| *** mriedem has quit IRC | 00:06 | |
| *** oomichi has joined #openstack-oslo | 00:47 | |
| *** harlowja is now known as harlowja_away | 00:58 | |
| *** shakayumi is now known as shakamunyi | 00:59 | |
| *** arnaud__ has joined #openstack-oslo | 01:20 | |
| *** morganfainberg is now known as morganfainberg_Z | 01:35 | |
| *** arnaud has quit IRC | 01:46 | |
| *** arnaud__ has quit IRC | 01:46 | |
| *** shakayumi has joined #openstack-oslo | 01:46 | |
| *** shakamunyi has quit IRC | 01:49 | |
| *** morganfainberg_Z is now known as morganfainberg | 02:37 | |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Raise an error if preversion versions are too low https://review.openstack.org/108269 | 02:47 |
|---|---|---|
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Look for and process sem-ver pseudo headers in git https://review.openstack.org/108270 | 02:47 |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Refactor oneline changelog walking. https://review.openstack.org/108271 | 02:47 |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Only consider tags that look like versions. https://review.openstack.org/114403 | 02:47 |
| *** stevemar has joined #openstack-oslo | 03:14 | |
| *** flashgordon is now known as jogo | 03:19 | |
| *** amotoki_ has joined #openstack-oslo | 03:51 | |
| *** amotoki_ has quit IRC | 03:52 | |
| *** Krast has joined #openstack-oslo | 03:56 | |
| *** stevemar has quit IRC | 04:44 | |
| *** amotoki_ has joined #openstack-oslo | 05:12 | |
| *** SridharG has joined #openstack-oslo | 05:27 | |
| *** SridharG has quit IRC | 05:33 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/oslo.vmware: Imported Translations from Transifex https://review.openstack.org/114732 | 06:08 |
| *** SridharG has joined #openstack-oslo | 06:39 | |
| *** AAzza_afk is now known as AAzza | 06:54 | |
| *** noelbk has joined #openstack-oslo | 06:54 | |
| *** ildikov has joined #openstack-oslo | 07:22 | |
| *** mrda is now known as mrda-away | 07:25 | |
| *** ihrachyshka has joined #openstack-oslo | 07:29 | |
| *** ihrachyshka_ has joined #openstack-oslo | 07:30 | |
| *** ihrachyshka has quit IRC | 07:34 | |
| *** viktors has joined #openstack-oslo | 07:39 | |
| *** AAzza is now known as AAzza_afk | 07:56 | |
| *** linkid has joined #openstack-oslo | 08:27 | |
| *** alexpilotti has joined #openstack-oslo | 08:55 | |
| *** pblaho has joined #openstack-oslo | 09:41 | |
| *** pblaho is now known as pblaho|afk | 09:59 | |
| *** pcm_ has joined #openstack-oslo | 10:04 | |
| *** pcm__ has joined #openstack-oslo | 10:07 | |
| *** pcm_ has quit IRC | 10:10 | |
| *** yamahata has joined #openstack-oslo | 10:16 | |
| *** yamahata has quit IRC | 10:21 | |
| *** pblaho|afk is now known as pblaho | 10:47 | |
| openstackgerrit | A change was merged to openstack/oslo.vmware: Add support for the DatastoreURL object https://review.openstack.org/114710 | 10:55 |
| *** sheeprine has quit IRC | 11:16 | |
| *** pblaho_ has joined #openstack-oslo | 11:16 | |
| *** pblaho has quit IRC | 11:20 | |
| *** sheeprine has joined #openstack-oslo | 11:29 | |
| *** sheeprine has joined #openstack-oslo | 11:29 | |
| *** Alexei_987 has joined #openstack-oslo | 11:44 | |
| *** gordc has joined #openstack-oslo | 12:08 | |
| *** pblaho_ is now known as pblaho|afk | 12:25 | |
| *** amotoki_ has quit IRC | 12:31 | |
| *** pblaho|afk is now known as pblaho_ | 12:37 | |
| *** AAzza_afk is now known as AAzza | 12:38 | |
| *** pblaho_ is now known as pblaho | 12:46 | |
| *** bknudson1 has quit IRC | 13:08 | |
| *** jeckersb_gone is now known as jeckersb | 13:13 | |
| *** pblaho_ has joined #openstack-oslo | 13:28 | |
| *** jecarey has joined #openstack-oslo | 13:29 | |
| *** pblaho has quit IRC | 13:31 | |
| *** dims has joined #openstack-oslo | 13:34 | |
| *** markmc has joined #openstack-oslo | 13:40 | |
| *** miqui has joined #openstack-oslo | 13:45 | |
| *** yamahata has joined #openstack-oslo | 13:45 | |
| *** jeckersb is now known as jeckersb_gone | 13:51 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 13:53 |
| *** jeckersb_gone is now known as jeckersb | 13:56 | |
| *** ildikov has quit IRC | 14:00 | |
| *** Krast_ has joined #openstack-oslo | 14:04 | |
| *** yamahata_ has joined #openstack-oslo | 14:04 | |
| *** miqui_ has joined #openstack-oslo | 14:04 | |
| *** zzzeek has joined #openstack-oslo | 14:06 | |
| *** hyakuhei_ has joined #openstack-oslo | 14:06 | |
| *** miqui has quit IRC | 14:11 | |
| *** markmc has quit IRC | 14:11 | |
| *** viktors has quit IRC | 14:11 | |
| *** Krast has quit IRC | 14:11 | |
| *** hyakuhei has quit IRC | 14:11 | |
| *** yamahata__ has quit IRC | 14:11 | |
| *** pblaho__ has joined #openstack-oslo | 14:12 | |
| *** pblaho__ is now known as pblaho | 14:12 | |
| *** yamahata has quit IRC | 14:13 | |
| *** yamahata has joined #openstack-oslo | 14:14 | |
| *** beekneemech is now known as bnemec | 14:15 | |
| *** pblaho_ has quit IRC | 14:16 | |
| *** oomichi has quit IRC | 14:16 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Reorganize DbTestCase to use provisioning completely https://review.openstack.org/110170 | 14:17 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Use testr instance provisioning to lazily create databases https://review.openstack.org/110486 | 14:17 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Add transactional test support to provisioning https://review.openstack.org/113153 | 14:17 |
| *** mriedem has joined #openstack-oslo | 14:17 | |
| *** viktors has joined #openstack-oslo | 14:18 | |
| *** markmc has joined #openstack-oslo | 14:18 | |
| zzzeek | hey viktors ! | 14:19 |
| viktors | zzzeek: hi! | 14:19 |
| zzzeek | how was vaca | 14:19 |
| *** SridharG has quit IRC | 14:20 | |
| viktors | zzzeek: vacation is always nice! :) | 14:20 |
| zzzeek | so ill be coming on to core review to help out | 14:20 |
| *** miqui_ is now known as miqui | 14:21 | |
| zzzeek | but until then I have a whole line of +2’s I need :) https://review.openstack.org/#/q/owner:%22Michael+Bayer%22+status:open,n,z | 14:21 |
| *** jecarey has quit IRC | 14:22 | |
| *** jecarey has joined #openstack-oslo | 14:22 | |
| viktors | zzzeek: I hope to look at oslo.db's patches tomorrow with the fresh head - today is my first working day ) | 14:22 |
| viktors | btw, dhellmann, should we wait for someone's vote, before we will add zzzeek to the core team? | 14:23 |
| zzzeek | viktors: OK. | 14:23 |
| viktors | zzzeek: by the way, I have a question about patch https://review.openstack.org/#/c/110486/ | 14:26 |
| zzzeek | yup | 14:27 |
| *** jecarey has quit IRC | 14:27 | |
| viktors | zzzeek: I think, that this is a bad idea - to depends on testr (or some another test runner) | 14:27 |
| viktors | zzzeek: have you tried to use something like this - https://docs.python.org/2/library/atexit.html ? | 14:28 |
| zzzeek | viktors: OK well there is also another change that will be to use scenarios, and for that we have to use testscenarios which is also part of the testr world | 14:28 |
| zzzeek | viktors: atexit is very unpredictable in my experience | 14:28 |
| zzzeek | viktors: all test systems support some form of “dispose after all tests” idea - so to support other runners, we can implement for those as well | 14:29 |
| zzzeek | viktors: here’s how SQLAlchemy does it for both nose and py.test: https://bitbucket.org/zzzeek/sqlalchemy/src/530d3f07e0c1e70e0f9b80d3b5986253e06dcaf2/lib/sqlalchemy/testing/plugin/?at=master | 14:30 |
| * viktors looking | 14:30 | |
| zzzeek | viktors: maybe there’s a way to make tox do it ? we are comfortable depending on tox right ? | 14:31 |
| viktors | zzzeek: well, sure we can do it via tox, but I'm not really sure about it | 14:32 |
| zzzeek | viktors: also what was proposed was that tests run via shell scripts | 14:32 |
| zzzeek | viktors: which i dont really like. it means we have to add a shell script to all projects and change how they all run tests | 14:32 |
| *** yamahata has quit IRC | 14:32 | |
| zzzeek | viktors: to me it seems like a failure of the test runners | 14:33 |
| viktors | zzzeek: my idea is - we should be able to run any test using any test runner | 14:33 |
| viktors | inside or outside of the virtual environment | 14:34 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 14:34 |
| zzzeek | viktors: there has to be a test collector of some kind in order to run tests | 14:34 |
| zzzeek | viktors: even if you use straight unittest, you need to call unittest.main() | 14:34 |
| viktors | zzzeek: thats true | 14:35 |
| zzzeek | viktors: so you need to solve the collection issue, and IMHO the setup/teardown bording that is part of that same issue | 14:36 |
| zzzeek | viktors: so IMHO the approach we use for resource setup/teardown should be integreated with the collector being used. | 14:36 |
| zzzeek | viktors: keep in mind the patch here in no way prevents tests from being run with any runner | 14:37 |
| zzzeek | viktors: I usually run with py.test, works great | 14:37 |
| zzzeek | viktors: this *only* applies to parallel test runs, which we only support with testr right now | 14:37 |
| viktors | zzzeek: yes, I agree that testr is default test runner, so we should use it's features. | 14:39 |
| viktors | zzzeek: I just want to be sure, that if I'll want to set pdb in a single test, I'll be able to do it without a lot of magic :) | 14:40 |
| viktors | for example | 14:40 |
| zzzeek | viktors: nothing is changing in that regard | 14:40 |
| zzzeek | viktors: using pdb with testrunner is already a pile of crap, nothing is changing there, I use py.test if im pdbing | 14:41 |
| viktors | zzzeek: ok, I should check it locally | 14:41 |
| zzzeek | viktors: all of these patches only alter the mechanics of when the “anonymous” database is created/dropped | 14:42 |
| *** tsekiyama has joined #openstack-oslo | 14:42 | |
| zzzeek | to be more efficient | 14:42 |
| viktors | zzzeek: ok | 14:43 |
| * viktors forgot about everything during vacation :) | 14:44 | |
| viktors | zzzeek: one more thing - have you some experience with db query retrying? | 14:46 |
| zzzeek | yup | 14:46 |
| *** markmcclain has joined #openstack-oslo | 14:49 | |
| viktors | zzzeek: can you provide some info about it? We just made a similar feature some time ago, but I'm not sure, that it in use | 14:50 |
| zzzeek | viktors: query retrying can become complex if we are talking about a connection that was dropped | 14:50 |
| zzzeek | viktors: because the transaction is gone | 14:50 |
| zzzeek | viktors: so I prefer to look at it as “transaction retrying" | 14:50 |
| zzzeek | viktors: and in the context of openstack and its web API methods, i think the simplest way to “retry” is just run the whoel web API method again in case of a dropped connection | 14:51 |
| *** rustlebee is now known as russellb | 14:52 | |
| *** bknudson has joined #openstack-oslo | 14:54 | |
| viktors | zzzeek: yes, we did it in a same way :) | 15:00 |
| *** gpocente1 is now known as gpocentek | 15:07 | |
| *** gpocentek has quit IRC | 15:07 | |
| *** gpocentek has joined #openstack-oslo | 15:07 | |
| zzzeek | viktors: there is a “retry the whole transcation” system I wrote, though i dont know that openstack needs it. you can see that here: https://bitbucket.org/zzzeek/sqlalchemy/issue/3104/transaction-replay-extension | 15:10 |
| *** jecarey has joined #openstack-oslo | 15:10 | |
| *** stevemar has joined #openstack-oslo | 15:11 | |
| *** ihrachyshka_ has quit IRC | 15:12 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.concurrency: lockutils: catch BusyError when locking with timeout https://review.openstack.org/114993 | 15:13 |
| viktors | zzzeek: well, we made it in a similar way - https://github.com/openstack/oslo.db/blob/master/oslo/db/api.py#L62-L120 | 15:14 |
| zzzeek | yes that patch is appropriate for wrapping an API method | 15:14 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Port to Python 3 https://review.openstack.org/108007 | 15:14 |
| zzzeek | the “transaciton replay” thing is actually much more elaborate and risky, it is recording all statements and replaying selected ones | 15:15 |
| zzzeek | viktors: and its 100% transparent too … :) | 15:15 |
| zzzeek | viktors: wrote it for a client. its nuts | 15:15 |
| viktors | zzzeek: :) | 15:16 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import notifier middleware from oslo-incubator https://review.openstack.org/105746 | 15:18 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator context module https://review.openstack.org/105745 | 15:18 |
| openstackgerrit | Julien Danjou proposed a change to openstack/oslo.messaging: Import oslo-incubator/middleware/base https://review.openstack.org/105744 | 15:18 |
| *** alexpilotti has quit IRC | 15:20 | |
| *** markmcclain has quit IRC | 15:35 | |
| *** ildikov has joined #openstack-oslo | 15:45 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 15:48 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 15:56 |
| *** bnemec has quit IRC | 15:57 | |
| amrith | hello oslo-core ... would you please review https://review.openstack.org/#/c/113407/, https://review.openstack.org/#/c/114656/, https://review.openstack.org/#/c/114659/, https://review.openstack.org/#/c/114664/ | 15:59 |
| *** AAzza is now known as AAzza_afk | 16:11 | |
| openstackgerrit | Bogdan Dobrelya proposed a change to openstack/oslo.messaging: Should not send replies for cast messages https://review.openstack.org/115010 | 16:14 |
| dhellmann | amrith: several of those failed their last jenkins run | 16:15 |
| dhellmann | amrith: and you need https://review.openstack.org/#/c/114614/ on that list :-) | 16:17 |
| openstackgerrit | Bogdan Dobrelya proposed a change to openstack/oslo.messaging: Should not send replies for cast messages https://review.openstack.org/115010 | 16:18 |
| *** viktors is now known as viktors|afk | 16:23 | |
| YorikSar | dhellmann: They're failing because of https://review.openstack.org/114647 | 16:29 |
| dhellmann | YorikSar: ah, ok | 16:30 |
| YorikSar | dhellmann: I wonder if it's ok if I approve it or should I wait for someone else?.. | 16:30 |
| dhellmann | YorikSar: +2a | 16:32 |
| YorikSar | dhellmann: Great :) | 16:33 |
| *** praneshp has joined #openstack-oslo | 16:33 | |
| dims | dhellmann: i need a oslo.utils release with mask_password to jump start the olso/nova integration (which also means updating global reqs) | 16:34 |
| YorikSar | dhellmann: But for the future - is it ok to not wait for second core to approve such changes? | 16:34 |
| dhellmann | dims: can we merge the mask_password fix first? | 16:36 |
| dims | dhellmann: yep | 16:36 |
| dhellmann | YorikSar: we tend to look at it case-by-case; it's better to ask for reviews here | 16:36 |
| dims | "with mask_password" :) | 16:36 |
| dhellmann | dims: there were 2 patches, right? add the function and enhance its security? | 16:36 |
| dims | digging up one sec. yes, there were 2 | 16:37 |
| dims | just one left - https://review.openstack.org/#/c/114614/ | 16:38 |
| dims | any nova cores around? (116414 has dhellmann's +2 already) | 16:39 |
| dims | oops oslo cores :) | 16:39 |
| dhellmann | :-) | 16:40 |
| dhellmann | time for lunch, bbiab | 16:40 |
| dims | you can see where my brain is wired to ask for reviews :) | 16:40 |
| *** zzzeek has quit IRC | 16:42 | |
| *** harlowja_away is now known as harlowja | 16:45 | |
| *** zzzeek has joined #openstack-oslo | 16:50 | |
| *** ildikov has quit IRC | 16:53 | |
| *** bnemec has joined #openstack-oslo | 17:04 | |
| *** markmcclain has joined #openstack-oslo | 17:18 | |
| *** markmcclain has quit IRC | 17:18 | |
| *** markmcclain has joined #openstack-oslo | 17:19 | |
| *** Alexei_987 has quit IRC | 17:33 | |
| *** harlowja is now known as harlowja_away | 17:46 | |
| *** SridharG has joined #openstack-oslo | 17:49 | |
| *** bnemec has quit IRC | 17:53 | |
| *** bnemec has joined #openstack-oslo | 17:56 | |
| *** alexpilotti has joined #openstack-oslo | 18:02 | |
| *** zzzeek has quit IRC | 18:03 | |
| *** zzzeek has joined #openstack-oslo | 18:05 | |
| *** arnaud has joined #openstack-oslo | 18:19 | |
| *** harlowja_away is now known as harlowja | 18:26 | |
| *** SridharG has quit IRC | 18:28 | |
| *** Alexei_987 has joined #openstack-oslo | 18:34 | |
| morganfainberg | so are we maintaining the incubator for a given release (e.g. stable/XXX) with the expectation that projects will sync on the stable tree? | 18:38 |
| morganfainberg | for their matching stable tree that is? | 18:38 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 18:39 |
| dhellmann | morganfainberg: yes, in case we find critical bugs in pieces that are still in the incubator | 18:40 |
| morganfainberg | dhellmann, cool. | 18:41 |
| morganfainberg | dhellmann, just checking :) | 18:41 |
| dims | dhellmann: need some stevedore help...getting this exception - http://paste.openstack.org/show/96964/ | 18:42 |
| dims | when i enable lines 167-171 in https://review.openstack.org/#/c/114551/6/oslo/vmware/selector.py,unified | 18:43 |
| dims | please take a look and let me know. brb | 18:43 |
| dhellmann | dims: it looks like the "policy" variable being passed in from the test is None | 18:43 |
| dhellmann | dims: that needs to be a name of a valid plugin from the 'selector.drivers' namespace | 18:44 |
| dhellmann | dims: I'm not sure that's a very good namespace, fwiw, it doesn't say anything about openstack or oslo or vmware in the name | 18:44 |
| dims | ah duh. thanks! i had set the default value to largest and forgot :) | 18:44 |
| *** pblaho has quit IRC | 18:44 | |
| dims | dhellmann: y, just toying with it i had it with oslo.vmware earlier | 18:44 |
| dhellmann | dims: ah | 18:45 |
| dims | whew! thanks | 18:45 |
| dhellmann | dims: https://pypi.python.org/pypi/entry_point_inspector | 18:45 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 18:47 |
| *** mrda-away is now known as mrda | 18:56 | |
| *** jecarey has quit IRC | 19:04 | |
| *** noelbk has quit IRC | 19:07 | |
| *** arnaud has quit IRC | 19:09 | |
| *** ildikov has joined #openstack-oslo | 19:16 | |
| dims | dhellmann: thanks | 19:26 |
| *** pblaho has joined #openstack-oslo | 19:28 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: port ds_util from nova to oslo.vmware https://review.openstack.org/114551 | 19:30 |
| *** jecarey has joined #openstack-oslo | 19:39 | |
| openstackgerrit | A change was merged to openstack/oslo.db: Handle DB2 SmallInteger type for change_deleted_column_type_to_boolean https://review.openstack.org/112030 | 19:42 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: VMWare: Support for pluggable Datastore selection https://review.openstack.org/114551 | 19:43 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: VMWare: Support for pluggable Datastore selection https://review.openstack.org/114551 | 19:44 |
| *** kgiusti has joined #openstack-oslo | 19:47 | |
| *** shakayumi has quit IRC | 19:53 | |
| dims | bnemec: ping | 19:55 |
| *** tsekiyam_ has joined #openstack-oslo | 19:55 | |
| *** gpocente1 has joined #openstack-oslo | 19:57 | |
| *** AAzza_afk is now known as AAzza | 19:58 | |
| *** kgiusti has quit IRC | 19:58 | |
| *** tsekiyama has quit IRC | 19:58 | |
| *** viktors|afk has quit IRC | 19:58 | |
| *** mriedem has quit IRC | 19:58 | |
| *** gpocentek has quit IRC | 19:58 | |
| *** mriedem has joined #openstack-oslo | 19:59 | |
| *** viktors has joined #openstack-oslo | 19:59 | |
| *** kgiusti has joined #openstack-oslo | 19:59 | |
| *** pblaho has quit IRC | 20:00 | |
| *** pcm__ has quit IRC | 20:01 | |
| *** pcm_ has joined #openstack-oslo | 20:02 | |
| *** noelbk has joined #openstack-oslo | 20:04 | |
| dims | dhellmann: we'll need a version of oslo.i18n with the test fixtures for use with oslo.serialization (https://review.openstack.org/#/c/113347/ is stuck because of that) | 20:15 |
| *** linkid has left #openstack-oslo | 20:17 | |
| openstackgerrit | A change was merged to openstack-dev/oslo-cookiecutter: Add common fixes for docs generation to template https://review.openstack.org/114667 | 20:25 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo-incubator: Fix MemcachedKeyCharacterError when keys have spaces https://review.openstack.org/113537 | 20:28 |
| dhellmann | markmc, flaper87|afk, jd__, dims, bnemec : can you take a look at the osprofiler spec? it seems to have a lot of support outside of oslo, but I'd like some feedback from inside, too: https://review.openstack.org/#/c/105096/ | 20:34 |
| jd__ | I suggest a mail so I don't forget because it's late here | 20:35 |
| dhellmann | dims: so we just need to cut a release? | 20:35 |
| dhellmann | jd__: sure | 20:35 |
| jd__ | thanks :) | 20:35 |
| dims | dhellmann: yes please | 20:35 |
| dhellmann | dims: ok | 20:36 |
| dims | jd__: can you please look at https://review.openstack.org/#/c/114614/ - need that to get oslo.utils into nova etc | 20:36 |
| *** kevinbenton has joined #openstack-oslo | 20:40 | |
| *** jeckersb is now known as jeckersb_gone | 20:40 | |
| kevinbenton | dhellmann: ping regarding a neuron CLI fix i’m working on. markmcclain sent me your way | 20:41 |
| kevinbenton | dhellmann: (sorry if you got that twice, first contacted you in infra) | 20:41 |
| dims | dhellmann: review #? (curious) | 20:42 |
| dhellmann | hi, kevinbenton | 20:43 |
| dhellmann | dims: I'm not sure what you mean? | 20:43 |
| dims | oops kevinbenton: review #? :) | 20:44 |
| dhellmann | dims: heh, ok, me, too :-) | 20:44 |
| dims | :) | 20:44 |
| kevinbenton | dhellmann: hi, we ran into an issue where the —timeout parameter was used by both the top-level neutron command and one of the subcommands. argparse would always eatup the timeout parameter for the top-level command | 20:44 |
| kevinbenton | https://review.openstack.org/#/c/114005/ | 20:44 |
| kevinbenton | dhellmann, dims: ^^ | 20:44 |
| kevinbenton | so that fixes it, but at the cost of not being able to specify any params to the top-level neutron command after the subcommand | 20:45 |
| dhellmann | kevinbenton: so the app value and the subcommand value use the same option name but mean different things? | 20:45 |
| kevinbenton | dhellmann: yes, load balancer creating a watcher with the timeout param is referring to the object | 20:45 |
| kevinbenton | dhellmann: top-level timeout is just for the command timeout | 20:45 |
| dhellmann | kevinbenton: frankly, I would change one of the option names to avoid confusion on the part of the user | 20:46 |
| dhellmann | kevinbenton: if the subcommand option isn't being honored now, then nothing can be relying on it and it's safe to change that one to something else (--watcher-timeout ?) | 20:46 |
| kevinbenton | dhellmann: actually the top-level one just merged and broke the subcommand | 20:47 |
| kevinbenton | dhellmann: so i’ll rename the top-level one | 20:47 |
| dhellmann | kevinbenton: ah, well, if you haven't released, then I'd roll that back | 20:47 |
| kevinbenton | dhellmann: we also revealed a hole in the unit tests :-) | 20:47 |
| dhellmann | kevinbenton: sounds like it | 20:48 |
| kevinbenton | dhellmann: do you think this patch is still worth looking into or should we even support duplication of option names? | 20:48 |
| dhellmann | kevinbenton: duplicating option names feels like a bad idea, since the user then has to be careful about how they order their command line | 20:49 |
| kevinbenton | dhellmann: ok | 20:51 |
| kevinbenton | dhellmann: thanks for your help | 20:51 |
| dhellmann | kevinbenton: sure thing | 20:51 |
| *** kgiusti has left #openstack-oslo | 20:53 | |
| *** AAzza is now known as AAzza_afk | 20:54 | |
| *** arnaud has joined #openstack-oslo | 20:55 | |
| *** miqui has quit IRC | 20:55 | |
| Alexei_987 | dhellmann: Hi could you please revisit this spec when you'll have time? https://review.openstack.org/#/c/105796/ | 20:57 |
| *** ihrachyshka has joined #openstack-oslo | 21:02 | |
| *** bknudson has quit IRC | 21:03 | |
| bnemec | dims: Sorry, missed your ping somehow. | 21:03 |
| bnemec | dhellmann: Will try to take a look asap. | 21:03 |
| *** arnaud has quit IRC | 21:03 | |
| dhellmann | bnemec: thanks | 21:04 |
| *** ihrachyshka has quit IRC | 21:04 | |
| *** yamahata has joined #openstack-oslo | 21:04 | |
| *** ihrachyshka has joined #openstack-oslo | 21:04 | |
| dims | bnemec: need help with https://review.openstack.org/#/c/114614/ | 21:05 |
| bnemec | dims: Okay, will try to hit that one too :-) | 21:06 |
| dims | thanks bnemec | 21:06 |
| dhellmann | Alexei_987: I think I'm still not getting the point of that proposal. The point of the executor is to hold the loop and manage receiving messages and calling the dispatcher. If you move that to the server, then you've got information about the execution model (threading or not) in the server and in the executor, since the eventlet stuff is left behind. | 21:08 |
| Alexei_987 | dhellmann: please see my comment regarding different drivers.. they usually contain their own main loop | 21:08 |
| Alexei_987 | dhellmann: server uses it's own thread to receive messages | 21:09 |
| dhellmann | Alexei_987: and so they could provide their own executor, no? | 21:09 |
| Alexei_987 | dhellmann: it's not how it happens in practice | 21:09 |
| Alexei_987 | dhellmann: now we have a bridge between 2 loops | 21:09 |
| Alexei_987 | dhellmann: we have it in 0MQ driver and proposed AMQP1.0 driver | 21:10 |
| dhellmann | Alexei_987: is that by design, or is that an artifact of us not having updated all of the drivers fully when we moved to the new public API for oslo.messaging? | 21:10 |
| Alexei_987 | dhellmann: from my POV it's an artifact of legacy design | 21:10 |
| Alexei_987 | dhellmann: in any case main loop in executor doesn't prevent us from using eventlet | 21:10 |
| dhellmann | Alexei_987: so rather than throwing out the new design, why not update the drivers? | 21:10 |
| *** gordc has quit IRC | 21:11 | |
| Alexei_987 | main loop in server* | 21:11 |
| Alexei_987 | dhellmann: update drivers to provide their own executor? | 21:11 |
| dhellmann | Alexei_987: we *want* the choice to use eventlet to be explicit, that's why we put it in a separate plugin layer | 21:11 |
| Alexei_987 | dhellmann: it still remains the same | 21:11 |
| Alexei_987 | dhellmann: we can use blocking executor or eventlet executor | 21:12 |
| dhellmann | Alexei_987: what remains the same? | 21:12 |
| Alexei_987 | dhellmann: choice | 21:12 |
| dhellmann | Alexei_987: if it's the same, what's the point? | 21:12 |
| Alexei_987 | dhellmann: main loop is moved to server... | 21:12 |
| dhellmann | Alexei_987: As I said, I think I'm missing something about the motivation for this change. | 21:12 |
| Alexei_987 | dhellmann: this allows us to remove bridge between 2 loops | 21:12 |
| Alexei_987 | dhellmann: and implement driver in a more efficient way | 21:12 |
| dhellmann | Alexei_987: why not remove that and leave the loop in the executor? | 21:13 |
| Alexei_987 | dhellmann: + it simplifies executor cause it doesn't have to care about connections and poolinng | 21:13 |
| Alexei_987 | dhellmann: to remove the bridge without moving loop 2 server we'll need driver to provide their own executors | 21:13 |
| dhellmann | Alexei_987: that layer does become simple if you move all of the work it's doing into another layer, but that doesn't seem like a good reason to move it | 21:13 |
| dhellmann | Alexei_987: if the driver has to provide its own server or its own executor, what's the difference? | 21:14 |
| Alexei_987 | dhellmann: now we have 2 executors - blocking + eventlet | 21:14 |
| Alexei_987 | driver will have to implement both | 21:14 |
| dhellmann | no | 21:14 |
| Alexei_987 | why not? | 21:14 |
| dhellmann | because deciding whether to use threads or eventlet or block is the job of the executor | 21:15 |
| *** markmc has quit IRC | 21:15 | |
| Alexei_987 | dhellmann: yes and we say that driver will provide it's own executor | 21:15 |
| dhellmann | the driver shouldn't worry about that | 21:15 |
| Alexei_987 | exactly | 21:15 |
| dhellmann | Alexei_987: well, they don't care about it now, do they? | 21:15 |
| Alexei_987 | dhellmann: cause of the bridge between 2 main loops | 21:15 |
| *** dims has quit IRC | 21:15 | |
| dhellmann | ok | 21:15 |
| Alexei_987 | I move receiving messages to server and dispatching messages stays in executor | 21:16 |
| Alexei_987 | cause driver knows how to receive messages | 21:16 |
| Alexei_987 | and executor knows about how to dispatch them | 21:16 |
| Alexei_987 | and right now executor does both things | 21:16 |
| dhellmann | you really need to be convincing the maintainers of the library, not me. I may just be asking dumb questions. I would like to hear what they say, and what haypo says about how this would affect the plans to get rid of eventlet and use asyncio in python 3. | 21:16 |
| Alexei_987 | IMHO this will help a lot | 21:17 |
| Alexei_987 | cause it makes executors simpler | 21:17 |
| Alexei_987 | (by moving receiving logic to driver | 21:17 |
| Alexei_987 | dhellmann: I'm explaining this cause I have -1 on spec so I'm trying to answer questions you've had | 21:18 |
| dhellmann | Alexei_987: maybe you can add some of this to the spec, then, and get them to take a look. It's not solely up to me, and I asked questions that it seems like you can answer without waiting. | 21:18 |
| dhellmann | Alexei_987: it's likely I'm just not seeing something because I'm not an expert on this particular library | 21:19 |
| Alexei_987 | dhellmann: I'm not sure what is the best way to attract attention to this spec | 21:20 |
| dhellmann | Alexei_987: I would start by answering the existing questions in the text of the spec, which will give you a new draft without a -1. Then you can ask some of the contributors to oslo.messaging to look at the updated draft. | 21:22 |
| Alexei_987 | dhellmann: I think that right now It maybe more usefull to focus on functional + performance tests | 21:22 |
| Alexei_987 | dhellmann: what do you think? | 21:22 |
| dhellmann | tests not related to this spec? | 21:22 |
| Alexei_987 | dhellmann: no.. however it would be easier to land such patches | 21:23 |
| Alexei_987 | dhellmann: cause functional tests will make sure that nothing is broken | 21:23 |
| Alexei_987 | and performance test will display positive impact of changes | 21:23 |
| Alexei_987 | dhellmann: should I also put comments and discussions in spec text? | 21:25 |
| dhellmann | Alexei_987: Well, there was some sort of thought put into the existing design that you want to change, so I would like to see that addressed when you give reasons for changing it. Maybe the design is bad, but maybe it meets requirements you're not seeing. So before you spend a lot of time working on code or tests, I would like to make sure there is some amount of agreement about the new approach being OK. | 21:26 |
| Alexei_987 | dhellmann: well tests will be for existing design | 21:26 |
| *** dims has joined #openstack-oslo | 21:26 | |
| dhellmann | Alexei_987: that's why we go through the spec process | 21:26 |
| Alexei_987 | dhellmann: so it's definitely worth spending some time on such things | 21:26 |
| dhellmann | Alexei_987: if you want to write tests, that's fine, I just don't think it's a way to prove your new approach is better | 21:26 |
| Alexei_987 | dhellmann: well if my approach will allow to implement driver that is working much faster with the same functionality it's not enough? | 21:27 |
| Alexei_987 | dhellmann: I don't think that current design is intended to be terribly slow on purpose | 21:28 |
| dhellmann | Alexei_987: no, of course not, but it was intended to have different layers responsible for different parts of the operation | 21:28 |
| dhellmann | Alexei_987: I need to go. | 21:28 |
| Alexei_987 | dhellmann: ok thanks for your time | 21:28 |
| openstackgerrit | A change was merged to openstack/oslo.concurrency: Fix docs generation https://review.openstack.org/114647 | 21:42 |
| *** dims has quit IRC | 21:43 | |
| *** dims has joined #openstack-oslo | 21:43 | |
| *** dims has quit IRC | 21:48 | |
| * bnemec hopes we never have a bug in mask_password. | 21:58 | |
| bnemec | The number of regexes has gotten so large that tracking down which one is causing a problem would be a nightmare. :-) | 21:58 |
| Alexei_987 | bnemec: why not adding some kind of subclass for private data? | 21:59 |
| Alexei_987 | bnemec: polymorphism would be able to handle it all | 22:00 |
| bnemec | Alexei_987: They're already private. I just suspect we've got some overlap in them and I doubt anyone actually understands what each one is doing anymore. | 22:01 |
| Alexei_987 | bnemec: I mean that password would be some special type and it would not be printed | 22:01 |
| *** stevemar2 has joined #openstack-oslo | 22:02 | |
| bnemec | Alexei_987: We don't have control over that. This is for masking passwords out of arbitrary strings. | 22:02 |
| *** zzzeek_ has joined #openstack-oslo | 22:03 | |
| *** pcm_ has quit IRC | 22:04 | |
| bnemec | It might be interesting to see if we could convince the other projects to use a special object for password strings though. | 22:04 |
| bnemec | Not 100% sure it would work in a way they'd like. | 22:04 |
| Alexei_987 | bnemec: It would be quite logical for auth_tokens for example | 22:05 |
| *** dims has joined #openstack-oslo | 22:07 | |
| *** ihrachyshka has quit IRC | 22:08 | |
| *** dims_ has joined #openstack-oslo | 22:08 | |
| *** stevemar2 has quit IRC | 22:09 | |
| *** toabctl_ has joined #openstack-oslo | 22:10 | |
| *** krotscheck_ has joined #openstack-oslo | 22:10 | |
| *** sheeprine has quit IRC | 22:11 | |
| *** toabctl has quit IRC | 22:11 | |
| *** zzzeek has quit IRC | 22:11 | |
| *** stevemar has quit IRC | 22:11 | |
| *** krotscheck has quit IRC | 22:11 | |
| *** sheeprine has joined #openstack-oslo | 22:11 | |
| *** sheeprine has quit IRC | 22:11 | |
| *** sheeprine has joined #openstack-oslo | 22:11 | |
| *** zzzeek_ is now known as zzzeek | 22:11 | |
| *** toabctl_ is now known as toabctl | 22:11 | |
| *** krotscheck_ is now known as krotscheck | 22:11 | |
| *** dims has quit IRC | 22:12 | |
| amrith | dhellmann, yt? | 22:30 |
| *** jecarey has quit IRC | 22:37 | |
| openstackgerrit | amrith proposed a change to openstack/oslo.concurrency: Mask passwords in exceptions and error messages https://review.openstack.org/114656 | 22:38 |
| amrith | dhellmann, https://review.openstack.org/#/c/114656/ is ready for you to re-review. Thx | 22:39 |
| *** dims_ has quit IRC | 22:42 | |
| *** arnaud has joined #openstack-oslo | 22:43 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/oslo.vmware: VMWare: Support for pluggable Datastore selection https://review.openstack.org/114551 | 22:44 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo.messaging: Moved all tests to unit folder. https://review.openstack.org/115136 | 22:45 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/oslo.messaging: Added functional test for RPC https://review.openstack.org/115137 | 22:45 |
| *** ajo has quit IRC | 22:47 | |
| *** arnaud__ has joined #openstack-oslo | 22:48 | |
| *** ajo has joined #openstack-oslo | 22:49 | |
| *** mriedem has left #openstack-oslo | 22:50 | |
| *** alexpilotti has quit IRC | 22:50 | |
| *** zzzeek has quit IRC | 22:53 | |
| *** zzzeek has joined #openstack-oslo | 22:53 | |
| *** zzzeek has quit IRC | 22:53 | |
| *** yamahata has quit IRC | 23:20 | |
| *** alexpilotti has joined #openstack-oslo | 23:26 | |
| *** bknudson has joined #openstack-oslo | 23:32 | |
| openstackgerrit | A change was merged to openstack/oslo.messaging: Sync jsonutils from oslo-incubator https://review.openstack.org/114454 | 23:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!