| *** achanda has quit IRC | 00:02 | |
| openstackgerrit | Merged openstack/cliff: Pass user command text to the Command object https://review.openstack.org/147381 | 00:06 |
|---|---|---|
| openstackgerrit | lifeless proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 00:16 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Honour MANIFEST.in excludes https://review.openstack.org/123597 | 00:16 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Better diagnostics on test failure https://review.openstack.org/173131 | 00:16 |
| *** sputnik13 has quit IRC | 00:23 | |
| *** mtanino_ has joined #openstack-oslo | 00:24 | |
| *** mtanino has quit IRC | 00:25 | |
| *** daniel3_ has quit IRC | 00:44 | |
| *** liusheng has joined #openstack-oslo | 01:01 | |
| *** daniel3_ has joined #openstack-oslo | 01:02 | |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 01:05 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Better diagnostics on test failure https://review.openstack.org/173131 | 01:05 |
| *** rjaiswal_ has quit IRC | 01:16 | |
| *** daniel3_ has quit IRC | 01:18 | |
| *** mfedosin has quit IRC | 01:21 | |
| *** jecarey has joined #openstack-oslo | 01:24 | |
| *** mfedosin has joined #openstack-oslo | 01:30 | |
| *** mfedosin has quit IRC | 01:42 | |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Silence logs + errors when stopping and group membership lost https://review.openstack.org/173030 | 01:45 |
| *** salv-orlando has joined #openstack-oslo | 01:46 | |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 01:50 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Honour MANIFEST.in excludes https://review.openstack.org/123597 | 01:50 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Better diagnostics on test failure https://review.openstack.org/173131 | 01:50 |
| *** mfedosin has joined #openstack-oslo | 01:50 | |
| *** salv-orlando has quit IRC | 01:53 | |
| *** mtanino_ has quit IRC | 02:04 | |
| *** achanda has joined #openstack-oslo | 02:21 | |
| *** yamahata has quit IRC | 02:23 | |
| *** dims_ has quit IRC | 02:28 | |
| *** stevemar has joined #openstack-oslo | 02:40 | |
| *** tsekiyam_ has joined #openstack-oslo | 02:52 | |
| *** tsekiyama has quit IRC | 02:55 | |
| *** tsekiyam_ has quit IRC | 02:56 | |
| *** stevemar has quit IRC | 03:06 | |
| *** gtt116_ has quit IRC | 03:06 | |
| *** ndipanov has quit IRC | 03:06 | |
| *** tristanC has quit IRC | 03:06 | |
| *** tristanC has joined #openstack-oslo | 03:13 | |
| *** stevemar has joined #openstack-oslo | 03:13 | |
| *** gtt116_ has joined #openstack-oslo | 03:13 | |
| *** ndipanov has joined #openstack-oslo | 03:13 | |
| *** achanda has quit IRC | 03:14 | |
| lifeless | https://review.openstack.org/173131 is green and needs +2's | 03:20 |
| *** yamahata has joined #openstack-oslo | 03:27 | |
| *** dims has joined #openstack-oslo | 03:28 | |
| *** dims has quit IRC | 03:33 | |
| *** Kennan has joined #openstack-oslo | 03:39 | |
| Kennan | hi :dhellmann there? | 03:39 |
| lifeless | probably too late in his evening | 03:40 |
| openstackgerrit | Dean Troyer proposed openstack/cliff: Defer interactive import https://review.openstack.org/173164 | 03:49 |
| *** jecarey has quit IRC | 03:54 | |
| *** sdake has joined #openstack-oslo | 03:56 | |
| *** salv-orlando has joined #openstack-oslo | 03:57 | |
| *** harlowja is now known as harlowja_away | 03:58 | |
| *** sdake_ has quit IRC | 03:58 | |
| *** sdake_ has joined #openstack-oslo | 04:02 | |
| *** salv-orlando has quit IRC | 04:04 | |
| *** sdake has quit IRC | 04:04 | |
| *** mfedosin has quit IRC | 04:41 | |
| *** achanda has joined #openstack-oslo | 04:52 | |
| *** sreshetn1 has joined #openstack-oslo | 05:14 | |
| *** salv-orlando has joined #openstack-oslo | 05:23 | |
| *** otherwiseguy has quit IRC | 05:25 | |
| *** otherwiseguy has joined #openstack-oslo | 05:28 | |
| *** salv-orlando has quit IRC | 05:30 | |
| *** rushiagr_away is now known as rushiagr | 05:33 | |
| *** mfedosin has joined #openstack-oslo | 05:39 | |
| *** salv-orlando has joined #openstack-oslo | 05:39 | |
| *** sreshetn1 has quit IRC | 05:46 | |
| *** salv-orlando has quit IRC | 05:50 | |
| *** salv-orlando has joined #openstack-oslo | 05:53 | |
| *** sreshetn1 has joined #openstack-oslo | 05:56 | |
| *** yamahata has quit IRC | 05:58 | |
| *** yamahata has joined #openstack-oslo | 05:59 | |
| *** mfedosin has quit IRC | 06:00 | |
| *** salv-orlando has quit IRC | 06:00 | |
| *** amotoki has joined #openstack-oslo | 06:03 | |
| *** Kennan2 has joined #openstack-oslo | 06:04 | |
| *** Kennan has quit IRC | 06:04 | |
| *** Kennan has joined #openstack-oslo | 06:07 | |
| *** Kennan2 has quit IRC | 06:09 | |
| *** ajo has joined #openstack-oslo | 06:27 | |
| *** mfedosin has joined #openstack-oslo | 06:43 | |
| *** kiseok7_ has joined #openstack-oslo | 06:44 | |
| *** mfedosin has quit IRC | 06:50 | |
| *** sreshetn1 has quit IRC | 06:52 | |
| *** sreshetn1 has joined #openstack-oslo | 07:00 | |
| *** andreykurilin has joined #openstack-oslo | 07:00 | |
| *** jamielennox is now known as jamielennox|away | 07:02 | |
| *** stevemar has quit IRC | 07:09 | |
| *** inc0 has joined #openstack-oslo | 07:09 | |
| *** andreykurilin has quit IRC | 07:13 | |
| *** arnaud___ has joined #openstack-oslo | 07:23 | |
| *** achanda has quit IRC | 07:26 | |
| *** yamahata has quit IRC | 07:30 | |
| *** dulek has joined #openstack-oslo | 07:36 | |
| *** ndipanov has quit IRC | 07:41 | |
| *** ndipanov has joined #openstack-oslo | 07:42 | |
| *** rpodolyaka1 has joined #openstack-oslo | 07:42 | |
| *** inc0 has quit IRC | 07:46 | |
| *** ajo has quit IRC | 07:47 | |
| *** haypo has joined #openstack-oslo | 07:53 | |
| *** ajo has joined #openstack-oslo | 07:54 | |
| *** e0ne has joined #openstack-oslo | 07:58 | |
| *** e0ne is now known as e0ne_ | 07:58 | |
| *** dims has joined #openstack-oslo | 08:07 | |
| *** e0ne_ has quit IRC | 08:08 | |
| *** e0ne has joined #openstack-oslo | 08:08 | |
| *** dulek has quit IRC | 08:08 | |
| *** dims has quit IRC | 08:12 | |
| *** ujjain has joined #openstack-oslo | 08:20 | |
| *** sdake has joined #openstack-oslo | 08:22 | |
| *** arnaud___ has quit IRC | 08:26 | |
| *** sdake_ has quit IRC | 08:26 | |
| *** sdake has quit IRC | 08:30 | |
| *** dulek has joined #openstack-oslo | 08:32 | |
| *** ozamiatin__ has joined #openstack-oslo | 08:40 | |
| *** e0ne is now known as e0ne_ | 08:45 | |
| *** shardy has joined #openstack-oslo | 08:46 | |
| *** ujjain has quit IRC | 08:49 | |
| *** ujjain has joined #openstack-oslo | 08:50 | |
| *** jaosorior has joined #openstack-oslo | 08:53 | |
| *** e0ne_ is now known as e0ne | 08:55 | |
| *** ujjain has quit IRC | 08:57 | |
| *** ihrachyshka has joined #openstack-oslo | 09:05 | |
| *** frickler_ is now known as frickler | 09:06 | |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 09:06 |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Honour MANIFEST.in excludes https://review.openstack.org/123597 | 09:06 |
| *** eezhova has joined #openstack-oslo | 09:08 | |
| *** cdent has joined #openstack-oslo | 09:20 | |
| *** ujjain has joined #openstack-oslo | 09:23 | |
| *** e0ne is now known as e0ne_ | 09:40 | |
| *** lifeless1 has joined #openstack-oslo | 09:54 | |
| *** ozamiatin__ has quit IRC | 09:57 | |
| *** e0ne_ is now known as e0ne | 09:58 | |
| *** lifeless has quit IRC | 10:02 | |
| *** harlowja_away has quit IRC | 10:02 | |
| *** pblaho has joined #openstack-oslo | 10:11 | |
| *** pblaho has quit IRC | 10:12 | |
| *** pblaho has joined #openstack-oslo | 10:12 | |
| *** sreshetn1 has quit IRC | 10:26 | |
| *** ozamiatin_ has joined #openstack-oslo | 10:40 | |
| *** dims has joined #openstack-oslo | 10:54 | |
| *** sreshetn1 has joined #openstack-oslo | 10:57 | |
| *** dims has quit IRC | 10:58 | |
| *** sreshetn1 has quit IRC | 11:08 | |
| openstackgerrit | Merged openstack/cliff: Change the argument passed to __init__ for help https://review.openstack.org/155784 | 11:15 |
| *** ujjain has quit IRC | 11:27 | |
| *** e0ne is now known as e0ne_ | 11:34 | |
| openstackgerrit | Kamil Rykowski proposed openstack/oslo.policy: Fix invalid indentation in _load_policy_file method https://review.openstack.org/173275 | 11:40 |
| *** sreshetn1 has joined #openstack-oslo | 11:44 | |
| *** kgiusti has joined #openstack-oslo | 11:50 | |
| *** sreshetn1 has quit IRC | 11:51 | |
| *** e0ne_ is now known as e0ne | 11:52 | |
| *** inc0 has joined #openstack-oslo | 11:55 | |
| *** salv-orlando has joined #openstack-oslo | 12:04 | |
| *** jungleboyj has joined #openstack-oslo | 12:14 | |
| *** ozamiatin_ has quit IRC | 12:14 | |
| *** sreshetn1 has joined #openstack-oslo | 12:16 | |
| *** flaper87 has quit IRC | 12:24 | |
| *** flaper87 has joined #openstack-oslo | 12:24 | |
| *** mfedosin has joined #openstack-oslo | 12:32 | |
| *** gordc has joined #openstack-oslo | 12:36 | |
| *** stpierre has joined #openstack-oslo | 12:45 | |
| *** ozamiatin_ has joined #openstack-oslo | 12:46 | |
| *** dims has joined #openstack-oslo | 12:52 | |
| *** inc0 has quit IRC | 12:55 | |
| *** jecarey has joined #openstack-oslo | 13:05 | |
| *** jecarey_ has joined #openstack-oslo | 13:06 | |
| *** e0ne is now known as e0ne_ | 13:06 | |
| *** jecarey has quit IRC | 13:07 | |
| *** jecarey__ has joined #openstack-oslo | 13:07 | |
| *** e0ne_ is now known as e0ne | 13:08 | |
| *** jecarey__ has quit IRC | 13:09 | |
| *** jecarey_ has quit IRC | 13:11 | |
| *** amuller has joined #openstack-oslo | 13:12 | |
| amuller | dims: ping about https://review.openstack.org/#/c/172127/ | 13:13 |
| *** mriedem has joined #openstack-oslo | 13:15 | |
| *** mriedem_away has quit IRC | 13:15 | |
| haypo | sileht, did you see https://review.openstack.org/#/c/172127/ ? do you know the purpose of the debug log "unpacked context: ..."? | 13:15 |
| haypo | amuller, your change looks safe, but i'm not sure that i understand your use case | 13:15 |
| sileht | haypo, this is the only log that allow us to debug I think | 13:16 |
| haypo | amuller, the message is only logged if you configured your application to log at DEBUG level no? | 13:16 |
| haypo | amuller, in debug mode, you should expect more noise especially in logs | 13:16 |
| dims | amuller: agree, it may not help the end user of openstack. but yes, as haypo says, we don't have another level for developer-y output | 13:16 |
| sileht | haypo, the logging of oslo.msg should be enhanced, to clearly understand the purpose of each like | 13:16 |
| sileht | *line* | 13:17 |
| haypo | dims, according to sileht, it can help if you really don't understand a bug | 13:18 |
| haypo | when sometimes doesn't work, any log is better than nothing :) | 13:18 |
| dims | wish there was another level finer than DEBUG :) | 13:18 |
| amuller | You should enable logging like that only if you suspect an issue with messaging | 13:19 |
| haypo | dims, the logging module allows you to define your own log level. you *can* use DEBUG+1, it's free | 13:19 |
| dims | may be a better message rather than just "unpacked context:"? | 13:19 |
| sileht | haypo, dims that help to track if a request is blocked on rabbit, or if the application is stuck | 13:19 |
| amuller | I think I can support that argument because that log is the *only* log when you put several Neutron services in DEBUG mode | 13:19 |
| haypo | amuller, you didn't reply to my question. why do you run your application in debug mode? | 13:19 |
| *** amotoki has quit IRC | 13:19 | |
| haypo | amuller, what's the relation between "users" and "debug mode" :-p | 13:20 |
| haypo | amuller, the logging module allows you to define a different log level per logger | 13:20 |
| haypo | it looks like openstack uses a ton of loggers :) | 13:20 |
| *** liusheng has quit IRC | 13:20 | |
| *** liusheng has joined #openstack-oslo | 13:21 | |
| amuller | haypo: Developers often run their enviornment in DEBUG mode, likewise for testing / pre-integration environments | 13:21 |
| haypo | (IMO the logging module is misused, getLogger(__name__) looks wrong to me) | 13:21 |
| amuller | haypo: Matter of fact I just supported a case where the issue ended up being messaging, but I can tell you that the unpacked context logging was of no assistance =D | 13:21 |
| amuller | haypo: Right now it's just spamming me with information that I perceive as useless. I can't think of a situation where it would help me to know that I'm getting random messages. It doesn't tell me what kind of message it is (Which would be more helpful) | 13:22 |
| haypo | amuller, IMO it's not really a bug, but more a configuration issue. oslo messaging should maybe use a different log level | 13:22 |
| amuller | haypo: I recently submitted a patch that was merged that reduced similar logging per message | 13:23 |
| amuller | haypo: I had a couple of weeks of log nirvana | 13:23 |
| amuller | with no periodic logging | 13:23 |
| amuller | just silent logs | 13:23 |
| amuller | unless something bad happened | 13:23 |
| amuller | now this log :( | 13:23 |
| haypo | amuller, when something happens, i must not be logged at debug level | 13:23 |
| sileht | haypo, dims I think we could hide this message, the only useful info is the request-id | 13:23 |
| sileht | haypo, dims and I pretty sure if someelse occurs this id is printed | 13:24 |
| haypo | sileht, when do you need this log? when debugging an application or when debugging oslo messaging itself | 13:24 |
| sileht | oslo.messaging | 13:24 |
| haypo | amuller, i'm not saying that you are wrong, i'm just trying to understand your use case | 13:24 |
| dims | +1 sileht | 13:24 |
| *** jecarey has joined #openstack-oslo | 13:25 | |
| amuller | haypo: We're in agreement then, I'm saying that the log is not useful when debugging a project | 13:25 |
| haypo | sileht, when hacking oslo messaging, is it possible to add the log temporary locally? | 13:25 |
| amuller | because projects have their own logging anyway for incoming messages | 13:25 |
| sileht | we need to rewrite something like that to debug oslo.messaging: https://review.openstack.org/#/c/109526/6 | 13:25 |
| ihrachyshka | sileht, request id? isn't it logged when REST request is coming by oslo.log itself? | 13:25 |
| sileht | haypo, yes, even in gate if needed so this is not really a probleme | 13:25 |
| amuller | with more contextual information | 13:25 |
| ihrachyshka | at least in neutron we rely on oslo.log adding needed context info on request | 13:26 |
| haypo | sileht, how do you add a patch to a gate? | 13:28 |
| haypo | sileht, or do you mean pushing a new version of a patch with the log? | 13:29 |
| sileht | haypo, yes | 13:29 |
| sileht | haypo, if you want to debug a nova patch, you can add the log in an oslo.messaging patch, | 13:30 |
| haypo | amuller, ok, so +2 for me https://review.openstack.org/#/c/172127/ | 13:30 |
| haypo | i tried to summarize our discussion in the issue | 13:30 |
| sileht | haypo, and creates a devstack patch that install oslo.messaging from source and and DependOn: with the id of the oslo.msg and the nova patch | 13:30 |
| amuller | haypo: thank you | 13:31 |
| haypo | sileht, ok, i understand the trick. it doesn't look too complex, and i hope that such bugs are rare :) | 13:31 |
| *** dims has quit IRC | 13:31 | |
| *** dims has joined #openstack-oslo | 13:32 | |
| haypo | dims, oh, you voted +2 and then removed your vote | 13:34 |
| dims | haypo: yes, waiting got the discussion here. | 13:34 |
| *** jungleboyj has quit IRC | 13:35 | |
| dims | still +2. marking it now | 13:35 |
| *** sdake has joined #openstack-oslo | 13:35 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:37 | |
| *** sdake has quit IRC | 13:40 | |
| *** rushiagr is now known as rushiagr_away | 13:44 | |
| *** sdake has joined #openstack-oslo | 13:57 | |
| *** dims has quit IRC | 14:00 | |
| *** dims has joined #openstack-oslo | 14:00 | |
| *** rushiagr_away is now known as rushiagr | 14:01 | |
| *** mtanino has joined #openstack-oslo | 14:04 | |
| *** jaypipes has joined #openstack-oslo | 14:06 | |
| openstackgerrit | Terry Howe proposed openstack/cliff: Sort the fuzzy matches https://review.openstack.org/173341 | 14:17 |
| *** stevemar has joined #openstack-oslo | 14:30 | |
| *** zzzeek has joined #openstack-oslo | 14:30 | |
| *** zz_jgrimm is now known as jgrimm | 14:32 | |
| openstackgerrit | Merged openstack/oslo.serialization: Add pypi download + version badges https://review.openstack.org/171269 | 14:38 |
| *** amuller has left #openstack-oslo | 14:38 | |
| *** jungleboyj has joined #openstack-oslo | 14:41 | |
| *** bknudson has joined #openstack-oslo | 14:43 | |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 14:47 |
| *** stevemar has quit IRC | 14:49 | |
| *** stevemar has joined #openstack-oslo | 14:49 | |
| *** e0ne is now known as e0ne_ | 14:55 | |
| *** e0ne_ is now known as e0ne | 14:56 | |
| *** dulek has quit IRC | 15:01 | |
| *** mtanino has quit IRC | 15:03 | |
| *** sreshetn1 has quit IRC | 15:05 | |
| openstackgerrit | Monty Taylor proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 15:06 |
| *** sreshetn1 has joined #openstack-oslo | 15:07 | |
| *** harlowja_at_home has joined #openstack-oslo | 15:13 | |
| *** mtanino has joined #openstack-oslo | 15:14 | |
| *** ozamiatin_ has quit IRC | 15:17 | |
| *** _amrith_ is now known as amrith | 15:18 | |
| *** jgrimm has quit IRC | 15:21 | |
| *** flaper87 has quit IRC | 15:23 | |
| *** yamahata has joined #openstack-oslo | 15:24 | |
| *** flaper87 has joined #openstack-oslo | 15:25 | |
| *** flaper87 has quit IRC | 15:25 | |
| *** flaper87 has joined #openstack-oslo | 15:25 | |
| *** amotoki has joined #openstack-oslo | 15:28 | |
| *** haypo has quit IRC | 15:38 | |
| *** amrith is now known as _amrith_ | 15:39 | |
| *** jgrimm has joined #openstack-oslo | 15:44 | |
| *** harlowja_at_home has quit IRC | 15:47 | |
| *** stevemar has quit IRC | 15:49 | |
| *** stevemar has joined #openstack-oslo | 15:49 | |
| * dhellmann SpamapS: I didn't want you to +1 it, I wanted you to review it. :-) | 15:50 | |
| *** arnaud___ has joined #openstack-oslo | 15:50 | |
| dhellmann | Kennan: hey, post your question and I'll try to reply but if it's long the ML might be better if we're as far out of sync in TZs as I think | 15:52 |
| *** ihrachyshka has quit IRC | 15:52 | |
| *** e0ne is now known as e0ne_ | 15:52 | |
| openstackgerrit | Merged openstack-dev/pbr: Better diagnostics on test failure https://review.openstack.org/173131 | 15:54 |
| openstackgerrit | Merged openstack/oslo.policy: Fix invalid indentation in _load_policy_file method https://review.openstack.org/173275 | 15:56 |
| openstackgerrit | Merged openstack/tooz: Handle errors that come out of pymemcache better https://review.openstack.org/173038 | 15:57 |
| openstackgerrit | Merged openstack/tooz: Make and use a thread safe pymemcache client subclass https://review.openstack.org/173101 | 16:01 |
| *** e0ne_ is now known as e0ne | 16:03 | |
| openstackgerrit | Michael Bayer proposed openstack/oslo.db: Implement new oslo.db.sqlalchemy.enginefacade module https://review.openstack.org/138215 | 16:08 |
| zzzeek | dhellmann: what do you think of merging ^^^^ | 16:08 |
| zzzeek | dhellmann: partially becuase it has a refactor of session.py into orm.py and engine.py, I can break that into a separate gerrit if that helps | 16:09 |
| dhellmann | zzzeek: my day is looking really busy, so I may not get to review that until tomorrow, but I'll put it high on my list for the morning | 16:09 |
| zzzeek | dhellmann: OK | 16:09 |
| dhellmann | zzzeek: is the spec approved, yet? | 16:10 |
| zzzeek | dhellmann: the controversial idea is that its an all new API that nobody is using :) | 16:10 |
| zzzeek | dhellmann: the spec was approved long ago sure | 16:10 |
| zzzeek | dhellmann: i think | 16:10 |
| dhellmann | ok, I couldn't remember | 16:10 |
| zzzeek | dhellmann: for kilo | 16:10 |
| dhellmann | ok, cool | 16:10 |
| zzzeek | dhellmann: though the current API does not match the spec exactly. same general idea | 16:10 |
| dhellmann | yeah, it doesn't have to match exactly | 16:10 |
| zzzeek | dhellmann: ok | 16:10 |
| zzzeek | dhellmann: it needed a lot more fluency whcih I’ve added. it will need lots of new things to work in all cases | 16:11 |
| zzzeek | dhellmann: well, new cases. id like to add isolation level support and integrate it more tightly with retry decorators and savepoints | 16:12 |
| *** sdake_ has joined #openstack-oslo | 16:12 | |
| *** _amrith_ is now known as amrith | 16:12 | |
| dhellmann | zzzeek: ack | 16:15 |
| *** sdake has quit IRC | 16:15 | |
| *** mtanino has quit IRC | 16:17 | |
| *** sreshetn1 has quit IRC | 16:17 | |
| *** sreshetn1 has joined #openstack-oslo | 16:18 | |
| *** sreshetn1 has quit IRC | 16:27 | |
| *** amotoki has quit IRC | 16:35 | |
| *** arnaud___ has quit IRC | 16:40 | |
| *** salv-orlando has quit IRC | 16:44 | |
| SpamapS | Hrm, looks like the next hurdle toward py3k is qpid-python | 16:49 |
| SpamapS | dhellmann: ^ have you looked into that yet? | 16:49 |
| dhellmann | SpamapS: no, I'm not sure how many folks use qpid | 16:52 |
| *** ozamiatin_ has joined #openstack-oslo | 16:54 | |
| *** amrith is now known as _amrith_ | 16:54 | |
| *** salv-orlando has joined #openstack-oslo | 16:55 | |
| SpamapS | dhellmann: My understanding was that RedHat wanted it to be used because their enterprise messaging offerings are all based around it. I don't know if they're still pushing that agenda. | 16:57 |
| krotscheck | Did the oslo.config release happen yet? | 16:58 |
| dhellmann | SpamapS: I've had conflicting reports. One of the things we should do is declare which drivers might not make the py3 cut, in Oslo and all the other projects where there are dependency issues. | 16:59 |
| *** harlowja has joined #openstack-oslo | 16:59 | |
| SpamapS | dhellmann: it looks like oslo.messaging is the only thing mentioning qpid in gate or check. | 17:00 |
| *** ihrachyshka has joined #openstack-oslo | 17:01 | |
| SpamapS | dhellmann: might be time to suggest its deprecation.. no testing in any gates == no idea if it actually works. ;) | 17:01 |
| dhellmann | SpamapS: I would support an action like that from the messaging lib team | 17:02 |
| dhellmann | SpamapS: though maybe start with a probe ("we think this might be broken, are you using it or should we deprecate it?") | 17:02 |
| SpamapS | dhellmann: thats really what I mean by "suggest its deprecation". :) | 17:03 |
| * dhellmann nods | 17:03 | |
| SpamapS | I'm still surprised every time I find a library that hasn't taken the 10 minutes that most things take to fix for python3. | 17:05 |
| *** sdake has joined #openstack-oslo | 17:07 | |
| SpamapS | dhellmann: anyway, I'll poke the dev and operators lists now. | 17:07 |
| dhellmann | SpamapS: great, thanks! | 17:07 |
| *** e0ne has quit IRC | 17:08 | |
| harlowja | SpamapS also ask about having kombu do the qpidd work/integration | 17:10 |
| openstackgerrit | Michael Bayer proposed openstack/oslo.db: WIP: Add reconnect_interval parameter to "ping" listener https://review.openstack.org/173421 | 17:10 |
| harlowja | *save openstack/oslo from doing it | 17:10 |
| harlowja | https://github.com/celery/kombu/blob/master/kombu/transport/qpid.py exists... | 17:10 |
| harlowja | perhapps we should help kombu + qpid (or its testing plans) instead of duplicating them... | 17:10 |
| *** sdake_ has quit IRC | 17:11 | |
| harlowja | ^ aka stop trying to reinvent messaging drivers and let someone else do it ;) | 17:11 |
| *** e0ne has joined #openstack-oslo | 17:13 | |
| SpamapS | harlowja: same Vikings, different village. | 17:14 |
| SpamapS | harlowja: the kombu driver is just as screwed as we are. ;) | 17:14 |
| *** e0ne has quit IRC | 17:14 | |
| *** sdake_ has joined #openstack-oslo | 17:14 | |
| harlowja | sure, put reinforcements at the lowest layer (kombu library itself); and then fight the vikings there? | 17:15 |
| harlowja | lol | 17:15 |
| * harlowja let's see how long we can keep the viking analogy alive, lol | 17:15 | |
| harlowja | battle commander harlowja at your service | 17:18 |
| harlowja | lol | 17:18 |
| *** sdake has quit IRC | 17:18 | |
| *** ndipanov has quit IRC | 17:22 | |
| SpamapS | doohhhhh | 17:23 |
| SpamapS | email client fail | 17:23 |
| * SpamapS double-posted to openstack-operators | 17:23 | |
| SpamapS | harlowja: we can just side-step this group of vikings if we give them a concession: we'll tie up all the qpid users and send them out on the beach. | 17:25 |
| harlowja | that seems fine | 17:25 |
| harlowja | is that like 1 or 2 users> | 17:25 |
| harlowja | ? | 17:25 |
| harlowja | i didn't think many people (maybe some RH enterprise customers?) used it | 17:25 |
| SpamapS | It might be 0 | 17:27 |
| SpamapS | http://superuser.openstack.org/articles/openstack-user-survey-insights-november-2014 doesn't even mention broker | 17:27 |
| harlowja | weird, lol | 17:28 |
| harlowja | i'm guessing though it has to be some RH enterprise people right | 17:28 |
| harlowja | i mean, righhhht, it can't be zero | 17:28 |
| harlowja | :-/ | 17:28 |
| SpamapS | It could be | 17:29 |
| SpamapS | It may have been non-zero at one time | 17:29 |
| SpamapS | and then they tried it | 17:29 |
| SpamapS | and got stuck on Grizzly or something | 17:29 |
| SpamapS | This is sort of like the PGSQL situation. Somebody has to step up and say "This is important enough for me to maintain" | 17:29 |
| harlowja | agreed | 17:29 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 17:30 | |
| SpamapS | btw once qpid is removed py34 tests fail because of mock (need pymock IIRC) | 17:30 |
| harlowja | i thought qpid + proton (or wahtever) was the future, lol | 17:30 |
| harlowja | but idk if anyone uses that either, lol | 17:31 |
| *** sdake has joined #openstack-oslo | 17:31 | |
| harlowja | :-/ | 17:31 |
| harlowja | * http://qpid.apache.org/proton/ | 17:31 |
| ozamiatin_ | SpamapS, Hi, I've tried to write to openstack-operators but it rejects my e-mail | 17:32 |
| SpamapS | harlowja: I don't think this is "qpid sucks and nobody uses it", this is "the testing matrix is too big to carry" | 17:33 |
| *** ihrachyshka has quit IRC | 17:33 | |
| * harlowja is waiting till the big-tent stuff turns openstack into https://projects.apache.org/indexes/category.html (where at that point openstack is pretty much the same as apache, lol) | 17:33 | |
| dims | ozamiatin_: subscribe first | 17:33 |
| SpamapS | ozamiatin_: don't cross-post like that | 17:33 |
| *** ihrachyshka has joined #openstack-oslo | 17:33 | |
| harlowja | SpamapS sure, understood, less testing matrix the better | 17:33 |
| SpamapS | ozamiatin_: You'll want a single thread per mailing list. | 17:33 |
| harlowja | *or less testing openstack has to do (and someone else can do for us...) | 17:33 |
| dims | harlowja: i am an apache guy, still prefer CTR over RTC | 17:34 |
| harlowja | whats CTR / RTC? | 17:34 |
| SpamapS | harlowja: Unless there is a real need for two options. For instance, RabbitMQ has serious scaling issues at high scale, but may be simpler for small/mid-size. | 17:34 |
| ozamiatin_ | SpamapS, dims, got it, thanks | 17:34 |
| dims | harlowja: Commit-Then-Review vs Review-Then-Commit | 17:34 |
| SpamapS | dims: ew | 17:35 |
| *** sdake_ has quit IRC | 17:35 | |
| dims | SpamapS: i am an odd duck here :) | 17:35 |
| SpamapS | dims: it's ok, I don't judge you. just, ew. ;) | 17:35 |
| dims | haha | 17:35 |
| SpamapS | Spent too much time doing post-commit code reviews in a medium sized PHP shop.. so.. much.. rage.. | 17:36 |
| harlowja | u mean post-commit-fix-that-shit-up-yourself, lol | 17:36 |
| * harlowja been there :-P | 17:36 | |
| dims | :) | 17:37 |
| SpamapS | It's tough when you're the one guy who understands that the database is not in fact powered by unicorns nor does it have infinite RAM and CPU for running queries in realtime. ;) | 17:37 |
| harlowja | damn, i've been lied to this whole time :( | 17:37 |
| SpamapS | I remember dragging one of the dev leads the 20 miles to the datacenter so he could see how small our server farm was. "You see! THIS IS WHAT $400,000 gets you. You can't have more!" | 17:38 |
| *** boris-42 has quit IRC | 17:38 | |
| dims | SpamapS: like this cloud stuff i keep hearing about :) | 17:38 |
| harlowja | next thing u'll tell us is that there isn't a santa claus either :( | 17:38 |
| SpamapS | dims: Oh this cloud thing is never going to work out. | 17:38 |
| dims | :) | 17:38 |
| SpamapS | How can I possibly convince developers to write decent queries if I can't actually show them the servers?! | 17:38 |
| SpamapS | Ran 3561 tests in 23.779s (-2.807s) | 17:39 |
| SpamapS | FAILED (id=1, failures=18, skips=137) | 17:39 |
| SpamapS | results of py34 tests w/o qpid | 17:40 |
| SpamapS | 18 fails are in mock | 17:40 |
| harlowja | hmmm, what mock issues u seeing | 17:40 |
| harlowja | i've seen some discrepeances where using the 'real' mock makes it go away | 17:40 |
| SpamapS | http://paste.ubuntu.com/10823123/ | 17:40 |
| harlowja | ya | 17:40 |
| harlowja | that one | 17:41 |
| harlowja | try | 17:41 |
| harlowja | try: | 17:41 |
| harlowja | # In python 3.3+ mock got included in the standard library... | 17:41 |
| harlowja | from unittest import mock | 17:41 |
| harlowja | except ImportError: | 17:41 |
| harlowja | import mock | 17:41 |
| harlowja | seem if that helps | 17:41 |
| harlowja | *see if that helps | 17:41 |
| SpamapS | interesting! | 17:41 |
| harlowja | i think thats a known bug | 17:41 |
| harlowja | from last time i looked | 17:41 |
| SpamapS | They should just do that in mock itself | 17:42 |
| harlowja | forgot which one it was in https://code.google.com/p/mock/issues/list | 17:42 |
| SpamapS | if thats possible | 17:42 |
| harlowja | agreed | 17:42 |
| SpamapS | ok that was 7 fails | 17:42 |
| harlowja | cool | 17:42 |
| SpamapS | oslo_messaging._drivers.common.Timeout: Timeout while waiting on RPC response - topic: "<unknown>", RPC method: "<unknown>" info: "<unknown>" | 17:42 |
| harlowja | https://code.google.com/p/mock/issues/detail?id=249&q=method-wrapper SpamapS or something like that | 17:43 |
| harlowja | so its sorta known/but not fixed... | 17:43 |
| harlowja | https://code.google.com/p/mock/issues/detail?id=234 (probably the same) | 17:43 |
| harlowja | hmm, timeouts are a tougher issue (is that just the VM's are slow...) | 17:44 |
| harlowja | *CI vms... | 17:44 |
| SpamapS | harlowja: this is my laptop now, so dunno | 17:44 |
| *** sdake_ has joined #openstack-oslo | 17:45 | |
| harlowja | kk | 17:45 |
| SpamapS | harlowja: it's failing during eventlet.hubs.hub when it runs a greenlet.switch() | 17:45 |
| harlowja | hmmm, that sux | 17:45 |
| harlowja | lol | 17:45 |
| SpamapS | so possible that we have new behavior in eventlet because python3 | 17:45 |
| *** sdake has quit IRC | 17:45 | |
| harlowja | force it to use the threading executor or something? | 17:45 |
| SpamapS | ok down to 4 fails | 17:46 |
| dims | SpamapS: there was a new eventlet release with python3 fixes a few days ago? | 17:46 |
| SpamapS | all are this RPC timeout | 17:46 |
| SpamapS | dims: yes, I'm testing with that. :) | 17:46 |
| SpamapS | otherwise it wouldn't work at all :) | 17:46 |
| dims | cool, just making sure :) | 17:47 |
| SpamapS | I should probably set this down. It's not what I'm supposed to be doing.. but my god does this itch need scratching | 17:47 |
| harlowja | what are u supposed to be doing? | 17:47 |
| SpamapS | harlowja: making caching work in shade. :) | 17:47 |
| harlowja | ah | 17:47 |
| sileht | harlowja, SpamapS we can already use kombu for qpid by setting the messaging url to kombu+qpid://..../ | 17:48 |
| *** sdake_ has quit IRC | 17:48 | |
| harlowja | sileht woot | 17:48 |
| sileht | but tests doesn't pass yet | 17:48 |
| harlowja | it'll be interesting to see who responds to SpamapS ML email(s) | 17:48 |
| *** sdake has joined #openstack-oslo | 17:48 | |
| SpamapS | sileht: that doesn't solve the problem | 17:48 |
| SpamapS | sileht: kombu imports python-qpid too | 17:48 |
| sileht | The main issue is that we can have a migration path between the legacy driver and the one from kombu | 17:48 |
| harlowja | well only need migration if someone is using it, haha | 17:49 |
| harlowja | :-P | 17:49 |
| sileht | SpamapS, no, but this is not more a oslo.msg issue :) | 17:49 |
| SpamapS | sileht: and that still means we don't have qpid in any gate. | 17:49 |
| harlowja | *at that point u go yell at kombu people for not having it in there gate :-P | 17:49 |
| harlowja | *problem solved ^ | 17:49 |
| SpamapS | kombu has a gate? | 17:49 |
| harlowja | see | 17:49 |
| harlowja | u can be the person to yell :-P | 17:49 |
| harlowja | it has travis, probably not very extensive gating there though | 17:50 |
| harlowja | but not sure | 17:50 |
| *** achanda has joined #openstack-oslo | 17:51 | |
| * harlowja doesn't know what travis VM(s)/containers come with | 17:51 | |
| *** sputnik13 has joined #openstack-oslo | 17:51 | |
| harlowja | i do know that the kazoo project sets up zookeeper in there travis stuff (but they just have scripts to do this) | 17:51 |
| harlowja | ^ so its not out of the realm of possiblity to gate on things there... | 17:52 |
| *** jaosorior has quit IRC | 17:52 | |
| *** sdake has quit IRC | 17:52 | |
| sileht | SpamapS, AFAIK qpid can't work when you enable multiple workers | 17:52 |
| *** sdake has joined #openstack-oslo | 17:52 | |
| SpamapS | https://github.com/celery/kombu/blob/master/.travis.yml | 17:52 |
| SpamapS | they use tox! | 17:53 |
| SpamapS | neat | 17:53 |
| sileht | SpamapS, python-qpid maintains a global object with a ton of FD open and when you fork your app everything is broken | 17:53 |
| sileht | kombu have a workaround for that (it monkey patch python-qpid...) | 17:53 |
| harlowja | SpamapS ya, travis is sorta differenet | 17:54 |
| harlowja | https://github.com/python-zk/kazoo/blob/master/.travis.yml has stuff that sets up zookeeper... | 17:54 |
| harlowja | https://github.com/python-zk/kazoo/blob/master/ensure-zookeeper-env.sh ... | 17:54 |
| harlowja | so its not impassible to gate on things there | 17:54 |
| openstackgerrit | Michael Bayer proposed openstack/oslo.db: Remove pre-SQLAlchemy-0.9.7 compat utilities https://review.openstack.org/173439 | 17:55 |
| SpamapS | sileht: ohnoes | 17:55 |
| harlowja | sileht that sucks, seeing as openstack likes to fork things, lol | 17:56 |
| SpamapS | harlowja: travis is post-commit tho | 17:56 |
| SpamapS | isn't it? | 17:56 |
| harlowja | SpamapS not exactly | 17:56 |
| harlowja | negative | 17:56 |
| harlowja | they hook into the PR mechanism | 17:56 |
| harlowja | ex: https://github.com/pinterest/pymemcache/pull/36 | 17:56 |
| harlowja | look at bottom there | 17:56 |
| harlowja | 'All is well — The Travis CI build passed '... | 17:56 |
| harlowja | so pre-commit | 17:56 |
| SpamapS | ok, so pre-commit, but not gate | 17:56 |
| SpamapS | cool | 17:57 |
| harlowja | sure, although might be able to tweak things more than i know about | 17:57 |
| * harlowja has only done the travis basics :-P | 17:57 | |
| SpamapS | well you can treat it like a gate if you are careful to refresh pull reqs before merging. | 17:57 |
| harlowja | ya | 17:57 |
| * SpamapS is a zuul fanboi.. so... ;) | 17:58 | |
| harlowja | :-P | 17:58 |
| harlowja | i do like the ease of just changing stuff in .travis.yml | 17:58 |
| harlowja | vs having to go to another repo... | 17:58 |
| harlowja | keeps the settings in the same repo... | 17:58 |
| SpamapS | 6 vs. 1/2 dozen | 17:58 |
| harlowja | ya ya | 17:58 |
| *** ozamiatin_ has quit IRC | 17:59 | |
| SpamapS | they must do something weird with tox | 18:00 |
| SpamapS | can't get any of kombu's tests to pass | 18:00 |
| harlowja | hmmm | 18:01 |
| harlowja | i don't think i had any issue when i did some commits | 18:01 |
| harlowja | but maybe its changed | 18:02 |
| SpamapS | File "/home/clint/src/kombu/kombu/five.py", line 14, in <module> | 18:02 |
| SpamapS | from amqp.five import __all__ # noqa | 18:02 |
| SpamapS | ImportError: cannot import name __all__ | 18:02 |
| SpamapS | you have to install requirements/dev.txt | 18:03 |
| SpamapS | so their trunk deps on unreleased py-amqp | 18:03 |
| harlowja | hmmm, od | 18:03 |
| harlowja | odd | 18:04 |
| SpamapS | yeah the tox is just weird | 18:05 |
| SpamapS | it uses nose from outside the tox venv too | 18:05 |
| *** amaurymedeiros has quit IRC | 18:05 | |
| SpamapS | because they just do 'nosetests' instead of using the correct pathing variables | 18:05 |
| harlowja | odd | 18:05 |
| SpamapS | which means it pulls in stuff from local python | 18:05 |
| *** jamielennox|away is now known as jamielennox | 18:05 | |
| harlowja | go yell at 'asksol' on #celery :-P | 18:06 |
| SpamapS | long uphill battle I think.. I'm just not used to working outside OpenStack's happy fun rainbow magical land where everything is gated and tox works. ;) | 18:07 |
| harlowja | :) | 18:09 |
| krotscheck | dhellmann: You mentioned having time for an oslo.config release this morning, is that still the case? | 18:09 |
| * harlowja will get u a hazmat suit that u can wear outside openstack then SpamapS | 18:09 | |
| harlowja | *in the big nasty other world, lol | 18:09 |
| *** amaurymedeiros has joined #openstack-oslo | 18:10 | |
| *** amaurymedeiros has quit IRC | 18:10 | |
| *** amaurymedeiros has joined #openstack-oslo | 18:10 | |
| dhellmann | krotscheck: it hasn't been so far, but let me see what I can do now :-) | 18:10 |
| krotscheck | dhellmann: Thanks! I am nothing if not compassionate of crazy schedules | 18:11 |
| krotscheck | :) | 18:11 |
| SpamapS | ok so if you force dev.txt deps in then kombu's tests pass in py34, but qpid is skipped | 18:14 |
| harlowja | ^ sounds like an oppurtunity to make kombu better :-P | 18:15 |
| harlowja | and/or open an issue to let someone else do it, ha | 18:15 |
| SpamapS | if there are qpid users yes | 18:15 |
| harlowja | :-P | 18:15 |
| dhellmann | krotscheck: done, 1.10.0 | 18:15 |
| krotscheck | dhellmann: WOOO, thanks :) | 18:16 |
| * krotscheck runs off to update global requirements. | 18:16 | |
| SpamapS | ozamiatin: btw thanks for sending a message to the operators list. | 18:16 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 18:16 | |
| dhellmann | SpamapS: "force dev.txt deps"? | 18:16 |
| SpamapS | dhellmann: .tox/py34/bin/pip install -r requirements/dev.txt | 18:17 |
| SpamapS | actually I think you need the -U as well | 18:17 |
| dhellmann | oh, kombu's tests -- I was looking for that file in oslo.messaging | 18:17 |
| * dhellmann goes back to doing one thing at a time | 18:17 | |
| krotscheck | dhellmann: So, when updating global-requirements, should I bump the minimum version as well? I wonder why oslo.config is pinned to >=1.9.3 is all. | 18:21 |
| *** yamahata has quit IRC | 18:22 | |
| *** salv-orlando has quit IRC | 18:25 | |
| *** sdake_ has joined #openstack-oslo | 18:25 | |
| *** sdake has quit IRC | 18:29 | |
| *** ozamiatin_ has joined #openstack-oslo | 18:29 | |
| *** _amrith_ is now known as amrith | 18:30 | |
| *** ozamiatin_ has quit IRC | 18:30 | |
| dhellmann | krotscheck: in your case, you need a new feature so you should raise the minimum value. that will have to wait until after all of the projects have stable/kilo branches cut, though | 18:31 |
| *** sreshetn1 has joined #openstack-oslo | 18:36 | |
| *** sdake has joined #openstack-oslo | 18:39 | |
| *** sdake_ has quit IRC | 18:43 | |
| openstackgerrit | Terry Howe proposed openstack/cliff: Add documentation for the value formatter https://review.openstack.org/173461 | 18:43 |
| *** yamahata has joined #openstack-oslo | 18:45 | |
| krotscheck | dhellmann: Seems fair, but argh impatient | 18:48 |
| krotscheck | Well, either way, patch submitted. I'll let them land it when appropriate. | 18:48 |
| *** amrith is now known as _amrith_ | 18:52 | |
| *** e0ne has joined #openstack-oslo | 19:00 | |
| dhellmann | krotscheck: yeah, it should just be another day or two | 19:01 |
| *** rushiagr is now known as rushiagr_away | 19:10 | |
| *** mtanino has joined #openstack-oslo | 19:18 | |
| *** _amrith_ is now known as amrith | 19:20 | |
| *** salv-orlando has joined #openstack-oslo | 19:26 | |
| *** ihrachyshka has quit IRC | 19:26 | |
| e0ne | Nakato: hi! is any update available about graduate-oslo-service (https://review.openstack.org/#/c/142659/) spec? | 19:28 |
| *** dguitarbite has quit IRC | 19:29 | |
| *** andreykurilin has joined #openstack-oslo | 19:32 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.utils: Add a rate limiting utility class https://review.openstack.org/172246 | 19:32 |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Updated from global requirements https://review.openstack.org/171403 | 19:34 |
| *** sdake_ has joined #openstack-oslo | 19:37 | |
| *** salv-orlando has quit IRC | 19:37 | |
| *** sreshetn1 has quit IRC | 19:40 | |
| *** sdake has quit IRC | 19:41 | |
| *** amrith is now known as _amrith_ | 19:46 | |
| *** mtanino has quit IRC | 19:49 | |
| *** sdake has joined #openstack-oslo | 19:51 | |
| *** sdake_ has quit IRC | 19:54 | |
| *** sreshetn1 has joined #openstack-oslo | 19:57 | |
| *** achanda has quit IRC | 20:01 | |
| *** kgiusti has left #openstack-oslo | 20:04 | |
| *** sdake_ has joined #openstack-oslo | 20:05 | |
| *** sdake has quit IRC | 20:07 | |
| *** jgrimm is now known as zz_jgrimm | 20:11 | |
| *** andreykurilin has quit IRC | 20:13 | |
| *** crc32 has joined #openstack-oslo | 20:22 | |
| *** sdake has joined #openstack-oslo | 20:23 | |
| *** achanda has joined #openstack-oslo | 20:25 | |
| *** sdake_ has quit IRC | 20:25 | |
| *** salv-orlando has joined #openstack-oslo | 20:33 | |
| *** _amrith_ is now known as amrith | 20:36 | |
| *** lifeless1 is now known as lifeless | 20:37 | |
| *** pblaho has quit IRC | 21:02 | |
| *** jecarey has quit IRC | 21:02 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 21:03 | |
| *** sreshetn1 has quit IRC | 21:03 | |
| *** sreshetn1 has joined #openstack-oslo | 21:07 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.serialization: Add docstring(s) to handler registry(s) https://review.openstack.org/173089 | 21:08 |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.serialization: More docstring cleanups/tweaks https://review.openstack.org/173095 | 21:09 |
| *** amrith is now known as _amrith_ | 21:11 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 21:17 | |
| openstackgerrit | lifeless proposed openstack-dev/pbr: Fixup semver https://review.openstack.org/124651 | 21:20 |
| *** jungleboyj has quit IRC | 21:21 | |
| *** jecarey has joined #openstack-oslo | 21:26 | |
| *** stpierre has quit IRC | 21:31 | |
| *** shardy is now known as shardy_z | 21:33 | |
| *** boris-42 has joined #openstack-oslo | 21:39 | |
| *** sreshetn1 has quit IRC | 21:44 | |
| *** cdent has quit IRC | 21:50 | |
| openstackgerrit | Merged openstack/tooz: Updated from global requirements https://review.openstack.org/171403 | 21:50 |
| mtreinish | dims, dhellmann: qq is there a way to get oslo logging to default to writing a log file? In my experience the stdout as default is not a really good default for tempest | 21:53 |
| *** mriedem is now known as mriedem_away | 21:56 | |
| *** e0ne has quit IRC | 21:56 | |
| dhellmann | mtreinish: set the --log-file option? | 21:58 |
| dhellmann | mtreinish: hmm, that's not exposed through set_defaults() though | 21:58 |
| mtreinish | yeah, I guess it was more of an oslo.config question is there a good way to set a local override | 21:58 |
| dhellmann | so it wouldn't be easy to do from within tempest code | 21:58 |
| mtreinish | ah ok | 21:58 |
| * dhellmann thinks | 21:59 | |
| dhellmann | mtreinish: we could add that to set_defaults(), I suppose | 21:59 |
| *** mriedem_away has quit IRC | 22:00 | |
| dhellmann | mtreinish: although that feels a little icky | 22:00 |
| dhellmann | mtreinish: could the thing that calls tempest set that instead? maybe even in tempest's config file? | 22:01 |
| dhellmann | or wherever tempest gets its config values | 22:01 |
| mtreinish | dhellmann: so because tempest gets exec'd through a test runner only means we really can only use a config file | 22:02 |
| mtreinish | I guess we could wrap things to force that in the config file | 22:03 |
| dhellmann | mtreinish: ok, that's fine, I just wasn't sure | 22:03 |
| dhellmann | we could also maybe do something hacky with an environment variable | 22:03 |
| dhellmann | I don't think I want most applications to be passing a default log file into set_defaults() though | 22:03 |
| mtreinish | I was hoping to do it in tempest's config.py somewhere so it was kinda seemless | 22:03 |
| dhellmann | os.environ['SOME_GOOD_NAME'] = 'path/to/file' | 22:04 |
| mtreinish | I mean for devstack we set it to log_file = tempest.log and it just writes that to the cwd where tempest is called | 22:04 |
| mtreinish | that's actually a sane default for tempest | 22:04 |
| dhellmann | maybe I should just be less pedantic, because that approach feels ugly, too | 22:04 |
| dhellmann | mtreinish: oh, cool, maybe do that then? | 22:04 |
| mtreinish | dhellmann: just do that in tempest's config.py? | 22:05 |
| dhellmann | mtreinish: let me look at what that file does | 22:06 |
| dhellmann | mtreinish: do you have a separate configuration file for tempest, or is it all in that .py module? | 22:07 |
| mtreinish | all the config opts get set in that module | 22:08 |
| *** jungleboyj has joined #openstack-oslo | 22:09 | |
| dhellmann | mtreinish: ok, it looks like we'll need to give you an API then | 22:09 |
| *** browne has joined #openstack-oslo | 22:09 | |
| mtreinish | yay, it makes me happy every time this happens :) | 22:09 |
| mtreinish | heh, I feel like I'm contributing | 22:09 |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Use oslo.serialization msgpackutils https://review.openstack.org/173058 | 22:11 |
| dhellmann | mtreinish: working on a patch now | 22:14 |
| mtreinish | cool, thanks | 22:16 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.log: fix pep8 errors https://review.openstack.org/173572 | 22:24 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.log: Provide an API to let tempest control the log file https://review.openstack.org/173573 | 22:24 |
| dhellmann | mtreinish: ^^ | 22:24 |
| mtreinish | dhellmann: oh I think there was a miscommunication. Tempest only uses a log file, it's just passed in through an env variable | 22:25 |
| mtreinish | that module was the only place options get defined | 22:26 |
| mtreinish | we actually can't use cli opts because tempest is only invoked through test runners | 22:26 |
| *** jecarey has quit IRC | 22:27 | |
| mtreinish | s/log file/config file | 22:27 |
| mtreinish | dhellmann: http://git.openstack.org/cgit/openstack/tempest/tree/tempest/config.py#n1178 | 22:28 |
| mtreinish | dhellmann: will this work? https://review.openstack.org/#/c/173581/ | 22:35 |
| *** achanda has quit IRC | 22:36 | |
| *** achanda has joined #openstack-oslo | 22:41 | |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Allow the acquired file to be closed manually https://review.openstack.org/171776 | 22:51 |
| dhellmann | mtreinish: well, the option names are not part of the API, so we don't want you messing with them directly | 22:54 |
| dhellmann | mtreinish: so the idea is you could just call this new function directly, and if we ever change the option definition out from under it we would have to support the API still | 22:55 |
| dhellmann | mtreinish: make sense? | 22:55 |
| mtreinish | dhellmann: yeah, sorry I just got confused by the commit msg on 173573 | 22:56 |
| mtreinish | teaches me to read them | 22:56 |
| dhellmann | mtreinish: heh | 22:56 |
| mtreinish | dhellmann: although I'm not sure why you want to mark this function as tempest exclusive | 22:57 |
| dhellmann | mtreinish: because in the general case applications should not do this, they should let the deployer control the logging behavior | 22:57 |
| dhellmann | if the filename is defined, we write to it, which is clunky but existing behavior we have to preserve | 22:58 |
| mtreinish | dhellmann: but it's just a default value right? If the deployer has something else in the config file that's loaded it'll override it | 22:58 |
| dhellmann | mtreinish: right, but now the default is None meaning "do not write to a file" so if nova, say, changes that to "nova.log" then nova's output would always go to that file | 22:59 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:59 | |
| mtreinish | sure, but I was just wondering if an app (something besides tempest) doesn't find logging to stdout useful they might want to set the default to something else | 23:00 |
| dhellmann | yeah, that's not usually a choice we want the apps to make for themselves | 23:01 |
| dhellmann | we want all of the apps to be consistent in that default behavior | 23:01 |
| mtreinish | fair enough, I guess there aren't too many things using oslo log that don't own there own stdout | 23:01 |
| dhellmann | right | 23:03 |
| *** mriedem has joined #openstack-oslo | 23:07 | |
| *** bknudson has quit IRC | 23:07 | |
| *** zzzeek has quit IRC | 23:10 | |
| *** gordc has quit IRC | 23:11 | |
| *** mriedem has quit IRC | 23:13 | |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Add helper to raise + chain exceptions https://review.openstack.org/173597 | 23:15 |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Add + use helper to raise + chain exceptions https://review.openstack.org/173597 | 23:15 |
| *** mriedem has joined #openstack-oslo | 23:17 | |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Add + use helper to raise + chain exceptions https://review.openstack.org/173597 | 23:23 |
| *** salv-orlando has quit IRC | 23:24 | |
| *** crc32 has quit IRC | 23:39 | |
| openstackgerrit | Merged openstack/taskflow: Include the 'dump_memory_backend' example in the docs https://review.openstack.org/172607 | 23:46 |
| *** mriedem1 has joined #openstack-oslo | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!