| openstackgerrit | Joshua Harlow proposed openstack/tooz: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192452 | 00:02 |
|---|---|---|
| *** crc32 has quit IRC | 00:02 | |
| *** ozialien has quit IRC | 00:04 | |
| *** enikanorov has quit IRC | 00:10 | |
| *** takedakn has joined #openstack-oslo | 00:11 | |
| *** takedakn has quit IRC | 00:16 | |
| *** amrith is now known as _amrith_ | 00:19 | |
| *** mtanino has quit IRC | 00:22 | |
| *** sdake has quit IRC | 00:24 | |
| vipul | :q | 00:29 |
| *** sputnik13 has joined #openstack-oslo | 00:36 | |
| openstackgerrit | Tony Breeds proposed openstack/oslo.concurrency: Add 2 callbacks to processutils.execute() https://review.openstack.org/190472 | 00:40 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/192381 | 00:41 |
| *** kgiusti1 has quit IRC | 00:44 | |
| *** sdake has joined #openstack-oslo | 00:48 | |
| *** sputnik13 has quit IRC | 00:54 | |
| *** crc32 has joined #openstack-oslo | 00:58 | |
| *** sdake_ has joined #openstack-oslo | 00:58 | |
| *** sdake has quit IRC | 01:01 | |
| openstackgerrit | Jeremy Stanley proposed openstack-dev/cookiecutter: Ignore .eggs cruft https://review.openstack.org/192480 | 01:02 |
| *** tobe has joined #openstack-oslo | 01:05 | |
| *** zz_jgrimm is now known as jgrimm | 01:06 | |
| *** jgrimm is now known as zz_jgrimm | 01:07 | |
| *** bknudson has joined #openstack-oslo | 01:11 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192458 | 01:19 |
| *** gordc has joined #openstack-oslo | 01:32 | |
| openstackgerrit | Merged openstack/oslo-specs: service: Adds Windows Oslo Service Workers spec https://review.openstack.org/191168 | 01:50 |
| openstackgerrit | Merged openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/192381 | 01:52 |
| openstackgerrit | Merged openstack/oslosphinx: Updated from global requirements https://review.openstack.org/192316 | 01:54 |
| openstackgerrit | Merged openstack/oslo.cache: Updated from global requirements https://review.openstack.org/192303 | 01:54 |
| openstackgerrit | Merged openstack/oslotest: Updated from global requirements https://review.openstack.org/192317 | 01:55 |
| *** vilobhmm has joined #openstack-oslo | 02:02 | |
| *** dims has quit IRC | 02:04 | |
| *** vilobhmm has quit IRC | 02:07 | |
| *** dims has joined #openstack-oslo | 02:08 | |
| *** dims has quit IRC | 02:19 | |
| openstackgerrit | Merged openstack/oslo.db: Updated from global requirements https://review.openstack.org/192306 | 02:35 |
| openstackgerrit | Merged openstack/oslo.policy: Updated from global requirements https://review.openstack.org/192310 | 02:36 |
| openstackgerrit | Merged openstack/oslo.i18n: Updated from global requirements https://review.openstack.org/192307 | 02:37 |
| openstackgerrit | Merged openstack/oslo.log: Updated from global requirements https://review.openstack.org/192308 | 02:40 |
| openstackgerrit | Merged openstack/oslo.context: Updated from global requirements https://review.openstack.org/192305 | 02:45 |
| openstackgerrit | Merged openstack/oslo.middleware: Updated from global requirements https://review.openstack.org/192309 | 02:45 |
| openstackgerrit | Merged openstack/oslo.serialization: Updated from global requirements https://review.openstack.org/192311 | 02:45 |
| openstackgerrit | Merged openstack/oslo.service: Updated from global requirements https://review.openstack.org/192312 | 02:45 |
| openstackgerrit | Merged openstack/oslo.versionedobjects: Updated from global requirements https://review.openstack.org/192314 | 02:45 |
| openstackgerrit | Merged openstack/oslo.concurrency: Updated from global requirements https://review.openstack.org/192304 | 02:47 |
| openstackgerrit | Merged openstack/oslo.utils: Updated from global requirements https://review.openstack.org/192313 | 02:50 |
| openstackgerrit | Merged openstack/oslo.vmware: Updated from global requirements https://review.openstack.org/192315 | 02:52 |
| *** sdake_ is now known as sdake | 03:00 | |
| *** ViswaV has quit IRC | 03:36 | |
| *** stevemar has quit IRC | 03:47 | |
| *** stevemar has joined #openstack-oslo | 03:48 | |
| *** vilobhmm has joined #openstack-oslo | 03:53 | |
| *** ozialien has joined #openstack-oslo | 04:03 | |
| *** jamespage has quit IRC | 04:06 | |
| *** jamespage has joined #openstack-oslo | 04:06 | |
| *** rushiagr_away is now known as rushiagr | 04:10 | |
| *** rjaiswal has quit IRC | 04:14 | |
| *** amotoki has joined #openstack-oslo | 04:19 | |
| *** achanda has joined #openstack-oslo | 04:23 | |
| *** r0j4z0 has joined #openstack-oslo | 04:25 | |
| *** gordc has quit IRC | 04:25 | |
| *** achanda has quit IRC | 04:32 | |
| *** rpodolyaka1 has joined #openstack-oslo | 04:54 | |
| *** ozialien has quit IRC | 05:01 | |
| *** achanda has joined #openstack-oslo | 05:02 | |
| *** rpodolyaka1 has quit IRC | 05:07 | |
| *** rpodolyaka1 has joined #openstack-oslo | 05:11 | |
| *** stevemar has quit IRC | 05:14 | |
| *** Kennan has quit IRC | 05:14 | |
| *** browne has quit IRC | 05:16 | |
| *** Kennan has joined #openstack-oslo | 05:20 | |
| *** nkrinner has joined #openstack-oslo | 05:31 | |
| *** achanda has quit IRC | 05:31 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192458 | 05:31 |
| *** tobe has quit IRC | 05:39 | |
| *** arnaud___ has joined #openstack-oslo | 05:41 | |
| *** rpodolyaka1 has quit IRC | 05:54 | |
| *** rpodolyaka1 has joined #openstack-oslo | 05:58 | |
| *** crc32 has quit IRC | 06:01 | |
| *** browne has joined #openstack-oslo | 06:03 | |
| *** browne has quit IRC | 06:14 | |
| *** rpodolyaka1 has quit IRC | 06:16 | |
| *** exploreshaifali has joined #openstack-oslo | 06:19 | |
| *** tobe has joined #openstack-oslo | 06:20 | |
| *** dulek has joined #openstack-oslo | 06:22 | |
| *** dulek has quit IRC | 06:25 | |
| *** browne has joined #openstack-oslo | 06:32 | |
| *** rjaiswal has joined #openstack-oslo | 06:33 | |
| *** salv-orl_ has joined #openstack-oslo | 06:41 | |
| *** boris-42 has quit IRC | 06:42 | |
| *** rpodolyaka1 has joined #openstack-oslo | 06:44 | |
| *** salv-orlando has quit IRC | 06:44 | |
| *** dulek has joined #openstack-oslo | 06:49 | |
| *** browne has quit IRC | 06:49 | |
| *** arnaud___ has quit IRC | 06:52 | |
| *** vilobhmm has quit IRC | 06:58 | |
| *** r0j4z0 has quit IRC | 07:05 | |
| *** sputnik13 has joined #openstack-oslo | 07:11 | |
| *** sdake has quit IRC | 07:20 | |
| *** haypo has joined #openstack-oslo | 07:28 | |
| *** ihrachyshka has joined #openstack-oslo | 07:29 | |
| *** jamespage_ has joined #openstack-oslo | 07:37 | |
| *** jamespage_ has quit IRC | 07:40 | |
| *** sputnik13 has quit IRC | 07:48 | |
| *** rpodolyaka1 has quit IRC | 07:51 | |
| *** sputnik13 has joined #openstack-oslo | 07:53 | |
| *** sputnik13 has quit IRC | 07:58 | |
| *** jaosorior has joined #openstack-oslo | 07:58 | |
| *** r0j4z0 has joined #openstack-oslo | 08:06 | |
| *** yassine has joined #openstack-oslo | 08:24 | |
| openstackgerrit | James Polley proposed openstack-dev/pbr: Fix everything by requiring setuptools>=16. https://review.openstack.org/187846 | 08:25 |
| *** rpodolyaka1 has joined #openstack-oslo | 08:32 | |
| *** dguitarbite has quit IRC | 08:38 | |
| openstackgerrit | Merged openstack/oslo.service: Use monotonic.monotonic and stopwatches instead of time.time https://review.openstack.org/190372 | 08:42 |
| *** mdbooth has quit IRC | 08:42 | |
| *** mdbooth has joined #openstack-oslo | 08:45 | |
| *** cdent has joined #openstack-oslo | 08:53 | |
| *** tobe has quit IRC | 08:58 | |
| *** tobe has joined #openstack-oslo | 08:59 | |
| *** cdent has quit IRC | 09:00 | |
| *** cdent has joined #openstack-oslo | 09:00 | |
| *** tobe_ has joined #openstack-oslo | 09:01 | |
| *** tobe_ has quit IRC | 09:01 | |
| *** tobe has quit IRC | 09:03 | |
| *** ozamiatin has joined #openstack-oslo | 09:09 | |
| openstackgerrit | Merged openstack/oslo.concurrency: Add 2 callbacks to processutils.execute() https://review.openstack.org/190472 | 09:13 |
| *** ndipanov has joined #openstack-oslo | 09:21 | |
| *** rpodolyaka1 has quit IRC | 09:23 | |
| *** tobe has joined #openstack-oslo | 09:23 | |
| *** rjaiswal has quit IRC | 09:24 | |
| *** rpodolyaka1 has joined #openstack-oslo | 09:25 | |
| *** e0ne has joined #openstack-oslo | 09:36 | |
| openstackgerrit | Merged openstack/tooz: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192452 | 09:37 |
| openstackgerrit | Merged openstack/tooz: Use the `excutils.raise_with_cause` after doing our type check https://review.openstack.org/192450 | 09:37 |
| openstackgerrit | Merged openstack/tooz: Use the 'driver_lock' around read operations https://review.openstack.org/192357 | 09:37 |
| *** e0ne is now known as e0ne_ | 09:43 | |
| *** e0ne_ is now known as e0ne | 09:47 | |
| *** ozamiatin has quit IRC | 09:49 | |
| *** arnaud___ has joined #openstack-oslo | 09:53 | |
| *** arnaud___ has quit IRC | 09:57 | |
| *** dguitarbite has joined #openstack-oslo | 10:06 | |
| *** davidlenwell has quit IRC | 10:12 | |
| *** davidlenwell has joined #openstack-oslo | 10:12 | |
| *** ozamiatin has joined #openstack-oslo | 10:14 | |
| *** dims has joined #openstack-oslo | 10:17 | |
| *** tobe has quit IRC | 10:17 | |
| *** dims has quit IRC | 10:17 | |
| *** dims has joined #openstack-oslo | 10:18 | |
| *** ativelkov has joined #openstack-oslo | 10:18 | |
| *** rpodolyaka1 has quit IRC | 10:19 | |
| ativelkov | Hi folks, may anybody answer a couple of questions on oslo.versioned_objects? | 10:19 |
| *** rpodolyaka1 has joined #openstack-oslo | 10:30 | |
| *** claudiub has joined #openstack-oslo | 10:31 | |
| *** rpodolyaka1 has quit IRC | 10:36 | |
| *** e0ne is now known as e0ne_ | 10:49 | |
| *** e0ne_ has quit IRC | 10:54 | |
| *** e0ne has joined #openstack-oslo | 10:56 | |
| eezhova | dims, ping | 11:04 |
| dims | hi eezhova | 11:05 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/oslo.messaging: Ensure consumer app don't use wait before stop https://review.openstack.org/192437 | 11:05 |
| eezhova | dims, can we try to make a release now? | 11:05 |
| dims | ativelkov: dansmith is not here yet. you can ask the question anyway and he'll respond when he gets in (west coast time) | 11:05 |
| dims | eezhova: i'd like to get the other 4 reviews in | 11:05 |
| dims | jd__: flaper87: can you please take a look at the 4 oslo.service reviews when you get a chance? https://review.openstack.org/#/q/status:open+project:openstack/oslo.service+branch:master,n,z | 11:06 |
| ativelkov | thanks dims | 11:06 |
| dims | sileht: dhellmann ^^ | 11:06 |
| dims | if you are around, please do as well, so we can try to get a test oslo.service release out | 11:06 |
| sileht | dims, I take a look | 11:07 |
| eezhova | dims, what has to be done to make a release? does the section "Tagging a Release" from https://wiki.openstack.org/wiki/Oslo/CreatingANewLibrary#Prepare_an_Initial_Release describe all steps? | 11:09 |
| ativelkov | dansmith: I'm looking at versioned_objects and it seems like CompoundFields do not coerce the values of their elements when the values are modified either by index access or by some method. I.e. the following is possible: http://paste.openstack.org/show/298141/ | 11:12 |
| ativelkov | So, the question is if this is intentional or not. If not, I may file a bug and submit a patch to fix this | 11:14 |
| *** rushiagr is now known as rushiagr_away | 11:24 | |
| dims | ativelkov: i'd file a bug for this one | 11:29 |
| *** e0ne is now known as e0ne_ | 11:29 | |
| ativelkov | dims, ok, I will | 11:30 |
| ativelkov | Another one (similar, but a bit different): if a compound field is "readonly", should we allow to modify the contents of the collection (e.g. pop elements from list or replace the, by index)? | 11:30 |
| toabctl | pypi says hacking 's latest version is 0.9.6 but seems that it's 0.10.2 ... | 11:34 |
| *** e0ne_ has quit IRC | 11:34 | |
| *** sdake has joined #openstack-oslo | 11:41 | |
| *** prad has quit IRC | 11:42 | |
| *** rpodolyaka1 has joined #openstack-oslo | 11:44 | |
| *** ganso_ has joined #openstack-oslo | 11:45 | |
| *** rpodolyaka1 has quit IRC | 11:46 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/oslo.messaging: Ensure consumer app don't use wait before stop https://review.openstack.org/192437 | 11:48 |
| *** rpodolyaka1 has joined #openstack-oslo | 11:48 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/oslo.messaging: Ensures that some assumptions are true. https://review.openstack.org/192437 | 11:50 |
| sileht | dhellmann, dims what do you think of https://review.openstack.org/192437 to ensure people don't run in some pitfalls (like Ritz hits yesterday) | 11:52 |
| openstackgerrit | Ilya Shakhat proposed openstack/oslo.config: Fix sorting of optional arguments https://review.openstack.org/192630 | 11:53 |
| *** dulek_ has joined #openstack-oslo | 11:57 | |
| *** dulek has quit IRC | 12:01 | |
| *** kgiusti has joined #openstack-oslo | 12:04 | |
| dims | sileht: thanks! | 12:04 |
| dims | sileht: LOOKING | 12:04 |
| *** sdake has quit IRC | 12:04 | |
| dims | oops! capslock | 12:04 |
| dims | sorry | 12:04 |
| dims | eezhova: watch the zuul page for the 4 oslo.service changes to merge - http://status.openstack.org/zuul/ - we can cut a release when those lang | 12:05 |
| eezhova | dims, okay, but according to zuul it seems there will be a merge conflict for https://review.openstack.org/#/c/190175/ | 12:07 |
| dims | eezhova: y, we can wait to see what merges and then run recheck or rebase | 12:07 |
| eezhova | dims, ack, I checked, it's gonna be a trivial conflict | 12:10 |
| dims | eezhova: k, rebase it and i'll fast approve it | 12:10 |
| dims | (please! :) | 12:10 |
| *** rushiagr_away is now known as rushiagr | 12:11 | |
| *** jlvillal has quit IRC | 12:11 | |
| dims | sileht: i see some changes in tests from blocking to eventlet..do we lose some test coverage because of that? | 12:12 |
| sileht | dims, the test itself is to test the dispatcher code so we can use any executor | 12:13 |
| sileht | dims, I have used the eventlet one, because blocking is the only executor that block on stop | 12:14 |
| sileht | s/on stop/on start | 12:14 |
| *** r0j4z0 has quit IRC | 12:19 | |
| *** e0ne has joined #openstack-oslo | 12:19 | |
| eezhova | dims, I can't rebase it now because it will involve rebasing a change it depends on and that one is already in a check queue and has no conflicts. Me or Marian can rebase when that one merges | 12:20 |
| dims | ack thanks | 12:20 |
| *** jlvillal has joined #openstack-oslo | 12:23 | |
| *** exploreshaifali has quit IRC | 12:27 | |
| sdague | dims: just kicking this in - https://review.openstack.org/#/c/189118/12 | 12:33 |
| *** dulek_ is now known as dulek | 12:34 | |
| *** dims has quit IRC | 12:36 | |
| *** dims has joined #openstack-oslo | 12:36 | |
| dims | sdague: nice thanx sean | 12:37 |
| sdague | we just needed the g-r sync so we didn't break any CD folks, but kevin landed that last night | 12:37 |
| dims | yes, i was tracking that | 12:38 |
| *** dulek has quit IRC | 12:39 | |
| *** dulek has joined #openstack-oslo | 12:40 | |
| *** bknudson has quit IRC | 12:41 | |
| dims | dhellmann: around? | 12:43 |
| dims | dhellmann: so fungi mentioned that gerrit admins can create a feature branch from external repo provided they shared the same commit in their history. i checked the ACL(s) of oslo.messaging and oslo.cache. since they looked similar i am guessing you may have karma to pull this off? please let me know. i have a git repo with the keystone stuff pulled in on top of oslo.cache ready to go | 12:45 |
| *** kgiusti has left #openstack-oslo | 12:45 | |
| *** exploreshaifali has joined #openstack-oslo | 12:46 | |
| *** e0ne is now known as e0ne_ | 12:55 | |
| *** e0ne_ is now known as e0ne | 12:57 | |
| openstackgerrit | Merged openstack/oslo.service: Add documentation for the service module https://review.openstack.org/191768 | 13:01 |
| *** bknudson has joined #openstack-oslo | 13:08 | |
| *** jamielennox is now known as jamielennox|away | 13:13 | |
| *** jecarey has joined #openstack-oslo | 13:13 | |
| *** amotoki_ has joined #openstack-oslo | 13:14 | |
| *** bknudson has quit IRC | 13:14 | |
| *** ozamiatin has quit IRC | 13:16 | |
| *** jamielennox|away is now known as jamielennox | 13:21 | |
| *** stpierre has joined #openstack-oslo | 13:26 | |
| *** rpodolyaka1 has quit IRC | 13:26 | |
| *** bknudson has joined #openstack-oslo | 13:27 | |
| *** stevemar has joined #openstack-oslo | 13:30 | |
| *** jamielennox is now known as jamielennox|away | 13:33 | |
| dansmith | ativelkov: yep | 13:35 |
| ativelkov | dansmith: I've submitted https://bugs.launchpad.net/oslo.versionedobjects/+bug/1466057 | 13:36 |
| openstack | Launchpad bug 1466057 in oslo.versionedobjects "Compound fields do not coerce values of their elements when modified" [Undecided,New] | 13:36 |
| dansmith | ativelkov: do you have a fix? | 13:37 |
| ativelkov | working on it right now | 13:37 |
| dansmith | okay | 13:37 |
| ativelkov | what about readonly attribute? | 13:37 |
| ativelkov | Another one (similar, but a bit different): if a compound field is "readonly", should we allow to modify the contents of the collection (e.g. pop elements from list or replace the, by index)? | 13:38 |
| *** rpodolyaka1 has joined #openstack-oslo | 13:39 | |
| dansmith | if you have an easy way to prevent it then sure, but I've never seen a readonly list attribute in practice :) | 13:39 |
| *** prad has joined #openstack-oslo | 13:40 | |
| ativelkov | we have them in Glance Artifacts | 13:41 |
| dansmith | okay | 13:42 |
| *** jamielennox|away is now known as jamielennox | 13:42 | |
| *** mriedem_away is now known as mriedem | 13:42 | |
| *** kbyrne has quit IRC | 13:46 | |
| openstackgerrit | Merged openstack/oslo.service: Eventlet service fixed https://review.openstack.org/192239 | 13:50 |
| * dhellmann notes a long backlog | 13:51 | |
| dhellmann | sileht: I like the general idea of enforcing those constraints; I'll put that in my review queue | 13:52 |
| dhellmann | dims: I see some jenkins -1 on those service patches, are those just the merge conflicts? | 13:52 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:58 | |
| dims | dhellmann: right. | 14:05 |
| dhellmann | ok | 14:05 |
| dims | eezhova: 2 of the reviews need rebases | 14:06 |
| dims | dhellmann: so fungi mentioned that gerrit admins can create a feature branch from external repo provided they shared the same commit in their history. i checked the ACL(s) of oslo.messaging and oslo.cache. since they looked similar i am guessing you may have karma to pull this off? please let me know. i have a git repo with the keystone stuff pulled in on top of oslo.cache ready to go | 14:06 |
| dims | dhellmann: https://github.com/dims/oslo.cache | 14:07 |
| fungi | chances are we'll need to update the acl to allow pushing a merge commit for review | 14:10 |
| *** kbyrne has joined #openstack-oslo | 14:15 | |
| dhellmann | dims: I'm going to need to spend some time figuring out how to do that | 14:15 |
| *** zzzeek has joined #openstack-oslo | 14:15 | |
| dhellmann | dims : I think I'm not a gerrit admin, so we probably need fungi or someone else from infra to assist | 14:16 |
| fungi | yeah, i can take care of the particulars, but be forewarned whatever gets pushed up is going to have to pass the currently configured jobs before it will be able to merge | 14:17 |
| dhellmann | fungi: wouldn't we just do "git merge" and then "git review" locally? or are merge commits different somehow in gerrit? | 14:17 |
| dhellmann | fungi: yep, we wouldn't want it any other way | 14:18 |
| dhellmann | dims: does the api for that code need to go through the incubator process? | 14:18 |
| fungi | right, i'd create a feature branch by pushing dims's custom fork into gerrit, then one of us (easiest if it's me since i'll already have my privs elevated for the branch push) will propose a merge commit of that feature branch to master for review, and that will get tests run on it | 14:20 |
| dims | fungi: yes, will work on it in the feature branch and finally merge when ready for jobs | 14:20 |
| dhellmann | fungi: it appears that dims' branch places the new files in a directory that isn't in the test path, so we'll end up testing things as we move the code into the right place to be picked up for packaging | 14:20 |
| fungi | though if the feature branch is going to need a while to bake in gerrit before merging to master, then we'll want to update the acl to allow one of you to propose a merge commit once you're ready | 14:20 |
| dims | dhellmann: that was the bare minimum import from keystone | 14:20 |
| dhellmann | dims: right, and I think that's a reasonable approach -- that lets us merge now, and then shuffle things around as normal commits | 14:21 |
| dims | cool | 14:21 |
| dhellmann | OTOH, do we need to do this in the incubator instead of oslo.cache? | 14:21 |
| dims | dhellmann: any advantages? | 14:22 |
| dhellmann | dims: if we expect lots of changes to the api, it's easier to test them in keystone if we're syncing code from the incubator | 14:22 |
| dhellmann | that ensures no backwards compatibility issues for keystone, without a lot of unstable releases of the lib | 14:23 |
| *** nkrinner has quit IRC | 14:23 | |
| dims | dhellmann: i think i have a way to do that dhellmann once we get past jobs into the master | 14:23 |
| dhellmann | ok | 14:23 |
| dims | dhellmann: we are testing oslo.service against nova, neutron, cinder for example | 14:23 |
| *** exploreshaifali has quit IRC | 14:24 | |
| dhellmann | dims: ok, if you have a way to do it that isn't too painful | 14:24 |
| dims | yes, we should not release the library until we know keystone would work | 14:24 |
| dims | y, should be ok | 14:24 |
| *** mdbooth has quit IRC | 14:27 | |
| *** ozamiatin has joined #openstack-oslo | 14:27 | |
| dims | fungi: thanks. is there another project that has the acl for merge commit i can copy from? | 14:28 |
| fungi | dims: swift, keystone... | 14:28 |
| dims | fungi: ack looking | 14:29 |
| * stevemar starts reading | 14:30 | |
| *** zz_jgrimm is now known as jgrimm | 14:30 | |
| stevemar | dims, which library should you not release until you know keystone would work? | 14:31 |
| dims | stevemar: oslo.cache | 14:31 |
| stevemar | is it available as a 0.1 now? | 14:32 |
| dims | stevemar: we are just getting the ball rolling with the initial import | 14:32 |
| stevemar | i'm happily push a patch for keystone that is trying to use oslo.cache 0.1 | 14:33 |
| stevemar | i'll* | 14:33 |
| stevemar | i guess i could do it from an external git repo too | 14:34 |
| dims | stevemar: thanks! | 14:39 |
| *** kgiusti has joined #openstack-oslo | 14:43 | |
| *** exploreshaifali has joined #openstack-oslo | 14:48 | |
| *** sdake_ has joined #openstack-oslo | 14:53 | |
| *** dulek has quit IRC | 14:53 | |
| *** Kevin_Bishop has joined #openstack-oslo | 14:54 | |
| *** browne has joined #openstack-oslo | 15:01 | |
| openstackgerrit | Marian Horban proposed openstack/oslo.service: Graceful shutdown WSGI/RPC server https://review.openstack.org/190175 | 15:03 |
| *** openstackgerrit has quit IRC | 15:05 | |
| *** openstackgerrit has joined #openstack-oslo | 15:05 | |
| *** amotoki_ has quit IRC | 15:06 | |
| *** amotoki_ has joined #openstack-oslo | 15:08 | |
| *** rpodolyaka1 has quit IRC | 15:08 | |
| Kevin_Bishop | Hello, could someone give me an approximate timeline for the oslo_service development? Thanks. | 15:12 |
| *** haypo has left #openstack-oslo | 15:12 | |
| dhellmann | Kevin_Bishop: we're working on it this cycle; are you waiting to be able to use it? | 15:13 |
| Kevin_Bishop | dhellmann, yeah we're trying to switch over some of our incubator code. | 15:14 |
| dhellmann | Kevin_Bishop: ok, dims probably has a better handle on dates, but the goal is to have a new lib this cycle | 15:15 |
| dhellmann | Kevin_Bishop: you're already using the incubated version? | 15:15 |
| Kevin_Bishop | dhellmann: that's great! Is there's anything we can do to help out? | 15:16 |
| dhellmann | Kevin_Bishop: I'll have to defer to dims on that, but I'm sure they would appreciate help developing/testing the adoption docs since there are those API changes | 15:16 |
| dims | Kevin_Bishop: which project? | 15:18 |
| Kevin_Bishop | dhellmann: Great. Yes, we are currently using the incubated version. We've been trying to move most of our incubated code to graduated modules. | 15:18 |
| Kevin_Bishop | Barbican | 15:18 |
| *** mtanino has joined #openstack-oslo | 15:18 | |
| dhellmann | Kevin_Bishop: that's good news, I'm really glad to hear it. The service stuff is pretty close. | 15:19 |
| dims | Kevin_Bishop: we are running some initial code against nova, neutron. we may have a pre-alpha/test release in a week or so. | 15:19 |
| dims | once we get it stable and we are happy with the API, we'll propose a global requirements change | 15:20 |
| Kevin_Bishop | dhellmann, dims: That's great to hear! | 15:20 |
| dims | Kevin_Bishop: redrobot is your liaison per https://wiki.openstack.org/wiki/CrossProjectLiaisons#Oslo - so if one of you are in the oslo meetings, you can follow along | 15:22 |
| dims | we'd love for you to try the initial releases and give us feedback if you are ok that we may break stuff | 15:23 |
| Kevin_Bishop | dims: Yeah, I'll be sure to let him know. | 15:23 |
| dims | Kevin_Bishop: thank you sir | 15:26 |
| *** yamahata has joined #openstack-oslo | 15:31 | |
| *** yamahata has quit IRC | 15:32 | |
| *** yamahata has joined #openstack-oslo | 15:32 | |
| *** ozamiatin has quit IRC | 15:35 | |
| *** jamielennox is now known as jamielennox|away | 15:38 | |
| *** harlowja_at_home has joined #openstack-oslo | 15:42 | |
| ativelkov | dansmith: I've found another one which looks like a bug to me: https://bugs.launchpad.net/oslo.versionedobjects/+bug/1466141 | 15:44 |
| openstack | Launchpad bug 1466141 in oslo.versionedobjects "Type inheritance is ignored when coercing object fields" [Undecided,New] | 15:44 |
| dansmith | ativelkov: well, the existing behavior is what we intended | 15:45 |
| dansmith | ativelkov: but if you want to allow inheritance, we probably need a different field, or a flag | 15:45 |
| *** viktors is now known as viktors|afk | 15:45 | |
| ativelkov | ah, I see | 15:45 |
| dansmith | ativelkov: the point of that field is sort of an interface definition... like "an object that implements the FixedIP model is placed here" | 15:46 |
| dansmith | which we wanted so that we could have alternate implementations of a given object | 15:46 |
| ativelkov | I actually found it in one of the tests which began to fail as soon as I fixed compound coercing | 15:46 |
| dansmith | we could have done it with inheritance, but we look them up via name in the registry | 15:46 |
| dansmith | okay | 15:46 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192458 | 15:47 |
| *** jamielennox|away is now known as jamielennox | 15:47 | |
| dansmith | ativelkov: I think maybe a different field that takes a class instead of a name would be good, | 15:47 |
| dansmith | and then we can make sure that what they provide has the defined class in its mro ancestry | 15:48 |
| ativelkov | https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/tests/test_objects.py#L285 - this fails, as MyList (which is actually ab ObjectListBase) expects its items to be VersionedObject, but in this tests they are MyObj - nad they always will be some subclass of VersionedObject | 15:49 |
| dansmith | ativelkov: yeah, just an oversight I think.. good catch, glad you caught that :D | 15:50 |
| *** ashleighfarnham has joined #openstack-oslo | 15:50 | |
| *** stevemar has quit IRC | 15:50 | |
| ativelkov | Classes are bad in this case, as you won't be able to define a class which reference themselves | 15:50 |
| ativelkov | e.g. http://paste.openstack.org/show/298801/ is impossible to interpret | 15:51 |
| dansmith | really? | 15:52 |
| *** stevemar has joined #openstack-oslo | 15:52 | |
| dansmith | I guess because it's run at parse time | 15:52 |
| ativelkov | yup | 15:52 |
| dansmith | object path/ | 15:52 |
| dansmith | ? | 15:52 |
| dansmith | nova.objects.Baz ? | 15:52 |
| *** jecarey has quit IRC | 15:53 | |
| *** exploreshaifali has quit IRC | 15:53 | |
| ativelkov | checking.. | 15:54 |
| ativelkov | No, seems like you cannot reference it - you'll have to import nova.objects from within nova.objects, which is obviously impossible | 15:57 |
| dansmith | ativelkov: no, I meant pass it a string | 15:57 |
| dansmith | ativelkov: and look up the ancestor class during coerce | 15:57 |
| *** jamielennox is now known as jamielennox|away | 15:58 | |
| ativelkov | ah, got it | 15:58 |
| *** tsekiyama has joined #openstack-oslo | 15:58 | |
| *** rpodolyaka1 has joined #openstack-oslo | 15:58 | |
| ativelkov | so, may be use the same obj_name instead, but - if some field (say, 'include_ancestors') is present - check obj_names of the bases as well? | 15:59 |
| dansmith | well, the problem with that is that we may have two objects with the same "name" that don't inherit from a common base | 15:59 |
| *** haypo has joined #openstack-oslo | 16:00 | |
| ativelkov | 'cause I like the concept of obj_name which allows to override the pythonic name, and it would be great to utilize it for inheritance lookups as well | 16:00 |
| dansmith | how about we just imply that the thing has to be a descendant of the name, if the name includes dots and is interpreted as a common ancestor class? | 16:00 |
| dansmith | and if no dots, we just match the obj_name() ? | 16:01 |
| ativelkov | probably I don't understand something | 16:01 |
| dansmith | hang on | 16:02 |
| ativelkov | So, in current implementation you may assign a field with any object which have the specified name, regardless of its inheritance hierarchy, right? | 16:03 |
| *** exploreshaifali has joined #openstack-oslo | 16:03 | |
| dansmith | ativelkov: https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_objects.py#L125-136 | 16:04 |
| dansmith | we intentionally make that compatible with MyObj, but not sharing a common ancestor | 16:04 |
| ativelkov | I got it, yes | 16:05 |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Delay interpolating the LOG string https://review.openstack.org/192749 | 16:05 |
| *** jamielennox|away is now known as jamielennox | 16:06 | |
| *** arnaud___ has joined #openstack-oslo | 16:06 | |
| *** vilobhmm has joined #openstack-oslo | 16:06 | |
| dhellmann | kgiusti, dims, sileht : I'm getting an error installing qpid because it can't find uuid/uuid.h -- do you know what ubuntu package I need for that? | 16:07 |
| ativelkov | But still, if you create a class which inherits from this MyObj2 and make an instance of it, it should be coercible to MyObj in the same way - even if it has some different obj_name | 16:07 |
| dhellmann | that's actually in the oslo.messaging unit tests | 16:07 |
| *** ihrachyshka has quit IRC | 16:07 | |
| *** rpodolyaka1 has quit IRC | 16:08 | |
| dims | dansmith: uuid-dev and swig was mentioned in https://review.openstack.org/#/c/191107/ | 16:09 |
| dhellmann | dims: thanks | 16:09 |
| *** rpodolyaka1 has joined #openstack-oslo | 16:09 | |
| dansmith | ativelkov: well, I dunno | 16:10 |
| * dansmith thinks | 16:10 | |
| *** ozamiatin has joined #openstack-oslo | 16:10 | |
| *** dims_ has joined #openstack-oslo | 16:11 | |
| dansmith | ativelkov: so you're thinking that coercion would take the thing you have, check its MRO and just make sure that something in the chain has an obj_name() equal to the one set on the field? | 16:11 |
| ativelkov | dansmith: yes, at least optionally if there is appropriate flaf | 16:11 |
| dansmith | ativelkov: yeah, okay, I think that will be okay.. it's not how I've been thinking about it so it takes me a minute to wrap my head around it | 16:12 |
| dansmith | ativelkov: but I think that'll be okay, especially if it's a flag | 16:12 |
| ativelkov | otherwise I don't know what to do with ObjectListBase - it's going to store its "objects" as a List of VersionedObject | 16:12 |
| *** achanda has joined #openstack-oslo | 16:12 | |
| dansmith | ativelkov: we define each of our lists specifically | 16:12 |
| ativelkov | which is an abstract class anyway | 16:12 |
| ativelkov | ah, I see. | 16:12 |
| dansmith | ativelkov: so we use that as the base, but we redefine the objects field as "a list of $these" | 16:13 |
| dansmith | but, we probably have a case coming up where allowing that to be more generic will be helpful actually | 16:13 |
| *** dims has quit IRC | 16:13 | |
| *** jecarey has joined #openstack-oslo | 16:13 | |
| ativelkov | got it | 16:13 |
| ativelkov | Yes, in Artifacts we'll need generic Artifact references which may get some specific artifact as its value | 16:14 |
| dansmith | yeah | 16:14 |
| ativelkov | So, I'll add a flag for that | 16:14 |
| *** ashleighfarnham has quit IRC | 16:18 | |
| *** ashleighfarnham has joined #openstack-oslo | 16:18 | |
| dansmith | cool, sounds good | 16:18 |
| openstackgerrit | Joshua Harlow proposed openstack/tooz: Remove file-driver special no-async abilities https://review.openstack.org/192758 | 16:21 |
| *** achanda_ has joined #openstack-oslo | 16:21 | |
| *** harlowja_at_home has quit IRC | 16:22 | |
| sileht | dhellmann, uuid-dev | 16:23 |
| sileht | dhellmann,and swig is needed too | 16:23 |
| sileht | hum I'm late | 16:24 |
| *** achanda has quit IRC | 16:24 | |
| *** sdake_ is now known as sdake | 16:25 | |
| kgiusti | sileht: dhellmann: yes - on ubuntu it's uuid-dev, fedora/rhel is libuuid-devel. swig should be the same on both. | 16:27 |
| kgiusti | sileht: dhellmann: flaper87 has submitted a patch to proton that will remove these requirements, but we won't see that until its next release | 16:28 |
| *** vilobhmm has quit IRC | 16:28 | |
| *** achanda_ has quit IRC | 16:31 | |
| *** openstackgerrit has quit IRC | 16:33 | |
| *** openstackgerrit has joined #openstack-oslo | 16:34 | |
| *** jamielennox is now known as jamielennox|away | 16:35 | |
| *** arnaud___ has quit IRC | 16:38 | |
| *** ozamiatin has quit IRC | 16:40 | |
| *** cdent_ has joined #openstack-oslo | 16:43 | |
| *** mtanino has quit IRC | 16:45 | |
| *** cdent has quit IRC | 16:45 | |
| *** cdent_ is now known as cdent | 16:45 | |
| *** mtanino has joined #openstack-oslo | 16:49 | |
| *** jamielennox|away is now known as jamielennox | 16:49 | |
| *** timsim has left #openstack-oslo | 16:51 | |
| *** alexpilotti has joined #openstack-oslo | 16:51 | |
| *** exploreshaifali has quit IRC | 16:52 | |
| *** yamahata has quit IRC | 16:54 | |
| *** yamahata has joined #openstack-oslo | 16:55 | |
| *** achanda has joined #openstack-oslo | 16:57 | |
| *** _amrith_ is now known as amrith | 16:58 | |
| *** vilobhmm has joined #openstack-oslo | 16:58 | |
| *** haypo has quit IRC | 16:59 | |
| *** e0ne has quit IRC | 17:00 | |
| dhellmann | kgiusti, sileht : thanks, it's working again | 17:01 |
| *** achanda has quit IRC | 17:02 | |
| *** exploreshaifali has joined #openstack-oslo | 17:03 | |
| *** rushiagr is now known as rushiagr_away | 17:04 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/oslo.service: Test for instantaneous shutdown fixed https://review.openstack.org/192427 | 17:04 |
| openstackgerrit | Merged openstack/oslo.service: Graceful shutdown WSGI/RPC server https://review.openstack.org/190175 | 17:05 |
| *** achanda has joined #openstack-oslo | 17:05 | |
| *** yamahata has quit IRC | 17:08 | |
| *** achanda has quit IRC | 17:11 | |
| *** ndipanov has quit IRC | 17:12 | |
| *** jamielennox is now known as jamielennox|away | 17:13 | |
| *** rpodolyaka1 has quit IRC | 17:14 | |
| *** dims_ has quit IRC | 17:16 | |
| *** rpodolyaka1 has joined #openstack-oslo | 17:16 | |
| *** dims has joined #openstack-oslo | 17:17 | |
| *** achanda has joined #openstack-oslo | 17:18 | |
| *** achanda_ has joined #openstack-oslo | 17:22 | |
| *** ViswaV has joined #openstack-oslo | 17:24 | |
| *** achanda has quit IRC | 17:25 | |
| *** ViswaV has quit IRC | 17:28 | |
| *** ViswaV has joined #openstack-oslo | 17:29 | |
| *** sdake_ has joined #openstack-oslo | 17:30 | |
| *** achanda_ has quit IRC | 17:33 | |
| *** sdake has quit IRC | 17:34 | |
| *** rpodolyaka1 has quit IRC | 17:34 | |
| *** rpodolyaka1 has joined #openstack-oslo | 17:37 | |
| *** yassine has quit IRC | 17:39 | |
| *** rpodolyaka1 has quit IRC | 17:41 | |
| *** amotoki_ has quit IRC | 17:43 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Use `encodeutils.exception_to_unicode` for exception -> string function https://review.openstack.org/192458 | 17:49 |
| *** e0ne has joined #openstack-oslo | 17:51 | |
| *** achanda has joined #openstack-oslo | 17:54 | |
| *** sdake has joined #openstack-oslo | 17:58 | |
| ativelkov | dansmith: there is one more issue with obj_name() and inheritance | 18:00 |
| *** browne has quit IRC | 18:01 | |
| *** sdake_ has quit IRC | 18:01 | |
| dansmith | ativelkov: okay | 18:03 |
| ativelkov | imagine there is a 'Base' class which overrides its obj_name to some value. What will happen to all its descendant "Derived"? It will have to override its obj_name's back, or it will be considered to be "Base" as well (i.e. there will be no way to create a field which coerces to "Derived") | 18:03 |
| *** browne has joined #openstack-oslo | 18:03 | |
| dansmith | ativelkov: well, that's kinda what I didn't do this in the first place, because if you change your obj_name() then you're "different" | 18:05 |
| ativelkov | yup | 18:05 |
| ativelkov | in Artifacts we do it slightly different | 18:06 |
| ativelkov | we do not inherit this name at all | 18:06 |
| ativelkov | it is always set to classname if it is not explicitly overridden at that particular class | 18:06 |
| ativelkov | Say, if "Base" overrides its name to "MyBase", then its descendant Derived will still be "Derived" unless explicitly renamed to something else | 18:08 |
| ativelkov | the renames are not inherited (yet the inheritance chain may be traced, so "Derived" may be used where "MyBase" is required) | 18:09 |
| *** sdake_ has joined #openstack-oslo | 18:10 | |
| *** sdake has quit IRC | 18:14 | |
| ativelkov | I'll try to think about it a bit more and come up with some solution. | 18:14 |
| *** achanda has quit IRC | 18:17 | |
| krotscheck | I'm pondering making a change to oslo config. The automatic configuration detection that I'm doing in the cors middleware requires a call to list_all_sections, which basically means that I can't really support duck typing. My question is: Is it feasible to make the config iteration work for non-registered groups, too? (and/or should we?) | 18:20 |
| *** cdent has quit IRC | 18:21 | |
| krotscheck | Relevant code: | 18:21 |
| krotscheck | https://review.openstack.org/#/c/191238/2/oslo_middleware/cors.py | 18:22 |
| krotscheck | Also, dhellmann - I modified that patch as you requested. https://review.openstack.org/#/c/190753/3 | 18:22 |
| dhellmann | krotscheck: you're already calling methods of the configopts class, why do you want to check the type? | 18:23 |
| krotscheck | dhellmann: Usability, mostly. | 18:23 |
| krotscheck | dhellmann: I know that as an engineer, I always appreciate having a sane error message. | 18:23 |
| dhellmann | krotscheck: we don't really do that anywhere else, though, do we? | 18:24 |
| krotscheck | dhellmann: ... I don't really accept "We don't do it anywhere else" as a reason not to improve usability, even in a small way. | 18:26 |
| *** isq_ has quit IRC | 18:27 | |
| dhellmann | krotscheck: python programmers do not do type checks, though, it's just not something we do | 18:27 |
| lifeless | dhellmann: https://review.openstack.org/#/c/192019/ and https://review.openstack.org/#/c/192024/ are both shallow and green if you have a few minutes | 18:28 |
| lifeless | dhellmann: the one above them requires integration work in other repos that I'm working through now | 18:28 |
| krotscheck | dhellmann: Didn't you just ask me to implement a type check to support backwards compatibility in https://review.openstack.org/#/c/190753/3 ? | 18:29 |
| dhellmann | lifeless: I've put them in my queue, but I'm monitoring release tags this afternoon so it may be a little while | 18:29 |
| dhellmann | krotscheck: I suggested 2 ways to solve that. Create a different base class, or look at what you're calling to see if it expects all of the arguments. That's not the same thing as saying "this argument must subclass from a specific class" | 18:29 |
| lifeless | dhellmann: ack | 18:29 |
| krotscheck | dhellmann: Fair point. | 18:30 |
| lifeless | dhellmann: opinion solicited: do we want our setup.py's to be executable ? | 18:31 |
| krotscheck | dhellmann: Actually, no. Just because one of the things you suggested is different, it doesn't negate the fact that the other one - inspecting a callable - is basically type checking. But then, it's all semantics. | 18:31 |
| dhellmann | lifeless: I don't usually expect them to be. Does it matter? | 18:31 |
| lifeless | dhellmann: or do we not care, since we're not ever telling people to just run them (its always in the context of pip, or a venv, where 'python setup.py' is the go) | 18:31 |
| lifeless | dhellmann: one of the side effects of moving from in-place file rewrites to atomic ones is that projects with +x on their setup.py are losing them in the next requirements sync | 18:32 |
| krotscheck | (Note: I don't really have an argument about removing the type checking) | 18:32 |
| *** mriedem has quit IRC | 18:32 | |
| dhellmann | krotscheck: I'm proposing a way to enforce an API, not look at the type. Someone could subclass ConfigOpts and override a method in a way that would not work with your code. | 18:32 |
| lifeless | dhellmann: I think a strong case can be made for consistency; and then its much more ambiguous about whether we want them to be +x or -x | 18:32 |
| krotscheck | Or a disagreement for that matter. | 18:32 |
| dhellmann | lifeless: I think -x | 18:33 |
| dims | lifeless: default "-x" | 18:33 |
| lifeless | dims: since its a sync job, its not really a 'default' thing - its going to be 'the thing' | 18:33 |
| krotscheck | dhellmann: Subclass Configopts? That's crazytalk! | 18:33 |
| krotscheck | Who would EVER do that. | 18:33 |
| * krotscheck is pointedly not pointing at himself. | 18:33 | |
| lifeless | I'll raise it on the dev list as the way we want it and see if there are screams | 18:33 |
| dims | lifeless: ++ | 18:33 |
| openstackgerrit | Jon Grimm proposed openstack/oslotest: Allow ``OS_DEBUG`` environment variable to specify log level. https://review.openstack.org/192055 | 18:34 |
| openstackgerrit | Michael Krotscheck proposed openstack/oslo.middleware: Support PasteDeploy https://review.openstack.org/191238 | 18:34 |
| dhellmann | krotscheck: yeah, I don't feel that strongly about that one line of code, but as a general rule I don't bother with type checking because it doesn't actually enforce an API and it makes passing fakes into the thing for testing harder. | 18:34 |
| *** ozamiatin has joined #openstack-oslo | 18:36 | |
| openstackgerrit | Jon Grimm proposed openstack/oslotest: Allow ``OS_DEBUG`` environment variable to specify log level. https://review.openstack.org/192055 | 18:36 |
| openstackgerrit | Jon Grimm proposed openstack/oslotest: Allow ``OS_DEBUG`` environment variable to specify log level. https://review.openstack.org/192055 | 18:43 |
| *** rpodolyaka1 has joined #openstack-oslo | 18:50 | |
| *** mriedem has joined #openstack-oslo | 18:54 | |
| openstackgerrit | Merged openstack/oslo.service: Test for instantaneous shutdown fixed https://review.openstack.org/192427 | 18:59 |
| *** amrith is now known as _amrith_ | 19:01 | |
| sdague | so... I noticed in looking at fungi's patch - https://review.openstack.org/#/c/184387/ and looking at the test results closely, that the oslo.db test code for opportunitistic db testing doesn't work if the url is mysql+pymysql | 19:03 |
| sdague | anyone addressing that on the oslo.db side? | 19:03 |
| fungi | zzzeek: ^ you had figured out the test fixes on that (thanks again, btw!) | 19:04 |
| zzzeek | fungi: np | 19:04 |
| *** ozamiatin has quit IRC | 19:05 | |
| sdague | zzzeek: cool, how far away from release are those fixes? | 19:05 |
| *** jaosorior has quit IRC | 19:05 | |
| fungi | i'm also looking now to see if i can suss out how that opportunistic detection works | 19:06 |
| sdague | because we definitely need them before landing https://review.openstack.org/#/c/184387/ | 19:06 |
| fungi | but i'm pretty new to that codebase, so more seasoned eyes are welcome | 19:06 |
| zzzeek | sdague: in…oslo.db ? um. looks like the patch is still bouncing around https://review.openstack.org/#/c/184392/ | 19:07 |
| fungi | i wonder if we need a depends-on in 184387 (and similar) to 184392 | 19:09 |
| fungi | oh! though i bet we need more than just landing 184392 | 19:09 |
| fungi | we probably also need to tag a new oslo.db release with that? | 19:10 |
| sdague | fungi: yeh, we'll need a new release | 19:10 |
| fungi | dims: dhellmann: might make sense to up the priority on 184392 since we need it landed _and_ released before we can sanely move forward on pymysql for a lot of the leaf projects (at least the ones using oslo.db opportunistic database detection in their tests) | 19:12 |
| openstackgerrit | Greg Hellings proposed openstack-dev/pbr: Add support for local versioning schemes https://review.openstack.org/192842 | 19:14 |
| *** openstackgerrit has quit IRC | 19:16 | |
| *** openstackgerrit has joined #openstack-oslo | 19:16 | |
| dhellmann | fungi: looks like it needs a rebase? | 19:18 |
| dims | dhellmann: when you get a chance, we can cut a test version of oslo.service now from HEAD now, all the changes are in. | 19:21 |
| dhellmann | dims: 0.1.0? | 19:22 |
| dims | dhellmann: sure | 19:23 |
| dims | thanks! | 19:23 |
| dhellmann | dims: done | 19:24 |
| *** jungleboyj has joined #openstack-oslo | 19:24 | |
| dhellmann | dims: as a note, the release notes script doesn't work quite right for the first release, so you have to edit the output before sending the announcement. | 19:25 |
| *** jungleboyj has quit IRC | 19:32 | |
| *** arnaud___ has joined #openstack-oslo | 19:39 | |
| *** ViswaV has quit IRC | 19:41 | |
| *** arnaud___ has quit IRC | 19:43 | |
| *** nkrinner has joined #openstack-oslo | 19:46 | |
| *** jungleboyj has joined #openstack-oslo | 19:50 | |
| *** rpodolyaka1 has quit IRC | 19:51 | |
| fungi | dhellmann: oh, probably. it was fine a few days ago but i will admit to having not had time to check back in on it | 19:52 |
| *** ozamiatin has joined #openstack-oslo | 19:53 | |
| jungleboyj | dims: I am here now. Sorry my connection timed out. | 19:54 |
| jungleboyj | I have jecarey lurking to help as he was going to look at fixing this. | 19:54 |
| jecarey | jungleboyj, there is a patch to oslo-incubator already: https://review.openstack.org/#/c/186533/ | 19:55 |
| jungleboyj | So, the issue is that we want to get the fix for this problem into stable/kilo as it is kidn-of a bad thing: | 19:55 |
| jungleboyj | jecarey: Oh, | 19:55 |
| jecarey | dims you -1'd it ... should we just skip it and make the change directly to stable/kilo if cinder needs it? | 19:56 |
| jecarey | dhellmann, ^^ | 19:56 |
| jungleboyj | jecarey: That is good. So, we need to convince dims we need it. :-) | 19:56 |
| dhellmann | jungleboyj, jecarey, dims : it should be ok to backport fixes to supported stable branches, even in the incubator | 19:56 |
| *** exploreshaifali has quit IRC | 19:57 | |
| *** Rockyg has joined #openstack-oslo | 19:57 | |
| dhellmann | jungleboyj, jecarey, dims: if the code is being removed from the master incubator, we could skip that step and go straight to the stable branch | 19:57 |
| *** Rockyg has quit IRC | 19:57 | |
| *** _amrith_ is now known as amrith | 19:57 | |
| jungleboyj | dhellmann: Ok ... | 19:58 |
| *** Rockyg has joined #openstack-oslo | 19:59 | |
| *** stevemar has quit IRC | 20:00 | |
| dhellmann | jungleboyj: we did that with one or two patches during kilo, I think | 20:02 |
| *** rpodolyaka1 has joined #openstack-oslo | 20:03 | |
| jecarey | dhellmann, patch is already ready for oslo-incubator https://review.openstack.org/#/c/186533/ | 20:04 |
| jecarey | seems cleaner in this case to get that in and then cherry pick it back to stable/kilo. | 20:04 |
| mtanino | jecarey: final version of that fix is here. https://review.openstack.org/#/c/190175/ 186533 needs a little update I think | 20:06 |
| *** kgiusti has quit IRC | 20:06 | |
| mtanino | and then cherry pick into stable/kilo right? | 20:07 |
| jecarey | mtanino, that's what I was thinking, but it shoulds like you could skip it and go straight to stable/kilo | 20:08 |
| mtanino | jecarey: you mean go straight to backport 190175 to stable/kilo, right? | 20:09 |
| dhellmann | jecarey: that patch is against master, for a module we're deleting from master | 20:15 |
| jecarey | dhellmann, ok ... so mtanino should just go staight to backport 190175 to stable/kilo. | 20:16 |
| mtanino | jecarey: It's make sense for me. | 20:17 |
| *** rpodolyaka1 has quit IRC | 20:17 | |
| *** ozamiatin has quit IRC | 20:18 | |
| *** rpodolyaka1 has joined #openstack-oslo | 20:19 | |
| *** achanda has joined #openstack-oslo | 20:19 | |
| *** achanda_ has joined #openstack-oslo | 20:23 | |
| dhellmann | jecarey, mtanino : that's my view, but we should get dims to confirm before you spend a bunch more effort on it. I would expect (a) a patch to oslo.service if there isn't one, (b) a patch to stable/kilo (c) patches to any other stable branches where this is a problem you want to fix | 20:23 |
| mtanino | dhellmann: I understand. | 20:24 |
| *** achanda has quit IRC | 20:26 | |
| *** ViswaV has joined #openstack-oslo | 20:32 | |
| *** e0ne has quit IRC | 20:32 | |
| *** e0ne has joined #openstack-oslo | 20:33 | |
| *** achanda_ has quit IRC | 20:38 | |
| *** e0ne has quit IRC | 20:41 | |
| openstackgerrit | Greg Hellings proposed openstack-dev/pbr: Add support for local versioning schemes https://review.openstack.org/192842 | 20:46 |
| dims | jecarey: i want to be sure that the change in 186533 exactly is the same that is in oslo.service first. then we can debate master oslo-incubator vs stable/kilo oslo-incubator | 20:47 |
| *** yamahata has joined #openstack-oslo | 20:54 | |
| *** haypo has joined #openstack-oslo | 20:54 | |
| *** nkrinner has quit IRC | 20:58 | |
| *** rpodolyaka1 has quit IRC | 20:59 | |
| *** jungleboyj has quit IRC | 21:02 | |
| *** ViswaV has quit IRC | 21:17 | |
| *** ViswaV has joined #openstack-oslo | 21:20 | |
| *** openstackgerrit has quit IRC | 21:24 | |
| *** openstackgerrit has joined #openstack-oslo | 21:24 | |
| *** ViswaV has quit IRC | 21:24 | |
| *** sdake_ is now known as sdake | 21:25 | |
| *** ViswaV has joined #openstack-oslo | 21:27 | |
| openstackgerrit | Jeremy Stanley proposed openstack/oslo.db: Switch from MySQL-python to PyMySQL https://review.openstack.org/184392 | 21:27 |
| fungi | dhellmann: dims: zzzeek: sdague: ^ rebased (minor conflict on one line of oslo_db/tests/sqlalchemy/test_sqlalchemy.py) | 21:28 |
| *** Kevin_Bishop has quit IRC | 21:30 | |
| Rockyg | dhellmann: I'll be back in about half an hour if you'll be around. Otherwise I'll look for you here tomorrow | 21:33 |
| haypo | fungi, hey. i'm porting openstack to python 3, and your patches helped me :) | 21:33 |
| fungi | viktors|afk: i'll let you and zzzeek debate the benefits of an additional tox env for legacy mysql-python testing in https://review.openstack.org/184392 | 21:34 |
| haypo | fungi, i now need your patch for glance, my latest patch https://review.openstack.org/#/c/192904/ (to fix tox -e py34) requires https://review.openstack.org/#/c/184373/ (PyMySQL for Glance) | 21:34 |
| haypo | fungi, Sabari wrote "DB migration tests are getting skipped for MySQL database with this patch" | 21:34 |
| fungi | haypo: great! i mostly just pushed them with a mass-proposal script. sdague pointed out the same on the nova change | 21:35 |
| *** jungleboyj has joined #openstack-oslo | 21:35 | |
| * jungleboyj is having a bad computer day. | 21:35 | |
| fungi | haypo: i _think_ a new oslo.db release with 184392 in it will correct that | 21:35 |
| haypo | fungi, for nova, i picked ironic tox.ini which uses PyMySQL on python 3, and MySQL-python on Python 2 | 21:35 |
| fungi | haypo: but it's still a bit voodoo and black magic for me | 21:36 |
| haypo | fungi, the "until we drop MySQL-python support" part is unclear to me :-p | 21:36 |
| *** achanda has joined #openstack-oslo | 21:36 | |
| fungi | haypo: right, i assume that's "so soon that there's no point in a second env to keep testing it" | 21:36 |
| *** ashleighfarnham has quit IRC | 21:36 | |
| fungi | but it's not my call | 21:37 |
| haypo | fungi, did you try to add a separated venv to test mysql-python? | 21:37 |
| *** bknudson has quit IRC | 21:37 | |
| fungi | nope | 21:37 |
| *** yamahata has quit IRC | 21:37 | |
| jecarey | mtanino, note comment from dims above ... update 186533. | 21:37 |
| *** Rockyg has quit IRC | 21:38 | |
| haypo | fungi, i missed the session where pymysql was discussed. i'm allowed to vote +2, but i don't feel able to take a strong decision in favor of dropping mysql-python support | 21:38 |
| *** ashleighfarnham has joined #openstack-oslo | 21:39 | |
| haypo | fungi, i would feel safer if you put a new venv to continue to test mysql-python. it puts less pressure on having to make a choice | 21:39 |
| haypo | i guess that a new gate will be needed to run this new venv, but it's usually fast and easy to get it | 21:40 |
| fungi | haypo: the upshot of the session was "it should be quick to just rip it out and replace it so let's not go for a long parallel support and deprecation period" | 21:40 |
| haypo | fungi, it's ok for applications, not for the specific case of oslo.db | 21:40 |
| fungi | haypo: i'm happy for someone else to take it over if the consensus shifts that direction, but it's not something i have time to implement | 21:40 |
| *** ViswaV has quit IRC | 21:40 | |
| haypo | for the other venv, it's something like 4 lines in tox. i just don't remember the syntax for connectors | 21:41 |
| dhellmann | Rockyg: I'm going offline for the evening in a bit, so maybe tomorrow | 21:41 |
| fungi | also, time for me to go pick up groceries, but i'll let the interested parties (i'm not an especially interested party, just didn't want it to stall for lack of volunteers) hash out the particulars | 21:42 |
| dhellmann | fungi: +2 on the patch, and if we need another tox env that can come in another patch | 21:42 |
| *** achanda has quit IRC | 21:42 | |
| fungi | dhellmann: sounds good. hopefully once this makes it into an oslo.db release we can more safely move forward on the corresponding changes in leaf projects | 21:44 |
| haypo | dhellmann, do you know the syntax to get the mysql-python connector? is it just "mysql://..." ? (it's "mysql+pymysql://..." for PyMySQL) | 21:44 |
| haypo | dhellmann, i'm not sure that a different patch is a good choice, fungi removes mysql-python dep | 21:45 |
| fungi | haypo: well, the mysql-python dep would stay removed, you would just tweak the tox env definition to explicitly install it along with the requirements list files | 21:46 |
| fungi | so that it only gets installed if you're invoking the env which would test mysql-python | 21:46 |
| haypo | fungi, i would prefer to keep mysql-python in .txt, it's updated by global requirements | 21:47 |
| haypo | the requirements script is not smart enough (yet?) to handle tox | 21:47 |
| fungi | but i'll leave it to the oslo.db cores to debate how many mysql libraries they want to support | 21:47 |
| haypo | it's better to make baby step. switching from mysql-python to pymysql by default for mysql is already a giant step :) | 21:48 |
| fungi | it just sounds like the scope will quickly escape what we discussed at the summit if we're going to let anyone use whatever mysql lib they like | 21:48 |
| dhellmann | fungi: if we land it this week, we can release monday | 21:48 |
| fungi | anyway, <- afk for a while | 21:48 |
| mtanino | jecarey: dims I compared between 186533(incubator) and 190175(oslo.service). There is little bit difference between these at service.py. | 21:52 |
| *** stpierre has quit IRC | 21:52 | |
| haypo | dhellmann, i have a local patch which is fungi's patch + "tox -e mysql-python" which uses "OS_TEST_DBAPI_ADMIN_CONNECTION=mysql://...". my patch keeps test-requirements-py2/py3, it only adds PyMySQL to py2 | 21:52 |
| haypo | dhellmann, is it ok to reuse the same change id for such variant of his patch? | 21:53 |
| haypo | ah. tox -e mysql-python fails :-p | 21:53 |
| haypo | cool, only one failure | 21:53 |
| dhellmann | haypo: I don't think we want to support both libraries, but I'll defer to dims and the oslo.db leads. | 21:55 |
| dhellmann | (rpodolyaka and viktors|afk) | 21:56 |
| *** mriedem has quit IRC | 21:58 | |
| dims | +1 to ask the leads first | 21:58 |
| haypo | dhellmann, viktors|afk already replied: "are we really going to drop ANY MySQL-python support from openstack?" | 22:01 |
| haypo | he replied with a question... :) | 22:01 |
| haypo | dhellmann, zzzeek looks to support the patch. he's a reference in the db world, no? | 22:02 |
| zzzeek | what is the question | 22:02 |
| haypo | zzzeek, hey! would you be ok to support mysql-python support in oslo.db, at least in tests? | 22:03 |
| zzzeek | sure | 22:03 |
| dhellmann | haypo: why do you want this? | 22:03 |
| zzzeek | ive been telling ppl for a few days now, the provising system should likely be fixed to support multiple drivers | 22:04 |
| zzzeek | which right now can only be done with this very awkward hack i gave to neutron | 22:04 |
| haypo | zzzeek, do you see any reason why a specific application would prefer to keep mysql-python? | 22:04 |
| haypo | dhellmann, sorry, wanting what? | 22:04 |
| zzzeek | haypo: because they are hitting race conditions wiht pymysql. again, neutron | 22:04 |
| haypo | zzzeek, wait. i'm very tired and i'm no more able to use a computer... | 22:04 |
| haypo | zzzeek, my question was: would you be ok to *drop* mysql-python support | 22:05 |
| dhellmann | haypo: I'm trying to understand why you care about oslo.db having support for more than one mysql driver | 22:05 |
| zzzeek | haypo: fine by me | 22:05 |
| zzzeek | haypo: that’s what the patch does , i thought | 22:05 |
| haypo | dhellmann, *i* have no opinion. as i wrote, i don't feel able to take a decision :) | 22:05 |
| dhellmann | ok. I think we're all probably reaching the end of our day, so it might not be the best time to talk about it | 22:06 |
| haypo | zzzeek, i was checking if you wrote the patch or not, i read "Michael Bayer: Uploaded patch set 3." | 22:06 |
| dhellmann | in general, I would rather we be consistent with what we support so we see consistent behavior in consuming projects | 22:06 |
| zzzeek | haypo: i only fixed it to make the tests work | 22:06 |
| dhellmann | be limited in what we support, that is | 22:06 |
| zzzeek | haypo: id like oslo.db to be open towards multiple drivers in general, but that’s not critical at the moemnt | 22:07 |
| haypo | zzzeek, just to make sure that i understand: with the patch, mysql-python is no more *tested*, but it still works, right? | 22:07 |
| *** jgrimm is now known as zz_jgrimm | 22:07 | |
| zzzeek | haypo: sure | 22:07 |
| haypo | zzzeek, ok | 22:07 |
| dhellmann | if we're not testing it, we shouldn't claim to support it | 22:07 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:08 | |
| haypo | dhellmann, "why you care about oslo.db having support for more than one mysql driver" it looks like any minor change in oslo* breaks all openstack | 22:08 |
| haypo | dhellmann, that's why i'm looking for more information to understand the risk :) | 22:09 |
| *** jungleboyj has quit IRC | 22:09 | |
| dhellmann | haypo: ok, that's fair. I think it might be confusing to distros if we aren't clear about what we really want them to be using, though. | 22:10 |
| haypo | zzzeek, neutron already uses PyMySQL: fungi's patch was merged, https://review.openstack.org/#/c/184383/ | 22:10 |
| haypo | glance will use PyMySQL too: https://review.openstack.org/#/c/184373/ (already a +2) | 22:12 |
| haypo | as glance, nova is waiting for oslo.db to change to PyMySQL: https://review.openstack.org/#/c/184387/ | 22:12 |
| haypo | cinder also switched to PyMySQL | 22:13 |
| haypo | dhellmann, for me, there is no more reason against switching to PyMySQL | 22:13 |
| dhellmann | haypo: I can't tell which side of which argument you're on any more :-) | 22:14 |
| haypo | dhellmann, i'm in favor of simply approving the patch as it is | 22:14 |
| dhellmann | haypo: which patch? | 22:15 |
| haypo | dhellmann, https://review.openstack.org/#/c/184392/ fungi's patch to switch to PyMySQL by default | 22:15 |
| haypo | in oslo.db | 22:15 |
| haypo | i will unblock glance & nova to do the same change | 22:15 |
| dhellmann | haypo: ok. I recommend voting +2 now, and leaving the W+1 until we can coordinate with rpodolyaka and viktors|afk just to be safe. | 22:16 |
| *** prad has quit IRC | 22:23 | |
| haypo | dhellmann, i voted +2, tried to summarize the discussion and i pinged viktors|afk & rpodolyaka | 22:24 |
| haypo | good night ;) | 22:24 |
| *** haypo has quit IRC | 22:24 | |
| *** dims has quit IRC | 22:26 | |
| *** jecarey has quit IRC | 22:30 | |
| *** zzzeek has quit IRC | 22:44 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 22:55 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:56 | |
| *** ganso_ has quit IRC | 22:58 | |
| *** yamahata has joined #openstack-oslo | 23:15 | |
| *** dims has joined #openstack-oslo | 23:24 | |
| *** zzzeek has joined #openstack-oslo | 23:25 | |
| *** dims has quit IRC | 23:28 | |
| *** dims has joined #openstack-oslo | 23:28 | |
| *** dims has quit IRC | 23:33 | |
| *** claudiub has quit IRC | 23:34 | |
| *** yamahata has quit IRC | 23:40 | |
| *** mriedem has joined #openstack-oslo | 23:42 | |
| *** mriedem1 has joined #openstack-oslo | 23:45 | |
| *** mriedem has quit IRC | 23:47 | |
| *** sdake_ has joined #openstack-oslo | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!