13:00:49 #startmeeting rpm_packaging 13:00:50 Meeting started Thu Nov 24 13:00:49 2016 UTC and is due to finish in 60 minutes. The chair is number80. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:55 The meeting name has been set to 'rpm_packaging' 13:01:00 glad 2 c u 13:01:00 #topic roll call 13:01:06 o/ 13:01:07 ping toabctl, dirk, apevec, aplanas, IgorYozhikov, jpena, jruzicka, number80, kaslcrof 13:01:12 #chair jpena IgorYozhikov toabctl 13:01:13 Current chairs: IgorYozhikov jpena number80 toabctl 13:02:42 o/ 13:03:24 hi 13:03:27 #chair dirk 13:03:28 Current chairs: IgorYozhikov dirk jpena number80 toabctl 13:03:32 let's start then 13:03:44 #topic keystone and mistral PRs 13:03:53 https://review.openstack.org/#/c/382196/ 13:04:00 https://review.openstack.org/#/c/396262/ 13:04:52 yes, just want to understand what should be done next to unblock dev core services 13:05:02 keystone fails in Suse because of deps that needs to be updated 13:05:07 *CI 13:05:14 I know 13:05:49 and about mistral dirk was promised to look into tests failures 13:06:01 okie 13:06:04 is there any progress? 13:06:19 or I can disable tests with bcond? 13:06:24 with_tests? 13:06:58 here we need feedback from SUSE side 13:07:18 IgorYozhikov: still wip 13:07:28 let me take a look now while the meeting is progressing 13:07:52 I have a look at the missing build deps now 13:08:02 * d0ugal wonders if he can help with Mistral test failures 13:08:03 dirk, that's great and do you have something to say about deps 4 keystone? 13:08:22 ah, looks like i'm late :) 13:08:43 )IgorYozhikov I'll fix that 13:08:47 ah, or toabctl 13:09:00 IgorYozhikov: we don't have most of the deps that you're referencing there.. 13:09:42 IgorYozhikov: e.g. python-hacking should not be in there 13:10:13 I think that we need to decide - build deps on vendors side and rebuild packages || add with_tests -> disable check and move further 13:10:26 thats independent 13:10:39 python-hacking is not a valid test dep, it shouldn#t be there 13:10:48 others like python-future and so on are just omissins in the ci repo 13:10:53 toabctl will fix that in a min 13:11:03 As I mentioned it used in tests 13:11:27 I tried to remove hacking and tests failed 13:14:38 so, dirk is watching into CI logs & elaborating next steps. Let's move further now? 13:14:54 try run_test.sh -P so it won't run flake8 13:15:21 it should pass 13:16:19 number80, are you suggesting to change %check section 4 keystone? 13:16:26 yes 13:17:03 %{__python2} setup.py testr -> run_test.sh -P ? 13:17:37 yes, let's see if it fixes the issue 13:18:36 number80, could you point me where is this file ? 13:18:49 IgorYozhikov: so looked at mistral.. the tests are doing a curl on http://www.example.com/ 13:18:56 which fails, because there is no network in a SUSE apckage build 13:19:05 no network -> tests fail 13:19:09 imho the test is broken 13:19:34 yes, we should have no network available 13:19:53 may be workaround? like local dummy http server? 13:20:22 fixing the test would be good 13:20:22 or there is a way of how exclude this test? 13:21:08 dirk: Would you like me to have a go at updating the test? 13:21:20 sure 13:21:29 there is one failure, it is mistral.tests.unit.engine.test_with_items.WithItemsEngineTest.test_with_items_and_adhoc_action 13:22:02 nice 13:22:02 oh, cool. I know that file. 13:22:28 #action d0ugal is going to fix mistral test mistral.tests.unit.engine.test_with_items.WithItemsEngineTest.test_with_items_and_adhoc_action 13:22:36 d0ugal: actually it is just a dns lookup it seems. there is also no dns 13:23:28 dirk: cool, should be fairly easy to resolve. 13:23:44 ok, seems like one problem less 13:24:29 and about runt_test.sh -P - I can't find this file here - https://github.com/openstack/keystone/tree/10.0.0 13:24:57 number80, could you please tell me where did you find it? 13:26:02 https://github.com/openstack/mistral/blob/master/run_tests.sh 13:26:15 it is master :( 13:26:20 ah yes 13:26:26 ah it is mistral 13:26:55 sorry for inconvenience 13:28:00 anyway it is exists in 3.0.2 https://github.com/openstack/mistral/blob/3.0.2/run_tests.sh 13:29:29 ah right 13:29:38 ok, so, should I try to change %check in mistral to use run_test.sh? 13:31:36 updated 13:31:41 https://review.openstack.org/#/c/382196/ 13:31:57 ack, for mistral after reading logs, I think we just need to exclude the failing tests 13:32:30 ok, will examine build logs 13:32:48 and about keystone 13:33:00 so here there are unmet deps right? 13:34:28 toabctl: is going to fix that momuntarily 13:34:33 lets move on 13:34:36 IgorYozhikov: ^^ 13:34:45 it is fixed. but not sure about that secxml lib thing 13:34:45 dirk, thanx :) 13:34:53 see coments in the review 13:34:55 it is also required 13:35:07 because it used in tests 13:35:19 IgorYozhikov, bindep has libxml2 mentioned 13:35:20 I add it according to tests logs 13:35:21 oh. 13:36:46 number80, https://packaging-ci.fuel-infra.org/job/master-rpm-packaging-build-centos7/1630/console 13:37:00 run_tests wants venv 13:37:47 IgorYozhikov: add -N 13:38:33 yep 13:38:39 done 13:38:47 I suggest that we look to the next topics? 13:38:54 toabctl, I will remove http || apache2 13:38:58 yes 13:39:09 let's move further 13:40:31 #topic XStatic 13:40:50 Here I want to clarify 2 things 13:41:26 1) should we package all xstatics or just required by horizon 13:42:14 2) should we build them as is - ie with embedded js,css,etc 13:43:28 As I remember Alan mentioned about 1 part a couple days ago 13:43:30 1) all required by horizon and that are on git.openstack.org/ 13:43:35 imho 13:43:39 actually, shouldn't they be treated like other dependencies? I'm not sure why we should build them 13:43:52 jpena: they're hosted on git.openstack.org 13:44:27 jpena: so its imho similar to other deps, like castellan, cliff.. 13:44:34 dirk: ack 13:45:52 ok, I already asked Andrii to start upload xstatics to master 13:46:02 ok 13:46:13 not sure if he is here now 13:46:34 I'll ask him to participate our meetings 13:46:45 in our meetings 13:46:47 :) 13:47:35 moving further? 13:48:59 +1 13:49:08 number80: can you ping apevec about the -1 ? 13:49:14 he was not so happy about xstatic 13:49:48 dirk, may be we set #agreed for xstatics just for history 13:50:38 it will be easier to track decisions 13:50:46 dirk: ack, I'm myself mixed about it, but it's logical to package whatever we have in git.openstack.org 13:51:00 #agreed package XStatic in rpm-packaging 13:52:17 next topic? 13:53:38 we have 7 mins to go 13:54:41 yes, go ahead 13:55:08 #topic - packages reviews (https://review.openstack.org/#/q/project:openstack/rpm-packaging+status:open ) 13:55:27 yes 13:55:39 any controversies? 13:56:01 covered already 13:56:08 ok 13:56:11 next topic 13:56:28 #topic pymod2pkg reviews (https://review.openstack.org/#/q/project:openstack/pymod2pkg+status:open ) 13:56:54 https://review.openstack.org/#/c/401214/4/pymod2pkg/__init__.py 13:57:03 I fixed the XStatic review, it should cover all cases now I think 13:57:15 both opened reviews are ready for merge 13:57:55 should we publish new version after merge of these 2 prs? 13:58:07 dirk: This may solve the Mistral test for you: https://review.openstack.org/#/c/402057/ 13:59:19 d0ugal: nice, that was quick 13:59:19 dirk, toabctl if you are fine with these 2 PRs, please merge https://review.openstack.org/#/q/project:openstack/pymod2pkg+status:open 13:59:53 IgorYozhikov, having a new version of pymod2pkg should help, currently all XStatic builds will fail because of that 14:00:18 yes, and that is why I'm asking about releasing of new version after merge 14:00:30 we r out of time :( 14:00:31 IgorYozhikov: done 14:00:36 jpena: any update on the rdo gate? 14:01:15 dirk: most reviews are getting CI results, only waiting for the networkx.drawing packages to make it report 14:01:49 jpena: ok, good. thanks 14:01:56 lets free the channel 14:01:58 #endmeeting