13:00:02 #startmeeting rpm_packaging 13:00:03 Meeting started Thu Jun 30 13:00:02 2016 UTC and is due to finish in 60 minutes. The chair is toabctl. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:08 The meeting name has been set to 'rpm_packaging' 13:00:12 hi 13:00:43 please add your agenda points as usual to https://etherpad.openstack.org/p/openstack-rpm-packaging 13:01:35 o/ 13:01:51 o/ 13:02:02 o/ 13:02:35 jpena: could you add a topic about your progress w/ the RDO gate? 13:02:58 number80: sure, just added 13:03:03 thanks :) 13:04:03 ok. let's start 13:04:08 #topic osl-lib as part for py*clients 13:04:11 IgorYozhikov, ^^ 13:04:39 toabctl, yes, just saw a bit time ago this new requirement 13:04:50 osl-lib? 13:04:51 most of clients are going to use it 13:04:55 osc-lib 13:05:00 ah yes 13:05:20 it'll simplify clients packaging 13:05:22 as a common part of functions extracted from osclient 13:05:37 #link https://pypi.python.org/pypi/osc-lib 13:06:09 I think that we also need to created project for it 13:06:34 IgorYozhikov, project? why not just another .spec.j2 ? 13:06:40 but for now osc-lib has strange behavior with version 13:06:57 toabctl, yes I mean folder & j2 13:07:02 yes, we experienced that but we can just ping version to latest known to work 13:07:30 that's why I'm not found of packaging trunk libs, they break too often 13:07:33 is there a bug for the "strange problem" ? 13:07:37 *release 13:07:57 toabctl: it's being addressed currently 13:08:03 https://github.com/openstack/osc-lib/blob/master/osc_lib/__init__.py#L18 13:08:46 when you are trying to import osc_lib without installed py*osclient -> it fails 13:09:18 I believe that version should be returned 0.2.0 as in egg# of osc_lib 13:09:50 otherwise it returns version of openstackclient not osc_lib 13:10:30 do we need it now? or can we wait for the next release which hopefully fixed the problem? 13:11:01 if we are not going to update py*clients in nearest future - so now, we can wait 13:12:24 ok. anything else about that topic? 13:12:29 nope 13:12:33 #topic packages reviews 13:12:43 #link https://review.openstack.org/#/q/project:openstack/rpm-packaging+status:open 13:12:53 a lot of versions bump today 13:13:12 yeah. great! 13:13:14 separate topic available: https://review.openstack.org/#/q/status:open+project:openstack/rpm-packaging+branch:master+topic:oslo-updates 13:14:08 https://review.openstack.org/#/c/333520/ && https://review.openstack.org/#/c/335923/ looks doing the same 13:14:08 some of them are trivial, others need to review CI logs 13:14:16 from the SUSE CI side, I hopefully fixed the problems with missing dependencies. but we have now a queue because our CI has currently only one worker 13:14:56 jpena: could you look at it ? ^ 13:15:15 number80: sure. I see 2 reviews failed to pass the SUSE CI, will have a look 13:15:40 thanks 13:16:28 toabctl, did you solve issue with oslo.cache and dogpile.cache? 13:16:49 IgorYozhikov, which issue? 13:17:06 toabctl: actually, that is causing one of the CI failures -> https://build.opensuse.org/package/live_build_log/home:suse-cloud-ci:rpm-packaging-openstack-4ad6ec3286ed04cd5811f46f9b0d1aa9777ae296/python-oslo.cache/SLE_12_SP1/x86_64 13:17:11 * toabctl was 2 weeks on vacation... 13:17:16 we need dogpile.cache 0.6 or later 13:17:19 dogpile >= 0.6 13:17:48 but oslo.cache 1.9 fails with this dogpile 13:18:07 also i see that 1.10 arrived - https://github.com/openstack/oslo.cache/tree/1.10.0 13:18:30 might be I can try bump it today? 13:18:31 jpena, yes. I updated dogpile.cache 13:18:56 IgorYozhikov, number80 did that already. see https://review.openstack.org/#/c/335928/ 13:19:39 toabctl, i see dependencies issues in SUSE CI - raise VersionConflict(dist, req).with_context(dependent_req) 13:19:39 [ 71s] ContextualVersionConflict: (iso8601 0.1.10 (/usr/lib/python2.7/site-packages), Requirement.parse('iso8601>=0.1.11'), set(['oslo.utils'])) 13:20:11 IgorYozhikov, where? 13:20:20 https://build.opensuse.org/package/live_build_log/home:suse-cloud-ci:rpm-packaging-openstack-a0fc365a69fddf0f9ded6482f4a089eec3ec1b7f/python-oslo.cache/SLE_12_SP1/x86_64 13:22:10 https://review.openstack.org/#/c/335915/1 +2.1 13:22:23 IgorYozhikov, I'll have a look with the iso8601 problem 13:22:44 toabctl, thanx 13:22:53 IgorYozhikov, merged https://review.openstack.org/#/c/335915/1 . I was just waiting for our CI 13:23:13 jpena, as IgorYozhikov mentioned, https://review.openstack.org/#/c/335923/ is a duplicate 13:23:32 jpena, can you have a look there, please? maybe close yours and comment on the other one? 13:23:48 toabctl: oh, sure. I didn't see that one 13:24:09 toabctl, https://review.openstack.org/#/c/330527/ could you please tel me - what is wrong with lint? 13:24:49 IgorYozhikov, see http://logs.openstack.org/27/330527/7/check/gate-rpm-packaging-tox-lint/b35cbb0/console.html . there's the diff output 13:25:01 IgorYozhikov, you can run it locally with "tox -elint" to check it 13:25:42 oh - %changelog 13:25:42 2016-06-27 13:45:28.620352 | \ No newline at end of file 13:26:07 IgorYozhikov, it's about the empty lines. 13:26:17 i.e. http://logs.openstack.org/27/330527/7/check/gate-rpm-packaging-tox-lint/b35cbb0/console.html#_2016-06-27_13_45_28_619822 13:27:12 toabctl, linter expecting 1 '\n' instead of 2? 13:27:14 right 13:27:44 yes 13:27:57 anything else about spec reviews? 13:28:12 but this should be reflected in wiki 13:28:24 going to update wiki page 13:28:38 IgorYozhikov, true. thanks! 13:29:21 #topic pymod2pkg reviews 13:29:33 #link https://review.openstack.org/#/q/project:openstack/pymod2pkg+status:open 13:30:07 both reviews listed there just need another +2 13:30:22 number80, do you need a pymod2pkg release on pypi when this got merged? 13:30:50 toabctl: likely but let's wait next week, I'll check if we need other changes first 13:31:47 ok 13:32:27 number80, might be better if this re-* rule will be documented 13:32:33 I'm about usage 13:32:48 IgorYozhikov: ack 13:33:02 may be 1 -2 examples 13:33:17 thanx :) 13:33:53 #topic RDO-based gate (WIP) 13:33:55 ^^ jpena 13:34:25 I've been working on making the RDO Trunk build tool (DLRN) support the rpm-packaging specs 13:34:49 so far I've got a working instance and I've collected the status at https://etherpad.openstack.org/p/DLRN-rpm-packaging 13:35:09 that caused some of my review-spam today :) 13:35:30 once we iron out the details, our plan is to use that for the RDO gate 13:35:53 cool 13:36:03 looks good 13:36:16 Yup, the nice touch is that we'd be able to build locally rpm-packaging spec for RDO very easily (so faster reviews) 13:36:33 number80: btw, there is one detail I just realized. DLRN always tries to build the latest master commit, while we have specific versions 13:36:43 as a gate job, that could create issues 13:37:19 I think we could solve that though 13:39:11 can I somewhere see the trunk builds? 13:39:45 toabctl: I'll set up an internet-facing instance, right now the one I used is inside the vpn 13:40:12 jpena, ok. and the one without rpm-packaging? is that public? 13:40:24 https://trunk.rdoproject.org/ 13:40:41 thx! 13:41:02 toabctl: we are building master, stable/mitaka and stable/liberty packages there (it's all in the index page) 13:41:38 hm. I just get 404 (https://trunk.rdoproject.org/f23/current/) 13:41:59 toabctl: that's waiting for an update :) Use f24 and f25 (for rawhide) 13:42:18 ah. centos works :) 13:42:31 well, reference platform for us :) 13:42:37 yes, centos - ok , fedora - 404 13:43:03 https://trunk.rdoproject.org/f24/current/ and https://trunk.rdoproject.org/f25/current/ work, although we have some non-built packages (trying to fix those) 13:43:27 :) 13:43:53 toabctl, https://review.openstack.org/#/c/330527/ passed linter 13:44:09 IgorYozhikov, great! :) 13:44:25 anything else about the RDO-based gate? 13:44:47 #topic open discussion 13:44:57 anthing else to discuss? 13:46:17 jpena, number80 looking at https://etherpad.openstack.org/p/DLRN-rpm-packaging , there are 2 problems mentioned (Source vs Source0 and version vs upstream_version). do you already have a plan how to solve these? 13:46:37 not now, after merging pymod2pkg - will start to do j2 4 renderspec 13:47:06 toabctl: I'm planning to work on renderspec to allow more customization per target 13:47:06 toabctl: number80 mentioned they could be fixed in renderspec. I think we might fix Source vs Source0 in the spec files themselves (only 3 specs use Source instead of Source0) 13:47:18 wfm 13:47:23 IgorYozhikov, I don't understand your comment. can you rephrase please? 13:48:19 I'm mean that I will start to work on spec.j2 for renderspec after https://review.openstack.org/#/c/330527/ been merged 13:48:41 IgorYozhikov, ah. ok 13:49:24 IgorYozhikov, which review was that with the iso8601 problem? 13:49:37 oslo.cache 13:49:46 ok. I guess we are done. we can continue in #openstack-rpm-packaging 13:49:52 thanks everybody! 13:49:54 #endmeeting