*** manishg has quit IRC | 00:18 | |
*** ozamiatin has joined #openstack-oslo | 00:38 | |
*** sputnik13 has quit IRC | 00:45 | |
*** ozamiatin has quit IRC | 01:00 | |
*** ozamiatin has joined #openstack-oslo | 01:01 | |
*** ozamiatin has quit IRC | 01:43 | |
*** r0j4z0 has quit IRC | 02:04 | |
*** dimsum__ has quit IRC | 02:22 | |
*** dimsum__ has joined #openstack-oslo | 02:36 | |
*** vilobhmm has joined #openstack-oslo | 02:55 | |
*** dimsum__ has quit IRC | 02:58 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo.concurrency: Updated from global requirements https://review.openstack.org/191613 | 03:12 |
---|---|---|
openstackgerrit | OpenStack Proposal Bot proposed openstack/taskflow: Updated from global requirements https://review.openstack.org/191614 | 03:13 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/tooz: Updated from global requirements https://review.openstack.org/191615 | 03:13 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/automaton: Updated from global requirements https://review.openstack.org/191616 | 03:13 |
*** tobe has joined #openstack-oslo | 03:20 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo.cache: Updated from global requirements https://review.openstack.org/191617 | 03:20 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo.reports: Updated from global requirements https://review.openstack.org/191618 | 03:20 |
*** salv-orl_ has quit IRC | 03:55 | |
*** vilobhmm has quit IRC | 03:55 | |
*** amotoki has joined #openstack-oslo | 03:56 | |
*** dimsum__ has joined #openstack-oslo | 03:58 | |
*** dimsum__ has quit IRC | 04:03 | |
*** amotoki has quit IRC | 04:04 | |
*** achanda has joined #openstack-oslo | 04:07 | |
*** amotoki has joined #openstack-oslo | 04:08 | |
*** achanda has quit IRC | 04:17 | |
*** amotoki has quit IRC | 04:31 | |
*** takedakn has joined #openstack-oslo | 04:35 | |
*** takedakn has quit IRC | 04:39 | |
*** rpodolyaka1 has joined #openstack-oslo | 04:40 | |
*** amotoki has joined #openstack-oslo | 04:41 | |
*** rpodolyaka1 has quit IRC | 04:49 | |
*** manishg has joined #openstack-oslo | 04:53 | |
*** arnaud___ has joined #openstack-oslo | 04:54 | |
*** sdake_ has joined #openstack-oslo | 04:56 | |
*** sdake has quit IRC | 04:59 | |
*** rpodolyaka1 has joined #openstack-oslo | 05:08 | |
*** nkrinner has joined #openstack-oslo | 05:20 | |
openstackgerrit | Merged openstack/taskflow: Updated from global requirements https://review.openstack.org/191614 | 05:28 |
*** manishg has quit IRC | 05:46 | |
*** rpodolyaka1 has quit IRC | 05:51 | |
*** r0j4z0 has joined #openstack-oslo | 05:52 | |
*** tteggel has quit IRC | 05:59 | |
*** tteggel has joined #openstack-oslo | 06:01 | |
*** Kennan2 has joined #openstack-oslo | 06:03 | |
*** Kennan has quit IRC | 06:04 | |
*** vilobhmm has joined #openstack-oslo | 06:12 | |
*** Kennan2 is now known as Kennan | 06:14 | |
*** vilobhmm has quit IRC | 06:16 | |
*** vilobhmm has joined #openstack-oslo | 06:17 | |
*** ihrachyshka has joined #openstack-oslo | 06:23 | |
openstackgerrit | Lei Zhang proposed openstack/oslotest: Avoid automatic mock replacements https://review.openstack.org/191665 | 06:26 |
*** shardy has joined #openstack-oslo | 06:29 | |
*** pblaho has joined #openstack-oslo | 06:31 | |
*** takedakn has joined #openstack-oslo | 06:39 | |
*** ihrachyshka has quit IRC | 06:42 | |
*** tobe has quit IRC | 06:44 | |
*** arnaud___ has quit IRC | 06:44 | |
*** takedakn has quit IRC | 06:46 | |
*** sdake has joined #openstack-oslo | 06:56 | |
*** sdake has quit IRC | 06:57 | |
*** sdake has joined #openstack-oslo | 06:57 | |
*** sdake_ has quit IRC | 07:00 | |
*** rpodolyaka1 has joined #openstack-oslo | 07:00 | |
*** tobe has joined #openstack-oslo | 07:07 | |
*** dulek has joined #openstack-oslo | 07:21 | |
*** prad has quit IRC | 07:27 | |
*** subscope has joined #openstack-oslo | 07:34 | |
*** ndipanov has joined #openstack-oslo | 07:42 | |
*** prad has joined #openstack-oslo | 07:42 | |
*** rpodolyaka1 has quit IRC | 07:47 | |
*** salv-orlando has joined #openstack-oslo | 07:48 | |
*** browne has quit IRC | 07:50 | |
*** r0j4z0 has quit IRC | 07:55 | |
*** pblaho has quit IRC | 07:59 | |
*** ozamiatin has joined #openstack-oslo | 08:01 | |
*** sdake has quit IRC | 08:06 | |
*** ihrachyshka has joined #openstack-oslo | 08:10 | |
*** haypo has joined #openstack-oslo | 08:11 | |
*** vilobhmm has quit IRC | 08:11 | |
*** haypo has quit IRC | 08:11 | |
*** pblaho has joined #openstack-oslo | 08:14 | |
*** yassine has joined #openstack-oslo | 08:23 | |
*** rpodolyaka1 has joined #openstack-oslo | 08:23 | |
*** pblaho has quit IRC | 08:29 | |
*** shardy_ has joined #openstack-oslo | 08:38 | |
*** shardy has quit IRC | 08:39 | |
*** takedakn has joined #openstack-oslo | 08:42 | |
*** Daviey has joined #openstack-oslo | 08:42 | |
*** ozamiatin has quit IRC | 08:43 | |
*** takedakn has quit IRC | 08:43 | |
*** shardy_ has quit IRC | 08:44 | |
*** shardy has joined #openstack-oslo | 08:44 | |
*** ozamiatin has joined #openstack-oslo | 08:45 | |
*** r0j4z0 has joined #openstack-oslo | 08:47 | |
*** subscope has quit IRC | 08:54 | |
*** rpodolyaka1 has quit IRC | 08:57 | |
*** rpodolyaka1 has joined #openstack-oslo | 08:59 | |
openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/oslo.messaging: Disable aioeventlet tests for py3 https://review.openstack.org/190993 | 09:01 |
*** ozamiatin has quit IRC | 09:03 | |
*** pblaho has joined #openstack-oslo | 09:03 | |
*** pblaho has quit IRC | 09:17 | |
*** e0ne has joined #openstack-oslo | 09:30 | |
*** rushiagr_away is now known as rushiagr | 09:34 | |
*** e0ne has quit IRC | 09:40 | |
*** ozamiatin has joined #openstack-oslo | 09:44 | |
*** e0ne has joined #openstack-oslo | 09:47 | |
*** salv-orl_ has joined #openstack-oslo | 09:51 | |
*** salv-orlando has quit IRC | 09:52 | |
*** f13o has joined #openstack-oslo | 09:56 | |
*** Kennan2 has joined #openstack-oslo | 10:03 | |
*** Kennan has quit IRC | 10:05 | |
*** dimsum__ has joined #openstack-oslo | 10:09 | |
*** ozamiatin has quit IRC | 10:10 | |
*** arnaud___ has joined #openstack-oslo | 10:11 | |
*** rpodolyaka1 has quit IRC | 10:15 | |
*** arnaud___ has quit IRC | 10:15 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/oslo-incubator: Just use os.urandom instead of Crypto.Random https://review.openstack.org/191581 | 10:27 |
*** ozamiatin has joined #openstack-oslo | 10:30 | |
*** e0ne is now known as e0ne_ | 10:32 | |
*** r0j4z0 has quit IRC | 10:36 | |
*** e0ne_ has quit IRC | 10:39 | |
*** e0ne has joined #openstack-oslo | 10:41 | |
*** jaypipes has joined #openstack-oslo | 11:00 | |
*** rushiagr is now known as rushiagr_away | 11:15 | |
*** dguitarbite has joined #openstack-oslo | 11:16 | |
*** pblaho has joined #openstack-oslo | 11:18 | |
*** rpodolyaka1 has joined #openstack-oslo | 11:20 | |
*** shardy_ has joined #openstack-oslo | 11:22 | |
*** shardy has quit IRC | 11:24 | |
*** tobe has quit IRC | 11:25 | |
*** rushiagr_away is now known as rushiagr | 11:27 | |
*** shardy_ has quit IRC | 11:28 | |
*** r0j4z0 has joined #openstack-oslo | 11:29 | |
*** shardy has joined #openstack-oslo | 11:29 | |
*** e0ne is now known as e0ne_ | 11:36 | |
openstackgerrit | Elena Ezhova proposed openstack/oslo.service: Add documentation for the service module https://review.openstack.org/191768 | 11:38 |
*** e0ne_ is now known as e0ne | 11:39 | |
openstackgerrit | Elena Ezhova proposed openstack/oslo.service: Add documentation for the service module https://review.openstack.org/191768 | 11:41 |
*** sdake has joined #openstack-oslo | 11:54 | |
*** sdake has quit IRC | 11:55 | |
*** sdake has joined #openstack-oslo | 11:55 | |
*** pblaho has quit IRC | 11:55 | |
*** stpierre has joined #openstack-oslo | 12:04 | |
*** rpodolyaka1 has quit IRC | 12:07 | |
*** rpodolyaka1 has joined #openstack-oslo | 12:11 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 12:16 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 12:16 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 12:17 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 12:17 |
*** amrith is now known as _amrith_ | 12:19 | |
*** e0ne is now known as e0ne_ | 12:28 | |
*** e0ne_ is now known as e0ne | 12:28 | |
*** prad has quit IRC | 12:30 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/oslo-incubator: Crypto.Random.atfork before fetching random bytes https://review.openstack.org/191581 | 12:31 |
*** gentux has joined #openstack-oslo | 12:36 | |
gentux | Hi there, I got a question from another team in my company : They want oslo.config to load variable from environment variable | 12:42 |
*** shardy_ has joined #openstack-oslo | 12:43 | |
gentux | is there anything is oslo to do something like that ? | 12:43 |
*** shardy has quit IRC | 12:47 | |
*** bknudson has joined #openstack-oslo | 12:48 | |
ihrachyshka | gentux, why not just reading the variable? | 12:50 |
*** shardy_ has quit IRC | 12:50 | |
ihrachyshka | I mean, directly | 12:50 |
*** shardy has joined #openstack-oslo | 12:51 | |
*** dimsum__ has quit IRC | 12:52 | |
*** f13o has quit IRC | 12:54 | |
*** dimsum__ has joined #openstack-oslo | 12:54 | |
*** dimsum__ is now known as dims | 12:55 | |
gentux | They want oslo.config to read variable directly from OS environment vars instead of configuration *.ini files | 12:55 |
dhellmann | gentux: we don't have anything like that right now. Which option do they want to control via the environment? | 12:55 |
*** exploreshaifali has joined #openstack-oslo | 12:55 | |
*** dims has quit IRC | 12:55 | |
*** dims has joined #openstack-oslo | 12:56 | |
gentux | that's the problem, they want to generalize this concept and be able to load any configuration value from OS instead of ini file. | 12:56 |
*** shardy has quit IRC | 12:56 | |
gentux | It seems some projet made a simple script to generate ini file from OS environment and then merge two ini files | 12:57 |
dhellmann | ok, the config parser supports some constructs that would be very difficult to express as environment variables | 12:57 |
dhellmann | you don't need to merge the files if you use the --config-dir option | 12:57 |
dhellmann | gentux: how are the environment variables set in the first place? | 12:58 |
gentux | I'll ask them, I'll be back in 5min | 13:00 |
*** shardy has joined #openstack-oslo | 13:01 | |
*** cdent has joined #openstack-oslo | 13:08 | |
*** jaosorior has joined #openstack-oslo | 13:10 | |
*** rushiagr is now known as rushiagr_away | 13:14 | |
gentux | well, it's not that clear. They are building a proof of concept close to what kolla does | 13:14 |
gentux | http://lists.openstack.org/pipermail/openstack-dev/2015-February/056144.html | 13:14 |
sdake | gentux i think we are abanonding that idea | 13:15 |
sdake | fyi :) | 13:15 |
gentux | sdake: I can understand that, the more I think about the more I find that quite hard to implement | 13:15 |
sdake | we have an implementation | 13:16 |
gentux | I'll dig it a little more, if I have some good information for you I'll share it here | 13:16 |
gentux | sdake: do you know where I can find this implementation ? | 13:17 |
*** yassine has quit IRC | 13:18 | |
*** yassine has joined #openstack-oslo | 13:19 | |
*** prad has joined #openstack-oslo | 13:23 | |
sdake | gentux moment | 13:25 |
*** mriedem has joined #openstack-oslo | 13:27 | |
sdake | gentux https://review.openstack.org/#/c/182168/ | 13:27 |
sdake | gentux for more details check: | 13:28 |
sdake | or maybe its less details: | 13:28 |
sdake | https://review.openstack.org/#/c/189157/4/specs/ansible-multi.rst | 13:28 |
gentux | ok, thanks sdake | 13:29 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 13:34 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 13:34 |
openstackgerrit | Oleksii Zamiatin proposed openstack/oslo-specs: New ZeroMQ driver implementation details https://review.openstack.org/187338 | 13:34 |
*** kgiusti has joined #openstack-oslo | 13:37 | |
*** _amrith_ is now known as amrith | 13:37 | |
*** ccrouch has joined #openstack-oslo | 13:37 | |
*** superdan is now known as dansmith | 13:49 | |
*** rpodolyaka1 has quit IRC | 13:50 | |
*** rpodolyaka1 has joined #openstack-oslo | 13:50 | |
*** jecarey has joined #openstack-oslo | 13:51 | |
*** claudiub has joined #openstack-oslo | 13:52 | |
*** amotoki has quit IRC | 13:54 | |
*** stevemar has joined #openstack-oslo | 14:12 | |
*** shardy_ has joined #openstack-oslo | 14:13 | |
*** shardy has quit IRC | 14:15 | |
*** arnaud___ has joined #openstack-oslo | 14:18 | |
*** shardy_ has quit IRC | 14:19 | |
*** shardy has joined #openstack-oslo | 14:19 | |
*** openstackgerrit has quit IRC | 14:24 | |
*** openstackgerrit has joined #openstack-oslo | 14:24 | |
*** bnemec has joined #openstack-oslo | 14:24 | |
*** ozamiatin has quit IRC | 14:25 | |
*** achanda has joined #openstack-oslo | 14:27 | |
*** sigmavirus24_awa is now known as sigmavirus24 | 14:28 | |
openstackgerrit | Vipin Balachandran proposed openstack/oslo.vmware: Exception hierarchy refactoring - part 1 https://review.openstack.org/114503 | 14:32 |
*** rpodolyaka1 has quit IRC | 14:35 | |
*** arnaud___ has quit IRC | 14:39 | |
*** achanda has quit IRC | 14:42 | |
openstackgerrit | Merged openstack/oslo.cache: Updated from global requirements https://review.openstack.org/191617 | 14:43 |
*** browne has joined #openstack-oslo | 14:52 | |
*** tsekiyama has joined #openstack-oslo | 14:53 | |
*** achanda has joined #openstack-oslo | 14:53 | |
*** dulek has quit IRC | 14:53 | |
*** ozamiatin has joined #openstack-oslo | 14:54 | |
*** rpodolyaka1 has joined #openstack-oslo | 14:56 | |
*** achanda_ has joined #openstack-oslo | 14:58 | |
*** zzzeek has joined #openstack-oslo | 14:58 | |
flaper87 | sileht: how would you feel with always installing uuid-devel/swig and the amqp1-requirements ? | 14:58 |
flaper87 | otherwise I'll have to put back the ImportError thing in the amqp1 tests | 14:59 |
flaper87 | which I really don't want because that misguides | 14:59 |
sileht | flaper87, I'm feeling good | 15:00 |
*** achanda has quit IRC | 15:00 | |
flaper87 | sileht: cool, will do that :D | 15:01 |
*** amotoki has joined #openstack-oslo | 15:01 | |
*** nkrinner has quit IRC | 15:01 | |
*** ozamiatin has quit IRC | 15:02 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 15:13 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 15:13 |
*** ihrachyshka has quit IRC | 15:13 | |
krotscheck | I made a change in https://review.openstack.org/#/c/190753/ to oslo_middleware that could use some discussion. | 15:13 |
sileht | flaper87, we cannot put the pyngus requirement into requirements.txt ? | 15:14 |
krotscheck | In particular, whether request should be optinoal or not. | 15:14 |
krotscheck | Not making it optional: Breaking change. | 15:14 |
krotscheck | But perhaps more consistent. | 15:14 |
krotscheck | Making it optional, not breaking change. | 15:15 |
flaper87 | sileht: apparently no (or not yet) | 15:20 |
flaper87 | ideally, we should have them in test-requirements | 15:20 |
flaper87 | but right now that'll make the requirements job fail | 15:20 |
flaper87 | I need to talk to lifeless | 15:20 |
sileht | flaper87, pyngus is still not into global-requirements ? | 15:20 |
*** e0ne is now known as e0ne_ | 15:22 | |
openstackgerrit | Merged openstack/oslo.concurrency: Updated from global requirements https://review.openstack.org/191613 | 15:23 |
*** sdake_ has joined #openstack-oslo | 15:23 | |
*** browne has quit IRC | 15:23 | |
sileht | flaper87, it should be now, the installation through pip works well now. | 15:23 |
sileht | flaper87, have you proposed a patch to raise the requirement to 1.3.1 ? | 15:24 |
flaper87 | sileht: wait, pyngus is in global-requirements but the version 1.2.0 is. We need 1.3.1, which requires changes in devstack to install those 2 packages | 15:25 |
flaper87 | which is what I was discussing with sdague earlier in -qa | 15:25 |
flaper87 | eventually, AFAIU, pyngus should not be in g-r | 15:25 |
sileht | flaper87, now we don't care of devstack | 15:26 |
sileht | flaper87, we can just raise the requirement | 15:26 |
*** Kevin_Bishop has joined #openstack-oslo | 15:26 | |
flaper87 | sileht: exactly, but pyngus 1.3.1 can't land in requirements since that change in devstack because the requirements test will try to install it | 15:26 |
*** sdake has quit IRC | 15:27 | |
flaper87 | Therefore, I'm forced to keep it in a separate requirements for now until that thing is cleared out | 15:27 |
flaper87 | sileht: does that make sense? | 15:27 |
flaper87 | Eventually, I'd guess pyngus should go to test-requirements.txt | 15:27 |
flaper87 | but again, I need to talk w/ lifeless about that | 15:28 |
sileht | flaper87, the requirements tests should not depends on devstack ... | 15:28 |
sileht | flaper87, ok | 15:28 |
Kevin_Bishop | Hi, I noticed that oslo_service is not listed in openstack's global-requirements.txt. Is there a reason for this? | 15:28 |
dims | Kevin_Bishop: we just started this work in liberty cycle, api is not yet ready | 15:28 |
Kevin_Bishop | dims ok great, thank you | 15:29 |
sileht | flaper87, got it, let's land that as-is, so and move it to requirements later, the requirements testing issue is fixed | 15:29 |
sileht | "later when the requirements testing issue is fixed" | 15:29 |
*** manishg has joined #openstack-oslo | 15:30 | |
*** rpodolyaka1 has quit IRC | 15:30 | |
*** e0ne_ has quit IRC | 15:32 | |
*** pradk has joined #openstack-oslo | 15:33 | |
*** e0ne has joined #openstack-oslo | 15:35 | |
*** isq_ has quit IRC | 15:39 | |
*** isq_ has joined #openstack-oslo | 15:39 | |
*** r0j4z0 has quit IRC | 15:41 | |
*** harlowja_at_home has joined #openstack-oslo | 15:43 | |
*** pradk has quit IRC | 15:43 | |
*** rpodolyaka1 has joined #openstack-oslo | 15:45 | |
*** jaosorior has quit IRC | 15:45 | |
*** ozamiatin has joined #openstack-oslo | 15:46 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 15:46 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 15:46 |
flaper87 | sileht: I'll get it right, I swear ^ | 15:46 |
dims | 5 mins to oslo meeting on #openstack-meeting-alt | 15:55 |
*** yassine has quit IRC | 15:57 | |
harlowja_at_home | who wants to review super-awesome example of a state-machine @ https://review.openstack.org/#/c/191477/ | 15:57 |
openstackgerrit | Qin Zhao proposed openstack/oslo.vmware: Add a locale param to allow locale specification. https://review.openstack.org/178104 | 15:57 |
harlowja_at_home | that oddly looks like the same machine as http://www.boost.org/doc/libs/1_55_0/libs/msm/doc/HTML/ch03s02.html#d0e358 ;) | 15:57 |
harlowja_at_home | ^ what a weird coincidence ;) | 15:58 |
*** ihrachyshka has joined #openstack-oslo | 15:58 | |
dims | harlowja_at_home: my kids were impressed when i told them if the just turned around the "cassette tape" there's a whole new set of songs :) they had never seen one before | 15:59 |
harlowja_at_home | lol | 15:59 |
harlowja_at_home | ah, the good ole days | 15:59 |
*** manishg has quit IRC | 15:59 | |
dims | :) will check it out after the meeting harlowja_at_home | 16:00 |
ihrachyshka | show them how to play music with a nail and a vynil | 16:00 |
harlowja_at_home | np :) | 16:00 |
ihrachyshka | dims, hm. is it openstack-meeting-3? | 16:01 |
*** manishg has joined #openstack-oslo | 16:01 | |
ihrachyshka | dims, cause there is chef meeting starting there | 16:01 |
harlowja_at_home | reminds me of the hipster stuff like @ https://imgur.com/ENeBi lol | 16:01 |
dims | #openstack-meeting-alt | 16:01 |
ihrachyshka | ah, openstack is about change, right! | 16:01 |
ihrachyshka | thanks | 16:02 |
*** ashleighfarnham has joined #openstack-oslo | 16:02 | |
ihrachyshka | hm, no change, it's my calendar showing some obsolete entries, sorry | 16:02 |
*** Guest67074 is now known as redrobot | 16:02 | |
*** exploreshaifali has quit IRC | 16:02 | |
*** jaypipes has quit IRC | 16:03 | |
amrith | ./ | 16:03 |
*** jaypipes has joined #openstack-oslo | 16:05 | |
*** ashleighfarnham has quit IRC | 16:09 | |
*** ashleighfarnham has joined #openstack-oslo | 16:09 | |
*** sdake has joined #openstack-oslo | 16:11 | |
*** salv-orlando has joined #openstack-oslo | 16:12 | |
*** hogepodge has quit IRC | 16:13 | |
*** sdake_ has quit IRC | 16:14 | |
*** salv-orl_ has quit IRC | 16:16 | |
*** ashleighfarnham has quit IRC | 16:17 | |
*** hogepodge has joined #openstack-oslo | 16:20 | |
*** ganso_ has joined #openstack-oslo | 16:22 | |
*** zz_jgrimm is now known as jgrimm | 16:22 | |
*** ashleighfarnham has joined #openstack-oslo | 16:25 | |
*** sputnik13 has joined #openstack-oslo | 16:30 | |
*** vilobhmm has joined #openstack-oslo | 16:30 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 16:36 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 16:36 |
ganso_ | Hello folks. I am trying to use processutils from oslo_concurrency to run a rsync command with --progress and display progress in stdout, but the command does not output anything to stdout while it is running. Is there another command or library in oslo that is able to accomplish this? | 16:42 |
openstackgerrit | Merged openstack/taskflow: Remove all 'lock_utils' now that fasteners provides equivalents https://review.openstack.org/188629 | 16:46 |
*** browne has joined #openstack-oslo | 16:51 | |
*** harlowja_at_home has quit IRC | 16:53 | |
*** exploreshaifali has joined #openstack-oslo | 17:00 | |
*** jdandrea has joined #openstack-oslo | 17:01 | |
*** e0ne has quit IRC | 17:02 | |
openstackgerrit | Ihar Hrachyshka proposed openstack/oslo.rootwrap: daemon: avoid raising UnboundLocalError to callers https://review.openstack.org/191895 | 17:02 |
*** rpodolyaka1 has quit IRC | 17:05 | |
*** ozamiatin has quit IRC | 17:14 | |
jdandrea | Is anyone aware of (and can point me toward) add-ons to OpenStack projects (e.g., heat, nova, cinder, neutron, et. al) that use oslo config files that are *separate* from the project config files? | 17:17 |
openstackgerrit | Merged openstack/oslo.log: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190906 | 17:20 |
openstackgerrit | Merged openstack/oslo.messaging: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190904 | 17:22 |
dhellmann | jdandrea: I'm not sure I understand what you mean. We have lots of drivers that define options not used directly by the applications, and we have a --config-dir option to allow an app to be run using more than one config file at a time. Can you give more detail about what you're trying to do? | 17:22 |
*** manishg has quit IRC | 17:23 | |
*** manishg has joined #openstack-oslo | 17:24 | |
jdandrea | I have a Heat plugin that isn't tied to any other project/server/etc. and has it's own config (just a few options). Right now the only thing I can seem to access via cfg.CONF is Heat's /etc/heat/heat.conf ... but me putting my options in there is reportedly not a good idea. | 17:24 |
openstackgerrit | Merged openstack/oslo.concurrency: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190905 | 17:24 |
openstackgerrit | Merged openstack/oslo.db: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190907 | 17:25 |
jdandrea | I also looked at the config docs and it wasn't clear to me how I'd use it for my own separate config. | 17:25 |
dhellmann | jdandrea: the deployer can launch the app with --config-dir (set to something like /etc/heat) and then you can use a separate file | 17:25 |
openstackgerrit | Merged openstack/stevedore: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190911 | 17:25 |
openstackgerrit | Merged openstack/oslo.serialization: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190903 | 17:26 |
jdandrea | I'm afraid I'm a bit naive here. Deployer? When I start heat I imagine the config dir is already /etc/heat by default (?). | 17:26 |
openstackgerrit | Merged openstack/taskflow: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190909 | 17:26 |
openstackgerrit | Merged openstack/debtcollector: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190910 | 17:26 |
openstackgerrit | Merged openstack/oslo.config: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190902 | 17:26 |
dims | jdandrea: here's a stripped down example - https://gist.github.com/dims/19cb6ee280c3177fbf52 | 17:26 |
jdandrea | Thanks, dims. Studying it ... | 17:27 |
*** dims has quit IRC | 17:27 | |
*** crc32 has joined #openstack-oslo | 17:27 | |
*** dims has joined #openstack-oslo | 17:28 | |
jdandrea | dims: How does that example know to use /etc/my.conf specifically? (Don't see it referenced in the app.py code) ... unless that should read /etc/app.conf ... and then as far as sys.argv[1:] I am not sure what is passed in on the cmd line that CONF will expect. *checking source* | 17:31 |
dhellmann | jdandrea: the --config-dir doesn't have a default, IIRC, only --config-file, so it may depend on how heat is launched | 17:33 |
dhellmann | jdandrea: once the config files are loaded, oslo.config doesn't differentiate on where the values come from, so it doesn't actually matter to your code which file the configuration settings live in | 17:33 |
dims | jdandrea: the conf file should have been named app.conf not my.conf (see line 10 of app.py) | 17:34 |
jdandrea | dims: Got it, thank you! | 17:34 |
jdandrea | dhellmann: Ooh, interesting. | 17:34 |
dhellmann | jdandrea: what you want to avoid, though, is having to update heat to tell it what files to load based on plugins. That's why we have --config-dir | 17:35 |
jdandrea | dhellmann: Yeah, I see what you mean, and I can't control *how* heat is launched in this case. Hmm ... | 17:35 |
*** crc32 has quit IRC | 17:35 | |
dims | jdandrea: you can create an instance of ConfigOpts() instead of using cfg.CONF | 17:35 |
dhellmann | jdandrea: right, so the actual location of the config settings is not something you should worry about in your code. | 17:35 |
openstackgerrit | Merged openstack/oslo.utils: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190901 | 17:36 |
jdandrea | dhellmann: Thank goodness. :) ... but I've been told "you should not put your third party plugin config settings inside heat.conf" | 17:36 |
dhellmann | jdandrea: if you register all of your options in a config group, that will provide a namespace to protect them from other heat options | 17:36 |
jdandrea | dhellmann: So, for instance, when I install my plugin I was thinking it would be nice to also install a default config file (or maybe that's not a good idea either). | 17:36 |
dims | jdandrea: if you use your own ConfigOptions, and pass it the name of the default_config_files to load from, you are all set | 17:36 |
dhellmann | jdandrea: see the driver options in oslo.messaging, for example | 17:36 |
dhellmann | jdandrea: you could drop a file in /etc/heat, but whether that's a good idea depends entirely on the distro packager rules | 17:37 |
jdandrea | dims: Oh, so in my plugin init I can use my own ConfigOptions and then amend default_config_files (but then to dhellmann's point that means I am worrying about the location of those settings huh). Sorry, trying not to confuse myself. :) | 17:37 |
*** salv-orlando has quit IRC | 17:37 | |
dhellmann | dims: I think you're making this harder than it needs to be? | 17:38 |
* dhellmann may not understand | 17:38 | |
jdandrea | dhellmann: Probably, but that's because I'm not sure how to make it easier. :) | 17:38 |
jdandrea | I have config options and I'm told "don't put them in heat.conf" ... so I'm all "OK, now what do I do." | 17:38 |
dims | dhellmann: from what i understood, heat folks want this stuff totally stand alone, so cfg.CONF of heat should never have stuff from jdandrea's conf files | 17:38 |
dhellmann | jdandrea: register your options in a group named in a way related to your plugin, and then don't worry about the source of any option settings. That's very explicitly not something application code should deal with. | 17:39 |
dhellmann | dims: that's not how config is meant to work | 17:39 |
dhellmann | dims: using a namespace group should be enough isolation | 17:39 |
jdandrea | dhellmann: Ok, that's good, and it turns out I already have that group. | 17:39 |
jdandrea | dhellmann: But ... the only place I seem to be able to place those options, then, is in heat.conf. Am I stuck or is there an easier way I'm missing? | 17:39 |
dhellmann | jdandrea: how are you running heat? | 17:40 |
jdandrea | Right now I am indeed using heat.conf | 17:40 |
*** arnaud___ has joined #openstack-oslo | 17:40 | |
jdandrea | dhellmann: I'm running it as-installed. No special mods. | 17:40 |
dhellmann | jdandrea: what actual command line do you use to start it? | 17:40 |
jdandrea | service heat-engine start | 17:41 |
dhellmann | jdandrea: what options are being given to heat by whatever startup script that command runs? | 17:41 |
jdandrea | dhellmann: Good question. I have no idea. *checking* | 17:41 |
jdandrea | dhellmann: I *do* see that, with devstack, it uses --config-file and expressly points to /etc/heat/heat.conf ... now checking the startup script version. | 17:42 |
dhellmann | dims: the idea with all of this config stuff is that the app developer shouldn't think too much about it, because how many config files and what they are called is a deployer concern -- different deployment styles will want different files in different places, and we don't want to have to code around all of them | 17:42 |
dhellmann | jdandrea: yes, so devstack may have to be changed to add --config-dir for your case | 17:43 |
jdandrea | dhellmann: By deployer are you referring to whomever installs and sets it up? | 17:43 |
dhellmann | jdandrea: yes, or whoever creates the packages with the startup scripts | 17:43 |
dims | dhellmann: true | 17:43 |
openstackgerrit | Merged openstack/tooz: Updated from global requirements https://review.openstack.org/191615 | 17:43 |
jdandrea | dhellmann: Ok, in this case I am not sure where the startup scripts came from (thinking they were provided with heat, but maybe not!). | 17:44 |
dhellmann | jdandrea: what os are you on? where did you get the packages? | 17:44 |
*** shardy_ has joined #openstack-oslo | 17:44 | |
jdandrea | Ubuntu 14.04 and I think (and will re-confirm) that everything was installed according to the online docs (vs. getting an all-in-one kind of thing). | 17:45 |
dhellmann | so you got packages from canonical? | 17:45 |
jdandrea | dhellmann: That's what I'm going to confirm. | 17:45 |
jdandrea | (suspecting probably) | 17:45 |
dhellmann | yeah, I think probably | 17:45 |
*** arnaud___ has quit IRC | 17:45 | |
dhellmann | I don't know what the startup script in those packages looks like, but it might also need to have a config-dir set | 17:45 |
*** shardy has quit IRC | 17:45 | |
jdandrea | So it might simply end up being a case of: "When you install my plugin, change your startup script to point to some_file.conf for the config and here's a sample config file." I was trying to make it more bootstrap-y in my setup.cfg so that they didn't need to do anything extra. | 17:46 |
jdandrea | Alas, sometimes those kind of changes end up taking a long time to get approved (anything apart from some already documented or approved method). :/ | 17:47 |
dhellmann | if your plugin is going to be packaged by multiple distros, you won't have a lot of control over that anyway | 17:47 |
dhellmann | so you should just document the options that need to be set and not worry about what file they are set in | 17:47 |
jdandrea | dhellmann: Not sure it will be packaged by any distros. | 17:47 |
dhellmann | is it public? | 17:48 |
jdandrea | dhellmann: Ok. So I can't write a sample. | 17:48 |
jdandrea | dhellmann: Not yet. I'd like it to be though. | 17:48 |
dhellmann | you could provide a partial sample file | 17:48 |
dhellmann | ah, well, if it's not public then yeah I wouldn't expect it to be packaged :-) | 17:48 |
jdandrea | dhellmann: Ok. I think, then, that my setup.cfg installing a cfg file is not advised. Instead I can provide a sample and say "put it wherever and reference it in your startup" | 17:49 |
dhellmann | yeah | 17:49 |
jdandrea | But when it's public I hope it will be! Just not yet. :) | 17:49 |
dhellmann | fwiw, it's completely legal to append that to the end of heat.conf | 17:49 |
jdandrea | dhellmann: it is? Oh. Well ... it already is. :) | 17:49 |
jdandrea | In that case I'm *really* in good shape. :D | 17:49 |
dhellmann | cool :-) | 17:50 |
*** shardy_ has quit IRC | 17:50 | |
*** shardy has joined #openstack-oslo | 17:51 | |
*** ozamiatin has joined #openstack-oslo | 17:58 | |
ihrachyshka | dhellmann, is there any alternative way to reach log wg except going to their meeting? | 17:58 |
jdandrea | dhellmann, dims: ty, appreciate the help | 17:59 |
dhellmann | ihrachyshka: I would expect the mailing list. I see some messages to openstack-dev with "[log]" for example | 17:59 |
dhellmann | jdandrea: np | 18:00 |
dhellmann | ihrachyshka: maybe I can help? | 18:00 |
*** ozamiatin has quit IRC | 18:02 | |
ihrachyshka | dhellmann, well, that email about oslo context/oslo log interactions... http://lists.openstack.org/pipermail/openstack-dev/2015-June/066295.html I wrote to both operators and -dev, not response so far. maybe I should try to add [log] too... | 18:02 |
dhellmann | ihrachyshka: yes, and I think that slipped off of my list of things to reply to so I'll look for it again | 18:03 |
ihrachyshka | dhellmann, thanks. | 18:06 |
*** stevemar2 has joined #openstack-oslo | 18:09 | |
*** stevemar has quit IRC | 18:09 | |
*** rpodolyaka1 has joined #openstack-oslo | 18:16 | |
*** sdake_ has joined #openstack-oslo | 18:30 | |
*** exploreshaifali has quit IRC | 18:33 | |
*** sdake has quit IRC | 18:34 | |
*** Kevin_Bishop has quit IRC | 18:36 | |
*** Kevin_Bishop has joined #openstack-oslo | 18:37 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 18:39 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 18:39 |
*** achanda_ has quit IRC | 18:40 | |
lifeless | morning | 18:45 |
lifeless | flaper87: hi | 18:45 |
flaper87 | lifeless: hey there :) | 18:45 |
flaper87 | I summoned you like too many times today | 18:46 |
flaper87 | :P | 18:46 |
flaper87 | lifeless: quick question: https://review.openstack.org/#/c/191107/ | 18:46 |
flaper87 | I've that requirement that also depends on that change on devstack. I talked to sdague and he explained that the goal is to remove non-rabbit things out of devstack | 18:47 |
flaper87 | my question is what's the right way to add that requirement to the requirements repo? | 18:47 |
flaper87 | Is it putting it in the test-requirements file? | 18:47 |
flaper87 | or should I wait until your spec is applied | 18:47 |
*** cdent has quit IRC | 18:49 | |
harlowja | zzzeek sooo for https://github.com/openstack/taskflow/blob/master/taskflow/persistence/backends/sqlalchemy/migration.py there is an easier way to run through the migration steps u say? | 18:49 |
harlowja | in a way that can be called via 'db_sync' (progromatically) | 18:50 |
zzzeek | harlowja: should be able to run command.upgrade() rather than re-implementing it | 18:50 |
harlowja | k | 18:50 |
zzzeek | http://alembic.readthedocs.org/en/latest/api.html#commands | 18:50 |
harlowja | thx | 18:50 |
zzzeek | someone else did this same thing some months ago | 18:50 |
harlowja | :-P | 18:50 |
*** stevemar2 is now known as stevemar | 18:51 | |
harlowja | looks good to me | 18:51 |
harlowja | will use that | 18:51 |
*** e0ne has joined #openstack-oslo | 18:51 | |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Enable amqp's protocol unit tests everywhere https://review.openstack.org/186671 | 18:53 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 18:53 |
*** crc32 has joined #openstack-oslo | 18:58 | |
*** Rockyg has joined #openstack-oslo | 18:59 | |
harlowja | zzzeek would https://bitbucket.org/zzzeek/alembic/pull-request/45/allow-for-providing-an-existing-connection/diff be ok with u? | 19:00 |
harlowja | or something similar/better | 19:00 |
zzzeek | harlowja: see http://alembic.readthedocs.org/en/latest/cookbook.html#sharing-a-connection-with-a-series-of-migration-commands-and-environments | 19:00 |
zzzeek | harlowja: added this just for openstack | 19:00 |
harlowja | ah | 19:00 |
harlowja | well there u go :-P | 19:00 |
zzzeek | this must be why ive seen this before | 19:00 |
harlowja | maybe, ha | 19:01 |
*** ViswaV has joined #openstack-oslo | 19:02 | |
*** rpodolyaka1 has quit IRC | 19:02 | |
flaper87 | sileht: around ? | 19:05 |
flaper87 | sileht: it passes now :P | 19:06 |
ViswaV | Hi, Question to olso.messaging team. Is there a future plan to support zaqar/marconi as an underlying messaging platform ? | 19:06 |
dhellmann | ViswaV: what's your use case for that? | 19:06 |
flaper87 | ViswaV: not right now | 19:06 |
flaper87 | and I honestly don't think it will | 19:06 |
flaper87 | not in o.m code-base at least | 19:07 |
flaper87 | but please, do share your thoughts | 19:07 |
flaper87 | :) | 19:07 |
ViswaV | We are using openstack.common.rpc (I know we are lagging way behind....! ). But we plan to move to oslo.messaging soon. | 19:07 |
flaper87 | ViswaV: may I ask who is we ? | 19:07 |
flaper87 | what project? | 19:07 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use alembic upgrade function/command directly https://review.openstack.org/191935 | 19:08 |
ViswaV | But we heard from another team that some Openstack projects including Nova is potentially considering to use Zaqar as the RPC/messaging transport.... so I was wondering if moving to oslo.messaging now is the right choice or wait until Openstack decides and settles on it's next gen RPC/messaging protocol... | 19:09 |
flaper87 | ViswaV: uh, no. | 19:09 |
ViswaV | 'We' is an internal team in eBay... We have teams that work on Trove and other yet not disclosed/opensourced openstack internal projects. | 19:10 |
flaper87 | Some projects are considering to use Zaqar for different use-cases but none of them is considering to use it as a replacement for rabbit/qpid/zmq | 19:10 |
flaper87 | ViswaV: ah so this probably came from the discussion with Trove folks | 19:10 |
flaper87 | the use case in the case of trove was to communicate with the guest agent and they mentioned they're using o.m in the guest agent | 19:11 |
ViswaV | ok. thx for the info. Yeah heard from another trove team member. Could you please elaborate on what kind of use cases? | 19:11 |
flaper87 | Therefore, they asked to have an o.m driver to make things simpler | 19:11 |
flaper87 | However, I really don't think that's going to happen (any time soon). The reason is precisely the question you asked | 19:11 |
flaper87 | Whether Zaqar can/cannot be used as an underlying message broker is a different discussion that can be taken elsewhere. However, it's not Zaqar's intention to do so | 19:12 |
harlowja | shouldn't zaquar make a nice library that provides a nice pub/sub api first before this? | 19:12 |
harlowja | *does that library exist? | 19:12 |
flaper87 | harlowja: zaqar* :P | 19:12 |
harlowja | ya, that | 19:13 |
flaper87 | harlowja: zaqarclient is what you're looking for | 19:13 |
harlowja | k | 19:13 |
flaper87 | and there's support for kombu | 19:13 |
flaper87 | well, I believe that landed | 19:13 |
flaper87 | but I gotta double check | 19:13 |
flaper87 | that said, there's still lot of work to do | 19:13 |
ViswaV | ok. got it. So o.m is not considering zaqar as a driver any time soon? I see kafka, zeromq etc in the BPs. | 19:13 |
harlowja | flaper87 thats a good start (kombu being more generic imho anyway) | 19:13 |
flaper87 | ViswaV: that's correct | 19:14 |
* flaper87 has serious doubts about kafka but... | 19:14 | |
harlowja | hmmm, i don't see zaqar https://github.com/celery/kombu/tree/master/kombu/transport in there | 19:14 |
*** exploreshaifali has joined #openstack-oslo | 19:14 | |
flaper87 | I'd love folks to start looking more into message federation for the underlying messaging later | 19:15 |
flaper87 | like the amqp1 protocol driver | 19:15 |
flaper87 | harlowja: aw https://github.com/celery/kombu/pull/315 | 19:15 |
flaper87 | it got stalled, I'll get it done | 19:15 |
harlowja | 'This pull request contains merge conflicts that must be resolved. ' sad face | 19:15 |
flaper87 | lol, it was called Marconi when that PR was propsoed | 19:15 |
ViswaV | When is zeromq landing in o.m ? .... a broker less messaging might be an attractive option for guestagent models... like trove and our product. | 19:15 |
flaper87 | proposed | 19:15 |
harlowja | ya, back in the olden days | 19:16 |
harlowja | time before time | 19:16 |
flaper87 | ViswaV: FWIW, there's a brokerless messaging driver already already ( http://docs.openstack.org/developer/oslo.messaging/AMQP1.0.html ) | 19:16 |
flaper87 | ViswaV: that said, there's also ZMQ but I believe it still needs some work | 19:17 |
* harlowja thought we were going to switch to all HTTP all the places ;) | 19:17 | |
flaper87 | ViswaV: I'll be writing more docs, posts and useful (TM) things for the amqp protocol soon | 19:17 |
flaper87 | harlowja: LOL | 19:17 |
ViswaV | flaper87: from that link you gave: "The driver also requires a broker that supports version 1.0 of the AMQP protocol. | 19:17 |
flaper87 | ok, going afk for a bit | 19:17 |
lifeless | flaper87: I don't understand the question, the pyngus thing is approved and seems straight forward | 19:18 |
*** crc32 has quit IRC | 19:18 | |
flaper87 | ViswaV: well, you can either have a broker for store-forward or just use a message router (qpid-dispatch) which is completely brokerless | 19:18 |
* flaper87 needs to fix those docs | 19:18 | |
flaper87 | ViswaV: the whole point is that it provides federation | 19:18 |
ViswaV | Got it. Thx. | 19:18 |
flaper87 | cases were you'd like to have store include notifications | 19:18 |
flaper87 | if you loose a message, you'll be sad and you won't be able to charge | 19:19 |
flaper87 | :P | 19:19 |
lifeless | flaper87: test-requirements are subject to the same logic as requirements.txt | 19:19 |
lifeless | flaper87: there's no difference | 19:19 |
flaper87 | lifeless: yeah, that's my current problem, I guess. | 19:19 |
harlowja | zzzeek i guess if u get a free moment or 2 https://review.openstack.org/#/c/191935/ :) | 19:19 |
flaper87 | lifeless: that patch was approved but it depends on a patch from devstack because pyngus 1.3.1 requires uuid-dev and swig | 19:19 |
zzzeek | harlowja: ok | 19:19 |
harlowja | thx | 19:19 |
flaper87 | lifeless: and those 2 are not isntalled by default | 19:20 |
flaper87 | installed* | 19:20 |
openstackgerrit | Merged openstack/tooz: Switch badges from 'pypip.in' to 'shields.io' https://review.openstack.org/190908 | 19:21 |
openstackgerrit | Mitsuhiro Tanino proposed openstack/oslo.service: Revert "service child process normal SIGTERM exit" https://review.openstack.org/189418 | 19:27 |
*** mtanino has joined #openstack-oslo | 19:28 | |
*** jaosorior has joined #openstack-oslo | 19:29 | |
*** amotoki has quit IRC | 19:33 | |
elarson | does anyone have tips for understanding why oslo_log is not logging for a module? | 19:33 |
elarson | the logs here don't work https://github.com/openstack/designate/blob/master/designate/hookpoints.py#L23 | 19:34 |
Kevin_Bishop | dims: Hey again, what is the timeline for oslo_service? Is there anything I can do to help out? | 19:35 |
lifeless | flaper87: so you need the glance module in devstack to install uuid-dev and swig ? | 19:35 |
lifeless | flaper87: I'm clearly still not understanding the policy problem. | 19:36 |
lifeless | flaper87: you say that oslo.messaging needs pyngus | 19:36 |
lifeless | flaper87: and we had pygnus in g-r | 19:37 |
lifeless | flaper87: sdague: perhaps step me through this? I can say what the requirements influence on any step is | 19:37 |
sdague | lifeless: there is a glance thing here? I thought it was just a proton thing? | 19:39 |
lifeless | sdague: I made an assumption based on flaper87 asking ;) | 19:39 |
sdague | I told flaper87 that honestly we want all the non rabbit rpc backends out of devstack, so instead of fixing things for proton in devstack, it should be moved out of tree. | 19:40 |
lifeless | so if its in an out of tree repo | 19:40 |
lifeless | that subscribes to projects.txt | 19:40 |
lifeless | then its still going to be in global-requirements.txt | 19:40 |
lifeless | if its in an unsynchronised repo, thats different | 19:41 |
stevemar | oh lifeless and sdague are around, how fortunate ... | 19:41 |
sdague | lifeless: it's not going to be a thing that subscribes to projects.txt I don't think, it would be an external devstack plugin | 19:41 |
stevemar | dhellmann, already sorta +2'ed this, but if you guys could take a quick look at https://review.openstack.org/#/c/190268/ | 19:42 |
lifeless | I'm not 100% clear on the rules for 'add to projects.txt' - the README.rst in requirements doesn't specify policy other than 'has accepted' | 19:42 |
*** Rockyg has quit IRC | 19:42 | |
*** Rockyg has joined #openstack-oslo | 19:43 | |
lifeless | but there's no harm to global-requirements.txt in having it there *as long as its installable for constraints generation* | 19:43 |
sdague | stevemar: it's python-dateutil right? | 19:43 |
lifeless | if we can't calculate an install of it on either py2.7 or 3.4 | 19:43 |
lifeless | inside a venv, on the periodic jobs node - then thats going to be an issue | 19:44 |
sdague | stevemar: pip search python_dateutil returns 0 hits | 19:44 |
*** jgrimm is now known as zz_jgrimm | 19:44 | |
sdague | stevemar: also the requirements review should mention the state of packaging | 19:45 |
lifeless | https://pypi.python.org/pypi/python-dateutil | 19:45 |
lifeless | stevemar: put that in the review | 19:46 |
*** ozamiatin has joined #openstack-oslo | 19:47 | |
*** rpodolyaka1 has joined #openstack-oslo | 19:47 | |
*** alexpilotti has quit IRC | 19:48 | |
*** yamahata has joined #openstack-oslo | 19:49 | |
lifeless | sdague: pypi's UI handles case and punctuation differences btw for doing searches | 19:49 |
*** achanda has joined #openstack-oslo | 19:50 | |
sigmavirus24 | elarson: define "dont' work" | 19:51 |
*** boris-42 has quit IRC | 19:52 | |
elarson | sigmavirus24: beyond things like LOG.info (when the debug is the default) not working, getLogger doesn't add the logger to oslo_log.log._loggers | 19:52 |
stevemar | lifeless, thx lifeless, i'll put that in there. sdague i'll see what i can get in terms of packaging info | 19:53 |
sigmavirus24 | Stupid question elarson, is that module even being imported? | 19:53 |
elarson | sigmavirus24: great question. it should be as I can throw an exception and stop the process. | 19:54 |
elarson | sigmavirus24: I also tried instantiating the logger lazily in an object property but that didn't do it either | 19:54 |
*** sdake has joined #openstack-oslo | 19:54 | |
elarson | my thought being that the loggers are already configured and mine wasn't included | 19:54 |
elarson | aka no handlers were added | 19:55 |
sigmavirus24 | elarson: so the only reason I can imagine that it would not add the handler would be if it was never actually imported | 19:55 |
sigmavirus24 | which would explain why nothing is logged (because nothing is executed) | 19:55 |
sigmavirus24 | elarson: fwiw: https://github.com/openstack/designate/search?utf8=%E2%9C%93&q=hookpoints | 19:55 |
sigmavirus24 | nothing's importing hookpoints at the moment | 19:55 |
elarson | sigmavirus24: right, I have a local branch with a hook point | 19:56 |
elarson | sigmavirus24: http://paste.openstack.org/show/294469/ | 19:56 |
sigmavirus24 | You know what my next question is, right? | 19:57 |
elarson | sigmavirus24: unfortunately no :( | 19:58 |
*** sdake_ has quit IRC | 19:58 | |
sigmavirus24 | What, if anything, imports designate.pool_manager.service | 19:58 |
sigmavirus24 | So far only https://github.com/openstack/designate/blob/6acb7191c8780880df3c923165a4a8b39c46c2a7/designate/cmd/pool_manager.py | 19:58 |
sigmavirus24 | (from https://github.com/openstack/designate/search?utf8=%E2%9C%93&q=pool_manager&type=Code) | 19:58 |
elarson | sigmavirus24: right, I'm running the pool manager | 19:59 |
elarson | but that does make me wonder what the logging.setup is getting there | 19:59 |
sigmavirus24 | elarson: Well failing all other things, I'd chuck some print statements in those modules to make sure they're actually being imported | 19:59 |
elarson | sigmavirus24: I am getting print statements, which is how I know things get imported. but I'm suspect of the config bits in the pool_manager. | 20:00 |
stevemar | sdague, lifeless danke, new patch should be up | 20:01 |
sigmavirus24 | elarson: interesting | 20:01 |
openstackgerrit | Joshua Harlow proposed openstack/automaton: Add `is_actionable_event` checking function (WIP) https://review.openstack.org/191955 | 20:01 |
*** sdake_ has joined #openstack-oslo | 20:01 | |
*** Rockyg has quit IRC | 20:04 | |
*** sdake has quit IRC | 20:05 | |
*** Rockyg has joined #openstack-oslo | 20:05 | |
lifeless | flaper87: so - summarising the pyngus thing: g-r has the union of all requirements, whether unconditional, conditional, optional, or test, for all projects that receive updates and sync. | 20:10 |
lifeless | flaper87: all projects that want to be co-installable should be in that list; if the proton driver needs to be co-installable, it needs to be in that list. | 20:10 |
lifeless | flaper87: there's very little incremental cost having things in there at this stage, if we run into computability issues, we'll raise and discuss, but the optionality or not of a backend doesn't affect it IMO | 20:11 |
lifeless | flaper87: HTH | 20:11 |
*** amrith is now known as _amrith_ | 20:12 | |
lifeless | flaper87: there are soft-update projects which you could be, if you want to have things that aren't acceptable for g-r, but I don't see anything about pyngus that makes it unacceptable | 20:12 |
*** ViswaV_ has joined #openstack-oslo | 20:14 | |
*** ViswaV has quit IRC | 20:14 | |
*** ihrachyshka has quit IRC | 20:14 | |
*** zzzeek has quit IRC | 20:16 | |
flaper87 | lifeless: that was my thought as well. I'll have to sync with sdague again since he things the changes I proposed for destack shouldn't land because those move away from devstack's goals for liberty | 20:16 |
*** Rockyg has quit IRC | 20:17 | |
openstackgerrit | Joshua Harlow proposed openstack/automaton: Add `is_actionable_event` checking function https://review.openstack.org/191955 | 20:22 |
*** zzzeek has joined #openstack-oslo | 20:24 | |
*** yamahata has quit IRC | 20:29 | |
*** rpodolyaka1 has quit IRC | 20:31 | |
*** achanda has quit IRC | 20:34 | |
*** achanda_ has joined #openstack-oslo | 20:34 | |
*** zzzeek has quit IRC | 20:38 | |
*** kgiusti has quit IRC | 20:41 | |
*** sdake has joined #openstack-oslo | 20:44 | |
*** achanda_ has quit IRC | 20:46 | |
*** ViswaV_ has quit IRC | 20:46 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/oslo-incubator: Crypto.Random.atfork before fetching random bytes https://review.openstack.org/191581 | 20:46 |
*** sdake_ has quit IRC | 20:48 | |
openstackgerrit | Merged openstack/oslotest: Create ConfigureLogging fixture https://review.openstack.org/190775 | 20:51 |
openstackgerrit | Merged openstack/oslotest: Add CreateFileWithContent fixture https://review.openstack.org/190795 | 20:51 |
*** ozamiatin has quit IRC | 20:54 | |
*** ozamiatin has joined #openstack-oslo | 20:55 | |
*** dims_ has joined #openstack-oslo | 20:56 | |
*** dims has quit IRC | 21:00 | |
*** dims_ has quit IRC | 21:00 | |
flaper87 | harlowja: https://review.openstack.org/#/c/186671/ pls ? | 21:03 |
*** stevemar2 has joined #openstack-oslo | 21:03 | |
*** stevemar has quit IRC | 21:04 | |
*** ViswaV has joined #openstack-oslo | 21:06 | |
*** bknudson has quit IRC | 21:09 | |
openstackgerrit | Marian Horban proposed openstack/oslo.service: Graceful shutdown WSGI/RPC server https://review.openstack.org/190175 | 21:14 |
*** manishg_ has joined #openstack-oslo | 21:16 | |
*** manishg has quit IRC | 21:17 | |
elarson | are there any strategies for lazy loading or reloading a the oslo logging config? | 21:19 |
elarson | my app starts and immediately complains that there are no handlers, which makes sense b/c at import time, they wouldn't have been configured. | 21:19 |
elarson | but then, after the module should have been imported, there is still no logging configured | 21:20 |
openstackgerrit | Flavio Percoco proposed openstack/oslo.messaging: Use `inferred=True` by default https://review.openstack.org/191778 | 21:22 |
* elarson pokes around for any sort of `after_config_loaded` event in oslo_config | 21:23 | |
dhellmann | elarson: either your app should be setting up logging sooner, or your lib should be adding null handlers | 21:24 |
elarson | this is nasty... http://paste.openstack.org/show/294502/ | 21:24 |
dhellmann | ugh, yeah, why does that exist? | 21:25 |
elarson | dhellmann: I think the problem is that this is loading something as import time (hooks via stevedore), which means no config could have been loaded at that time | 21:25 |
dhellmann | elarson: what is logging on import though? | 21:25 |
dhellmann | and couldn't the app set up logging before loading the plugins? | 21:26 |
elarson | dhellmann: I added some .debug messages to show what plugins are found and what get applied | 21:26 |
dhellmann | elarson: is stevedore being invoked as part of importing a module? | 21:27 |
elarson | dhellmann: the plugins end up applied as decorators, so they need to return a function when applied | 21:27 |
elarson | dhellmann: for reference https://github.com/openstack/designate/blob/master/designate/hookpoints.py | 21:27 |
dhellmann | elarson: I'm trying to understand the startup order of the application, because it should read the config file, set up logging, then do anything else that is app-specific and it doesn't sound like that's what's happening | 21:28 |
*** ViswaV has quit IRC | 21:29 | |
elarson | dhellmann: right, so the idea is to decorate a function with a hook point (@hookpoints.hook_point()). then a package can be installed that provides a decorator for that hook point. so by the time the config has been loaded, we've already applied the decorator | 21:29 |
*** exploreshaifali has quit IRC | 21:30 | |
elarson | I could move towards applying the hookpoints more lazily, but I think that would make them more complicated. | 21:30 |
elarson | sounds like it might be better to call a loaded_hooks() after the config has been loaded in order to provide the necessary debug info | 21:31 |
dhellmann | yes, I think moving that import to a point after the config is loaded and the logging is set up is what you want to do | 21:31 |
*** zzzeek has joined #openstack-oslo | 21:32 | |
dhellmann | elarson: you could also let the decorator build a cache of functions from plugins the first time it is called, instead of on import | 21:32 |
elarson | dhellmann: yeah, I think that is how it would have to work if I want to keep the application of the decorator reasonable | 21:33 |
dhellmann | elarson: a generic version of that might be a useful addition to stevedore | 21:33 |
elarson | dhellmann: the lazy binding or a helpful debug messsage that can logged after logging is configure? | 21:34 |
elarson | configured* | 21:34 |
dhellmann | elarson: the lazy binding decorator that uses a hook manager | 21:35 |
*** sputnik13 has quit IRC | 21:35 | |
*** dims has joined #openstack-oslo | 21:35 | |
*** e0ne has quit IRC | 21:36 | |
elarson | dhellmann: would it be reasonable to only target decorators? ie any plugin found needs to accept a function? | 21:37 |
dhellmann | elarson: probably a class that takes a namespace as argument and then provides a decorator method that takes a hook name. Then when the decorated function is called it makes sure the plugins are loaded and invokes the right hooks before & after calling the function. | 21:37 |
*** sputnik13 has joined #openstack-oslo | 21:37 | |
dhellmann | I guess it depends on what you want the hooks to do. Are all of the hook functions decorators, too? | 21:38 |
elarson | dhellmann: I actually started down that route, but preferred the decorator method as it lets users completely replace the function | 21:39 |
dhellmann | but yeah, if that's the thing you need, I think a version that didn't rely on oslo.config could live in stevedore and be used elsewhere | 21:39 |
dhellmann | elarson: oh, the interop folks are going to hate that :-/ | 21:39 |
elarson | dhellmann: why do you say that? | 21:39 |
dhellmann | "I installed this plugin and now the app doesn't pass defcore's tests" | 21:39 |
elarson | dhellmann: well the plugins are meant to be org specific. | 21:40 |
dhellmann | elarson: ok, I guess it's up to the org to make the choice about interop then | 21:41 |
elarson | exactly | 21:41 |
dhellmann | anyway, don't let me derail the real work you're doing, but if you have a pattern you want to add to stevedore we would welcome the contribution | 21:41 |
elarson | dhellmann: definitely. | 21:41 |
*** bknudson has joined #openstack-oslo | 21:43 | |
*** ozamiatin has quit IRC | 21:43 | |
elarson | I know the rax nova folks have implemented a similar system for applying org patches and I imagine there are similar patterns everyone uses to merge org with upstream, so I'm hoping there is some fruit to be shared there. | 21:43 |
dhellmann | elarson: feel free to add me to the reviewer list to make sure I see it | 21:43 |
elarson | time will tell... | 21:43 |
elarson | dhellmann: ah ok, will do | 21:43 |
*** stevemar2 is now known as stevemar | 21:43 | |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use alembic upgrade function/command directly https://review.openstack.org/191935 | 21:43 |
dhellmann | elarson: yeah, I think rax relies heavily on a monkeypatching tool, IIUC, and that's another approach but I like having it baked in as a feature | 21:43 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use alembic upgrade function/command directly https://review.openstack.org/191935 | 21:44 |
mtanino | dims: hi, may I ask about the Marian's fix 190175? | 21:46 |
dims | mtanino: sure | 21:46 |
mtanino | dims: After merged the patch, do we need to backport that fix into each component such as cinder, nova, etc? | 21:46 |
mtanino | or are there plan to use oslo.service directly at each component? | 21:47 |
dims | mtanino: i'd like to do the latter. (oslo.service directly at each component). we should be able to get the first revision of oslo.service API soon enough (a couple of weeks) | 21:48 |
dims | would that work? | 21:48 |
*** manishg_ has quit IRC | 21:48 | |
mtanino | dims: sure. It would work. | 21:48 |
dims | awesome! | 21:48 |
mtanino | thanks for letting me know :) | 21:49 |
dims | yw. is there a cinder bug? mtanino | 21:49 |
dims | looks like you have one https://bugs.launchpad.net/oslo.service/+bug/1463132 | 21:50 |
openstack | Launchpad bug 1463132 in oslo.service "Graceful shutdown was broken at Kilo" [Undecided,New] - Assigned to Mitsuhiro Tanino (mitsuhiro-tanino) | 21:50 |
mtanino | dims: https://bugs.launchpad.net/cinder/+bug/1464822 | 21:50 |
openstack | Launchpad bug 1464822 in Cinder "Cinder does not shutdown processes gracefully" [Low,In progress] - Assigned to Mitsuhiro Tanino (mitsuhiro-tanino) | 21:50 |
mtanino | yes. this one. | 21:50 |
mtanino | oh, different | 21:51 |
dims | cool lets make sure they refer to reach other so you get notified | 21:51 |
mtanino | dims: OK. so, I will duplicate my oslo.service bug for Marian's bug and add refecence into Cinder's bug. | 21:53 |
dims | sounds good mtanino | 21:53 |
mtanino | thanks :) | 21:53 |
*** boris-42 has joined #openstack-oslo | 22:02 | |
*** yamahata has joined #openstack-oslo | 22:03 | |
*** claudiub has quit IRC | 22:04 | |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Notify on the individual engine steps https://review.openstack.org/189554 | 22:07 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Handle conductor ctrl-c more appropriately https://review.openstack.org/189536 | 22:07 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add indestructible 99 bottles of beer example https://review.openstack.org/189526 | 22:07 |
*** mriedem is now known as mriedem_away | 22:09 | |
*** exploreshaifali has joined #openstack-oslo | 22:13 | |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Address concurrent mutation of sqlalchemy backend https://review.openstack.org/98689 | 22:13 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Address concurrent mutation of sqlalchemy backend https://review.openstack.org/98689 | 22:14 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Rename logbook module -> models module https://review.openstack.org/189995 | 22:15 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Rename logbook module -> models module https://review.openstack.org/189995 | 22:16 |
openstackgerrit | Joshua Harlow proposed openstack/taskflow: Add ability to use tooz to find and advertise workers https://review.openstack.org/151495 | 22:17 |
*** stevemar has quit IRC | 22:21 | |
*** david-lyle has quit IRC | 22:31 | |
*** Kevin_Bishop has quit IRC | 22:32 | |
*** yamahata has quit IRC | 22:33 | |
*** dims has quit IRC | 22:34 | |
*** ViswaV has joined #openstack-oslo | 22:35 | |
*** exploreshaifali has quit IRC | 22:35 | |
*** jecarey has quit IRC | 22:37 | |
openstackgerrit | Joshua Harlow proposed openstack/futurist: Allow adding periodic callables at runtime (WIP) https://review.openstack.org/191998 | 22:42 |
*** hogepodge has quit IRC | 22:52 | |
*** zzzeek has quit IRC | 22:56 | |
*** alexpilotti has joined #openstack-oslo | 23:01 | |
bknudson | has anyone used http://pythonhosted.org/testfixtures/mocking.html ? | 23:02 |
*** dims has joined #openstack-oslo | 23:03 | |
*** alexpilotti has quit IRC | 23:03 | |
*** dims has quit IRC | 23:03 | |
bknudson | this looks handy: http://pythonhosted.org/testfixtures/datetime.html#datetimes | 23:04 |
*** dims has joined #openstack-oslo | 23:04 | |
*** jaosorior has quit IRC | 23:05 | |
*** zzzeek has joined #openstack-oslo | 23:08 | |
*** david-lyle has joined #openstack-oslo | 23:11 | |
openstackgerrit | Joshua Harlow proposed openstack/automaton: Add a bigger CD-player state-machine doctest(ed) example https://review.openstack.org/191477 | 23:16 |
*** ganso_ has quit IRC | 23:18 | |
*** hogepodge has joined #openstack-oslo | 23:19 | |
openstackgerrit | Merged openstack/taskflow: Use hash path lookup vs path finding https://review.openstack.org/182132 | 23:22 |
*** EmilienM|afk is now known as EmilienM | 23:24 | |
*** zzzeek has quit IRC | 23:25 | |
*** ashleighfarnham has quit IRC | 23:42 | |
*** zzzeek has joined #openstack-oslo | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!