| *** odyssey4me has quit IRC | 00:14 | |
| *** odyssey4me has joined #openstack-requirements | 00:14 | |
| *** hongbin has joined #openstack-requirements | 01:05 | |
| prometheanfire | dhellmann: false positive in lower-constraints check? http://logs.openstack.org/17/558217/5/check/requirements-check/a380698/job-output.txt.gz#_2018-04-02_21_30_50_271245 | 02:18 |
|---|---|---|
| * prometheanfire knows it's late | 02:18 | |
| *** udesale has joined #openstack-requirements | 03:53 | |
| *** hongbin has quit IRC | 03:56 | |
| *** andreas_s has joined #openstack-requirements | 04:39 | |
| *** andreas_s has quit IRC | 04:43 | |
| *** snapiri has joined #openstack-requirements | 05:37 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/558392 | 06:23 |
| *** florianf has joined #openstack-requirements | 06:42 | |
| *** andreas_s has joined #openstack-requirements | 07:39 | |
| *** andreas_s has quit IRC | 07:44 | |
| *** jpich has joined #openstack-requirements | 08:06 | |
| *** andreas_s has joined #openstack-requirements | 08:20 | |
| *** ralonsoh has joined #openstack-requirements | 08:29 | |
| *** florianf has quit IRC | 10:19 | |
| *** florianf has joined #openstack-requirements | 10:21 | |
| *** florianf has quit IRC | 10:54 | |
| adam__ | What is the rule of thumb for sizing your control nodes? Particularly for memory? Someone gave me something that just seems way too high of requirements. | 11:00 |
| *** florianf has joined #openstack-requirements | 11:12 | |
| *** adam__ has quit IRC | 11:13 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for oslotest to new release 3.4.1 https://review.openstack.org/558478 | 11:19 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for pbr to new release 4.0.1 https://review.openstack.org/558479 | 11:21 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for oslo.config to new release 6.0.1 https://review.openstack.org/558480 | 11:23 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for oslo.cache to new release 1.30.0 https://review.openstack.org/558481 | 11:24 |
| *** ralonsoh_ has joined #openstack-requirements | 12:01 | |
| *** ralonsoh has quit IRC | 12:04 | |
| *** odyssey4me has quit IRC | 12:15 | |
| *** odyssey4me has joined #openstack-requirements | 12:15 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/requirements master: Add OpenStack-Ansible projects to list of projects https://review.openstack.org/558492 | 12:18 |
| *** edmondsw has quit IRC | 12:28 | |
| *** edmondsw has joined #openstack-requirements | 12:28 | |
| *** edmondsw has quit IRC | 12:32 | |
| *** udesale has quit IRC | 12:50 | |
| *** lbragstad has joined #openstack-requirements | 12:50 | |
| *** udesale has joined #openstack-requirements | 12:50 | |
| *** edmondsw has joined #openstack-requirements | 12:58 | |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: remove debug prints from lower constraint checks https://review.openstack.org/558501 | 13:24 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: pick requirements directory based on operating mode https://review.openstack.org/558504 | 13:30 |
| dhellmann | prometheanfire : I was out yesterday. I'm looking at issues today. | 13:35 |
| *** snapiri has quit IRC | 13:46 | |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: remove dead code https://review.openstack.org/558513 | 13:58 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: show which property does not match exactly https://review.openstack.org/558514 | 13:58 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: avoid testing exact matches twice https://review.openstack.org/558515 | 13:58 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: report all attributes that do not match exactly https://review.openstack.org/558516 | 13:58 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: stop requiring comments to match global-requirements.txt https://review.openstack.org/558517 | 13:58 |
| dhellmann | prometheanfire : the *comment* part of the line doesn't match, which I exposed with some more debugging stuff in the series I just posted for review. I also proposed a patch to skip checking the comments. | 13:59 |
| prometheanfire | I'm more in favor of skipping comments for this | 14:02 |
| *** snapiri has joined #openstack-requirements | 14:10 | |
| *** ralonsoh_ has quit IRC | 14:11 | |
| *** vpickard has joined #openstack-requirements | 14:27 | |
| *** cjloader has joined #openstack-requirements | 14:45 | |
| *** andreas_s has quit IRC | 15:29 | |
| *** andreas_s has joined #openstack-requirements | 15:34 | |
| *** andreas_s has quit IRC | 15:38 | |
| openstackgerrit | Dirk Mueller proposed openstack/requirements master: Remove XStatic-Magic-Search https://review.openstack.org/558544 | 15:55 |
| *** udesale has quit IRC | 16:12 | |
| *** jpich has quit IRC | 16:33 | |
| openstackgerrit | Merged openstack/requirements master: remove debug prints from lower constraint checks https://review.openstack.org/558501 | 16:47 |
| openstackgerrit | Merged openstack/requirements master: update constraint for pbr to new release 4.0.0 https://review.openstack.org/557009 | 17:02 |
| prometheanfire | heh, pbr merged just in time for a new pbr | 17:22 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: skip complex comparison for constraints without markers https://review.openstack.org/558566 | 17:23 |
| *** mugsie_ has joined #openstack-requirements | 18:41 | |
| *** EmilienM has quit IRC | 18:47 | |
| *** mordred has quit IRC | 18:47 | |
| *** masayukig[m] has quit IRC | 18:47 | |
| *** mugsie has quit IRC | 18:47 | |
| *** EmilienM_ has joined #openstack-requirements | 18:47 | |
| *** EmilienM_ is now known as EmilienM | 18:48 | |
| *** EmilienM has quit IRC | 18:49 | |
| *** EmilienM has joined #openstack-requirements | 18:49 | |
| smcginnis | Is global-requirements.txt not being updated anyore? | 18:50 |
| prometheanfire | it is, but individual projects can fall outside it now | 18:50 |
| prometheanfire | just because project a needs one version of a lib doesn't mean project b needs to update it | 18:51 |
| smcginnis | OK, some of these lower-constraint patches are setting packages higher than what is in global-requirement.txt. | 18:52 |
| *** masayukig[m] has joined #openstack-requirements | 18:53 | |
| prometheanfire | I don't think that should be good... | 18:53 |
| prometheanfire | dhellmann: ^ | 18:53 |
| dhellmann | that's why I wanted to remove the minimums from global-requirements.txt | 18:54 |
| *** mordred has joined #openstack-requirements | 18:54 | |
| dhellmann | if we're going to let projects diverge, I don't see why some can't claim they need a newer version as long as the upper-constraints.txt entry is at least that high | 18:54 |
| smcginnis | Which is fine. Once projects have moved over to the new setup. | 18:55 |
| dhellmann | if they don't *need* those newer versions, then we should lower the constraints | 18:55 |
| smcginnis | It appears to just be causing problems right now with trying to do that. | 18:55 |
| dhellmann | I noticed several that were like that, especially with sphinx | 18:56 |
| smcginnis | dhellmann: The problems I'm seeing is the project's lower-constraint file has higher versions than the openstack/requirenents lower-constraints.txt and global-requirements.txt files. | 18:56 |
| smcginnis | And I checked history and it isn't a matter of something changing after the patch was submitted. | 18:56 |
| dhellmann | ok | 18:56 |
| dhellmann | so that theory is short | 18:56 |
| dhellmann | shot | 18:56 |
| smcginnis | How were these lower-constraint files generated for the per-repo patches? | 18:57 |
| dhellmann | the lower-constraints.txt file in the requirements repo isn't involved in any checks, and shouldn't be -- if we constrain everything to the same lower bound then we can't diverge | 18:57 |
| prometheanfire | ya, that's not the intention | 18:57 |
| dhellmann | I used that global list that dirk prepared and took the relevant entries for each repo out to create the initial patches | 18:57 |
| smcginnis | Well, I think we should start everything the same, then once that's level set, then allow divergence. | 18:57 |
| smcginnis | Ah, OK. | 18:58 |
| smcginnis | It's just going to take some manual clean up to get at least the one I'm looking at to pass. | 18:58 |
| dhellmann | yeah | 18:58 |
| dhellmann | how far off are they? | 18:59 |
| smcginnis | It's a "not insignificant" list: http://logs.openstack.org/83/555483/2/check/requirements-check/aa12794/job-output.txt.gz#_2018-04-03_02_14_19_042756 | 18:59 |
| dhellmann | hmm the requirements.txt entry for glance-store is lower than the constraint | 19:00 |
| dhellmann | I wonder if that means glance didn't land some g-r update? | 19:00 |
| dhellmann | or if the original constraint list was just wrong | 19:00 |
| smcginnis | g-r matches it fine: https://github.com/openstack/requirements/blob/master/global-requirements.txt#L69 | 19:01 |
| dhellmann | ok, so the original list of constraints was off | 19:01 |
| smcginnis | Yeah, it appears so. | 19:01 |
| dhellmann | and we didn't notice that in some projects that landed their patch with the new job before we had the new tests in place | 19:01 |
| dhellmann | sigh | 19:01 |
| dhellmann | I guess the question is do we want to change the requirement or the constraint list? | 19:02 |
| dhellmann | it's probably more accurate to update the constraint list | 19:02 |
| smcginnis | So should I make the g-r match the lower-constraints, or raise their requirements.txt to match what we're now declaring as lower? Yeah... | 19:02 |
| dhellmann | let me see if I can script the constraint file update | 19:02 |
| smcginnis | I've started going through and updating requirements.txt. But just a little concerned that their requirements.txt will not match g-r for now. | 19:03 |
| dhellmann | that no longer matters | 19:03 |
| dhellmann | that's the point of all of this | 19:03 |
| smcginnis | But until we've moved it all over, it may be a point of confusion. | 19:04 |
| prometheanfire | we need to turn gr.txt into a pull method instead of the push method that it was | 19:04 |
| dhellmann | the sync job is turned off and the new check rules are in place. what more is there to being moved over? | 19:04 |
| dhellmann | prometheanfire : if we pull changes into that list, it's going to set things up so one repo can break another | 19:05 |
| prometheanfire | dhellmann: 'that list'? | 19:05 |
| smcginnis | Well, all projects updated to be using lower-constraints for one. :) | 19:05 |
| dhellmann | prometheanfire : g-r | 19:05 |
| prometheanfire | dhellmann: how so? | 19:05 |
| dhellmann | smcginnis : actually testing the lower constraints is orthogonal to allowing them to be different | 19:05 |
| dhellmann | prometheanfire : suppose we have a dependency in g-r and projects A and B with version 1.0 | 19:06 |
| smcginnis | Except when them being different prevents the project from testing the lower constraints. | 19:06 |
| dhellmann | then project A updates its dependency to 1.2 and updates its lower constraint at the same time | 19:06 |
| dhellmann | now B wants to change some other dependency | 19:06 |
| dhellmann | but it can't change its file because the g-r no longer matches -- hmm, well, that only matters if the thing we're testing is an exclusion, not a minimum | 19:07 |
| dhellmann | so ignore changing a min, and consider the same steps for adding an exclusion | 19:07 |
| dhellmann | smcginnis : the values need to match *within the repo* they shouldn't need to match g-r except for the exclusion list | 19:07 |
| prometheanfire | dhellmann: the main change we'd have to do is allow exclusions out of the >= range | 19:07 |
| dhellmann | smcginnis : and exclusions just need to be a subset | 19:08 |
| dhellmann | prometheanfire : why so? | 19:08 |
| prometheanfire | do we still need to track global exclusions? | 19:09 |
| dhellmann | we do for != but not for >= | 19:09 |
| prometheanfire | if we also are tracking the highest global >= (which we basically do because our project cross gates with others), then if we update foo from >=1.0.0 to >=2.0.0, we loose the exclusion tracking for !=1.1.0,!=1.2.0 etc | 19:11 |
| dhellmann | that restriction is why we cannot track >= values in the g-r list any more | 19:11 |
| prometheanfire | so projects that still are on >=1.0.0 (because per-project reqs), will not have inherit those exclusions | 19:11 |
| prometheanfire | that's one solution | 19:11 |
| dhellmann | if we need a "highest global" value we need to compute it from some set of individual requirement lists | 19:12 |
| prometheanfire | I think we need to figure out if we need a highest global, if not then we remove the min, it was only there for lc generation iirc | 19:13 |
| prometheanfire | as long as the highest global is able to be generated I think we are good | 19:13 |
| dhellmann | yeah, we can compute the highest global by scanning a bunch of input files | 19:15 |
| prometheanfire | yep | 19:15 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: script to fix lower constraints entries that do not match local requirements https://review.openstack.org/558610 | 19:38 |
| dhellmann | smcginnis : I used https://review.openstack.org/558610 to update the glance patch in https://review.openstack.org/555483 | 19:38 |
| dhellmann | smcginnis : oh, sorry, it looks like I stomped on the version of that patch you proposed :-/ | 19:39 |
| smcginnis | Err, or I stomped on yours. | 19:40 |
| smcginnis | dhellmann: Yours has a lot more changes than mine. | 19:41 |
| smcginnis | dhellmann: Mind resubmitting it? | 19:41 |
| dhellmann | hmm, git review doesn't want to do that. let me see if I can make it | 19:41 |
| dhellmann | some of the changes may just be sorting | 19:44 |
| dhellmann | but some are versions | 19:44 |
| *** cjloader has quit IRC | 19:52 | |
| *** cjloader_ has joined #openstack-requirements | 19:52 | |
| *** cjloader_ has quit IRC | 19:53 | |
| *** cjloader has joined #openstack-requirements | 19:54 | |
| openstackgerrit | Doug Hellmann proposed openstack/requirements master: script to fix lower constraints entries that do not match local requirements https://review.openstack.org/558610 | 19:57 |
| *** cjloader has quit IRC | 19:58 | |
| *** cjloader has joined #openstack-requirements | 20:05 | |
| *** florianf has quit IRC | 20:13 | |
| smcginnis | dhellmann: Not bad, down to just one - http://logs.openstack.org/83/555483/7/check/requirements-check/593f021/job-output.txt.gz#_2018-04-03_20_00_13_634253 | 20:29 |
| *** cjloader has quit IRC | 20:41 | |
| *** cjloader has joined #openstack-requirements | 20:42 | |
| *** cjloader has quit IRC | 20:43 | |
| *** cjloader has joined #openstack-requirements | 20:43 | |
| *** cjloader has quit IRC | 20:44 | |
| *** cjloader has joined #openstack-requirements | 20:45 | |
| *** vpickard is now known as vpickard_ | 20:58 | |
| *** edmondsw has quit IRC | 21:09 | |
| *** edmondsw has joined #openstack-requirements | 21:09 | |
| *** edmondsw has quit IRC | 21:14 | |
| *** cjloader has quit IRC | 21:30 | |
| openstackgerrit | Merged openstack/requirements master: Updating minimum version of os-brick https://review.openstack.org/558221 | 21:31 |
| openstackgerrit | Merged openstack/requirements master: Add OpenStack-Ansible projects to list of projects https://review.openstack.org/558492 | 21:43 |
| dhellmann | smcginnis : weird | 22:01 |
| dhellmann | smcginnis : those are the same in my local repo | 22:01 |
| dhellmann | we should abandon all of those bot proposed updates | 22:03 |
| dhellmann | updated https://review.openstack.org/555483 | 22:03 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for python-neutronclient to new release 6.8.0 https://review.openstack.org/558644 | 22:23 |
| *** lbragstad has quit IRC | 22:36 | |
| *** lbragstad has joined #openstack-requirements | 22:53 | |
| *** cjloader has joined #openstack-requirements | 23:15 | |
| *** cjloader has quit IRC | 23:21 | |
| *** cjloader has joined #openstack-requirements | 23:22 | |
| *** cjloader has quit IRC | 23:26 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!