| *** stevemar has joined #openstack-oslo | 00:26 | |
| *** dimsum__ has joined #openstack-oslo | 00:34 | |
| *** dims_ has joined #openstack-oslo | 00:35 | |
| *** dims_ is now known as dims | 00:36 | |
| *** dimsum__ has quit IRC | 00:39 | |
| *** achanda has joined #openstack-oslo | 00:46 | |
| *** achanda has quit IRC | 00:51 | |
| *** sputnik13 has quit IRC | 00:57 | |
| *** sputnik13 has joined #openstack-oslo | 01:08 | |
| *** sputnik13 has quit IRC | 01:18 | |
| *** mriedem has quit IRC | 01:18 | |
| *** sputnik13 has joined #openstack-oslo | 01:23 | |
| *** sputnik13 has quit IRC | 01:25 | |
| *** sputnik13 has joined #openstack-oslo | 01:32 | |
| *** sputnik13 has quit IRC | 01:32 | |
| *** amrith is now known as _amrith_ | 01:41 | |
| *** _amrith_ is now known as amrith | 01:51 | |
| *** himangi has joined #openstack-oslo | 01:55 | |
| *** amrith is now known as _amrith_ | 02:12 | |
| *** _amrith_ is now known as amrith | 03:01 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a fully functional orderedset (WIP) https://review.openstack.org/160219 | 03:18 |
|---|---|---|
| openstackgerrit | Merged openstack/oslosphinx: Speed up blueprint checking with naming convention https://review.openstack.org/158922 | 03:26 |
| openstackgerrit | Merged openstack/oslo-incubator: Simplify boolean expression https://review.openstack.org/158948 | 03:29 |
| *** dims has quit IRC | 03:35 | |
| *** amotoki has joined #openstack-oslo | 03:44 | |
| *** achanda has joined #openstack-oslo | 03:49 | |
| *** achanda has quit IRC | 03:54 | |
| *** david-lyle_afk has quit IRC | 04:02 | |
| *** diga has joined #openstack-oslo | 04:06 | |
| *** himangi has quit IRC | 04:18 | |
| openstackgerrit | Merged openstack/oslo.config: Add expose_opt to CfgFilter https://review.openstack.org/115565 | 04:26 |
| *** himangi has joined #openstack-oslo | 04:30 | |
| *** dims has joined #openstack-oslo | 04:35 | |
| *** dims has quit IRC | 04:40 | |
| *** noelbk has joined #openstack-oslo | 04:43 | |
| *** achanda has joined #openstack-oslo | 04:49 | |
| openstackgerrit | Merged openstack-dev/pbr: "packages" might list multiple directories https://review.openstack.org/159918 | 04:52 |
| *** sputnik13 has joined #openstack-oslo | 05:58 | |
| *** sputnik13 has quit IRC | 06:01 | |
| *** sputnik13 has joined #openstack-oslo | 06:02 | |
| *** achanda has quit IRC | 06:03 | |
| *** sputnik1_ has joined #openstack-oslo | 06:05 | |
| *** sputnik13 has quit IRC | 06:07 | |
| *** achanda has joined #openstack-oslo | 06:27 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo.vmware: Imported Translations from Transifex https://review.openstack.org/160083 | 06:35 |
| *** himangi has quit IRC | 06:44 | |
| *** hogepodge has quit IRC | 06:55 | |
| *** himangi has joined #openstack-oslo | 06:58 | |
| *** takedakn has joined #openstack-oslo | 07:13 | |
| *** achanda has quit IRC | 07:13 | |
| *** achanda has joined #openstack-oslo | 07:15 | |
| *** achanda has quit IRC | 07:23 | |
| openstackgerrit | Merged openstack/oslo.utils: Utility API to generate EUI-64 IPv6 address https://review.openstack.org/137774 | 07:26 |
| *** achanda has joined #openstack-oslo | 07:29 | |
| *** jaosorior has joined #openstack-oslo | 07:31 | |
| *** SridharGaddam has joined #openstack-oslo | 07:36 | |
| openstackgerrit | Abhijeet Malawade proposed openstack/tooz: fix mysql driver url parsing https://review.openstack.org/159835 | 07:43 |
| *** hogepodge has joined #openstack-oslo | 07:49 | |
| *** achanda has quit IRC | 07:52 | |
| ttx | jogo: right. StoryBoard has n:n project groups so you can just put a project in as many projectgroups as needed | 07:52 |
| *** achanda has joined #openstack-oslo | 07:57 | |
| *** ihrachyshka has joined #openstack-oslo | 07:58 | |
| *** achanda has quit IRC | 08:16 | |
| *** Guest78669 is now known as d0ugal | 08:20 | |
| *** d0ugal has joined #openstack-oslo | 08:21 | |
| *** i159 has joined #openstack-oslo | 08:22 | |
| *** dtantsur|afk is now known as dtantsur | 08:26 | |
| *** himangi has quit IRC | 08:27 | |
| *** stevemar has quit IRC | 08:36 | |
| *** salv-orlando has joined #openstack-oslo | 08:44 | |
| *** viktors|afk has quit IRC | 08:50 | |
| openstackgerrit | Victor Sergeyev proposed openstack/oslo.db: POC: Add a public DB fixture https://review.openstack.org/159883 | 09:04 |
| *** sputnik1_ has quit IRC | 09:05 | |
| *** ihrachyshka has quit IRC | 09:07 | |
| *** dulek has joined #openstack-oslo | 09:13 | |
| *** salv-orlando has quit IRC | 09:13 | |
| *** salv-orlando has joined #openstack-oslo | 09:14 | |
| *** takedakn has quit IRC | 09:39 | |
| *** ajo has joined #openstack-oslo | 09:44 | |
| *** e0ne has joined #openstack-oslo | 09:49 | |
| *** noelbk has quit IRC | 09:59 | |
| *** salv-orlando has quit IRC | 10:11 | |
| openstackgerrit | Merged openstack/oslo.vmware: Imported Translations from Transifex https://review.openstack.org/160083 | 10:23 |
| *** fnordahl has quit IRC | 10:26 | |
| *** dtantsur is now known as dtantsur|bbl | 10:29 | |
| *** ihrachyshka has joined #openstack-oslo | 10:46 | |
| openstackgerrit | Julien Danjou proposed openstack-dev/hacking: Add a check for staticmethod https://review.openstack.org/151952 | 11:14 |
| openstackgerrit | Julien Danjou proposed openstack-dev/hacking: tests: allow to run the same test for different code https://review.openstack.org/152108 | 11:14 |
| *** achanda has joined #openstack-oslo | 11:17 | |
| *** achanda has quit IRC | 11:21 | |
| *** denis_makogon has joined #openstack-oslo | 11:52 | |
| *** e0ne is now known as e0ne_ | 12:03 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: rabbit: heartbeat implementation https://review.openstack.org/146047 | 12:06 |
| *** dims has joined #openstack-oslo | 12:09 | |
| *** takedakn has joined #openstack-oslo | 12:12 | |
| *** e0ne_ has quit IRC | 12:13 | |
| dims | hi jd__, flaper87, dhellmann, beekneemech, i pushed a priority fix in oslo.log for Nova folks, details here - https://bugs.launchpad.net/oslo.log/+bug/1426544 | 12:22 |
| openstack | Launchpad bug 1426544 in OpenStack Compute (nova) "Nova switching to oslo_log blows up instance object repr in logs" [High,Confirmed] - Assigned to Davanum Srinivas (DIMS) (dims-v) | 12:22 |
| *** e0ne has joined #openstack-oslo | 12:27 | |
| *** tedross has joined #openstack-oslo | 12:40 | |
| *** dulek_ has joined #openstack-oslo | 12:57 | |
| *** dulek has quit IRC | 12:57 | |
| *** himangi has joined #openstack-oslo | 13:01 | |
| sdague | dims: per -nova conversation, the logs are actually *really* broken after oslo.log merged. All the request_id, user_id, project_id information is gone | 13:07 |
| sdague | even with your proposed fix | 13:08 |
| *** e0ne is now known as e0ne_ | 13:20 | |
| dims | sdague: ack | 13:21 |
| sdague | for those playing along at home - here is the critical bug - https://bugs.launchpad.net/oslo.log/+bug/1427209 | 13:24 |
| openstack | Launchpad bug 1427209 in oslo.log "oslo.log doesn't log request_id, project_id, user_id in nova" [Critical,New] | 13:24 |
| *** e0ne_ is now known as e0ne | 13:24 | |
| dims | sdague: so i picked the same exact message in both logs | 13:25 |
| dims | http://logs.openstack.org/10/151010/1/check/check-tempest-dsvm-full/80d6fab/logs/screen-n-api.txt.gz#_2015-01-29_00_25_24_550 | 13:25 |
| dims | http://logs.openstack.org/49/160149/1/check/check-tempest-dsvm-full/fc9e3c7/logs/screen-n-api.txt.gz#_2015-03-01_14_04_25_461 | 13:25 |
| dims | they look the same | 13:26 |
| dims | so at least some paths are working | 13:26 |
| sdague | so it's just missing in the actually wsgi log message> | 13:27 |
| sdague | ? | 13:27 |
| sdague | which is a really important one | 13:27 |
| sdague | http://logs.openstack.org/10/151010/1/check/check-tempest-dsvm-full/80d6fab/logs/screen-n-api.txt.gz#_2015-01-29_00_25_24_512 | 13:27 |
| *** inc0 has joined #openstack-oslo | 13:27 | |
| dims | trying to get a sense of how may are missing | 13:29 |
| *** gordc has joined #openstack-oslo | 13:30 | |
| *** amrith is now known as _amrith_ | 13:30 | |
| sdague | so, the wsgi path is going through eventlet | 13:31 |
| sdague | that might be the only one that's missed, but it's a big break from the n-api side | 13:34 |
| dims | sdague: digging through http://paste.openstack.org/show/185052/ now | 13:35 |
| sdague | dims: looks like the writable logger is to blame | 13:36 |
| dims | is there a line in nova that logs this? | 13:37 |
| dims | i mean calls .write() | 13:38 |
| dims | looking | 13:38 |
| sdague | https://github.com/openstack/nova/blob/master/nova/wsgi.py#L224 | 13:39 |
| dims | self._wsgi_logger = loggers.WritableLogger(self._logger) | 13:39 |
| sdague | it's all passed to eventlet, then we use eventlet's request logging | 13:39 |
| sdague | anyway, I need to deal with snow for a bit. back later. | 13:41 |
| dims | ack thanks | 13:41 |
| *** salv-orlando has joined #openstack-oslo | 13:43 | |
| *** exploreshaifali has joined #openstack-oslo | 13:44 | |
| *** bknudson has quit IRC | 13:46 | |
| *** trown has left #openstack-oslo | 13:48 | |
| *** amotoki has quit IRC | 13:55 | |
| *** dtantsur|bbl is now known as dtantsur | 13:59 | |
| *** tedross has quit IRC | 14:04 | |
| *** cdent has joined #openstack-oslo | 14:06 | |
| *** inc0 has quit IRC | 14:08 | |
| *** bknudson has joined #openstack-oslo | 14:10 | |
| *** mriedem has joined #openstack-oslo | 14:10 | |
| *** mriedem has quit IRC | 14:11 | |
| *** mriedem has joined #openstack-oslo | 14:11 | |
| *** dulek_ has quit IRC | 14:14 | |
| *** dulek has joined #openstack-oslo | 14:14 | |
| *** tedross has joined #openstack-oslo | 14:19 | |
| *** dims has quit IRC | 14:25 | |
| *** dims has joined #openstack-oslo | 14:26 | |
| mriedem | dhellmann: is the release_library.sh script supposed to work for python-novaclient? | 14:29 |
| *** ccrouch has joined #openstack-oslo | 14:30 | |
| dims | mriedem: dhellmann may not be around today | 14:31 |
| mriedem | ok | 14:32 |
| *** tedross has quit IRC | 14:32 | |
| dims | beekneemech: will you be able make it to the oslo meeting today? i am supposed to run it today, but working this nova/oslo.log show stopper | 14:32 |
| *** takedakn has quit IRC | 14:36 | |
| *** viktors has joined #openstack-oslo | 14:36 | |
| *** jungleboyj has quit IRC | 14:37 | |
| *** inc0 has joined #openstack-oslo | 14:37 | |
| *** denis_makogon has quit IRC | 14:43 | |
| *** jecarey has joined #openstack-oslo | 14:49 | |
| *** dulek has quit IRC | 14:52 | |
| *** _amrith_ is now known as amrith | 14:55 | |
| *** dulek has joined #openstack-oslo | 14:58 | |
| *** pblaho_ has joined #openstack-oslo | 15:00 | |
| sdague | .deprecated was removed from oslo.log? | 15:01 |
| sdague | what's the replacement pattern? | 15:02 |
| dhellmann | mriedem: it *should* but you have to set up your launchpad project in a particular way with naming conventions for milestones. Did it fail, or are you asking before trying? | 15:03 |
| dhellmann | dims: my house-guest plans changed so I'm around today. You can still chair the meeting, but I'll be able to be there. | 15:03 |
| *** pblaho has quit IRC | 15:03 | |
| dims | dhellmann: thanks! | 15:03 |
| dhellmann | sdague: there's a function in versionutils | 15:03 |
| dhellmann | dims: oh, I just saw the thing about the nova show-stopper, so if I need to run the meeting I can. Just let me know, I didn't want to step on any toes. :-) | 15:04 |
| sdague | dhellmann: so what's the new invocation? | 15:04 |
| dims | dhellmann: yes, please go ahead! | 15:04 |
| mriedem | dhellmann: there are a few issues, i have a storyboard thing open | 15:04 |
| mriedem | dhellmann: will probably push some changes up to release-tools | 15:05 |
| sdague | I have 2 patches in flight with LOG.deprecated in them | 15:05 |
| dhellmann | sdague: you call versionutils.report_deprecated_feature directly | 15:05 |
| dhellmann | we had to remove a circular dependency between those 2 modules :-/ | 15:05 |
| dhellmann | dims: ok, I'll take care of it | 15:06 |
| dhellmann | mriedem: let me know when you push and I'll help ttx review them | 15:06 |
| dhellmann | sdague: the docs haven't landed yet, https://review.openstack.org/#/c/147312/2/doc/source/usage.rst,cm | 15:07 |
| sdague | does debcollector have something for this yet? | 15:08 |
| sdague | it seems odd to revert back to incubator, then have to change again shortly | 15:09 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 15:11 | |
| *** daniel3_ has joined #openstack-oslo | 15:13 | |
| *** mtanino has joined #openstack-oslo | 15:14 | |
| dims | dhellmann: things logged through WritableLogger does not seem to have context information | 15:16 |
| dhellmann | dims: hmm | 15:16 |
| dhellmann | sdague: debtcollector is not going to include version info like versionutils does -- it's for developer-facing deprecation messages for API changes | 15:17 |
| dims | i suspected a disconnect between oslo.log and nova context | 15:17 |
| dhellmann | dims: likely, or maybe the ContextFormatter isn't being set up correctly for the writable loggers? | 15:17 |
| dims | dhellmann: hmm | 15:19 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.log: Add migration notes https://review.openstack.org/147312 | 15:19 |
| dims | dhellmann: https://github.com/openstack/nova/blob/master/nova/wsgi.py#L118 | 15:22 |
| *** jungleboyj has joined #openstack-oslo | 15:23 | |
| dhellmann | dims: let me look at the writable logger code so I can see what it is doing... | 15:23 |
| dhellmann | ok, as I thought that's just a thin shim | 15:23 |
| dhellmann | does something in the wsgi app set up the request context using oslo.context? | 15:24 |
| dhellmann | dims: ^^ | 15:24 |
| dims | checking | 15:25 |
| *** pblaho__ has joined #openstack-oslo | 15:26 | |
| mriedem | dhellmann: https://review.openstack.org/#/c/160378/ | 15:27 |
| *** e0ne is now known as e0ne_ | 15:28 | |
| *** e0ne_ is now known as e0ne | 15:28 | |
| dims | dhellmann: looks like here - https://github.com/openstack/nova/blob/master/nova/api/auth.py#L150 | 15:29 |
| dims | sdague: mriedem: can you please confirm that's where the nova RequestContext is minted? | 15:29 |
| *** pblaho_ has quit IRC | 15:30 | |
| mriedem | dims: looks right | 15:30 |
| dhellmann | mriedem: ah, yeah, good point | 15:30 |
| sdague | dims: to the best of my knowledge | 15:31 |
| mriedem | dhellmann: still doesn't work for python-novaclient b/c of the milestone rename stuff assuming the milestone next-$version scheme which novaclient doesn't follow - looks like a lot of the clients don't follow that | 15:31 |
| mriedem | so was thinking about putting some skip logic in the script for that | 15:32 |
| *** exploreshaifali has quit IRC | 15:32 | |
| *** stevemar has joined #openstack-oslo | 15:32 | |
| dhellmann | mriedem: or start using that convention so we're handling the libs consistently? | 15:32 |
| mriedem | dhellmann: i don't have access to create series and milestones manually in lp for novaclient | 15:32 |
| mriedem | but yeah.. | 15:32 |
| *** beekneemech is now known as bnemec | 15:32 | |
| *** achanda has joined #openstack-oslo | 15:33 | |
| dhellmann | mriedem: yeah, you need to be on the drivers team, so you'd need to talk to someone there about adding you -- I think you can request access through lp but it'll be faster on irc | 15:33 |
| mriedem | you'd think :) | 15:33 |
| mriedem | but probably not | 15:33 |
| dhellmann | heh | 15:33 |
| dhellmann | dims: nova.context.RequestContext doesn't appear to subclass oslo_context.context.RequestContextx | 15:34 |
| *** yamahata has joined #openstack-oslo | 15:35 | |
| mriedem | it never has | 15:35 |
| mriedem | nova has it's own request context | 15:35 |
| dhellmann | mriedem: it needs to now to work properly with oslo.log | 15:35 |
| mriedem | there was an attempt to use oslo.context a few releases ago for domains i think? | 15:35 |
| *** zz_jgrimm is now known as jgrimm | 15:35 | |
| dhellmann | the oslo_context class was copied from nova | 15:35 |
| mriedem | but the thing that required that was dropped | 15:36 |
| dims | dhellmann: mriedem: trying https://review.openstack.org/#/c/160348/ | 15:36 |
| *** dulek has quit IRC | 15:38 | |
| dhellmann | dims: left a comment about to_dict | 15:39 |
| dhellmann | mriedem: https://blueprints.launchpad.net/oslo.log/+spec/app-agnostic-logging-parameters | 15:39 |
| mriedem | dhellmann: ok, so oslo_context already processes the instance kwarg, but since we're not using the right thing oslo.log isn't formatting it correctly? | 15:42 |
| mriedem | i should say, oslo.log already processes the kwarg | 15:42 |
| dhellmann | mriedem: what's being passed in? | 15:42 |
| mriedem | something like, LOG.warning(_('foo happened'), instance=instance) | 15:43 |
| mriedem | where instance is a nova object | 15:43 |
| mriedem | that implements __repr__ | 15:43 |
| *** achanda has quit IRC | 15:44 | |
| dhellmann | mriedem: ok, so that's going to passed through to the logging context just as it used to be, and the default for the format string hasn't changed yet so it will keep working | 15:45 |
| mriedem | so why are we seeing the entire instance object logged rather than just the uuid like we did before? | 15:45 |
| mriedem | or are you saying it will go back to 'normal' after we start using oslo.context? | 15:45 |
| dims | mriedem: i already patched oslo.log for the instance stuff | 15:45 |
| dhellmann | oh, I didn't realize it was being logged incorrectly | 15:46 |
| dims | am chasing the wsgi/context problem now | 15:46 |
| mriedem | dims: link? | 15:46 |
| dims | dhellmann: mriedem: one sec | 15:46 |
| sdague | mriedem: https://bugs.launchpad.net/oslo.log/+bug/1427209 | 15:46 |
| openstack | Launchpad bug 1427209 in oslo.log "oslo.log doesn't log request_id, project_id, user_id in nova" [Critical,Confirmed] - Assigned to Davanum Srinivas (DIMS) (dims-v) | 15:46 |
| mriedem | https://review.openstack.org/#/c/160007/ | 15:46 |
| dhellmann | so we need an oslo.log release for that | 15:46 |
| dims | yep and yep | 15:46 |
| dims | dhellmann: yep, only if we can nail the wsgi/context | 15:47 |
| sdague | mriedem: there are multiple issues in flight here, I only discovered the second one when looking to verify the formatting fix | 15:47 |
| dhellmann | dims: eventually we'll want to convert that to "resource_uuid" somewhere (maybe inside oslo.log?) to allow for the format string change | 15:49 |
| dims | yep | 15:49 |
| dims | dhellmann: note that the instance is specified as kwargs in LOG calls, not from context | 15:50 |
| dhellmann | yeah, still, when we change the default format string there won't be anywhere for that value to go if we don't store it as resource_* | 15:50 |
| sdague | unrelated question, which I'm running into while working on tests, why is policy_dirs a configuration variable? | 15:50 |
| dims | dhellmann: right, that's why nova sets its own context logging string | 15:51 |
| dhellmann | sdague: I think I lost that argument | 15:51 |
| sdague | as in, in what world do we believe that needs to be deployment changable | 15:51 |
| sdague | dhellmann: is there history somewhere, because this seems like a very odd knob | 15:51 |
| dhellmann | dims: oh, if that's the case, then why was this failing and why do we need to do anything special? | 15:51 |
| dhellmann | sdague: I'd have to go find the review in gerrit. | 15:52 |
| dhellmann | I may have just given up on that to get the feature into the library. | 15:52 |
| dims | dhellmann: instance was getting picked up, but we were not applying the instance_uuid_format/instance_format to it | 15:54 |
| dhellmann | actually, it looks like I was just assuming we did want it configurable: https://review.openstack.org/#/c/105362/ | 15:54 |
| dhellmann | sdague: ^^ | 15:54 |
| sdague | dhellmann: yeh, just found it | 15:54 |
| dhellmann | dims: ah, got it | 15:54 |
| dims | dhellmann: so the entire __repr__ was printed | 15:54 |
| sdague | that links to a blueprint that doesn't exist | 15:54 |
| *** salv-orlando has quit IRC | 15:55 | |
| dhellmann | dims: maybe we should think about generalizing that "how to format a special object" stuff so it's not just useful for instances | 15:55 |
| dims | dhellmann: ack | 15:55 |
| *** salv-orlando has joined #openstack-oslo | 15:55 | |
| dhellmann | sdague: the blueprint links in gerrit don't work for oslo because they aren't under the openstack project group | 15:56 |
| dims | sdague: https://review.openstack.org/#/c/160149/ should pick up the nova Depends-On: change and latest oslo.log trunk right? | 15:56 |
| *** harlowja_at_home has joined #openstack-oslo | 15:56 | |
| dims | sdague: am pulling in oslo.log from trunk using "LIBS_FROM_GIT=oslo.log" and the Depends-On points to nova chance | 15:57 |
| sdague | dims: so that assumes the changes are landed in oslo.log, right? | 15:58 |
| dims | sdague: the instance explosion, i landed it | 15:58 |
| sdague | ok | 15:58 |
| *** salv-orlando has quit IRC | 16:00 | |
| sdague | dhellmann: yeh, so it looks like the original blueprint only cared about .d support, not that it's configurable - https://blueprints.launchpad.net/oslo-incubator/+spec/policy-configuration-directories - maybe we could remove that config var, as it seems like knob that we're highly unlikely to want. | 16:02 |
| *** sputnik13 has joined #openstack-oslo | 16:06 | |
| *** kgiusti has joined #openstack-oslo | 16:06 | |
| *** sputnik13 has quit IRC | 16:07 | |
| dhellmann | sdague: makes sense | 16:07 |
| *** zzzeek has joined #openstack-oslo | 16:08 | |
| *** tsekiyama has joined #openstack-oslo | 16:23 | |
| *** daniel3_ has quit IRC | 16:24 | |
| *** boris-42 has quit IRC | 16:32 | |
| openstackgerrit | Sean Dague proposed openstack/oslo.policy: remove policy_dirs option https://review.openstack.org/160407 | 16:37 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a fully functional orderedset (WIP) https://review.openstack.org/160219 | 16:37 |
| sdague | dhellmann: well, proposed there | 16:37 |
| *** viktors is now known as viktors|afk | 16:43 | |
| openstackgerrit | Michael Bayer proposed openstack/oslo.db: Implement generic update-on-match feature https://review.openstack.org/146228 | 16:46 |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: NotifyPublisher need handle amqp_auto_delete https://review.openstack.org/138723 | 16:52 |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: NotifyPublisher need handle amqp_auto_delete https://review.openstack.org/138723 | 16:52 |
| *** e0ne is now known as e0ne_ | 16:55 | |
| *** e0ne_ is now known as e0ne | 16:56 | |
| dims | dhellmann: could use your eyes again - https://review.openstack.org/#/c/160348/ when you get a chance please | 17:00 |
| *** harlowja_at_home has quit IRC | 17:00 | |
| *** inc0 has quit IRC | 17:01 | |
| ihrachyshka | dhellmann, we could come up with some API that would generate something 'similar' to datetime, but not exactly the one :) | 17:01 |
| ihrachyshka | dhellmann, still, that would be a new API | 17:01 |
| ihrachyshka | but easier to adopt | 17:01 |
| *** prad has joined #openstack-oslo | 17:08 | |
| *** daniel3_ has joined #openstack-oslo | 17:11 | |
| *** e0ne has quit IRC | 17:14 | |
| *** yamahata has quit IRC | 17:16 | |
| jogo | ttx: nice | 17:16 |
| *** salv-orlando has joined #openstack-oslo | 17:16 | |
| *** pblaho__ is now known as pblaho | 17:17 | |
| dhellmann | ihrachyshka: I really don't think we want to reinvent datetime. Can we fix this upstream? | 17:21 |
| ihrachyshka | dhellmann, I suspect there is a reason for the module not to support it till now | 17:22 |
| *** noelbk has joined #openstack-oslo | 17:22 | |
| ihrachyshka | dhellmann, viktors|afk ideas on why datetime is not leap seconds aware (and documented as such)? ^^ | 17:22 |
| ihrachyshka | dhellmann, see first answer http://stackoverflow.com/questions/21027639/python-datetime-not-accounting-for-leap-second-properly | 17:23 |
| ihrachyshka | dhellmann, maybe we can use it as a workaround? :) | 17:24 |
| ihrachyshka | though it would just mean we don't ValueError (while still generating incorrect datetime object that does not correspond to real 'leaped' time) | 17:24 |
| *** vigneshvar has joined #openstack-oslo | 17:25 | |
| dhellmann | ihrachyshka: if the problem is just in the datetime class with the validation of seconds, that seems like a relatively easy fix? | 17:26 |
| ihrachyshka | dhellmann, it depends on whether we consider timeutils returning a 'slightly' incorrect datetime object as a problem | 17:27 |
| *** sputnik13 has joined #openstack-oslo | 17:27 | |
| dims | sdague: looks like the context stuff did the trick - http://logs.openstack.org/49/160149/6/check/check-devstack-dsvm-cells/3277ac0/logs/screen-n-api.txt.gz#_2015-03-02_17_18_44_426 | 17:27 |
| ihrachyshka | dhellmann, if we're ok with it consistently leaping for +1 sec for leap sec case, then ok | 17:27 |
| ihrachyshka | dhellmann, I guess the hack around is a good short term fix, but we may need to rethink the usage of datetime objects in long term (spec?) | 17:28 |
| dhellmann | ihrachyshka: whether we're ok with that would depend on the context, which is why I think we should just fix the stdlib module instead of working around it | 17:28 |
| dhellmann | that documentation message says the module "does not" support leap seconds not "will never" | 17:29 |
| ihrachyshka | dhellmann, I don't feel like diving into python dev community, but we'll see :) | 17:30 |
| dhellmann | ihrachyshka: as it happens, haypo is a python developer | 17:31 |
| ihrachyshka | right, someone with hands dirty in cpython would be good to have on board | 17:32 |
| haypo | ihrachyshka, dhellmann : datetime is not my favorite module :) too many users complain about it, because each user has a different use case :-p | 17:32 |
| dhellmann | haypo: in this case, we have a leap-second bug coming up later this year: https://bugs.launchpad.net/oslo.utils/+bug/1427212 | 17:32 |
| openstack | Launchpad bug 1427212 in oslo.utils "timeutils is not leap second safe" [Undecided,New] | 17:32 |
| ihrachyshka | haypo, "too many users complain about it" isn't it openstack way? | 17:32 |
| ihrachyshka | "each user has a different use case" that one is also pretty openstack | 17:32 |
| ihrachyshka | :) | 17:32 |
| dhellmann | haypo: we're trying to decide whether (and how) to work around it, or if we can/should fix it upstream (my preference) | 17:32 |
| haypo | last issue related to leap second: http://bugs.python.org/issue18718 -- doc fix to mention that datetime doesn't support leap seconds :-p | 17:33 |
| haypo | + Unlike the :mod:`time` module, the :mod:`datetime` module does not support | 17:33 |
| haypo | + leap seconds. | 17:33 |
| dhellmann | haypo: yeah, that won't help us here because we might end up with a crashing server | 17:33 |
| ihrachyshka | that's an easy fix :) | 17:33 |
| dhellmann | haypo: do you know if it doesn't support them out of principle, or just because no one has made it work yet? | 17:34 |
| haypo | dhellmann: no idea | 17:34 |
| dhellmann | haypo: ok | 17:34 |
| ihrachyshka | haypo, do you have people on your mind to ask? | 17:35 |
| haypo | dhellmann: which result do you expect for datetime(2012,6,30,23,59,60)-datetime(2012,6,30,23,59,59) ? | 17:36 |
| haypo | ihrachyshka: sure, alexander beloposky aka belopolsky | 17:36 |
| haypo | ihrachyshka: you may start by opening a issue to request the feature | 17:36 |
| haypo | ihrachyshka: belopolsky is our time expect :) | 17:36 |
| ihrachyshka | haypo, hm, 1 sec diff, isn't it?? | 17:37 |
| ihrachyshka | haypo, ack, will do | 17:37 |
| dhellmann | haypo: I expect it to be what's passed in, I guess. I mean, that's the actual time value, right? | 17:37 |
| otherwiseguy | dhellmann: Just added you as a reviewer for https://review.openstack.org/#/c/160130/ . I'm mplementing rootwrap daemon mode support for neutron and running into some tricky environment variable stuff and would appreciate your input. | 17:37 |
| dhellmann | otherwiseguy: ok, ttx is our expert for rootwrap | 17:37 |
| *** vigneshvar has quit IRC | 17:37 | |
| haypo | ihrachyshka: ok, so what do you expect for datetime(2012,7,1,0,0,0)-datetime(2012,6,30,23,59,59) ? 1 or 2 seconds? | 17:37 |
| dhellmann | haypo: oh, your message wrapped and I didn't notice that it was a computation not just one time | 17:38 |
| ihrachyshka | I guess 2 secs (assuming it's the moment the leap second wasadded) | 17:38 |
| haypo | i don't know how other time modules handle leap seconds | 17:38 |
| otherwiseguy | dhellmann: great, thanks. just did a git blame on daemon.py and it showed all you, didn't notice it was a code move. :) | 17:38 |
| ihrachyshka | haypo, I suspect leap seconds are stoped in e.g. tzdata (?) and the data should be crawled each time we do such calculations | 17:39 |
| ihrachyshka | *stored | 17:39 |
| *** alexpilotti has joined #openstack-oslo | 17:39 | |
| haypo | ihrachyshka: datetime has a poor support for timezones | 17:39 |
| ihrachyshka | or we rely on proper leap aware functions from e.g. POSIX under the hood | 17:39 |
| dhellmann | otherwiseguy: yuriy taraday is another good reviewer for that, since he did the work on daemon mode | 17:40 |
| haypo | ihrachyshka: it's not a good idea to rely on the system if you want portability. we had a *lot* of issues with windows, solaris and aix | 17:40 |
| ihrachyshka | haypo, I suppose they also have some platform specific way to get the data | 17:40 |
| ihrachyshka | or screw them :P | 17:40 |
| haypo | example: https://hg.python.org/cpython/file/b22755f8ab5f/Modules/timemodule.c#l582 | 17:40 |
| otherwiseguy | dhellmann: you wouldn't happen to know if he hangs out on IRC/how to get ahold of him? | 17:41 |
| dhellmann | otherwiseguy: I don't remember his irc nick :-/ | 17:41 |
| ihrachyshka | otherwiseguy, email? | 17:41 |
| dhellmann | otherwiseguy: gerrit should do tab completion on his name for you, though | 17:41 |
| dims | dhellmann: sdague: mriedem: we should roll out the oslo.log release for that "instance" show stopper | 17:41 |
| haypo | ihrachyshka: the old https://www.python.org/dev/peps/pep-0431/ may enhance timezone support, but it's still a draft | 17:41 |
| haypo | ihrachyshka: seriously, datetime is a just a can of warms. be careful :) | 17:42 |
| haypo | of worms | 17:42 |
| otherwiseguy | dhellmann: Yeah, I added him to the review first and hadn't heard anything. Just wondered about some real-time communication. ;) | 17:42 |
| dhellmann | dims: how about a +2a on https://review.openstack.org/#/c/147312/ | 17:42 |
| ihrachyshka | SCARY | 17:42 |
| dhellmann | otherwiseguy: ah, ok | 17:42 |
| dims | dhellmann: ack | 17:43 |
| otherwiseguy | ihrachyshka: good point. I always forget that email thing. | 17:43 |
| otherwiseguy | :p | 17:43 |
| ihrachyshka | :) | 17:43 |
| ihrachyshka | yeah, it's so 90s | 17:43 |
| ihrachyshka | we're in TDD world (twitter driven development), right? | 17:44 |
| haypo | ihrachyshka: i may help you to support leap seconds in datetime, but it doesn't look to be easy to add it today. it might break the backward compatibility... | 17:44 |
| *** yamahata has joined #openstack-oslo | 17:44 | |
| dhellmann | dims: oh, I see jecarey had comments on that patch. We can merge it after the release | 17:44 |
| ihrachyshka | haypo, sounds like a piece too huge for me to bite on the first sight, though I'll check later myself | 17:45 |
| dims | dhellmann: y, i was reading the rendered document, reads well | 17:45 |
| dhellmann | dims: I'll go ahead and tag 0.4.0 | 17:45 |
| haypo | ihrachyshka: just open an issue at bugs.python.org ;) | 17:45 |
| ihrachyshka | haypo, and thanks in advance for help | 17:45 |
| dhellmann | then update that doc patch | 17:45 |
| *** ChuckC has quit IRC | 17:45 | |
| dims | dhellmann: ok | 17:46 |
| *** vigneshvar has joined #openstack-oslo | 17:46 | |
| * ihrachyshka leaves home. cu all | 17:46 | |
| dims | dhellmann: do you know if other projects have custom RequestContext classes? we may want to add a note about that too | 17:47 |
| dhellmann | dims: good point. I mean, the whole point of providing the context library was for this, but yeah. | 17:47 |
| dims | y | 17:47 |
| dhellmann | dims, mriedem, sdague : oslo.log 0.4.0 released | 17:49 |
| *** ihrachyshka has quit IRC | 17:50 | |
| *** i159 has quit IRC | 17:51 | |
| openstackgerrit | Doug Hellmann proposed openstack/oslo.log: Add migration notes https://review.openstack.org/147312 | 17:56 |
| *** boris-42 has joined #openstack-oslo | 18:05 | |
| *** stpierre has joined #openstack-oslo | 18:07 | |
| openstackgerrit | Eric Brown proposed openstack/oslo.config: Typo in StrOpt docstring: Integer to String https://review.openstack.org/159698 | 18:14 |
| *** harlowja_away is now known as harlowja_ | 18:21 | |
| therve | dims, Heat has one at least | 18:27 |
| *** dtantsur is now known as dtantsur|afk | 18:28 | |
| dims | therve: thanks, we should make sure it inherits from the oslo_context one | 18:29 |
| *** achanda has joined #openstack-oslo | 18:31 | |
| therve | Yeah it does | 18:31 |
| dims | therve: cool | 18:33 |
| openstackgerrit | Sean Dague proposed openstack/oslo.policy: remove policy_dirs option https://review.openstack.org/160407 | 18:37 |
| *** achanda has quit IRC | 18:46 | |
| *** achanda has joined #openstack-oslo | 18:46 | |
| *** crc32 has joined #openstack-oslo | 18:55 | |
| *** devlaps has joined #openstack-oslo | 19:01 | |
| dhellmann | sdague: I'm not sure how to go about deprecating policy_dirs. If we just remove it, there's no warning to the user. | 19:01 |
| dhellmann | bnemec, you did some work on deprecating opts recently, thoughts? ^^ | 19:03 |
| *** crc32 has quit IRC | 19:03 | |
| bnemec | Hmm, this is a little different. That was for renamed opts. | 19:06 |
| dhellmann | so what's the policy for options we're no longer using at all? | 19:06 |
| bnemec | But we could put some logic in oslo.policy to log if the opt is set. | 19:06 |
| dhellmann | it seems especially important given that this is the option potentially telling us where to look for a file that we might now not be loading at all | 19:07 |
| bnemec | Yeah | 19:07 |
| bnemec | As a one-off we could just default the opt to None in oslo.policy and log if someone overrides that, and use policy.d otherwise. | 19:09 |
| bnemec | A formal "this opt is deprecated for removal" in oslo.config might be good though. | 19:09 |
| dhellmann | it would be useful to have a "deprecated and no longer used" feature in oslo.config, too | 19:09 |
| bnemec | Heh | 19:09 |
| *** ChuckC has joined #openstack-oslo | 19:11 | |
| bnemec | dhellmann: I'll take a look at adding something like that to oslo.config. Shouldn't be too hard. | 19:11 |
| bnemec | Of course, famous last words and all that. :-) | 19:11 |
| *** crc32 has joined #openstack-oslo | 19:12 | |
| dhellmann | bnemec: sounds good, thanks | 19:12 |
| *** cdent has quit IRC | 19:20 | |
| *** shakamunyi has joined #openstack-oslo | 19:20 | |
| *** devlaps has quit IRC | 19:29 | |
| *** e0ne has joined #openstack-oslo | 19:32 | |
| *** e0ne is now known as e0ne_ | 19:32 | |
| *** e0ne_ is now known as e0ne | 19:32 | |
| *** himangi has left #openstack-oslo | 19:35 | |
| *** stpierre has quit IRC | 19:44 | |
| *** stpierre has joined #openstack-oslo | 19:46 | |
| *** diga has quit IRC | 19:49 | |
| *** crc32 has quit IRC | 19:56 | |
| openstackgerrit | Ben Nemec proposed openstack/oslo.config: Add ability to deprecate opts for removal https://review.openstack.org/160475 | 19:57 |
| bnemec | dhellmann: ^Should do the trick | 19:57 |
| dhellmann | bnemec: should we tie into versionutils there to say when it will be removed? | 19:58 |
| *** e0ne is now known as e0ne_ | 19:59 | |
| bnemec | dhellmann: I wasn't sure. Libs don't have a concept of "deprecated in K, will be removed in L". | 19:59 |
| dhellmann | bnemec: now that we're capping stable branches, the libs can know about versions | 20:00 |
| dhellmann | we can make that enhancement later, because we still have to figure out what we're doing with verisonutils | 20:00 |
| *** e0ne has joined #openstack-oslo | 20:00 | |
| bnemec | Yeah, and we can't use versionutils in oslo.config. | 20:01 |
| bnemec | So we'd have to reimplement all of that here. | 20:01 |
| dhellmann | well, we can if we always leave it in the incubator, which is looking like what we'll have to do | 20:01 |
| bnemec | Ah :-( | 20:01 |
| dhellmann | that's not a big deal, it's going to change once or twice per cycle so it's pretty darn stable | 20:02 |
| dhellmann | but as a matter of dependency management, there's no useful way to put it in a library | 20:02 |
| bnemec | dhellmann: Even if we don't use it in oslo.config? | 20:03 |
| dhellmann | no, but that's one of the main places we want to be able to use it | 20:03 |
| sdague | dhellmann: it's only because of the fatal deprecations option right? | 20:03 |
| dhellmann | sdague: that's what introduces the circular dependency, yes | 20:04 |
| openstackgerrit | Doug Hellmann proposed openstack/oslo-incubator: Add liberty release name to versionutils https://review.openstack.org/160479 | 20:05 |
| sdague | so what about combining a couple of these foundational libraries to get around that? | 20:05 |
| sdague | or having fatal deprecations be env triggered for version utils, and put the conf option in oslo.config that just sets the env variable | 20:07 |
| *** achanda has quit IRC | 20:07 | |
| dhellmann | it's not just fatal configuration deprecations though, right? it's any deprecation? | 20:08 |
| dhellmann | so then we have oslo.config controlling the behavior of versionutils | 20:08 |
| sdague | sure | 20:08 |
| *** achanda has joined #openstack-oslo | 20:08 | |
| dhellmann | this isn't the problem I'm trying to solve right now, so I don't want to unload what I'm working on to think about this | 20:08 |
| sdague | ok, no prob | 20:08 |
| *** saikrishna has joined #openstack-oslo | 20:09 | |
| dhellmann | bnemec: that deprecation warning is only going to be triggered when some code tries to read the config value, right? | 20:09 |
| dhellmann | but that won't be the case for an option we're removing, will it? or is it going to be triggered when the options are logged? | 20:09 |
| bnemec | dhellmann: Yeah, it's the same problem as the previous change - at parse time we don't have the Opt definition so we don't know whether it's deprecated or not. | 20:10 |
| dhellmann | yeah, we relied on the logging behavior for that, too | 20:10 |
| bnemec | It will be triggered when the options are logged in debug, but that requires the debug firehose to be turned on. | 20:10 |
| dhellmann | ok, just verifying | 20:10 |
| dhellmann | oh, hrm | 20:10 |
| dhellmann | and I guess we can't trigger when the option is registered because the file hasn't been loaded yet | 20:11 |
| bnemec | Right. | 20:11 |
| dhellmann | :-| | 20:11 |
| dhellmann | I suppose logging for debug is better than nothing | 20:11 |
| bnemec | Basically we can't rely on the file being loaded when the opt is created, or vice versa because they can both happen asynchronously. | 20:11 |
| bnemec | I suppose we could always call the debug logging thing and give it a silent flag or something. | 20:13 |
| dhellmann | when would we call it? :-) | 20:13 |
| bnemec | In fact, I wonder if that happens anyway. It should be safe to call LOG.debug when the log level is set higher than that? | 20:13 |
| dhellmann | yeah, the messages are just ignored | 20:13 |
| dhellmann | oh, right, so maybe it will log deprecations anyway | 20:13 |
| * dhellmann hates side-effects | 20:14 | |
| bnemec | You'd have to do it any time an opt is registered or we run conf() | 20:14 |
| dhellmann | we could track deprecated options before __call__() happens and look for them when the file is read, then when new options are registered we could notice that we have the file and a deprecated option registration and do something special | 20:15 |
| dhellmann | but that's going to be more work | 20:15 |
| dhellmann | small steps, let's take this version first | 20:15 |
| bnemec | It's amazing how complicated it is to parse an ini file. :-) | 20:15 |
| dhellmann | it's amazing how complicated *we make it* to parse an ini file | 20:16 |
| bnemec | Heh, true | 20:16 |
| dhellmann | if we didn't let code anywhere in the system register a configuration option we'd be better off in this regard | 20:16 |
| dhellmann | of course other things would be harder :-) | 20:16 |
| bnemec | Probably also worth noting that if someone manages to run a service without ever referencing the deprecated opt, it probably doesn't actually matter to them if it goes away. | 20:17 |
| * bnemec is way overdue for lunch | 20:18 | |
| dhellmann | bnemec: but if we remove the opt, as in this case, then nothing will refer to it | 20:20 |
| bnemec | dhellmann: Right, but we can't do that if we're going to deprecate it first. | 20:26 |
| *** vigneshvar has quit IRC | 20:33 | |
| dhellmann | bnemec: true | 20:38 |
| dims | dhellmann: requirements bump please - https://review.openstack.org/#/c/160435/ | 20:41 |
| *** achanda has quit IRC | 20:52 | |
| *** shakamunyi has quit IRC | 21:07 | |
| *** shakamunyi has joined #openstack-oslo | 21:08 | |
| *** pmathews has joined #openstack-oslo | 21:19 | |
| *** saikrishna has quit IRC | 21:20 | |
| *** jungleboyj has quit IRC | 21:20 | |
| *** SergeyLukjanov has quit IRC | 21:24 | |
| *** amrith is now known as _amrith_ | 21:25 | |
| dims | dhellmann: something weird in amqp logging - http://logs.openstack.org/48/160348/9/gate/gate-tempest-dsvm-large-ops/ad9ae21/logs/screen-n-cond.txt.gz#_2015-03-02_20_04_09_069 | 21:32 |
| *** kgiusti has quit IRC | 21:41 | |
| *** ChuckC has quit IRC | 21:44 | |
| *** e0ne is now known as e0ne_ | 21:46 | |
| *** achanda has joined #openstack-oslo | 21:46 | |
| *** e0ne_ is now known as e0ne | 21:48 | |
| openstackgerrit | Michael Bayer proposed openstack/oslo.db: Implement new oslo.db.sqlalchemy.enginefacade module https://review.openstack.org/138215 | 21:49 |
| *** achanda has quit IRC | 21:52 | |
| *** shakamunyi_ has joined #openstack-oslo | 21:53 | |
| *** shakamunyi has quit IRC | 21:54 | |
| *** jaosorior has quit IRC | 22:02 | |
| *** jecarey has quit IRC | 22:08 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a fully functional orderedset https://review.openstack.org/160219 | 22:10 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a fully functional orderedset https://review.openstack.org/160219 | 22:12 |
| *** andreykurilin_ has joined #openstack-oslo | 22:18 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a frozen checking decorator https://review.openstack.org/160536 | 22:21 |
| *** e0ne has quit IRC | 22:23 | |
| *** jungleboyj has joined #openstack-oslo | 22:26 | |
| *** e0ne has joined #openstack-oslo | 22:27 | |
| harlowja_ | dhellmann yt | 22:29 |
| harlowja_ | for https://review.openstack.org/#/c/156711/ and https://review.openstack.org/#/c/153298/ (involving haypo ) what the heck is expected there :-P | 22:29 |
| harlowja_ | meta-spec ;-/ | 22:29 |
| harlowja_ | whats that, lol | 22:30 |
| * harlowja_ doesn't quite understand why 2 specs are bad | 22:30 | |
| harlowja_ | they are sorta different :-P | 22:30 |
| harlowja_ | *if u have any idea, if not thats ok, will bother thierry | 22:31 |
| *** jecarey has joined #openstack-oslo | 22:32 | |
| *** andreykurilin_ has quit IRC | 22:33 | |
| *** stevemar has quit IRC | 22:33 | |
| *** andreykurilin_ has joined #openstack-oslo | 22:34 | |
| *** e0ne has quit IRC | 22:34 | |
| *** SergeyLukjanov has joined #openstack-oslo | 22:37 | |
| *** e0ne has joined #openstack-oslo | 22:38 | |
| *** mtanino has quit IRC | 22:39 | |
| *** _amrith_ is now known as amrith | 22:39 | |
| *** alexpilotti has quit IRC | 22:41 | |
| *** achanda has joined #openstack-oslo | 22:42 | |
| *** mtanino has joined #openstack-oslo | 22:47 | |
| *** e0ne has quit IRC | 22:49 | |
| *** e0ne has joined #openstack-oslo | 22:53 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add a fully functional orderedset https://review.openstack.org/160219 | 22:59 |
| *** andreykurilin_ has quit IRC | 23:00 | |
| *** daniel3_ has quit IRC | 23:03 | |
| *** shakamunyi_ has quit IRC | 23:06 | |
| dhellmann | dims: +2a | 23:07 |
| dhellmann | dims: are you talking about the log line or the traceback that follows it? | 23:08 |
| *** bknudson has quit IRC | 23:08 | |
| dims | dhellmann: the traceback after that. i think it be because of the dict<->object round trip, hoping this will fix it dhellmann - https://review.openstack.org/#/c/160535/ | 23:08 |
| dhellmann | dims: I wonder if the request is not be reconstructed properly after it comes over the wire | 23:08 |
| dims | y | 23:08 |
| dhellmann | dims: I'll bet that will fix it | 23:10 |
| dims | cross my fingers | 23:10 |
| dims | ttyl | 23:10 |
| *** salv-orlando has quit IRC | 23:20 | |
| openstackgerrit | Merged openstack/taskflow: Bring over pretty_tox.sh from nova/heat/others https://review.openstack.org/158106 | 23:20 |
| *** salv-orlando has joined #openstack-oslo | 23:21 | |
| openstackgerrit | Merged openstack/taskflow: Stick to one space after a period https://review.openstack.org/158980 | 23:22 |
| *** jgrimm is now known as zz_jgrimm | 23:26 | |
| dhellmann | harlowja_: I was hoping to get pretty_tox.sh into oslotest so we don't have to copy it all over the place https://review.openstack.org/#/c/151757/ | 23:28 |
| harlowja_ | ah | 23:29 |
| harlowja_ | cool | 23:29 |
| harlowja_ | might as well | 23:29 |
| harlowja_ | good to know | 23:29 |
| dhellmann | although that apparently needs a rebase and to pull the right version of the script body | 23:29 |
| harlowja_ | :-P | 23:30 |
| * dhellmann has had an unusually high number of interruptions lately | 23:31 | |
| *** prad has quit IRC | 23:31 | |
| *** gordc has quit IRC | 23:31 | |
| harlowja_ | damn HP people leaving cores and such | 23:32 |
| harlowja_ | lol | 23:32 |
| harlowja_ | and interrupting dhellmann | 23:32 |
| harlowja_ | lol | 23:32 |
| dhellmann | harlowja_: heh | 23:32 |
| openstackgerrit | Doug Hellmann proposed openstack/oslotest: Import nova's wrapper for nice testr output https://review.openstack.org/151757 | 23:35 |
| dhellmann | mtreinish: what's the status of splitting out the subunit-trace command from tempest-lib to make it possible to pip install it separately? can I help with that? | 23:37 |
| mtreinish | dhellmann: I had completely forgotten about that | 23:38 |
| dhellmann | mtreinish: I'd be happy to do the work, if you like, but you're more familiar with the code. | 23:38 |
| *** openstackgerrit has quit IRC | 23:38 | |
| *** openstackgerrit has joined #openstack-oslo | 23:38 | |
| dhellmann | mtreinish: I could also file a bug/blueprint if this is a bad time to do the work (dunno what your work load is like right now) | 23:39 |
| *** stpierre has quit IRC | 23:39 | |
| mtreinish | dhellmann: well I'm not oppossed to moving it into it's own thing. BUt what was the motivation for doing it again | 23:41 |
| *** stpierre has joined #openstack-oslo | 23:41 | |
| dhellmann | mtreinish: I would like to be able to deliver that wrapper script from ^^ in a package of some sort instead of copying it around into various projects. I had planned to put it into oslotest, but it could also go into the package with the lib itself, I guess. However, no oslo projects can depend on tempest-lib because tempest-lib depends on some of the oslo projects... | 23:42 |
| dhellmann | so I was hoping to be able to carve that one piece of functionality out of tempest-lib to make it installable on its own so we can use it in oslo | 23:42 |
| mtreinish | dhellmann: just oslo.log I think, but sure | 23:42 |
| dhellmann | yeah, I wasn't sure about config and frankly didn't want to have to keep up with the list :-) | 23:42 |
| mtreinish | fwiw, the one source of pretty_tox.sh might not work, because there are local differences between some projects' versions of it | 23:43 |
| dhellmann | there don't need to be, though, right? | 23:43 |
| dhellmann | a version could be shared as long as the projects were willing to share it? | 23:44 |
| mtreinish | well, for the most part. At it's very basic it's just a pipe testr. Some projects have a workaround for until-failure. Others add some output to the end of the run | 23:44 |
| dhellmann | I don't care about one-offs. We can just put the script in oslotest if we have to, but I don't want 27 copies in all of the oslo repos being different. | 23:45 |
| mtreinish | dhellmann: it really doesn't need to be a separate script to just run tox jobs, because you can just do bash -c "testr run --subunit | subunit-trace" | 23:45 |
| dhellmann | other projects can have their own, or use ours, or not use it at all | 23:45 |
| mtreinish | dhellmann: ok sure | 23:45 |
| *** mriedem is now known as mriedem_away | 23:46 | |
| mtreinish | dhellmann: I can try to do the subunit-trace split tomorrow (if you ping me in the morning to remind me :) ) | 23:47 |
| dhellmann | mtreinish: will do | 23:47 |
| *** jungleboyj has quit IRC | 23:47 | |
| dhellmann | thanks! | 23:47 |
| mtreinish | it shouldn't be hard, just spin up a cookiecutter and copy the one file and add an entry point | 23:47 |
| dhellmann | mtreinish: sure, or you can use oslo-incubator/tools/filter_git_history.sh if you want to preserve the history | 23:48 |
| dhellmann | you still need the cookiecutter stuff after that, of course | 23:48 |
| dhellmann | mtreinish: ok, time for me to sign off, I'll talk to you about it more in the morning | 23:49 |
| mtreinish | dhellmann: oh yeah, I guess I should try to preserve the history | 23:49 |
| mtreinish | dhellmann: ok cya | 23:49 |
| mtreinish | dhellmann: oh also fwiw, the issue I put in the earlier review I've since fixed in subunit-trace so you didn't have to update it :) | 23:50 |
| dhellmann | mtreinish: oh, well :-) | 23:51 |
| *** salv-orlando has quit IRC | 23:51 | |
| dhellmann | mtreinish: does this look like what we need? https://github.com/dhellmann/subunit-trace-import | 23:52 |
| dhellmann | mtreinish: obviously the setup files need to be modified, and we need the cookiecutter stuff applied, but I thought that could happen after an import | 23:53 |
| dhellmann | I don't know if we need any other files besides that one script | 23:53 |
| mtreinish | dhellmann: nope that should be it | 23:55 |
| mtreinish | that looks like all that's needed | 23:55 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use ordered(set/dict) to retain ordering of requires/optional https://review.openstack.org/160558 | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!