| *** mriedem_away has quit IRC | 00:15 | |
| *** lei-zh has joined #openstack-placement | 02:06 | |
| *** lei-zh has quit IRC | 02:31 | |
| *** lei-zh has joined #openstack-placement | 02:32 | |
| *** takashin has left #openstack-placement | 03:03 | |
| *** tetsuro has joined #openstack-placement | 04:10 | |
| *** lei-zh has quit IRC | 04:10 | |
| *** tetsuro has quit IRC | 05:03 | |
| *** tetsuro has joined #openstack-placement | 05:03 | |
| *** tetsuro_ has joined #openstack-placement | 05:04 | |
| *** tetsuro has quit IRC | 05:08 | |
| *** lei-zh has joined #openstack-placement | 05:11 | |
| *** lei-zh has quit IRC | 05:18 | |
| *** tetsuro_ has quit IRC | 05:29 | |
| *** tetsuro has joined #openstack-placement | 05:34 | |
| *** tetsuro has quit IRC | 05:38 | |
| *** lei-zh has joined #openstack-placement | 05:42 | |
| *** tetsuro has joined #openstack-placement | 06:05 | |
| *** tetsuro has quit IRC | 06:09 | |
| *** tetsuro has joined #openstack-placement | 06:10 | |
| *** mnaser has quit IRC | 06:16 | |
| *** mnaser has joined #openstack-placement | 06:17 | |
| *** vdrok has quit IRC | 06:22 | |
| *** vdrok_ has joined #openstack-placement | 06:23 | |
| *** rubasov has quit IRC | 07:20 | |
| *** rubasov has joined #openstack-placement | 07:25 | |
| *** tetsuro has quit IRC | 07:50 | |
| *** helenafm has joined #openstack-placement | 08:19 | |
| *** e0ne has joined #openstack-placement | 08:41 | |
| *** tetsuro has joined #openstack-placement | 09:07 | |
| openstackgerrit | Takashi NATSUME proposed openstack/placement master: Add a link to "Add Generation to Consumers" spec https://review.openstack.org/614150 | 09:13 |
|---|---|---|
| *** lei-zh has quit IRC | 09:33 | |
| *** tetsuro has quit IRC | 09:39 | |
| *** cdent has joined #openstack-placement | 09:42 | |
| cdent | aw: how nice https://docs.openstack.org/placement/latest/ | 09:47 |
| *** tetsuro has joined #openstack-placement | 09:48 | |
| *** tetsuro has quit IRC | 09:51 | |
| *** ttsiouts has joined #openstack-placement | 10:22 | |
| *** e0ne has quit IRC | 10:32 | |
| openstackgerrit | wangqiang-bj proposed openstack/nova-specs master: fix wrong spelling of "unnecessary" https://review.openstack.org/614173 | 10:35 |
| openstackgerrit | wangqiang-bj proposed openstack/nova-specs master: fix wrong spelling of "unnecessary" https://review.openstack.org/614173 | 10:37 |
| *** cdent has quit IRC | 10:42 | |
| *** helenafm has quit IRC | 10:45 | |
| openstackgerrit | John Garbutt proposed openstack/nova-specs master: Add Unified Limits Spec https://review.openstack.org/602201 | 10:53 |
| *** e0ne has joined #openstack-placement | 10:57 | |
| openstackgerrit | wanghui proposed openstack/placement master: Fix the error package name in comment https://review.openstack.org/614184 | 10:59 |
| openstackgerrit | wanghui proposed openstack/nova-specs master: Fix a typo of 'keystoneauth1' https://review.openstack.org/614187 | 11:02 |
| *** cdent has joined #openstack-placement | 11:13 | |
| openstackgerrit | wangqiang-bj proposed openstack/placement master: fix wrong spelling of "explicit" https://review.openstack.org/614192 | 11:16 |
| jroll | edleafe: around now, I see you found your answer, anything else I can help with? | 11:24 |
| openstackgerrit | Merged openstack/placement master: Fix the error package name in comment https://review.openstack.org/614184 | 12:07 |
| *** ttsiouts has quit IRC | 12:16 | |
| *** tssurya has joined #openstack-placement | 12:20 | |
| *** e0ne has quit IRC | 12:21 | |
| openstackgerrit | Merged openstack/placement master: fix wrong spelling of "explicit" https://review.openstack.org/614192 | 12:28 |
| *** ttsiouts has joined #openstack-placement | 12:32 | |
| edleafe | jroll: just wondering how those version tests work in ironic. Wouldn't they always be skipped, since they're running as unit tests? | 12:36 |
| jroll | edleafe: we configure a database in CI, so they always run, and developers can do the same locally if they want. | 12:38 |
| edleafe | jroll: ah, I see. | 12:38 |
| cdent | edleafe: did that --autogenerate stuff make any sense and is it of any use? | 12:47 |
| cdent | the tests suggest I got it a bit wrong, but the concept is here | 12:48 |
| edleafe | cdent: just getting started on my day. What --autogenerate stuff are you referring to? | 12:49 |
| cdent | edleafe: I left comments on https://review.openstack.org/#/c/614024/ about how the described way is not as powerful as it could be and then demo that in: https://review.openstack.org/#/c/614030/1 | 12:50 |
| edleafe | cdent: ok, I'll take a look at it in a bit | 12:51 |
| *** helenafm has joined #openstack-placement | 13:06 | |
| *** cdent has quit IRC | 13:07 | |
| *** e0ne has joined #openstack-placement | 13:08 | |
| *** mriedem has joined #openstack-placement | 13:13 | |
| *** cdent has joined #openstack-placement | 13:27 | |
| efried | cdent: At a glance, the toc depth on the index page could be reduced a tad. IMO we don't need the microversion history expanded. | 13:30 |
| cdent | efried: yeah, agree | 13:31 |
| *** ttsiouts has quit IRC | 13:40 | |
| *** ttsiouts has joined #openstack-placement | 13:42 | |
| edleafe | cdent: ok, read your comments. I guess I'm not sure what --autogenerate has to do with it | 13:52 |
| cdent | one reason for using alembic at all is so you don't have to write migrations, alembic figures them out for you | 13:53 |
| cdent | you use alembic to create them | 13:53 |
| cdent | as described in the commit message of the demo | 13:53 |
| edleafe | Well, it's sort of a cart/horse situation | 13:53 |
| edleafe | You have to make the changes *somewhere*. If you do it in sqla, alembic *might* be able to autogenerate the migration | 13:54 |
| cdent | if you do it in any other way then in the models, then the migration is too complicated and we should reject it out of hand because migrations that can't be identfiied by alembic are clearly too complicated and not something we want | 13:55 |
| edleafe | After playing around with it a bit, and reading all the disclaimers, I don't find autogenerate to be that reliable | 13:55 |
| * cdent says tautologically | 13:55 | |
| edleafe | Not true - a rename isn't seen by alembic | 13:55 |
| cdent | I'm not suggesting that we use it as a general rule, but rather than it is an available too we might want to document | 13:55 |
| cdent | of a table or a column? | 13:55 |
| cdent | why would we ever do a rename? | 13:56 |
| cdent | mind you, I come from that strange alien planet that says that migrations should almost never happen | 13:56 |
| edleafe | https://alembic.zzzcomputing.com/en/latest/autogenerate.html#what-does-autogenerate-detect-and-what-does-it-not-detect | 13:56 |
| cdent | yeah, the list of things it can't detect are things that shouldn't be done :) | 13:56 |
| edleafe | I used autogenerate to create the initial migration, and you saw that it still needed tweaking | 13:57 |
| cdent | right, but it gave you a starting point | 13:57 |
| cdent | that's sort of the point | 13:57 |
| cdent | it's a tool | 13:57 |
| cdent | tools are nice | 13:57 |
| edleafe | And we shouldn't be doing migrations that are so complex that the tool saves you anything | 13:57 |
| cdent | except you can save a step: you have to make the changes in the models anyway | 13:58 |
| cdent | alembic can detect anything sane in those | 13:58 |
| edleafe | In the case of creating an entire database, sure. In the case of adding an index or a column - nope | 13:58 |
| cdent | "Column additions, removals." yes "Basic changes in indexes and explcitly-named unique constraints" yes | 13:58 |
| cdent | Like I said: we don't have to indicate any of this _yet_, but it would be useful for two things to be true: | 13:59 |
| edleafe | Look, the end result is a revision file, no? How it is created is up to the person making that change. | 13:59 |
| cdent | a) the online migration part of env.py works so that people in the know can use the alembic commands to their fullest | 14:00 |
| cdent | b) we eventually document that availability | 14:00 |
| cdent | it sounds like we are in agreement, so I'm not sure why "Look," is necessary here | 14:00 |
| cdent | in your changes a) was not true, last I checked | 14:01 |
| edleafe | Because I don't feel confident enough in autogenerate to encourage its use | 14:01 |
| cdent | which is why I wrote the demo code, to help you gain some confidence | 14:01 |
| edleafe | didn't work :) | 14:02 |
| cdent | I don't care if you document autogenerate | 14:02 |
| cdent | but I do care that you fix a) | 14:02 |
| edleafe | Sure, I definitely want to fix a). But it goes back to my original statement: I don't see what --autogenerate has to do with it | 14:03 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add a placement-manage CLI https://review.openstack.org/600161 | 14:04 |
| cdent | you can't run --autogenerate without a working | 14:04 |
| cdent | so one way to test that env.py is correct, while experimenting, is to use revision --autogenerate from the command line | 14:04 |
| cdent | as demostrated in my DNM | 14:04 |
| cdent | if you do 'alembic revision' without autogenerate it doesn't tickly the a)-ness | 14:05 |
| cdent | does that make more sense? | 14:06 |
| edleafe | No, still sounds the same to me as creating the revision manually. Especially with the "verify and tweak the resulting file" | 14:08 |
| edleafe | Saves a little typing, for sure | 14:08 |
| cdent | okay: | 14:08 |
| cdent | that's not my point | 14:08 |
| cdent | let me try to explain in a different way: | 14:08 |
| cdent | when you run 'alembic revision' with no args it doesn't need to consult the database, therefore the code in env.py in the method 'run_migrations_online' is not run | 14:09 |
| cdent | when you run 'alembic revision --autogenerate' it does need access to an existing database | 14:09 |
| cdent | therefore 'run_migrations_online' is called and as the code is currently written it fails | 14:10 |
| cdent | thus: | 14:10 |
| cdent | the command 'alembic revision --autogenerate' is a tool for doing you to do TDD to make that code more correct while trying to resolve a) | 14:10 |
| cdent | s/doing/allowing/ | 14:10 |
| cdent | or to put it another way: | 14:11 |
| cdent | I figure out that env.py was incomplete because I tried to use 'alembic revision --autogenerate' | 14:11 |
| cdent | we want env.py to be right | 14:11 |
| cdent | the goal is env.py to be right | 14:11 |
| cdent | EOF | 14:12 |
| edleafe | Yay! Something I can agree with!! | 14:12 |
| edleafe | :) | 14:12 |
| edleafe | I get that autogenerate is helpful in figuring out how to correct env.py. | 14:13 |
| cdent | I'm not sure if I should be relieved yet or not | 14:13 |
| edleafe | I just don't see it being tremendously useful afterwards | 14:13 |
| cdent | that's fine. we'll let the future decide that for us. | 14:14 |
| cdent | biab | 14:20 |
| *** cdent has quit IRC | 14:21 | |
| *** e0ne has quit IRC | 14:33 | |
| *** e0ne has joined #openstack-placement | 14:37 | |
| openstackgerrit | Jack Ding proposed openstack/nova-specs master: High Precision Event Timer (HPET) on x86 guests https://review.openstack.org/607989 | 14:42 |
| *** cdent has joined #openstack-placement | 15:08 | |
| openstackgerrit | Theodoros Tsioutsias proposed openstack/nova-specs master: Enable rebuild for instances in cell0 https://review.openstack.org/554218 | 15:19 |
| *** e0ne has quit IRC | 15:25 | |
| *** e0ne has joined #openstack-placement | 15:35 | |
| openstackgerrit | Merged openstack/placement master: Add a link to "Add Generation to Consumers" spec https://review.openstack.org/614150 | 16:29 |
| *** e0ne has quit IRC | 16:30 | |
| sean-k-mooney | cdent: im going to have to respin the nova change again because we have flak8 set for a 79 charater limit instead of 80.. | 16:31 |
| sean-k-mooney | sorry of the noise | 16:31 |
| cdent | sean-k-mooney: no problem | 16:31 |
| sean-k-mooney | am i have almost finished the placment port too. | 16:31 |
| sean-k-mooney | however i notice a circualr depenancy if i add the decorator in util.py | 16:32 |
| sean-k-mooney | http://paste.openstack.org/show/733655/ | 16:32 |
| sean-k-mooney | i can move the definition before the import and leave a comment but is there somewhere else i should add it instead | 16:33 |
| cdent | sean-k-mooney: the right fix for that is to remove the db using methods somewhere else | 16:33 |
| cdent | I'll do that now, and you can base your changes on that | 16:34 |
| sean-k-mooney | ok ill push up what i have and then ill rebase on your change to keep a copy of it | 16:34 |
| sean-k-mooney | i noticed you comment regarding cyclical imports so was not sure if you planned to fix it or were we just living with it | 16:35 |
| cdent | there are some which are sort okay, but the db one I don't like | 16:37 |
| sean-k-mooney | cdent: ya i was honestly surprised to see the placment.object and placement.microversion imports. generall util files are imported everhwere and rarely import thing form the current project to avoid this | 16:43 |
| cdent | sean-k-mooney: much of that was the result of haste | 16:43 |
| sean-k-mooney | cdent: with the split out. ya its not a big deal its more of a code smell | 16:44 |
| sean-k-mooney | it works but it has maitaince issues like this | 16:44 |
| cdent | yes | 16:44 |
| cdent | the haste I'm talking about is not the haste of extraction, the haste of the original creation of placement | 16:44 |
| edleafe | cdent: I just tried playing with autogenerate again, this time with the changes you made to env.py. When I run the revision command with autogenerate, I get "ImportError: No module named placement" | 16:45 |
| cdent | edleafe: you need to run the alembic command from within the virtenv .tox/py??/bin/alembic | 16:45 |
| cdent | if you already did that, then not sure | 16:45 |
| edleafe | oh yeah - I was just doing from the straight python env | 16:46 |
| edleafe | I'm adding an autogen section to the doc, and wanted to test it out first | 16:46 |
| edleafe | Because I wanted to make you say 'huzzah!' | 16:47 |
| cdent | I'll save it up and use two ! | 16:47 |
| openstackgerrit | sean mooney proposed openstack/placement master: Harden placement init under wsgi https://review.openstack.org/612444 | 16:50 |
| sean-k-mooney | ille rebase ^ when the cycle is broken of you can if you like. it should just work form the gerrit ui hopefully | 16:51 |
| *** helenafm has quit IRC | 17:05 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Move ensure_consumer to a new placement.db.util https://review.openstack.org/614307 | 17:06 |
| cdent | sean-k-mooney: ^ that might do the trick | 17:07 |
| sean-k-mooney | cool im just going to grab something to eat. ill pull it down and rebase locally when i get backin an hour or os | 17:09 |
| *** tssurya has quit IRC | 17:19 | |
| *** ttsiouts has quit IRC | 17:28 | |
| *** ttsiouts has joined #openstack-placement | 17:29 | |
| *** ttsiouts has quit IRC | 17:33 | |
| *** e0ne has joined #openstack-placement | 18:10 | |
| *** e0ne has quit IRC | 18:25 | |
| openstackgerrit | sean mooney proposed openstack/placement master: Harden placement init under wsgi https://review.openstack.org/612444 | 18:37 |
| sean-k-mooney | cdent: edleafe have either of ye had issues running placement.tests.functional.db.test_resource_provider.ResourceClassTestCase.test_create_duplicate_id_retry_failing | 18:39 |
| sean-k-mooney | even on master that functional test is failing for me locally | 18:40 |
| cdent | sean-k-mooney: i haven't seen that. what kind of error is it giving you? | 18:40 |
| sean-k-mooney | i have two failing functional tests the relevent bits are here http://paste.openstack.org/show/733661/ | 18:41 |
| * cdent tries | 18:42 | |
| sean-k-mooney | actully you might need the full output http://paste.openstack.org/show/733662/ | 18:42 |
| cdent | I haven't run the functional tests locally with 2.7 in a while | 18:42 |
| sean-k-mooney | oh i can try them with 3.6 too | 18:42 |
| sean-k-mooney | these are passing upstream so im assuming its a centos thing or somthing is not mocked correctly | 18:42 |
| edleafe | I'm running them locally on 2.7 and they are passing | 18:43 |
| sean-k-mooney | ya its weird im assuming its something to do with my env but not sure what | 18:44 |
| sean-k-mooney | ill try recloning placement once teh 3.6 version finishes | 18:44 |
| cdent | sean-k-mooney: are your tox env's recently -r ? | 18:44 |
| sean-k-mooney | well the py36 functional test where a first run and yes 5 minutes ago to try and fix the issue :) | 18:45 |
| cdent | I assumed as much, but figured better check everything | 18:46 |
| sean-k-mooney | just did a fresh clone and running tox again | 18:46 |
| cdent | what version of sqlite have you got? | 18:47 |
| sean-k-mooney | good question i assume thats a bindep so its not using one installed in the tox env? | 18:48 |
| sean-k-mooney | ill check | 18:48 |
| sean-k-mooney | [stack@devstack2 placement-test]$ sqlite3 --version | 18:48 |
| sean-k-mooney | 3.7.17 2013-05-20 00:56:22 118a3b35693b134d56ebd780123b7fd6f1497668 | 18:48 |
| sean-k-mooney | 2013 damb centos packages are old | 18:48 |
| cdent | I've got 3.22 | 18:49 |
| sean-k-mooney | so ya same behavior with a clean clone. i can try and install a newer sqlite3 package if you think that will help | 18:49 |
| cdent | could be that | 18:49 |
| cdent | it's the only thing I can think of | 18:50 |
| sean-k-mooney | ill grab the latest for upstream one sec | 18:50 |
| cdent | but not sure why it would be a problem | 18:50 |
| *** e0ne has joined #openstack-placement | 19:13 | |
| *** cdent has quit IRC | 19:56 | |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Added alembic environment https://review.openstack.org/611441 | 20:51 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Delete the old migrations https://review.openstack.org/611440 | 20:51 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Add a document for creating DB revisions https://review.openstack.org/614024 | 20:51 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: WIP - Show an alembic migration https://review.openstack.org/614025 | 20:51 |
| openstackgerrit | Ed Leafe proposed openstack/placement master: Rename the database files https://review.openstack.org/614350 | 20:51 |
| *** tssurya has joined #openstack-placement | 20:58 | |
| *** e0ne has quit IRC | 21:18 | |
| *** mriedem has quit IRC | 21:30 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Add a placement-manage CLI https://review.openstack.org/600161 | 21:32 |
| *** ttsiouts has joined #openstack-placement | 21:55 | |
| *** ttsiouts has quit IRC | 22:04 | |
| *** ttsiouts has joined #openstack-placement | 22:05 | |
| *** ttsiouts has quit IRC | 22:10 | |
| *** e0ne has joined #openstack-placement | 22:49 | |
| *** tssurya has quit IRC | 22:51 | |
| *** e0ne has quit IRC | 23:13 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!