | Tim Kelsey proposed a change to openstack/oslo.messaging: Adds message security plugins https://review.openstack.org/109806 | 09:28 |
| openstackgerrit | A change was merged to openstack/oslo.db: Specify raise_on_warnings=False for mysqlconnector https://review.openstack.org/107221 | 09:35 |
| *** Alexei_987 has quit IRC | 09:37 | |
| openstackgerrit | Oleksii Chuprykov proposed a change to openstack/oslo.db: Add create_database_permission_exist function https://review.openstack.org/109259 | 09:49 |
| *** alexpilotti has joined #openstack-oslo | 10:19 | |
| *** Alexei_987 has joined #openstack-oslo | 10:27 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/110104 | 10:29 |
| *** pcm_ has joined #openstack-oslo | 10:36 | |
| *** pcm__ has joined #openstack-oslo | 10:37 | |
| *** pcm_ has quit IRC | 10:41 | |
| *** HenryG_afk is now known as HenryG | 10:53 | |
| *** pblaho has quit IRC | 11:06 | |
| *** pcm__ has quit IRC | 11:14 | |
| *** pcm_ has joined #openstack-oslo | 11:15 | |
| ihrachyshka | I wonder whether there are plans for a new oslo.db version | 11:21 |
| *** GheRiver1 is now known as GheRivero | 11:32 | |
| *** nacim has quit IRC | 11:35 | |
| *** pblaho has joined #openstack-oslo | 11:36 | |
| *** nacim has joined #openstack-oslo | 11:37 | |
| *** SridharG has joined #openstack-oslo | 11:49 | |
| *** SridharG has quit IRC | 12:03 | |
| *** nacim has quit IRC | 12:23 | |
| *** nacim has joined #openstack-oslo | 12:26 | |
| *** ihrachyshka_ has joined #openstack-oslo | 12:30 | |
| *** ihrachyshka_ has quit IRC | 12:31 | |
| *** ihrachyshka_ has joined #openstack-oslo | 12:32 | |
| *** ihrachyshka has quit IRC | 12:32 | |
| *** gordc has joined #openstack-oslo | 12:41 | |
| *** ihrachyshka__ has joined #openstack-oslo | 12:56 | |
| *** lbragstad has joined #openstack-oslo | 12:57 | |
| *** ihrachyshka_ has quit IRC | 12:59 | |
| *** pblaho_ has joined #openstack-oslo | 13:03 | |
| *** pblaho has quit IRC | 13:03 | |
| *** dims has joined #openstack-oslo | 13:31 | |
| *** bknudson has joined #openstack-oslo | 13:41 | |
| *** jokke__ is now known as jokke_ | 13:53 | |
| *** lbragsta_ has joined #openstack-oslo | 13:56 | |
| *** lbragsta_ has quit IRC | 13:57 | |
| *** lbragsta_ has joined #openstack-oslo | 13:58 | |
| *** lbragstad has quit IRC | 13:58 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo-incubator: network_utils: make tests portable https://review.openstack.org/110293 | 14:04 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo-incubator: log: make tests portable https://review.openstack.org/110297 | 14:07 |
| *** pblaho_ has quit IRC | 14:09 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo-incubator: network_utils: make tests portable https://review.openstack.org/110293 | 14:15 |
| *** dhellmann has quit IRC | 14:17 | |
| *** dhellmann has joined #openstack-oslo | 14:20 | |
| *** ildikov has quit IRC | 14:22 | |
| *** ihrachyshka__ is now known as ihrachyshka | 14:24 | |
| ihrachyshka | i159: am I right that you're somehow involved in oslo.db? | 14:24 |
| *** ajo has quit IRC | 14:24 | |
| i159 | ihrachyshka: Hi. Yes I'am. I'm a member of Mirantis oslo.db team. | 14:26 |
| ihrachyshka | good, great. so I have a question to an expert then :) | 14:26 |
| ihrachyshka | I'm working on mysqldb -> mysqlconnector switch | 14:26 |
| ihrachyshka | some of projects require changes here and there in migration rules | 14:27 |
| *** ajo has joined #openstack-oslo | 14:27 | |
| ihrachyshka | now I'm struggling with glance | 14:27 |
| ihrachyshka | and it fails with (InterfaceError) Use multi=True when executing multiple statements when a glance migration rule is attempted | 14:27 |
| ihrachyshka | the rule is not a python script, but a SQL file | 14:27 |
| ihrachyshka | see: https://github.com/openstack/glance/blob/master/glance/db/sqlalchemy/migrate_repo/versions/006_mysql_upgrade.sql | 14:28 |
| ihrachyshka | my question is 1) WTF (other projects don't have those) | 14:28 |
| ihrachyshka | 2) what to do with it? it fails inside /usr/lib/python2.7/site-packages/migrate/versioning/script/sql.py, where the contents of the file is executed on cursor | 14:29 |
| ihrachyshka | I guess we could modify that migrate library to pass multi=True, but 1) it does not belong to openstack 2) I don't know whether this multi=True thing is portable | 14:29 |
| ihrachyshka | i159: ^^ | 14:30 |
| *** dims has quit IRC | 14:31 | |
| *** amotoki has quit IRC | 14:34 | |
| gordc | dhellmann: was wondering, is there something special that needs to be done to get oslo.config/messaging>=1.4.0.0a3 to work? i get distributions not found errors when i run unit tests | 14:36 |
| i159 | ihrachyshka: I think your problem is glance-migrate specific. multi=True is multi databases mode, I think so. I never had a deal with multi=True mode, so I'm sorry, I can't help you. You can ping rpodolyaka or zzzeek, they are muuuuch more experienced than me and it is possible that they can help you. | 14:39 |
| ihrachyshka | multi=True is for 'allow multiple statements being executed in one go' | 14:40 |
| ihrachyshka | rpodolyaka1: ^^ | 14:40 |
| ihrachyshka | i159: thanks anyway :) | 14:40 |
| ihrachyshka | yeah, I also think it's glance specific. we don't generally execute pre-crafted mysql scripts | 14:41 |
| rpodolyaka1 | ihrachyshka: hmm, interesting | 14:43 |
| rpodolyaka1 | ihrachyshka: I think, multi database mode is a red herring, but what we have here is that we pass a few SQL statements to cursor.execute() call | 14:44 |
| rpodolyaka1 | ihrachyshka: somehow, it raises a warning with mysqlconnector DB API driver | 14:44 |
| rpodolyaka1 | InterfaceError: (InterfaceError) Use multi=True when executing multiple statements 'select 1; select 2;' {} | 14:45 |
| rpodolyaka1 | In [7]: eng.execute('select 1; select 2;') | 14:45 |
| ihrachyshka | it comes from really old times: https://github.com/openstack/glance/commit/a2c9a1eb5ce3b5abd650811ad9c97d5aa307af8a | 14:45 |
| rpodolyaka1 | ihrachyshka: ^^ | 14:45 |
| ihrachyshka | rpodolyaka1: is it connector issue? | 14:46 |
| rpodolyaka1 | In [11]: eng = sa.create_engine('mysql://root:devel@localhost/devel') | 14:46 |
| rpodolyaka1 | In [12]: eng.execute('select 1; select 2;') | 14:46 |
| rpodolyaka1 | Out[12]: <sqlalchemy.engine.result.ResultProxy at 0x7f7cc8c593d0> | 14:46 |
| rpodolyaka1 | this is with mysql-python | 14:46 |
| rpodolyaka1 | ihrachyshka: I'm not sure, it's a bug, it might also be a feature :P | 14:47 |
| rpodolyaka1 | ihrachyshka: I'd ask on mysqlconnector irc channel, if they have one :) | 14:47 |
| ihrachyshka | looks like a feature, they have a special handling for this in their code :) | 14:47 |
| ihrachyshka | they don't, that's a problem :) | 14:48 |
| ihrachyshka | though I can reach the author thru email | 14:48 |
| ihrachyshka | I'll probably go that route | 14:48 |
| *** tsufiev has quit IRC | 14:55 | |
| dhellmann | gordc: you might need a newer version of tox/virtualenv/pip -- let me know if you're still having issues | 14:58 |
| dhellmann | rpodolyaka1, ihrachyshka : is that flag for multiple SQL statements in one call vs. multiple databases? | 14:59 |
| gordc | dhellmann: hmm. ok, i'll give that a try | 14:59 |
| rpodolyaka1 | dhellmann: for multiple SQL statements | 14:59 |
| dhellmann | gordc: which versions do you have installed? | 15:00 |
| ihrachyshka | dhellmann: yes | 15:00 |
| dhellmann | rpodolyaka1: interesting. seems like something we should support. | 15:00 |
| gordc | tox==1.6.1, pip==1.5.6 | 15:00 |
| ihrachyshka | dhellmann: yeah, otherwise we need to modify projects like in https://review.openstack.org/#/c/110271/ | 15:00 |
| rpodolyaka1 | dhellmann: it seems this is mysqlconnector DB API driver specific, as we haven't seen this issue so far (we've been testing mysql-python/psycopg2/pysqlite drivers) | 15:01 |
| *** tsufiev_ has joined #openstack-oslo | 15:01 | |
| *** morganfainberg_Z is now known as morganfainberg | 15:01 | |
| gordc | dhellmann: venv==1.10.1 | 15:02 |
| dhellmann | rpodolyaka1: so those other connectors don't accept this option, or the default is different? | 15:02 |
| ihrachyshka | rpodolyaka1: well, we could make our library more smart and handling that case for us, passing multi=True when needed (?) | 15:02 |
| ihrachyshka | dhellmann: I guess the former, but need to check | 15:02 |
| dhellmann | gordc: I'm on 1.11.2 in at least one project, so maybe that's it then | 15:03 |
| dhellmann | ihrachyshka: ok | 15:03 |
| gordc | dhellmann: ok, i'll try updating venv | 15:04 |
| *** tsufiev_ has quit IRC | 15:05 | |
| *** tsufiev_ has joined #openstack-oslo | 15:06 | |
| gordc | dhellmann: yep, that fixed it. thanks! | 15:06 |
| dhellmann | \o/ | 15:06 |
| *** tsufiev_ has quit IRC | 15:07 | |
| *** jecarey has joined #openstack-oslo | 15:07 | |
| ihrachyshka | dhellmann: yeah, doesn't look like mysql-python has it: https://github.com/arnaudsj/mysql-python/blob/master/MySQLdb/cursors.py | 15:07 |
| dhellmann | ihrachyshka: so we need our own connector compatibility layer? | 15:08 |
| ihrachyshka | dhellmann: maybe, I don't know whether it would be easier to fix that case by case though, as in the patch I've posted above | 15:08 |
| dhellmann | gordc: did you write a spec for the oslo.middleware graduation? | 15:08 |
| *** tsufiev has joined #openstack-oslo | 15:09 | |
| ihrachyshka | dhellmann: I've sent an email to the author of connector to see how he feels about that | 15:10 |
| *** lbragsta_ is now known as lbragstad | 15:10 | |
| dhellmann | ihrachyshka: I think we'll be better off if we can make oslo.db adapt to the projects as they are now, without having to change them | 15:11 |
| ihrachyshka | dhellmann: for that specific issue, we already have some relevant hack inside migrate library for sqlite: https://github.com/stackforge/sqlalchemy-migrate/blob/master/migrate/versioning/script/sql.py#L37 | 15:11 |
| gordc | dhellmann: i want to say yes, but i can't find any... should i write one up? and i guess another for notifiermiddleware? | 15:11 |
| dhellmann | gordc: yes, please -- I just approved the wrong bp, realized my mistake, went to look for the spec for the other to use the tool to do it (instead of doing it all by hand again), and didn't find one | 15:12 |
| gordc | dhellmann: ok. will do a quick write up | 15:13 |
| dhellmann | gordc: I'll go ahead and target the bp, but if you could get a spec put together this week that would let me approve it before my next sync with ttx | 15:13 |
| gordc | dhellmann: will do | 15:13 |
| dhellmann | gordc: thanks! | 15:13 |
| jd__ | gordc: did you port/integrate the notifier middleware in oslo.middleware? | 15:14 |
| jd__ | I think it was missing in your original porting | 15:14 |
| jd__ | s/porting/run of graduate.sh/ | 15:14 |
| *** nacim has quit IRC | 15:14 | |
| ihrachyshka | dhellmann: is that migrate library handled by oslo? | 15:14 |
| gordc | jd__: no. i believe in original discussion with dhellmann we wanted to keep notifier middleware separate so we didn't have to bring in oslo.messaging into middleware lib | 15:15 |
| jd__ | haha | 15:15 |
| jd__ | and dhellmann just commented that we may not want to put the notifier middleware in oslo.messaging either | 15:15 |
| jd__ | so WAT? | 15:16 |
| dhellmann | ihrachyshka: no, sqlalchemy-migrate isn't owned by oslo | 15:16 |
| gordc | jd__: yeah, we're going to have it as a lib by itself | 15:16 |
| ihrachyshka | dhellmann: but still, it's under stackforge (?) | 15:16 |
| jd__ | I don't really like that idea | 15:16 |
| jd__ | like we already have largely enough repository/packages | 15:17 |
| jd__ | :/ | 15:17 |
| dhellmann | jd__, gordc : yeah, the dependencies there were tricky, but we've been merging some of the other libs I thought we needed to keep together. If you can work out how to have them all in one place, that's fine, just write it up. :-) | 15:17 |
| dhellmann | ^keep together^keep apart | 15:17 |
| dhellmann | I didn't want oslo.messaging to require webob, and I wasn't sure about the middleware library requiring oslo.messaging. :-/ | 15:17 |
| gordc | jd__, dhellmann, is the webob package big? i would think it'd be best to move the notifier middleware into which lib that brings in smaller dependency | 15:19 |
| dhellmann | gordc: I would have to look at the dependency list, but you might be right. | 15:20 |
| jd__ | gordc: I'd say webob is lighter than oslo.messaging actually, wdyt dhellmann? | 15:30 |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo.db: oslo.db.exceptions module documentation https://review.openstack.org/93435 | 15:31 |
| gordc | jd__: yeah, that's what i think as well... if that's the case, i'm ok to with your initial patch for moving notifiermiddleware to oslo.messaging if everyone else is cool with it. | 15:32 |
| jd__ | the one I abandoned? /me cries | 15:33 |
| *** zzzeek has joined #openstack-oslo | 15:38 | |
| dhellmann | jd__: so oslo.messaging would depend on webob for that notifier? That might be OK. Are there any other dependencies? | 15:42 |
| jd__ | dhellmann: no | 15:43 |
| dhellmann | jd__: ? | 15:44 |
| jd__ | dhellmann: sorry, yes and no | 15:44 |
| jd__ | cf the patches :) | 15:44 |
| jd__ | https://review.openstack.org/#/q/status:open+project:openstack/oslo.messaging+branch:master+topic:bp/graduate-notifier-middleware,n,z | 15:44 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import notifier middleware from oslo-incubator https://review.openstack.org/105746 | 15:44 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator context module https://review.openstack.org/105745 | 15:44 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator/middleware/base https://review.openstack.org/105744 | 15:44 |
| *** ihrachyshka has quit IRC | 15:46 | |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo.db: Utility function for robust schema drop https://review.openstack.org/106419 | 15:47 |
| dhellmann | jd__: ok, I'll leave it up to you and gordc to work out the right home and write up the spec for it. I don't expect to have time to review any of that until next week (I'm out for the next 3 days) | 15:47 |
| jd__ | I don't want to write specs, I've already wrote the patches | 15:48 |
| *** HenryG_ has joined #openstack-oslo | 15:48 | |
| jd__ | I say screw that bureaucracy | 15:48 |
| jd__ | I'm starting #openstack-rebel | 15:48 |
| i159 | zzzeek: Hi. Have you seen my pull request to sa? | 15:50 |
| zzzeek | i159: have I ever ! | 15:51 |
| i159 | zzzeek: is it ok? | 15:51 |
| dhellmann | jd__: you can write it once, or every time someone asks why we have the dependencies set up that way :-/ | 15:51 |
| zzzeek | i159: i added lots of comments | 15:51 |
| zzzeek | i159: b.c. i know you like code reviews! :) | 15:52 |
| * zzzeek loves code review | 15:52 | |
| *** HenryG has quit IRC | 15:52 | |
| * dhellmann has adopted the mantra "write it down" | 15:52 | |
| *** HenryG_ is now known as HenryG_afk | 15:52 | |
| *** HenryG_afk is now known as HenryG | 15:52 | |
| jd__ | dhellmann: nobody's gonna dare ask me that :p | 15:52 |
| dhellmann | hahahahahahaha | 15:52 |
| jd__ | but yeah I'll write that at the bottom of my todo list | 15:52 |
| i159 | zzzeek: oh, good! yes, I really like it. It can teach me a tricks that are not in books =) | 15:54 |
| *** i159 has quit IRC | 15:55 | |
| openstackgerrit | gordon chung proposed a change to openstack/oslo-specs: graduate-oslo-middleware https://review.openstack.org/110353 | 16:01 |
| openstackgerrit | gordon chung proposed a change to openstack/oslo-specs: graduate-oslo-middleware https://review.openstack.org/110353 | 16:03 |
| gordc | jd__: hey... soo.... you writing up at that spec? :) | 16:05 |
| *** dims has joined #openstack-oslo | 16:08 | |
| *** SridharG has joined #openstack-oslo | 16:11 | |
| *** markmcclain has joined #openstack-oslo | 16:15 | |
| *** nacim has joined #openstack-oslo | 16:16 | |
| *** ildikov has joined #openstack-oslo | 16:19 | |
| gordc | zzzeek: when you have time, could you take a quick look at last comment: https://review.openstack.org/#/c/101009/ | 16:22 |
| zzzeek | gordc: ok | 16:22 |
| gordc | zzzeek: thanks! | 16:23 |
| zzzeek | gordc: oh is this back to what we talked about 2 months ago ? | 16:26 |
| dims | howdy folks | 16:28 |
| dims | dhellmann, ping | 16:28 |
| gordc | zzzeek: yep... kind of an undestatment, but i procrastinate... | 16:28 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 16:29 |
| *** gordc is now known as gordc_lunch | 16:32 | |
| *** yamahata has joined #openstack-oslo | 16:36 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 16:38 |
| dims | dhellmann, i don't seem to have privileges to push tags - http://paste.openstack.org/show/88956/ | 16:40 |
| *** Alexei_987 has quit IRC | 16:41 | |
| dims | dhellmann, looks like i don't have oslo-ptl karma. can you please add me if appropriate | 16:48 |
| *** dims has quit IRC | 16:49 | |
| *** dims has joined #openstack-oslo | 16:50 | |
| dims | dhellmann, looks like i don't have oslo-ptl karma. can you please add me if appropriate | 16:50 |
| *** pblaho_ has joined #openstack-oslo | 16:52 | |
| dhellmann | dims: done | 16:58 |
| dims | thanks, looks like it is in progress | 16:59 |
| *** tsekiyama has joined #openstack-oslo | 17:01 | |
| dhellmann | dims: woo! | 17:02 |
| dims | dhellmann, hmm, don't see oslo.utils on pypi | 17:04 |
| *** SridharG has quit IRC | 17:04 | |
| dhellmann | dims: did you not register the project? that has to be done by hand, first | 17:04 |
| dims | dhellmann, i "inherited" oslo.utils and assumed it was there already. will register | 17:05 |
| *** tsekiyama has quit IRC | 17:05 | |
| dhellmann | dims: ok, that may mean tagging again as 0.1.1 | 17:06 |
| dims | understood | 17:06 |
| *** praneshp has joined #openstack-oslo | 17:06 | |
| dims | dhellmann, dropping email to the -dev list | 17:23 |
| dhellmann | dims: cool, thanks | 17:24 |
| *** gmurphy has quit IRC | 17:26 | |
| *** gmurphy has joined #openstack-oslo | 17:29 | |
| *** harlowja_away is now known as harlowja | 17:36 | |
| *** gordc_lunch is now known as gordc | 17:48 | |
| *** tsekiyama has joined #openstack-oslo | 17:56 | |
| *** Alexei_987 has joined #openstack-oslo | 17:58 | |
| *** SridharG has joined #openstack-oslo | 18:04 | |
| *** markmcclain has quit IRC | 18:12 | |
| *** markmcclain has joined #openstack-oslo | 18:16 | |
| *** markmcclain1 has joined #openstack-oslo | 18:17 | |
| *** markmcclain has quit IRC | 18:21 | |
| harlowja | woot, oslo.utils, nice | 18:21 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 18:23 |
| *** noelbk has joined #openstack-oslo | 18:41 | |
| *** pabelanger has joined #openstack-oslo | 18:43 | |
| pabelanger | Are questions about oslo.messaging usage appropriate for this room? | 18:44 |
| gordc | zzzeek: just to confirm, you would recommend building tables similar to api models, even if it means more duplicate data... (and assuming we have some performance baseline to reference) | 18:46 |
| zzzeek | gordc: I’d not recommend duplicate data unless a clear need is demonstrated | 18:46 |
| zzzeek | gordc: id go for normalized first | 18:47 |
| zzzeek | e.g. user_id is on resource | 18:47 |
| gordc | zzzeek: ok, but our api model of 'Meter' also references a few resource attributes as well. so to build a meter we'd need meter->sample->resource | 18:49 |
| zzzeek | gordc: yeah, the prposed change you have in https://review.openstack.org/#/c/101009/8/specs/juno/big-data-sql-v2.rst w/ three tables looks OK to me | 18:49 |
| *** gmurphy has quit IRC | 18:50 | |
| gordc | zzzeek: ok. i'll give that a try. i wasn't sure how much we should try to avoid joins... the one thing i remembered was joins can be a problem. | 18:50 |
| zzzeek | gordc: they can, but subqueries are worse | 18:51 |
| zzzeek | gordc: starting with normalized is always a good idea | 18:51 |
| gordc | zzzeek: ah, got it. i'll give the proposal a try. | 18:51 |
| gordc | zzzeek: thanks for the review | 18:51 |
| zzzeek | gordc: with MySQL you sometimes can get away with pulling a small table into a list of ids, then doing a second query with IN, that’s a last resort kind of thing | 18:51 |
| zzzeek | gordc: joins aren’t terrible in MySQL if subqueries aren’t used | 18:52 |
| zzzeek | gordc: we just need to do an EXPLAIN and see where the bottlenecks are | 18:52 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement a dialect-level function dispatch system https://review.openstack.org/109436 | 18:52 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 18:52 |
| gordc | zzzeek: i see. i'll try building up proposed solution and see what EXPLAIN shows. | 18:52 |
| zzzeek | gordc: also INNER JOINs, vs. outer joins. avoid outer joins | 18:53 |
| gordc | zzzeek: yeah, i noticed that. i tried outer joins on another set of tables... that blew up. | 18:54 |
| *** pblaho_ is now known as pblaho | 18:56 | |
| *** gmurphy has joined #openstack-oslo | 18:56 | |
| dhellmann | pabelanger: yes | 19:12 |
| pabelanger | dhellmann, Thanks. I think I figured out my question for the moment | 19:15 |
| dhellmann | pabelanger: even better :-) | 19:16 |
| Alexei_987 | noelbk: Hi you can ping me here if needed | 19:31 |
| *** lbragsta_ has joined #openstack-oslo | 19:31 | |
| Alexei_987 | noelbk: expiry fix is definetely better in terms of HA deployment since it doesn't cause any errors on the sender side so maybe we should combine both approaches | 19:32 |
| *** lbragst__ has joined #openstack-oslo | 19:32 | |
| *** lbragst__ is now known as lbragstad_ | 19:33 | |
| *** lbragsta_ has quit IRC | 19:33 | |
| noelbk | Alexei_987: Yes, the expiry is good too for other cases where the receiver dies. Let's do both. My remaining task in mine is to get the server to time out if the client fails to recreate the reply queue | 19:34 |
| Alexei_987 | noelbk: +1 sounds like a good plan | 19:34 |
| *** lbragstad has quit IRC | 19:34 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement a dialect-level function dispatch system https://review.openstack.org/109436 | 19:35 |
| noelbk | Alexei_987: Could you make your expiry timeout a config variable? | 19:36 |
| noelbk | Alexei_987: I'm currently looking for how to wire a config variable in to the server reply timeout | 19:37 |
| Alexei_987 | noelbk: Yeah you can mention it in patch. and we'll update it | 19:37 |
| noelbk | Alexei_987: I did mention the config variable in my comment to the expiry patch... did you mean mention it somewhere else? | 19:38 |
| Alexei_987 | noelbk: oh sorry.. must have missed it | 19:38 |
| Alexei_987 | noelbk: we'll update the patch | 19:38 |
| noelbk | Alexei_987: also the unit test, although my own test seems a little hackish | 19:38 |
| Alexei_987 | noelbk: IMHO unit tests are useless for rabbit driver | 19:39 |
| Alexei_987 | noelbk: we don't actually test anything | 19:39 |
| Alexei_987 | noelbk: I'm working on functional tests | 19:39 |
| noelbk | Alexei_987: but making sure the right parameter gets passed to kombu is important | 19:39 |
| noelbk | Alexei_987: and protects against someone else breaking it later | 19:39 |
| Alexei_987 | noelbk: IMHO it's much more important to test functionality | 19:40 |
| Alexei_987 | noelbk: and not each parameter | 19:40 |
| Alexei_987 | noelbk: functional tests for failover will cover that | 19:40 |
| Alexei_987 | noelbk: we are using inmemory transport for kombu that is different from real rabbitmq | 19:41 |
| Alexei_987 | noelbk: so unit test can pass and driver will still be broken | 19:41 |
| *** gordc has quit IRC | 19:41 | |
| noelbk | Alexei_987: point taken. I prefer to add the unit test as well as functional. Even though that doesn't really confirm that rabbit will behave the right way, it prevents someone from breaking things in a future pacth | 19:43 |
| Alexei_987 | noelbk: Ok :) | 19:43 |
| *** dims has quit IRC | 19:47 | |
| *** openstackgerrit has quit IRC | 19:48 | |
| dhellmann | dims: https://review.openstack.org/#/c/107690/ takes care of adding oslo.utils to the governance repo for you | 19:50 |
| noelbk | Alexei_987: so, the timeout argument in publisher_send is pretty mu unused. There's no timeout argument to ensure() | 19:51 |
| *** lbragstad_ has quit IRC | 19:56 | |
| *** lbragstad has joined #openstack-oslo | 19:56 | |
| *** arnaud has joined #openstack-oslo | 19:56 | |
| *** lbragstad has quit IRC | 19:59 | |
| *** lbragstad has joined #openstack-oslo | 19:59 | |
| *** gordc has joined #openstack-oslo | 19:59 | |
| *** markmcclain1 has quit IRC | 19:59 | |
| *** markmcclain has joined #openstack-oslo | 20:00 | |
| *** dims has joined #openstack-oslo | 20:19 | |
| *** pblaho has quit IRC | 20:46 | |
| Alexei_987 | zzzeek: Hi are you available? | 20:46 |
| *** lbragstad has quit IRC | 20:48 | |
| *** lbragstad has joined #openstack-oslo | 20:49 | |
| *** lbragstad has quit IRC | 20:53 | |
| *** openstackgerrit has joined #openstack-oslo | 20:58 | |
| zzzeek | Alexei_987: whats up | 20:58 |
| Alexei_987 | zzzeek: wanted to ask you about dogpile.cache | 21:02 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 21:02 |
| zzzeek | Alexei_987: yah | 21:02 |
| Alexei_987 | zzzeek: we need to set socket timeout somehow | 21:02 |
| Alexei_987 | for all cache operations | 21:02 |
| zzzeek | for what backend | 21:02 |
| Alexei_987 | zzzeek: and it seems that it depends on backend | 21:02 |
| Alexei_987 | zzzeek: that's the question | 21:03 |
| zzzeek | Alexei_987: that is true | 21:03 |
| zzzeek | Alexei_987: not all backends have a timeout | 21:03 |
| zzzeek | Alexei_987: for example, dbm, memory | 21:03 |
| Alexei_987 | zzzeek: ah.. I see so we need to do fix backend specific | 21:03 |
| Alexei_987 | zzzeek: and ensure we use proper backend | 21:03 |
| Alexei_987 | zzzeek: we are using memcached | 21:04 |
| Alexei_987 | zzzeek: but it also has 3 different connectors | 21:04 |
| zzzeek | Alexei_987: OK, so not redis, and a single memcached python lib ? | 21:04 |
| *** lbragstad has joined #openstack-oslo | 21:04 | |
| Alexei_987 | zzzeek: and only 1 supports timeout | 21:04 |
| zzzeek | Alexei_987: well all but pylibmc suck | 21:04 |
| *** haypo has left #openstack-oslo | 21:04 | |
| zzzeek | Alexei_987: id recommend stick with one backend, assuming its pylibmc :) | 21:05 |
| Alexei_987 | zzzeek: from what I see it doesn't support timeout | 21:06 |
| Alexei_987 | zzzeek: but maybe I'm wrong | 21:06 |
| zzzeek | Alexei_987: http://sendapatch.se/projects/pylibmc/behaviors.html | 21:06 |
| *** lbragstad has quit IRC | 21:07 | |
| Alexei_987 | zzzeek: ah true | 21:07 |
| zzzeek | Alexei_987: connect_timeout, receive_timeout, send_timeout, timeouts galore | 21:07 |
| Alexei_987 | zzzeek: I was looking in the wrong file :( | 21:07 |
| Alexei_987 | zzzeek: thanks | 21:07 |
| *** lbragstad has joined #openstack-oslo | 21:07 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Implement a dialect-level function dispatch system https://review.openstack.org/109436 | 21:08 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely. https://review.openstack.org/110170 | 21:09 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use dialect dispatch for engine initiailization. https://review.openstack.org/110446 | 21:10 |
| *** flaper87 is now known as flaper87|afk | 21:12 | |
| *** SridharG has quit IRC | 21:18 | |
| *** gordc has quit IRC | 21:49 | |
| *** lbragsta_ has joined #openstack-oslo | 21:51 | |
| *** lbragstad has quit IRC | 21:54 | |
| *** lbragsta_ has quit IRC | 21:55 | |
| *** HenryG is now known as HenryG_afk | 22:18 | |
| *** jaosorior has quit IRC | 22:22 | |
| *** jecarey has quit IRC | 22:25 | |
| *** bknudson has quit IRC | 22:30 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use testr instance provisioning to lazily create databases https://review.openstack.org/110486 | 22:36 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use testr instance provisioning to lazily create databases https://review.openstack.org/110486 | 22:38 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely https://review.openstack.org/110170 | 22:39 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use testr instance provisioning to lazily create databases https://review.openstack.org/110486 | 22:39 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use dialect dispatch for engine initiailization. https://review.openstack.org/110446 | 22:42 |
| *** tsekiyam_ has joined #openstack-oslo | 22:43 | |
| *** tsekiyama has quit IRC | 22:47 | |
| *** tsekiyam_ has quit IRC | 22:48 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/pycadf: revise readme with a project description https://review.openstack.org/110490 | 22:51 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/pycadf: add CONTRIBUTING doc https://review.openstack.org/110491 | 22:53 |
| *** arnaud has quit IRC | 23:04 | |
| *** alexpilotti has quit IRC | 23:05 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/pycadf: clean up license headers https://review.openstack.org/110498 | 23:09 |
| *** dims has quit IRC | 23:14 | |
| *** yamahata_ has quit IRC | 23:15 | |
| *** dims has joined #openstack-oslo | 23:16 | |
| *** dims is now known as Guest13918 | 23:16 | |
| *** Guest13918 is now known as dims_ | 23:17 | |
| *** dims_ has left #openstack-oslo | 23:19 | |
| *** bknudson has joined #openstack-oslo | 23:20 | |
| *** dims_ has joined #openstack-oslo | 23:20 | |
| *** bknudson has quit IRC | 23:25 | |
| gus | zzzeek: mysqlconnector doesn't accept execute("sql stmt1; sql stmt 2"), and the sqlalchemy docs are unclear on whether that should be allowed or not | 23:30 |
| zzzeek | there was discusson of this somewhere, oh in an email. there’s a silly flag the DBAPI wants to allow that. | 23:31 |
| gus | Should mysqlconnector use cursor.execute(multi=True), or should I track down and fix callers issuing multiple statements at once? | 23:31 |
| zzzeek | Id go for the latter | 23:31 |
| gus | the latter? | 23:31 |
| zzzeek | mutliple statemnts in execute() is pretty amateur | 23:31 |
| gus | ack. | 23:31 |
| zzzeek | :) | 23:31 |
| gus | I'll go on a semicolon hunt :P | 23:31 |
| zzzeek | we have to do it in *one* case, in the pyodbc SQL server driver | 23:31 |
| gus | oh wait, what? | 23:31 |
| gus | I found this in the context of mysqldb->mysqlconnector. | 23:32 |
| zzzeek | thats SQLAclhemy. openstack deosnt work on SQL Server :) | 23:32 |
| zzzeek | yeah i know | 23:32 |
| gus | so mysqldb (and presumably postgres and anywhere else we've ever deployed keystone) apparently supports multiple statements. | 23:33 |
| gus | mysqlconnector does not. | 23:33 |
| gus | (and possibly other non-default driver choices) | 23:33 |
| gus | so where does pyodbc come into this? | 23:34 |
| *** bknudson has joined #openstack-oslo | 23:34 | |
| gus | https://github.com/openstack/keystone/blob/master/keystone/common/sql/migrate_repo/versions/036_havana.py | 23:36 |
| gus | ^ is the example of multiple statements I ran across first. | 23:36 |
| zzzeek | pyodbc comes into this b.c. wihtin all of sqlalchemy it is the only place where a “;” is unavoidable in one particular case | 23:37 |
| gus | aha. | 23:37 |
| gus | .. so does that mean mysqlconnector needs to be fixed to always use multi=True anyway? | 23:38 |
| zzzeek | no, why? | 23:40 |
| zzzeek | mysqlconnevtor has nothing to do with pyodbc | 23:40 |
| gus | oh ok - then I don't know how pyodbc and sqlalchemy interact (if at all) ;) | 23:41 |
| zzzeek | its one of many DBAPIs we support | 23:41 |
| zzzeek | in sqlalchemy | 23:41 |
| gus | oh right, as a backend driver - yeah that has nothing to do with mysqlconnector then. | 23:42 |
| *** nacim has quit IRC | 23:42 | |
| gus | .. so sqla's execute() only supports a single statement and I should fix any callers who try otherwise. should be easy enough to exercise the migration paths with a fresh devstack install. | 23:43 |
| *** HenryG_afk is now known as HenryG | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!