| *** ijw has quit IRC | 00:00 | |
| *** Swami has quit IRC | 00:01 | |
| tonyb | How do I set required for all jobs in a repo/project (without creating my own jobs?) I thought I could do somethign like: https://review.openstack.org/#/c/609823/2/.zuul.yaml but that didn't work | 00:09 |
|---|---|---|
| ianw | tonyb: i think it's a job-specific option | 00:14 |
| ianw | tonyb: to be more specific, the answer is probably to make a base template with that, and inherit | 00:19 |
| tonyb | ianw: Okay I think I understand | 00:26 |
| *** gyee has quit IRC | 00:29 | |
| pabelanger | ianw: mind looking at https://review.openstack.org/609855/ so when clarkb gets back we can get a +3 | 00:30 |
| pabelanger | currently nodepool release jobs are failing | 00:30 |
| pabelanger | http://lists.openstack.org/pipermail/release-job-failures/2018-October/000972.html | 00:31 |
| * fungi looks too | 00:32 | |
| pabelanger | thanks! | 00:32 |
| *** longkb has joined #openstack-infra | 00:33 | |
| fungi | do we need to reenqueue a tag once that merges? | 00:36 |
| fungi | or do you have it covered? | 00:36 |
| *** mattoliverau has joined #openstack-infra | 00:37 | |
| pabelanger | I can, if needed. Other wise, happy to defer to someone else | 00:39 |
| openstackgerrit | Merged openstack-infra/project-config master: Fix syntax error with zuul-tarball job https://review.openstack.org/609855 | 00:48 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Minor documentation updates https://review.openstack.org/609636 | 00:52 |
| pabelanger | zuul enqueue-ref --tenant=openstack --trigger=gerrit --pipeline=release --project=openstack-infra/nodepool --ref=refs/tags/3.3.0 --newrev=907251740903ed8c44eb9556d7a34a1e32abd476 | 00:54 |
| pabelanger | fungi: clarkb: ianw: how does ^ look | 00:54 |
| *** eernst has quit IRC | 00:55 | |
| fungi | lgtm. you can also feed the inventory.yaml url from the failed run to the reenqueue.py script in my homedir on the zuul server | 00:56 |
| pabelanger | oh, neat. let me try that | 00:56 |
| fungi | and then run the command it spits out | 00:57 |
| pabelanger | ++ | 00:58 |
| pabelanger | thanks | 00:58 |
| pabelanger | okay, jobs reenqueued | 00:59 |
| ianw | sorry, missed this, but LGTM :) | 00:59 |
| pabelanger | yah, reenqueue.py works great | 01:00 |
| pabelanger | okay, worked. nodepool 3.3.0 on tarballs.o.o and pypi | 01:03 |
| pabelanger | tristanC: ^ | 01:03 |
| pabelanger | Oh | 01:04 |
| pabelanger | Hmm | 01:04 |
| pabelanger | fungi: clarkb: I think we build a python2 | 01:04 |
| pabelanger | http://logs.openstack.org/90/907251740903ed8c44eb9556d7a34a1e32abd476/release/release-zuul-python/dbd3fb2 | 01:05 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: Add a pre-finalise.d phase https://review.openstack.org/609863 | 01:06 |
| clarkb | you have to set setup.cfg info to build a python3 wheel with pbr iirc | 01:06 |
| clarkb | pabelanger: you should be able to install thr sdist though | 01:06 |
| pabelanger | clarkb: http://git.openstack.org/cgit/openstack-infra/project-config/tree/playbooks/zuul-tarball/run.yaml#n5 | 01:07 |
| pabelanger | I guess that is python2 on the node | 01:07 |
| fungi | [wheel] universal = 1 | 01:07 |
| clarkb | the python version that makes the wheel shouldnt matter | 01:08 |
| clarkb | its an output problem | 01:08 |
| pabelanger | yah, we don't seem to set that in nodepool | 01:08 |
| clarkb | but the sdist wont care | 01:08 |
| fungi | oh, or 3-only | 01:08 |
| clarkb | so your pip install will still work | 01:08 |
| pabelanger | nodepool-3.3.0-py2-none-any.whl | 01:08 |
| pabelanger | I would expect that to be nodepool-3.3.0-py3-none-any.whl | 01:09 |
| pabelanger | like nodepool-3.2.0 | 01:09 |
| fungi | if you don't set universal then need to build wheel under a python3 interpreter | 01:09 |
| pabelanger | okay, if that is the case, then our run.yaml playbook is not correct | 01:09 |
| pabelanger | we likey should use python3 | 01:09 |
| clarkb | pabelanger: the versiom of python that makes the wheel doesnt matter iirc | 01:11 |
| clarkb | pbr uses config to determine what the output is | 01:11 |
| clarkb | oh fungi says the version does matter | 01:12 |
| clarkb | in any case the release that was just made should work via sdist | 01:12 |
| fungi | yeah, sdist will be fine | 01:12 |
| openstackgerrit | Paul Belanger proposed openstack-infra/project-config master: Force python3 for zuul-tarball job https://review.openstack.org/609866 | 01:13 |
| pabelanger | fungi: clarkb: okay, so if I follow, ^ should build the wheel as python3. And today, if we did pip3 nodepool, we won't get the wheel since it only is python2 | 01:14 |
| clarkb | right you get the sdist | 01:14 |
| pabelanger | okay | 01:14 |
| clarkb | so you can install nodepool just fine | 01:14 |
| clarkb | but it will take a little more time | 01:14 |
| *** bobh has joined #openstack-infra | 01:24 | |
| fungi | sorry, wasn't really able to type earlier | 01:25 |
| fungi | the major interpreter version determines whether wheel (or setup.py bdist_wheel) creates a whl filename with py or py3 | 01:26 |
| fungi | py2 or py3 | 01:26 |
| fungi | unless you set [bdist_wheel] universal = 1 in setup.cfg then it will say any instead (implying it works with both) | 01:27 |
| fungi | formerly [wheel] but that's deprecated in favor of [bdist_wheel] in the most recent wheel release | 01:27 |
| *** rh-jelabarre has joined #openstack-infra | 01:29 | |
| *** mrsoul has joined #openstack-infra | 01:30 | |
| tonyb | Can I Depends-On a chnage in openstack-zuul-jobs and have it do what I expect? or is openstack-zuul-jobs a trusted repo list project-config? | 01:36 |
| fungi | you should be able to depends-on a change for it and see the effects | 01:40 |
| tonyb | fungi: Thanks. | 01:42 |
| openstackgerrit | Tony Breeds proposed openstack-infra/openstack-zuul-jobs master: [DNM] Testing a new project-template for ceilometer https://review.openstack.org/609901 | 01:43 |
| openstackgerrit | Corey Bryant proposed openstack-infra/zuul-jobs master: Add ensure-python role https://review.openstack.org/609761 | 01:45 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement a Kubernetes driver https://review.openstack.org/535557 | 01:47 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement an OpenShift resource provider https://review.openstack.org/570667 | 01:49 |
| *** aidin has joined #openstack-infra | 01:55 | |
| *** eernst has joined #openstack-infra | 01:58 | |
| *** apetrich has quit IRC | 02:15 | |
| *** zigo has quit IRC | 02:29 | |
| *** yamamoto has quit IRC | 02:31 | |
| *** yamamoto has joined #openstack-infra | 02:31 | |
| *** roman_g has quit IRC | 02:39 | |
| *** aidin has quit IRC | 02:44 | |
| *** hongbin has joined #openstack-infra | 02:45 | |
| *** lbragstad has joined #openstack-infra | 02:45 | |
| *** aidin has joined #openstack-infra | 02:46 | |
| *** armax has joined #openstack-infra | 02:48 | |
| *** psachin has joined #openstack-infra | 02:53 | |
| *** rlandy|bbl is now known as rlandy | 03:00 | |
| *** rlandy has quit IRC | 03:02 | |
| *** bobh has quit IRC | 03:02 | |
| *** armax_ has joined #openstack-infra | 03:03 | |
| *** armax has quit IRC | 03:04 | |
| *** bobh has joined #openstack-infra | 03:05 | |
| *** dave-mccowan has quit IRC | 03:05 | |
| *** armax has joined #openstack-infra | 03:07 | |
| *** armax_ has quit IRC | 03:08 | |
| *** aidin has quit IRC | 03:10 | |
| *** bobh has quit IRC | 03:16 | |
| *** aidin has joined #openstack-infra | 03:37 | |
| *** armax has quit IRC | 03:49 | |
| *** udesale has joined #openstack-infra | 03:54 | |
| *** hongbin has quit IRC | 04:01 | |
| *** aidin has quit IRC | 04:02 | |
| *** dayou has quit IRC | 04:06 | |
| *** smcginnis has quit IRC | 04:06 | |
| *** lathiat has quit IRC | 04:06 | |
| *** lathiat has joined #openstack-infra | 04:06 | |
| *** felipemonteiro has joined #openstack-infra | 04:06 | |
| *** dayou has joined #openstack-infra | 04:06 | |
| *** EmilienM is now known as EvilienM | 04:13 | |
| openstackgerrit | Merged openstack-infra/zuul-sphinx master: Add attr-overview directive https://review.openstack.org/604980 | 04:16 |
| *** bhavikdbavishi has joined #openstack-infra | 04:17 | |
| openstackgerrit | Tony Breeds proposed openstack-infra/project-config master: Use root user for editing pip.conf https://review.openstack.org/609909 | 04:18 |
| tonyb | ianw: Can you look over ^^ | 04:19 |
| *** carl_cai has joined #openstack-infra | 04:20 | |
| *** lbragstad has quit IRC | 04:21 | |
| *** ykarel__ has joined #openstack-infra | 04:22 | |
| *** abhishekk has joined #openstack-infra | 04:27 | |
| abhishekk | hi, in gate environment, while executing gate jobs is httpd server is running, if yes then which port it uses? | 04:28 |
| *** icey has quit IRC | 04:30 | |
| *** mpeterson has quit IRC | 04:30 | |
| *** icey has joined #openstack-infra | 04:30 | |
| *** mpeterson has joined #openstack-infra | 04:30 | |
| fungi | infra-root: (and anyone else) just a heads up, i may not be available tomorrow through the weekend | 04:34 |
| *** ykarel__ is now known as ykarel | 04:38 | |
| tonyb | abhishekk: I'm far from an expert but I don't think there is a web server running. | 04:46 |
| tonyb | abhishekk: I guess a dsvm job with wsgi must have a virtual server for each endpoint | 04:47 |
| tonyb | abhishekk: what do you need one for? and in which jobs? | 04:47 |
| *** ramishra has joined #openstack-infra | 04:48 | |
| openstackgerrit | Merged openstack-infra/zuul master: Add a quick-start test job https://review.openstack.org/609514 | 04:50 |
| *** lathiat_ has joined #openstack-infra | 04:51 | |
| chkumar|off | Shrews: go ahead, I am happy to review the code | 04:53 |
| *** lathiat has quit IRC | 04:54 | |
| abhishekk | tonyb, thank you | 04:54 |
| abhishekk | tonyb, I am doing something like downloading image from external network for functional test | 04:54 |
| abhishekk | but those tests fails most of the time due to network issues | 04:55 |
| abhishekk | SotK, I am doing something like this https://review.openstack.org/#/c/608856/3/glance/tests/functional/v2/test_images.py | 04:55 |
| tonyb | abhishekk: Ahh okay. | 04:55 |
| tonyb | abhishekk: So yeah short term you'd be best placed to provide your own (either as you're doing) or as a stand-alone role/task in your functional setups | 04:57 |
| *** eernst has quit IRC | 04:57 | |
| tonyb | abhishekk: There are per provider mirrors (for apt/yum/images) but I'm not sure exactly how you'd plumb them into your functaional jobs | 04:57 |
| tonyb | abhishekk: It'd be worth discussin this on os-dev | 04:58 |
| abhishekk | tonyb, yes, I am doing that but those tests are timing out now, where as they are working properly in local environment :( | 04:58 |
| tonyb | abhishekk: which 'that' do you mean in the above statement? | 04:58 |
| abhishekk | that means, running local http server and downloading file from that | 04:59 |
| tonyb | abhishekk: for example the the link you gave you're pulling from https://www.openstack.org/assets/openstack-logo/ which isn't local to the provider | 05:00 |
| abhishekk | tonyb, sorry look at the PS 5 from that patch | 05:00 |
| tonyb | abhishekk: so even talkign to 127.0.0.1:8111 they're still failing? | 05:00 |
| abhishekk | tonyb, yes | 05:00 |
| *** eernst has joined #openstack-infra | 05:00 | |
| tonyb | abhishekk: Oh :( No idea then | 05:02 |
| abhishekk | tonyb, ok, thank you for inputs though | 05:03 |
| tonyb | abhishekk: I guess requests-mock isn't an option? | 05:04 |
| abhishekk | tonyb, yes that's the last one, I guess in functional tests of glance we are not mocking anything | 05:04 |
| tonyb | abhishekk: Yeah having a real sever is the best option | 05:05 |
| abhishekk | tonyb, yes, lets see, I will post it on os-dev if not able to encounter the issue, thank you again | 05:05 |
| tonyb | abhishekk: I havea half formed idea abiyt providing a really basic one and launching that before the stestr run. | 05:06 |
| * tonyb will stall on that until you post to os-dev to see if I can solidify it | 05:06 | |
| abhishekk | tonyb, sounds good | 05:06 |
| *** toabctl has joined #openstack-infra | 05:07 | |
| *** ssbarnea has quit IRC | 05:11 | |
| *** janki has joined #openstack-infra | 05:15 | |
| *** pcaruana has joined #openstack-infra | 05:15 | |
| *** rh-jelabarre has quit IRC | 05:22 | |
| openstackgerrit | Tony Breeds proposed openstack-infra/openstack-zuul-jobs master: Add ceilometer specific templates https://review.openstack.org/609901 | 05:23 |
| *** ssbarnea has joined #openstack-infra | 05:30 | |
| *** olivierb has joined #openstack-infra | 05:31 | |
| *** felipemonteiro has quit IRC | 05:31 | |
| abhishekk | tonyb, looks like, I manage to pulled it off, I will make sure by recheking 3-4 times | 05:31 |
| tonyb | \o/ | 05:32 |
| abhishekk | :D, | 05:32 |
| *** chkumar|off is now known as chandankumar | 05:34 | |
| *** kjackal has joined #openstack-infra | 05:34 | |
| openstackgerrit | Simon Westphahl proposed openstack-infra/zuul master: Check paused parent on node failure of child job https://review.openstack.org/609911 | 05:38 |
| *** jesusaur has quit IRC | 05:41 | |
| *** eernst has quit IRC | 05:45 | |
| *** aidin has joined #openstack-infra | 05:49 | |
| *** ykarel has quit IRC | 06:01 | |
| *** ykarel has joined #openstack-infra | 06:02 | |
| *** kopecmartin|ruck has quit IRC | 06:04 | |
| *** apetrich has joined #openstack-infra | 06:04 | |
| *** kopecmartin has joined #openstack-infra | 06:05 | |
| *** aidin has quit IRC | 06:10 | |
| *** jesusaur has joined #openstack-infra | 06:11 | |
| AJaeger | tonyb: you can do something like http://git.openstack.org/cgit/openstack-infra/project-config/tree/zuul.d/projects.yaml#n5226 as well - no need for a new template, just add the individual jobs | 06:12 |
| prometheanfire | if people have time, this dib change would be nice https://review.openstack.org/609863 | 06:14 |
| prometheanfire | just a couple lines | 06:14 |
| *** aidin has joined #openstack-infra | 06:15 | |
| *** janki has quit IRC | 06:19 | |
| *** dpawlik has joined #openstack-infra | 06:30 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Small cleanups for zuul.d/projects.yaml https://review.openstack.org/609778 | 06:31 |
| *** aidin has quit IRC | 06:33 | |
| *** janki has joined #openstack-infra | 06:34 | |
| AJaeger | config-core, three small cleanups for review, please: https://review.openstack.org/609793 https://review.openstack.org/609795 https://review.openstack.org/609778 | 06:39 |
| tonyb | AJaeger: I did try that but is seems like it'd be more prone to drift ? | 06:47 |
| tonyb | AJaeger: Given we can use the new templates in multiple repos do you still think that adding the overritgh per job is ther right way? | 06:48 |
| AJaeger | tonyb: if you have multiple users, I'm fine taking the template | 06:49 |
| AJaeger | tonyb: I would like to give jd_ time to review https://review.openstack.org/609901 as well so that the telemetry team is aware of it. | 06:50 |
| AJaeger | tonyb: I added jd_ now to the review... | 06:50 |
| tonyb | AJaeger: Okay cool. I'll work on it next week. There isn't a great rush. The powervm team have a work around | 06:51 |
| tonyb | AJaeger: so did I ;P | 06:51 |
| AJaeger | tonyb: so, go ahead and change neutron to ceilometer - and then let's wait a few days... | 06:51 |
| AJaeger | ;) | 06:51 |
| tonyb | AJaeger: Will do. | 06:51 |
| *** kopecmartin is now known as kopecmartin|ruck | 06:52 | |
| openstackgerrit | Tony Breeds proposed openstack-infra/openstack-zuul-jobs master: Add ceilometer specific templates https://review.openstack.org/609901 | 06:53 |
| AJaeger | tonyb: still a few neutron left in your change ^ | 06:54 |
| tonyb | AJaeger: Next week I'll sync up with dhellmann to find out if/when he;ll migreat the ceilometer-zvm zuul config in repo so I can update it | 06:54 |
| tonyb | AJaeger: :( | 06:54 |
| AJaeger | tonyb: he will not update it - it's not an official repo | 06:55 |
| AJaeger | tonyb: so do it yourself, please. | 06:55 |
| AJaeger | tonyb: and dhellmann is on PTO next week | 06:55 |
| tonyb | AJaeger: Ahh okay. I'll find the tools and run them myself then | 06:56 |
| AJaeger | tonyb: repo openstack/goal-tools - this checks governance repo for release teams, no idea whether it works with single repo... | 06:56 |
| openstackgerrit | Tony Breeds proposed openstack-infra/openstack-zuul-jobs master: Add ceilometer specific templates https://review.openstack.org/609901 | 06:57 |
| tonyb | AJaeger: Thanks, and I think I got them all this time | 06:57 |
| tonyb | AJaeger: I can probably hack it to work for a single repo as needed ;P | 06:58 |
| tonyb | AJaeger: Thanks for your help I think I'm going knock off now before I mess anythign else up ;P | 06:58 |
| AJaeger | np, enjoy your evening/night! | 06:59 |
| *** rcernin has quit IRC | 07:03 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Install docker from distro properly https://review.openstack.org/609844 | 07:03 |
| *** jesusaur has quit IRC | 07:08 | |
| *** ykarel_ has joined #openstack-infra | 07:12 | |
| *** ykarel_ has quit IRC | 07:13 | |
| *** jesusaur has joined #openstack-infra | 07:13 | |
| *** ykarel_ has joined #openstack-infra | 07:14 | |
| *** ykarel has quit IRC | 07:14 | |
| *** bhavikdbavishi has quit IRC | 07:16 | |
| *** apetrich has quit IRC | 07:17 | |
| *** shardy has joined #openstack-infra | 07:18 | |
| *** jtomasek has joined #openstack-infra | 07:19 | |
| *** ykarel_ is now known as ykarel | 07:22 | |
| openstackgerrit | Merged openstack-infra/zuul master: Fixes for quick-start playbooks https://review.openstack.org/609551 | 07:23 |
| openstackgerrit | Merged openstack-infra/zuul master: Remove disable-recommends in quick-start test https://review.openstack.org/609553 | 07:23 |
| *** ginopc has joined #openstack-infra | 07:27 | |
| *** ginopc has quit IRC | 07:30 | |
| *** apetrich has joined #openstack-infra | 07:30 | |
| *** tosky has joined #openstack-infra | 07:45 | |
| *** jpich has joined #openstack-infra | 07:46 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Retry jobs failed with MERGER_FAILURE https://review.openstack.org/609945 | 07:53 |
| *** e0ne has joined #openstack-infra | 07:59 | |
| *** electrofelix has joined #openstack-infra | 08:04 | |
| *** d0ugal has quit IRC | 08:07 | |
| *** d0ugal has joined #openstack-infra | 08:08 | |
| *** ginopc has joined #openstack-infra | 08:08 | |
| *** rpittau has quit IRC | 08:09 | |
| *** rpittau has joined #openstack-infra | 08:10 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Use tempest-pg-full https://review.openstack.org/609958 | 08:10 |
| *** rpittau has quit IRC | 08:10 | |
| *** rpittau has joined #openstack-infra | 08:11 | |
| *** roman_g has joined #openstack-infra | 08:28 | |
| *** bauzas is now known as PapaOurs | 08:30 | |
| *** zigo has joined #openstack-infra | 08:34 | |
| *** derekh has joined #openstack-infra | 08:41 | |
| *** e0ne has quit IRC | 08:44 | |
| *** dpawlik has quit IRC | 08:46 | |
| *** e0ne has joined #openstack-infra | 08:46 | |
| *** slaweq has quit IRC | 08:46 | |
| *** dpawlik has joined #openstack-infra | 08:51 | |
| *** psachin has quit IRC | 08:55 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add ensure-python role https://review.openstack.org/609761 | 08:56 |
| *** janki has quit IRC | 08:58 | |
| *** janki has joined #openstack-infra | 08:58 | |
| *** janki has quit IRC | 09:03 | |
| *** janki has joined #openstack-infra | 09:04 | |
| *** janki has quit IRC | 09:05 | |
| *** janki has joined #openstack-infra | 09:05 | |
| *** tomaw has quit IRC | 09:07 | |
| *** tomaw has joined #openstack-infra | 09:08 | |
| *** ramishra has quit IRC | 09:09 | |
| *** ramishra has joined #openstack-infra | 09:09 | |
| *** e0ne has quit IRC | 09:13 | |
| *** e0ne has joined #openstack-infra | 09:16 | |
| *** dtantsur|afk is now known as dtantsur | 09:33 | |
| *** gnuoy has quit IRC | 09:35 | |
| *** gnuoy has joined #openstack-infra | 09:37 | |
| *** gnuoy has quit IRC | 09:38 | |
| *** gnuoy has joined #openstack-infra | 09:38 | |
| *** xinliang has quit IRC | 09:42 | |
| *** psachin has joined #openstack-infra | 09:45 | |
| *** gnuoy has quit IRC | 09:50 | |
| openstackgerrit | Merged openstack-infra/openstack-zuul-jobs master: Ensure python3.7 installed for openstack-tox-py37 https://review.openstack.org/609766 | 09:51 |
| *** imacdonn has quit IRC | 09:53 | |
| *** imacdonn has joined #openstack-infra | 09:53 | |
| *** gnuoy has joined #openstack-infra | 09:55 | |
| *** xinliang has joined #openstack-infra | 09:56 | |
| *** slaweq has joined #openstack-infra | 09:56 | |
| *** ykarel is now known as ykarel|lunch | 10:04 | |
| *** rkukura has quit IRC | 10:13 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: wip: add status graph rendering https://review.openstack.org/610001 | 10:27 |
| *** psachin has quit IRC | 10:33 | |
| *** e0ne has quit IRC | 10:34 | |
| *** udesale has quit IRC | 10:36 | |
| *** pabelanger has quit IRC | 10:40 | |
| *** jpena|off has quit IRC | 10:41 | |
| *** rascasoft has quit IRC | 10:42 | |
| *** jpena|off has joined #openstack-infra | 10:42 | |
| *** agopi has quit IRC | 10:46 | |
| *** abhishekk has quit IRC | 10:46 | |
| *** psachin has joined #openstack-infra | 10:49 | |
| *** sshnaidm is now known as sshnaidm|off | 10:57 | |
| *** dayou has quit IRC | 11:09 | |
| *** apetrich has quit IRC | 11:10 | |
| *** ykarel_ has joined #openstack-infra | 11:11 | |
| *** ykarel|lunch has quit IRC | 11:14 | |
| *** dayou has joined #openstack-infra | 11:14 | |
| *** longkb has quit IRC | 11:20 | |
| *** e0ne has joined #openstack-infra | 11:27 | |
| *** psachin has quit IRC | 11:28 | |
| *** apetrich has joined #openstack-infra | 11:29 | |
| *** psachin has joined #openstack-infra | 11:33 | |
| *** bhavikdbavishi has joined #openstack-infra | 11:37 | |
| *** dave-mccowan has joined #openstack-infra | 11:38 | |
| *** bhavikdbavishi has quit IRC | 11:46 | |
| *** rh-jelabarre has joined #openstack-infra | 11:46 | |
| *** bhavikdbavishi has joined #openstack-infra | 11:46 | |
| *** bobh has joined #openstack-infra | 12:01 | |
| *** kjackal has quit IRC | 12:04 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Fix skipping of max-servers if not set https://review.openstack.org/610019 | 12:06 |
| *** kjackal has joined #openstack-infra | 12:07 | |
| *** bobh has quit IRC | 12:09 | |
| *** sthussey has joined #openstack-infra | 12:11 | |
| dhellmann | tonyb , AJaeger : I think I made the migration tools take a list of repos as argument, but if not patches to do so would be fine | 12:15 |
| *** e0ne has quit IRC | 12:16 | |
| AJaeger | dhellmann: indeed, you do - the scripts handle teams, but underlying tool uses repos. Great! | 12:17 |
| *** e0ne has joined #openstack-infra | 12:17 | |
| dhellmann | cool | 12:17 |
| dhellmann | in other news, until my isp comes back I have IRC and web but my dev box can't see the internet :-/ | 12:18 |
| AJaeger | dhellmann: start your vacation a day earlier ;) | 12:18 |
| dhellmann | I like the way you think, AJaeger | 12:18 |
| ssbarnea | dhellmann AJaeger : I need some help with one infra issue, I need a permalink to the fedora qcow image, one that current has a floating name at https://nb02.openstack.org/images/ -- how can I get one? | 12:19 |
| AJaeger | ssbarnea: both of us are the wrong persons to help you with that. | 12:20 |
| dhellmann | yeah, unfortunately I don't know much about how images are managed | 12:20 |
| dhellmann | what causes the name to keep changing? | 12:20 |
| ssbarnea | i don't know who to ask,... | 12:20 |
| AJaeger | I doubt that there's an easy way - so, best explain the use case... | 12:20 |
| AJaeger | ssbarnea: any of the admins - fungi, clarkb, pabelanger, mordred, Shrews might be able to help | 12:20 |
| ssbarnea | i don't want to add extra heuristics for looking for the last image, is better to update the process that is udpating it to create a symlink. | 12:21 |
| ssbarnea | thanks, lets hope they will be able to tell me where it can be done. usually is not hard to create a symlink. | 12:21 |
| dhellmann | how do we pick the latest image for the other types of images? | 12:21 |
| AJaeger | ssbarnea: explain them your use case ;) The names are only used by nodepool and are not meant for you to consume at all... | 12:22 |
| *** smarcet has joined #openstack-infra | 12:23 | |
| *** smarcet has quit IRC | 12:24 | |
| *** ansmith has joined #openstack-infra | 12:24 | |
| ssbarnea | AJaeger: dhellmann : we are trying to use the same image for undercloud deploymentm see https://review.openstack.org/#/c/591540/57/config/environments/standalone_fedora_libvirt.yml | 12:25 |
| AJaeger | ssbarnea: and those images are available uploaded in our clouds - we push them from nb02 inside each cloud | 12:26 |
| AJaeger | ssbarnea: are you aware of http://nl01.openstack.org/image-list ? | 12:27 |
| AJaeger | ssbarnea: might be best to write an email to openstack-infra explaining the use case, showing the review and asking for help. Using nb02.openstack.org/images/ is the wrong answer ;) | 12:28 |
| ssbarnea | AJaeger: I think that we already know that but i need a temporary solution, so we can continue with f28 enablement. Still, if the image is changing every day, it becomes impossible to test. | 12:30 |
| ssbarnea | weshay: ^^ any ideas? | 12:31 |
| AJaeger | infra-root, please see above ^ | 12:32 |
| AJaeger | ssbarnea: can't help constructively... | 12:32 |
| *** ramishra has quit IRC | 12:34 | |
| *** rlandy has joined #openstack-infra | 12:40 | |
| weshay | ssbarnea, we don't need to use that image locally man | 12:41 |
| weshay | however the links changing all the time w/o softlinks to a perm link is understandable | 12:41 |
| weshay | hard | 12:41 |
| weshay | ssbarnea, use the fedora cloud qcow man | 12:41 |
| Shrews | ssbarnea: AJaeger: the image name format contains a unique identifier (generated by zookeeper) that's necessary for nodepool operations. That's how the name is stored there. I'm not inclined to make large changes in nodepool to alter that for an external entity | 12:41 |
| *** psachin has quit IRC | 12:45 | |
| frickler | ssbarnea: I may have missed some earlier discussion on that, but can you remind us why you prefer an infra-built image to the default upstream cloud image? | 12:49 |
| *** jtomasek has quit IRC | 12:49 | |
| Shrews | we might be able to offer a new web endpoint to expose the latest. still seems rather fragile to depend on our images (we could remove it at any time, and break you) | 12:50 |
| frickler | only possible approach would be to create a job that builds an image and publishes it on tarballs.o.o, independent from what nodepool does | 12:51 |
| *** kgiusti has joined #openstack-infra | 12:51 | |
| *** yamamoto has quit IRC | 12:54 | |
| *** yamamoto has joined #openstack-infra | 12:54 | |
| Shrews | ssbarnea: here are the dib elements we use in nodepool to build our fedora-28 image: http://git.openstack.org/cgit/openstack-infra/project-config/tree/nodepool/nodepool.yaml#n162 | 12:56 |
| ssbarnea | Shrews: thanks! | 12:57 |
| *** mriedem has joined #openstack-infra | 12:58 | |
| *** yamamoto has quit IRC | 12:59 | |
| *** agopi has joined #openstack-infra | 13:02 | |
| *** bobh has joined #openstack-infra | 13:06 | |
| *** elod has quit IRC | 13:06 | |
| *** elod has joined #openstack-infra | 13:06 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Ensure that completed handlers are removed frequently https://review.openstack.org/610029 | 13:07 |
| *** dtantsur is now known as dtantsur|afk | 13:07 | |
| *** agopi has quit IRC | 13:07 | |
| *** carl_cai has quit IRC | 13:10 | |
| *** yamamoto has joined #openstack-infra | 13:10 | |
| *** eharney has joined #openstack-infra | 13:10 | |
| *** bhavikdbavishi1 has joined #openstack-infra | 13:13 | |
| mordred | ssbarnea: yah - I think bulding one using the same elements for your use would be better - because then you can exclude the element that includes every git repo in infra - which should make the resulting images a much more reasonable size without having a functional impact | 13:13 |
| *** bhavikdbavishi has quit IRC | 13:13 | |
| *** markvoelker has quit IRC | 13:15 | |
| *** bhavikdbavishi1 has quit IRC | 13:17 | |
| *** ssbarnea_ has quit IRC | 13:18 | |
| *** mriedem is now known as hansmoleman | 13:22 | |
| *** munimeha1 has joined #openstack-infra | 13:22 | |
| *** lbragstad has joined #openstack-infra | 13:23 | |
| odyssey4me | folks, I see network configuration was added to glean in https://github.com/openstack-infra/glean/commit/fa417426ae4c1892f6a855be142b0060fe946a77 - except it's only added for gentoo... is there a reason to hold that back, rather than have it apply to a broader set of distributions? | 13:24 |
| odyssey4me | when using glean on bionic we're finding that the DNS isn't configured, so this is a bit of a hindrence | 13:25 |
| odyssey4me | the issue doesn't apply in infra because infra is using unbound and disables any other way to configure DNS on the nodes | 13:25 |
| *** ijw has joined #openstack-infra | 13:26 | |
| *** ijw has quit IRC | 13:26 | |
| *** ijw has joined #openstack-infra | 13:27 | |
| *** kjackal has quit IRC | 13:28 | |
| *** kjackal has joined #openstack-infra | 13:31 | |
| *** agopi has joined #openstack-infra | 13:32 | |
| *** munimeha1 has quit IRC | 13:38 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Implement a Kubernetes driver https://review.openstack.org/535557 | 13:39 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Add tox functional testing for drivers https://review.openstack.org/609515 | 13:39 |
| *** boden has joined #openstack-infra | 13:40 | |
| mordred | odyssey4me: we should be doing network config for ubuntu as well ... OH - networkd, not network | 13:40 |
| *** dansmith is now known as SteelyDan | 13:41 | |
| mordred | odyssey4me: no, I don't think there is a specific reason, we just haven't had anybody interested in it yet | 13:41 |
| *** ijw has quit IRC | 13:41 | |
| *** stephenfin is now known as finucannot | 13:41 | |
| odyssey4me | mordred: well, I think I've figured out the actual issue - networkd only does the ip bits, no dns | 13:42 |
| odyssey4me | currently glean puts down a resolv.conf for the resolvers | 13:42 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Ensure that completed handlers are removed frequently https://review.openstack.org/610029 | 13:42 |
| odyssey4me | but if resolved is enabled at all, it will overwrite that, and DNS resolution fails because resolved doesn't have any DNS configuration done | 13:42 |
| odyssey4me | so I guess the solution here would be to allow glean to configure resolved or modify resolve.conf | 13:43 |
| Shrews | ianw: i hope the port thing *is* a neutron bug. are you suggesting we wait to find out if it is before continuing with the nodepool change? | 13:44 |
| Shrews | i guess that's probably the sanest thing to do | 13:45 |
| mordred | odyssey4me: yeah. although - avoiding resolved is the thing I'd personally rather do, given that it doesn't know how to properly do dns | 13:45 |
| *** felipemonteiro has joined #openstack-infra | 13:46 | |
| mordred | but if it's unavoidable for some reason, I imagine the better thing to do would be to have glean actually configure it | 13:46 |
| *** ijw has joined #openstack-infra | 13:47 | |
| mordred | strigazi: heya - if you have a sec, the magnum functional test job for openstacksdk has stopped working and I don't know why ... http://logs.openstack.org/17/608317/3/check/openstacksdk-functional-devstack-magnum/11eaab9/job-output.txt.gz | 13:52 |
| *** rpittau has quit IRC | 13:54 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Add tox functional testing for drivers https://review.openstack.org/609515 | 13:54 |
| *** munimeha1 has joined #openstack-infra | 13:57 | |
| *** bdodd has quit IRC | 14:03 | |
| *** efried is now known as fried_rice | 14:05 | |
| *** ijw has quit IRC | 14:05 | |
| *** bdodd has joined #openstack-infra | 14:06 | |
| *** ijw has joined #openstack-infra | 14:07 | |
| *** eharney has quit IRC | 14:09 | |
| *** smcginnis has joined #openstack-infra | 14:09 | |
| *** ijw_ has joined #openstack-infra | 14:10 | |
| *** ijw has quit IRC | 14:12 | |
| prometheanfire | odyssey4me: mordred, so, add resolvd support to gleans dns config for networkd? | 14:14 |
| prometheanfire | mordred: I think we should define supported OS versions, my question is what we should enable networkd on | 14:15 |
| prometheanfire | also, does this look like it's checking the wrong file to anyone else (not looking at /etc/systemd/network) https://github.com/openstack-infra/glean/commit/fa417426ae4c1892f6a855be142b0060fe946a77#diff-1b42769ab0f2150100d2c2c1c090b512R540 | 14:16 |
| *** pabelanger has joined #openstack-infra | 14:16 | |
| *** yamamoto has quit IRC | 14:18 | |
| *** yamamoto has joined #openstack-infra | 14:18 | |
| *** lewo has joined #openstack-infra | 14:19 | |
| *** ykarel__ has joined #openstack-infra | 14:20 | |
| *** ykarel_ has quit IRC | 14:22 | |
| openstackgerrit | Sorin Sbarnea proposed openstack/gertty master: Change the default location for gertty config file https://review.openstack.org/581868 | 14:22 |
| *** yamamoto has quit IRC | 14:24 | |
| *** yamamoto has joined #openstack-infra | 14:24 | |
| *** lbragstad is now known as elbragstad | 14:25 | |
| *** munimeha1 has quit IRC | 14:32 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool master: Add universal flag to setup.cfg https://review.openstack.org/610057 | 14:35 |
| *** ijw_ has quit IRC | 14:35 | |
| *** ijw has joined #openstack-infra | 14:35 | |
| *** xinliang has quit IRC | 14:36 | |
| *** haleyb has quit IRC | 14:38 | |
| *** smarcet has joined #openstack-infra | 14:39 | |
| *** ansmith has quit IRC | 14:40 | |
| *** markvoelker has joined #openstack-infra | 14:41 | |
| *** bnemec is now known as beekneemech | 14:44 | |
| *** eharney has joined #openstack-infra | 14:44 | |
| openstackgerrit | Matthew Thode proposed openstack-infra/glean master: check networkd files https://review.openstack.org/610064 | 14:44 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add opendev nameservers (1/2) https://review.openstack.org/605092 | 14:49 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add opendev nameservers (2/2) https://review.openstack.org/610066 | 14:49 |
| *** felipemonteiro has quit IRC | 14:56 | |
| *** ansmith has joined #openstack-infra | 14:57 | |
| *** e0ne has quit IRC | 14:58 | |
| prometheanfire | mordred: so, for networkd and resolved, we can just enable resolved if networkd is enabled and it will just work (if the symlink exists), should I make the glean-networkd unit file depend/start resolved? | 14:59 |
| openstackgerrit | Merged openstack-infra/project-config master: Force python3 for zuul-tarball job https://review.openstack.org/609866 | 15:02 |
| clarkb | why wouldnt you let the image build decide if resolvd should be enabled? | 15:03 |
| clarkb | I dont think glean should do that? | 15:03 |
| prometheanfire | clarkb: glean writes out /etc/resolv.conf, which overrides resolved | 15:04 |
| prometheanfire | see the last bullet here https://www.freedesktop.org/software/systemd/man/systemd-resolved.service.html#/etc/resolv.conf | 15:04 |
| *** ansmith has quit IRC | 15:05 | |
| pabelanger | --skip-dns is a thing for glean | 15:05 |
| pabelanger | that shouldn't touch resolv.conf | 15:05 |
| clarkb | ya glean doesnt set dns for infra | 15:05 |
| clarkb | we configure nameserver to localhost and run unbound and glean doesnt touch it | 15:05 |
| clarkb | I think you can do similar with resolved | 15:06 |
| prometheanfire | true, can just add that to ExecStart=%%GLEANSH_PATH%%/glean.sh --distro networkd | 15:06 |
| *** e0ne has joined #openstack-infra | 15:06 | |
| prometheanfire | though resolved would interfere if you wish to use unbound locally | 15:07 |
| pabelanger | http://git.openstack.org/cgit/openstack-infra/project-config/tree/nodepool/elements/nodepool-base/glean@.service.d/override.conf | 15:07 |
| pabelanger | is the default ARGS we use for glean | 15:07 |
| pabelanger | I guess gentoo isn't using --interface? | 15:07 |
| clarkb | prometheanfire: yes I dont think infra would want resolved | 15:08 |
| clarkb | Im saying configuring dns like that should be image build choice not glean decision | 15:08 |
| clarkb | then I can choose to use unbound or resolved | 15:08 |
| clarkb | or point at google or cloudflare directly | 15:09 |
| prometheanfire | so it'd have to be optional, somehow, maybe two diferent glean-networkd.service files, one disabling dns/resolved and one enabling it, toggled by the --skip-dns flag | 15:10 |
| clarkb | (I think default is whatever dhcp says to do and however your dhcp client configures it) | 15:10 |
| clarkb | prometheanfire: why does glean have to know? I guess the case where dhcp isnt used? | 15:10 |
| clarkb | can we jus say that isnt supported and more people should use dhcp? | 15:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add __init__.py to web directory https://review.openstack.org/610080 | 15:11 |
| prometheanfire | I doubt it, iirc some providers only do static still | 15:11 |
| AJaeger | config-core, please review https://review.openstack.org/609958 https://review.openstack.org/609795 https://review.openstack.org/609778 | 15:12 |
| clarkb | prometheanfire: actually it would work to set the statically configured resolvers in resolv.conf right? | 15:14 |
| clarkb | you dont get resolved but dns will work via libc | 15:14 |
| *** rpioso|afk is now known as rpioso | 15:14 | |
| prometheanfire | glean-networkd as it esists now prepares stuff for resolved, if the end user wishes to use resolved they need to both set an override.conf for glean-networkd.service and symlink /etc/resolv.conf to /run/systemd/resolve/stub-resolv.conf | 15:15 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add __init__.py to web directory https://review.openstack.org/610080 | 15:15 |
| prometheanfire | clarkb: ya, if /etc/resolv.conf is a file then it 'wins' and resolvd just uses the contents from /etc/resolv.conf | 15:15 |
| clarkb | then it is already functional? | 15:16 |
| prometheanfire | yes, as long as we are fine with my previous comment | 15:16 |
| prometheanfire | glean-networkd as it esists now prepares stuff for resolved, if the end user wishes to use resolved they need to both set an override.conf for glean-networkd.service and symlink /etc/resolv.conf to /run/systemd/resolve/stub-resolv.conf | 15:16 |
| *** bhavikdbavishi has joined #openstack-infra | 15:16 | |
| clarkb | I dont really care as long as dns is functional on boot which it should be | 15:17 |
| *** dpawlik has quit IRC | 15:17 | |
| prometheanfire | ya, it would be | 15:17 |
| clarkb | if you want something more that seems out of scope for glean | 15:17 |
| clarkb | like infra using unbound is independnt of glean | 15:17 |
| *** jtomasek has joined #openstack-infra | 15:18 | |
| prometheanfire | my point with bringing this up is that using glean means you need to do even more extra work to allow resolved to work | 15:18 |
| prometheanfire | but it sounds like everyone is fine with that | 15:18 |
| prometheanfire | so meh | 15:18 |
| *** ansmith has joined #openstack-infra | 15:18 | |
| *** gyee has joined #openstack-infra | 15:19 | |
| *** adriancz has quit IRC | 15:20 | |
| *** ijw has quit IRC | 15:20 | |
| prometheanfire | odyssey4me: so, looks like glean doesn't need to care about resolved, at least for now it will set up dns for /etc/resov.conf but not systemd-resolved | 15:21 |
| *** ijw has joined #openstack-infra | 15:21 | |
| clarkb | prometheanfire: one of the reasons glean exists is that cloud-init does far too much for you. Things that should be handled by your image build or configuration management. I'm trying to sort out the balance of working networking (and dns) with glean in this context and I personally err on the side of dns will work on boot so glean has done its job | 15:23 |
| clarkb | prometheanfire: if users want something different (like infra using unbound) that is an image build/configuration management responsibility so that you can make the correct decisions for you | 15:24 |
| prometheanfire | clarkb: and that's fine, it's easier for me to not write code than to write code :P | 15:25 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use zuul/web/static in setup_hook https://review.openstack.org/609764 | 15:26 |
| clarkb | prometheanfire: one thing we might be able to assume is that if you are explicitly using networkd (so gentoo today and not the other distros glean supports) maybe we write the symlink with glean (and don't make it a set of options? | 15:30 |
| clarkb | prometheanfire: I think the distros that aren't resolved by default would find it really odd to get rsolved support generally | 15:30 |
| clarkb | but I also don't know how common resolved is if networkd is used? I know resolved requires networkd but not the other way around iirc | 15:32 |
| odyssey4me | clarkb prometheanfire I was thinking more along the lines of if systemd-resolved is enabled, then write the right config for it. If not, then do what it does now (write /etc/resolv.conf). And do much the same for systemd-networkd. | 15:32 |
| clarkb | odyssey4me: that also seems like a reasonable approach if glean can determine that easily (I think it can at least via systemctl) | 15:33 |
| clarkb | then your image build or config mgmt can explicitly enable a feature and glean will just honor that | 15:33 |
| odyssey4me | It's more to make glean work with whatever is in place, rather than have it compete - which is exactly what we just found happened on bionic - in bionic resolved is on by default. | 15:33 |
| clarkb | (really what we want to avoid doing is surprising the user like cloud init can do) | 15:33 |
| odyssey4me | *systemd-resolved | 15:33 |
| clarkb | odyssey4me: oh is it? I may avoid upgrading to bionic at home then :P | 15:33 |
| * clarkb grumbles something about dns over dbus being bad | 15:34 | |
| clarkb | at least they fixed the known security bugs with it :( | 15:34 |
| prometheanfire | odyssey4me: the config for which nameservers to use is already set up if using systemd-networkd | 15:34 |
| odyssey4me | yep, that's pretty much what I'm thinking - glean should just respect whatever's in the image... for now we've disabled systemd-resolved in our image, because meh | 15:34 |
| odyssey4me | prometheanfire: alright, in that case then writing out the systemd-networks config should be based on whether that service is enabled IMO | 15:35 |
| prometheanfire | the main question is if glean should use resolved if also using networkd, I think people don't want resolved to be tied to networkd like that | 15:35 |
| prometheanfire | odyssey4me: more words (less pronouns) :P | 15:35 |
| odyssey4me | prometheanfire: no, I think it should only write the config if the service is present and enabled | 15:35 |
| prometheanfire | odyssey4me: right now it only writes the networkd files if networkd is enabled | 15:36 |
| prometheanfire | it already does what you ask | 15:36 |
| prometheanfire | the config for resolved is just the DNS= entries in the network files, glean already writes that info to /etc/resov.conf | 15:37 |
| *** e0ne has quit IRC | 15:37 | |
| prometheanfire | if resolved is disabled that DNS= info does nothing iirc | 15:37 |
| odyssey4me | prometheanfire: well, on a bionic host it wrote the config, and DNS was totally broken - I had to mask systemd-resolved and re-run glean to get it operational again | 15:37 |
| prometheanfire | so we are already doing the right thing | 15:38 |
| prometheanfire | odyssey4me: what is 'the config' | 15:38 |
| prometheanfire | and what is 'it' | 15:38 |
| prometheanfire | pronoun games | 15:38 |
| odyssey4me | prometheanfire: this was the rsulting set for eth0 on the bionic host (note no DNS present): http://paste.openstack.org/show/731976/ | 15:39 |
| clarkb | odyssey4me: was /etc/resolv.conf written though? | 15:40 |
| *** e0ne has joined #openstack-infra | 15:40 | |
| prometheanfire | odyssey4me: oh, I guess we don't write out the DNS= lines then | 15:40 |
| *** e0ne has quit IRC | 15:40 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add line to setup.cfg limiting to python >=3.5 https://review.openstack.org/610089 | 15:42 |
| odyssey4me | clarkb: it was written, then systemd-resolved realised that /etc/resolv.conf was different to what it liked and changed it... and DNS resolution broke because there were no resolvers configured any more | 15:42 |
| clarkb | ah | 15:42 |
| prometheanfire | clarkb: odyssey4me mordred, do you think that the networkd config should write out those lines then? | 15:42 |
| prometheanfire | odyssey4me: that part (overwriting resolv.conf) is the OS's fault, not glean | 15:42 |
| prometheanfire | it's not overwritten on gentoo at least | 15:42 |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool master: Add line to setup.cfg limiting to python >=3.5 https://review.openstack.org/610090 | 15:42 |
| odyssey4me | prometheanfire: it's systemd-resolved's fault, yes | 15:43 |
| odyssey4me | systemd-resolved changes /etc/resolv.conf to contain 'nameserver 127.0.0.53', and then expects '/etc/systemd/resolved.conf' to contain the resolvers I think... although it may get those from systemd-networkd's interface configs, I don't know | 15:45 |
| *** rlandy is now known as rlandy|biab | 15:45 | |
| mordred | clarkb: yeah - I think having glean write the info to all the various places is a good idea | 15:45 |
| clarkb | odyssey4me: ok given that (dns not working on boot as I expected) I think the best option is to attempt to configure resolved if enabled | 15:46 |
| odyssey4me | ++ | 15:46 |
| prometheanfire | odyssey4me: it can read from networkd config too, let me check | 15:46 |
| *** dpawlik has joined #openstack-infra | 15:47 | |
| prometheanfire | which is what I'd prefer to configure, leave the distro resolved.conf file alone | 15:47 |
| *** kjackal_v2 has joined #openstack-infra | 15:48 | |
| *** kjackal has quit IRC | 15:48 | |
| openstackgerrit | Doug Hellmann proposed openstack-infra/project-config master: update the branching script to include a semver bump https://review.openstack.org/609827 | 15:49 |
| *** markvoelker has quit IRC | 15:51 | |
| *** dpawlik has quit IRC | 15:51 | |
| *** ansmith has quit IRC | 15:51 | |
| *** markvoelker has joined #openstack-infra | 15:52 | |
| prometheanfire | mordred: clarkb it looks like most of the time dns info is given globally, should those dns servers be configured for all interfaces? | 15:53 |
| clarkb | ssbarnea: following up on testing images, you should be able to add jobs in zuul that run on the fedora-28 image and have zuul do much of the heavy lifting for you | 15:54 |
| clarkb | prometheanfire: I'm not sure I understand since dns is a layer 7 service that runs on top of the network stack | 15:55 |
| ssbarnea | clarkb: yep, i know. we already have one CR that is creating such job. Still, there is a long chain to fixes we need to do till this will merge. | 15:55 |
| prometheanfire | clarkb: it's based on how the info is handed off to glean in /mnt/config/openstack/latest/network_data.json | 15:55 |
| clarkb | prometheanfire: right and dns info isn't interface specific there. And the dns service on the host isn't interface specific either? | 15:56 |
| prometheanfire | while openstack CAN specify the dns service per interface, it usually doesn't | 15:56 |
| *** ansmith has joined #openstack-infra | 15:57 | |
| prometheanfire | and resolved can also do dns per interface | 15:57 |
| prometheanfire | see https://paste.pound-python.org/show/gM5ezN6lqofVD2aFbq5t/ for an example of dns per interface | 15:57 |
| clarkb | how does that work for a process resolving a name? | 15:57 |
| clarkb | routes should determine how you talk to the resolver that is configured, interfaces don't directly come into it | 15:58 |
| clarkb | I guess the analog is dhcp on interface foo says dns server is xyz and dhcp on another interface cna say somethin different | 15:59 |
| prometheanfire | ya | 15:59 |
| clarkb | in the case of global config from config-drive I would set it globally if possible otherwise setting it on each interface is probably the closest behavior match? | 15:59 |
| prometheanfire | it also allows for having multiple default routes and allows for complex routing | 15:59 |
| clarkb | with debuntu you can set the dns config globally iirc | 16:00 |
| *** ssbarnea has quit IRC | 16:00 | |
| clarkb | not sure if resolvd/networkd do that | 16:00 |
| prometheanfire | clarkb: ya, globally is done in /etc/systemd/resolved.conf though (which is likely fine if we only replace #FallbackDNS= with the values from glean) | 16:00 |
| prometheanfire | networkd doesn't do dns on it's own | 16:01 |
| *** derekh has quit IRC | 16:01 | |
| *** finucannot is now known as stephenfin | 16:01 | |
| *** panda has quit IRC | 16:01 | |
| prometheanfire | actually it has a DNS= there, we could use that | 16:01 |
| prometheanfire | A space-separated list of IPv4 and IPv6 addresses to use as system DNS servers. DNS requests are sent to one of the listed DNS servers in parallel to suitable per-link DNS servers acquired from systemd-networkd.service(8) or set at runtime by external applications. For compatibility reasons, if this | 16:02 |
| prometheanfire | setting is not specified, the DNS servers listed in /etc/resolv.conf are used instead, if that file exists and any servers are configured in it. This setting defaults to the empty list. | 16:02 |
| *** panda has joined #openstack-infra | 16:02 | |
| clarkb | amorin are you still around? (it is late on a friday I know). Wondering if we should disable nodepool launches on ovh gra1 while its a bit unstable and if that is a known issue on your side (happy to help debug from our side if we can) | 16:02 |
| prometheanfire | so, write to both /etc/resolv.conf and the DNS= line in /etc/systemd/resolved.conf | 16:02 |
| clarkb | prometheanfire: that sounds like what we want for the global values | 16:03 |
| *** ginopc has quit IRC | 16:04 | |
| prometheanfire | yep | 16:04 |
| *** janki has quit IRC | 16:04 | |
| prometheanfire | too bad I can't do lineinfile in glean :P | 16:04 |
| clarkb | the systemd ini configs are probably readable by config parser? (I have never tested this) | 16:05 |
| clarkb | if that works its stdlib so no deps and should make editing the configs simpler than direct string manipulation | 16:05 |
| pabelanger | clarkb: at one point, devstack jobs (I think) had a custom index.html file on logs.o.o, but I cannot seem to find that. Is that still a thing in zuulv3? | 16:06 |
| prometheanfire | hmm, probably, I'll have to see | 16:08 |
| clarkb | pabelanger: yes it is still athing, looks like devstack-gate jobs have it but not the new jobs | 16:08 |
| clarkb | pabelanger: I'm guessing that was missed in the zuulv3ification | 16:09 |
| pabelanger | thanks | 16:09 |
| clarkb | pabelanger: I think it has to do with file paths | 16:09 |
| pabelanger | http://logs.openstack.org/53/606853/2/check/legacy-tempest-dsvm-neutron-full-centos-7/8fa2241/ | 16:09 |
| clarkb | pabelanger: the legacy jobs write logs to different paths than the new jobs so the conditional sin't hit? I'm still looking for where we do that | 16:09 |
| openstackgerrit | Merged openstack-infra/zuul master: Check paused parent on node failure of child job https://review.openstack.org/609911 | 16:10 |
| clarkb | pabelanger: the content comes from devstack-gate/help/* | 16:10 |
| openstackgerrit | Merged openstack-infra/zuul master: Optionally disable service worker in zuul-web https://review.openstack.org/609626 | 16:10 |
| pabelanger | clarkb: got it thanks! I think once we switch openstack to use upload-logs-swift role, we maybe look to see how we can allow jobs to add their own html content too. | 16:11 |
| clarkb | pabelanger: http://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/manifests/static.pp#n210 is the config that consumes those "logs" | 16:11 |
| pabelanger | ++ | 16:12 |
| *** bobh has quit IRC | 16:12 | |
| *** eernst has joined #openstack-infra | 16:14 | |
| hogepodge | I was hoping to get some infra eyes on this patch. It changes behavior in the gate that's mean to "save time and bandwidth". I want to make sure it's not putting undue load on infra if it merges. https://review.openstack.org/#/c/609646/8 | 16:15 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Run dstat and generate graphs in unit tests https://review.openstack.org/610100 | 16:17 |
| *** jpich has quit IRC | 16:17 | |
| *** hansmoleman is now known as mriedem_away | 16:18 | |
| clarkb | hogepodge: it should be fine particularly if you are using the pypi caching mirrors in the local cloud region | 16:19 |
| *** ykarel__ has quit IRC | 16:19 | |
| evrardjp | clarkb: infra has reverse proxy for PyPI now, or do we still mirror? | 16:22 |
| evrardjp | maybe the reverse proxy are caching? I am confused by your sentence :) | 16:23 |
| clarkb | evrardjp: we switched to a reverse proxy for pypi when it became untenable to keep mirroring pypi with the gigabytes of machine learning packages being pushed there every day | 16:23 |
| evrardjp | but yeah it seems it is using the infra mirrors anyway | 16:23 |
| clarkb | evrardjp: and yes the reverse proxies cache | 16:23 |
| *** bobh has joined #openstack-infra | 16:23 | |
| evrardjp | ok that clarifies my understanding :) | 16:24 |
| *** rlandy|biab is now known as rlandy | 16:24 | |
| evrardjp | or confirms whatever the right word is. | 16:24 |
| clarkb | evrardjp: mirroring pypi requires more than 2TB of disk now and is growing at many gigabytes per day. | 16:24 |
| clarkb | so we gave up | 16:24 |
| *** tknp has joined #openstack-infra | 16:24 | |
| *** tknp has left #openstack-infra | 16:24 | |
| *** ssbarnea has joined #openstack-infra | 16:25 | |
| clarkb | https://pypi.org/project/tf-nightly-gpu/#files packages like that | 16:25 |
| clarkb | that one package is about a gigabyte per day | 16:25 |
| *** fried_rice is now known as fried_rolls | 16:27 | |
| evrardjp | internet scale! | 16:29 |
| *** eernst has quit IRC | 16:32 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add reference to svg icon in the PWA manifest https://review.openstack.org/610105 | 16:33 |
| *** eernst has joined #openstack-infra | 16:34 | |
| *** eernst has quit IRC | 16:34 | |
| *** eernst has joined #openstack-infra | 16:34 | |
| *** irclogbot_2 has joined #openstack-infra | 16:35 | |
| prometheanfire | ok, got the networkd side done | 16:35 |
| *** shardy has quit IRC | 16:37 | |
| openstackgerrit | Matthew Thode proposed openstack-infra/glean master: write dns info to networkd https://review.openstack.org/610107 | 16:42 |
| *** irclogbot_2 has quit IRC | 16:42 | |
| prometheanfire | clarkb, odyssey4me: part one ^ | 16:42 |
| clarkb | prometheanfire: that doesn't handle the global case? | 16:44 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add icons in the PWA manifest https://review.openstack.org/610105 | 16:46 |
| prometheanfire | clarkb: not yet, I just wanted to get some code up, that's why I -W'd it | 16:47 |
| clarkb | gotcha | 16:47 |
| prometheanfire | that's also the hard part, but if config parser works it should be easier | 16:48 |
| clarkb | looks like packethost mirror crashed again. This implies it isn't a disk issue, or the attempt to fully provision the disk did not work | 16:56 |
| clarkb | the last thing in syslog is ansible running | 16:58 |
| openstackgerrit | Merged openstack-infra/nodepool master: Add universal flag to setup.cfg https://review.openstack.org/610057 | 17:04 |
| clarkb | jobs are failing in pre when that happens so I think the impact is not catastrophic, but still would be good to understand it | 17:05 |
| *** Bhujay has joined #openstack-infra | 17:07 | |
| *** SpamapS has quit IRC | 17:13 | |
| *** SpamapS has joined #openstack-infra | 17:15 | |
| *** trown is now known as trown|lunch | 17:16 | |
| prometheanfire | I suppose I should only write out to /etc/systemd/resolved.conf if that file already exists (aka if resolved is installed) | 17:17 |
| *** e0ne has joined #openstack-infra | 17:28 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/zuul master: WIP: Run dstat and generate graphs in unit tests https://review.openstack.org/610100 | 17:32 |
| *** lujinluo has joined #openstack-infra | 17:38 | |
| lujinluo | Hi, does anyone know who I should contact regarding my account in review.openstack.org? please kindly advise, thanks | 17:44 |
| clarkb | lujinluo: I can help. What would you like to know? | 17:44 |
| lujinluo | thanks clarkb! | 17:45 |
| lujinluo | i have an account with email address lujinluo at gmail.com | 17:45 |
| lujinluo | but people are having trouble adding that account as reviewers on review.openstack.org | 17:45 |
| lujinluo | the error message is "Lujin Luo <lujinluo@gmail.com> does not identify a registered user or group" | 17:46 |
| clarkb | lujinluo: the issue is you have two accounts with that email address so that key isn't unique | 17:46 |
| clarkb | lujinluo: if you give me a few minutes I can dig up details on the two accounts. While I am doing that can you tell me what account id number you have at https://review.openstack.org/#/settings/ | 17:47 |
| lujinluo | I see. on my side, the one that i am using is with Account ID 28581 | 17:47 |
| lujinluo | thanks! clarkb | 17:48 |
| *** Bhujay has quit IRC | 17:50 | |
| clarkb | lujinluo: is the username that you push code with 'lujin' ? | 17:51 |
| lujinluo | yes | 17:51 |
| clarkb | ok 28765 is the other account. What we can do is update the openid on that account to be for 28581, then disable 28765 | 17:52 |
| clarkb | I'll work on getting that done | 17:52 |
| lujinluo | thank you so much for your help! clarkb | 17:52 |
| clarkb | mordred: ^ maybe you can stick around for a few minute to be extra eyeballs on that since its sql update. I am going to get prepped and have a thing for you in a few minutes | 17:52 |
| AJaeger | config-core, please review https://review.openstack.org/609909 https://review.openstack.org/609958 https://review.openstack.org/609795 https://review.openstack.org/609778 | 17:52 |
| *** armax has joined #openstack-infra | 17:55 | |
| clarkb | mordred: corvus: UPDATE account_external_ids SET account_id = '28581' WHERE account_id = '28765'; | 17:56 |
| clarkb | does that look right? | 17:56 |
| mordred | clarkb: yes, I thnk so? | 17:58 |
| clarkb | AJaeger: on https://review.openstack.org/#/c/609958/1 the new zuulv3 job shouldn't use devstack-gate at all can we just remove that job from d-g? | 17:59 |
| clarkb | mordred: its cool to update the column being selected on? | 17:59 |
| *** ssbarnea_ has joined #openstack-infra | 18:00 | |
| Shrews | clarkb: yup | 18:00 |
| clarkb | Shrews: mordred ok cool running that now | 18:00 |
| clarkb | looks like it worked and only updated the one row | 18:02 |
| clarkb | lujinluo: I think you should be all set now | 18:03 |
| lujinluo | awesome. clarkb let me do a quick double check | 18:03 |
| lujinluo | cool clarkb yeah! i can now add that email address | 18:04 |
| *** armax has quit IRC | 18:05 | |
| lujinluo | clarkb: if it is not too much, would it be possible to delete/hide my old account? i no long have access to that account | 18:05 |
| lujinluo | it would be fine if it is too hard | 18:05 |
| clarkb | lujinluo: yes I disabled the old account, that was necessary to get the email to work as a key | 18:07 |
| clarkb | lujinluo: I moved the openid from the old ccount to the current account with the db update, then disabled the old account | 18:07 |
| lujinluo | sorry i was not clear enough. i have another account with another email address, luo.lujin at jp.fujitsu.com | 18:08 |
| lujinluo | all three of them share the same "Lujin Luo" | 18:08 |
| lujinluo | entry | 18:08 |
| clarkb | oh | 18:08 |
| lujinluo | so people always get confused which one to pick | 18:09 |
| openstackgerrit | Merged openstack-infra/project-config master: Use root user for editing pip.conf https://review.openstack.org/609909 | 18:09 |
| clarkb | lujinluo: that account is 17491 | 18:09 |
| lujinluo | since i no longer have access to the previous company email address, i miss out all the review requests go to that account | 18:09 |
| lujinluo | yes clarkb | 18:10 |
| clarkb | lujinluo: do we need to move the openid for that one too? that one is a bit harder because the email addresses differ | 18:10 |
| clarkb | or is it enough to just disable the account? we can reenable it later if necessary too | 18:10 |
| cgoncalves | hi. looking for a DIB core to review https://review.openstack.org/#/c/609169/ :) | 18:10 |
| lujinluo | if it is possible let's move the openid | 18:10 |
| clarkb | lujinluo: its possible just not sure if I have to sanitize the email too | 18:11 |
| *** eernst has quit IRC | 18:11 | |
| clarkb | fungi probably knows but isn't around today. Maybe we can double check with him when he is back (hopefully next week) | 18:12 |
| lujinluo | clarkb: ok, got it. | 18:12 |
| lujinluo | thanks for your help today clarkb! :-) | 18:12 |
| lujinluo | have a good day! | 18:13 |
| clarkb | lujinluo: you too | 18:13 |
| *** eernst has joined #openstack-infra | 18:13 | |
| *** trown|lunch is now known as trown | 18:14 | |
| *** eernst has quit IRC | 18:14 | |
| clarkb | not sure if anyone else has noticed this but firefox seems to be rejecting ssl certs on some sites and it isn't clear why. The UI renders a green padlock. I haven't noticed this on any of our sites, but if found should probably let us know | 18:14 |
| pabelanger | clarkb: which version of firefox? I remember seeing something about latest version dropping old cert providers | 18:15 |
| clarkb | 63 beta 11 | 18:16 |
| AJaeger | clarkb: sure, can remove it... | 18:17 |
| openstackgerrit | Merged openstack-infra/zuul master: Add __init__.py to web directory https://review.openstack.org/610080 | 18:20 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Remove legacy-tempest-dsvm-neutron-pg-full https://review.openstack.org/609958 | 18:20 |
| AJaeger | clarkb: done ^ | 18:20 |
| clarkb | AJaeger: thanks | 18:20 |
| *** olivierb has quit IRC | 18:21 | |
| *** eernst has joined #openstack-infra | 18:22 | |
| *** eernst has quit IRC | 18:23 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Simplify Builds page https://review.openstack.org/610137 | 18:30 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add more information to build page https://review.openstack.org/610138 | 18:30 |
| openstackgerrit | Merged openstack-infra/openstack-zuul-jobs master: Remove obsolete legacy-tempest-dsvm-py35-full-devstack-plugin-ceph https://review.openstack.org/609795 | 18:30 |
| *** eharney has quit IRC | 18:34 | |
| *** e0ne has quit IRC | 18:37 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Cleanup devstack-gate jobs https://review.openstack.org/609958 | 18:40 |
| AJaeger | clarkb: found some more jobs that we can remove and update my change ^ | 18:40 |
| *** smarcet has quit IRC | 18:41 | |
| *** eharney has joined #openstack-infra | 18:45 | |
| *** ijw has quit IRC | 18:45 | |
| *** kmalloc is now known as needscoffee | 18:48 | |
| *** bhavikdbavishi has quit IRC | 18:51 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Run dstat and generate graphs in unit tests https://review.openstack.org/610100 | 18:51 |
| clarkb | needscoffee: ++ to healthcheck by default rather than configurable addon. That is all | 18:55 |
| needscoffee | clarkb: ^_^ | 18:57 |
| needscoffee | clarkb: i hate it when things are "configurable" like that. and it seemed useful. | 18:58 |
| needscoffee | that email thread just confirmed my choice was a good one. | 18:58 |
| clarkb | it seems odd to support such a feature and not turn it on by default | 18:59 |
| needscoffee | well we had it in our default paste-ini | 19:02 |
| needscoffee | 100% agree | 19:02 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Create ansible-galaxy-import role https://review.openstack.org/608472 | 19:02 |
| AJaeger | clarkb: could you review https://review.openstack.org/#/c/609778/ as well, please? | 19:08 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Remove unneeded cruft from bdist_wheel fix attempts https://review.openstack.org/610143 | 19:08 |
| *** fried_rolls is now known as fried_rice | 19:10 | |
| *** rkukura has joined #openstack-infra | 19:16 | |
| AJaeger | thanks, clarkb | 19:18 |
| openstackgerrit | David Moreau Simard proposed openstack-infra/zuul-jobs master: Add a role to install and configure dstat to run in the background https://review.openstack.org/518374 | 19:20 |
| openstackgerrit | David Moreau Simard proposed openstack-infra/zuul-jobs master: Add a role to install and configure dstat to run in the background https://review.openstack.org/518374 | 19:20 |
| openstackgerrit | Merged openstack-infra/project-config master: Small cleanups for zuul.d/projects.yaml https://review.openstack.org/609778 | 19:21 |
| *** lujinluo has quit IRC | 19:27 | |
| *** eernst has joined #openstack-infra | 19:32 | |
| mwhahaha | I'm looking to convert a legacy job to v3, what's the prefered way to copy the repo data into the workspace or consume the stuff that was checked out? Or is it best to just keep using zuul-cloner | 19:34 |
| clarkb | mwhahaha: zuul does that automatically for you in v3 | 19:35 |
| AJaeger | mwhahaha: get rid of zuul-cloner! | 19:35 |
| mwhahaha | clarkb: yea i know it checks it out, but does it drop in in /home/zuul/workspace/<repo> ? | 19:35 |
| clarkb | mwhahaha: /home/zuul/src/$hostname/path/to/repo (it is encoded in a zuul ansible var too) | 19:35 |
| * mwhahaha is trying to understand the pathing | 19:35 | |
| *** eernst has quit IRC | 19:35 | |
| clarkb | it might be workspace/src. Its a ansible variable /me looks it up | 19:35 |
| mordred | mwhahaha: it's done in the src dir of the zuul home dir on the remote node in golang format | 19:36 |
| mordred | mwhahaha: so - /home/zuul/src/git.openstack.org/openstack/nova home/zuul/src/git.openstack.org/openstack/swift etc | 19:36 |
| clarkb | mwhahaha: http://logs.openstack.org/07/599607/2/check/openstack-zuul-jobs-linters/a3ecca5/zuul-info/inventory.yaml I loke to look at an example inventory file. zuul.project.src_dir | 19:37 |
| *** electrofelix has quit IRC | 19:37 | |
| AJaeger | "{{ ansible_user_dir }}/{{ zuul.projects['git.openstack.org/openstack/tempest'].src_dir }}" | 19:37 |
| mwhahaha | yea it's kinda annoying that you have to know the full url in the zuul.projects[] data | 19:37 |
| *** dhill_ has quit IRC | 19:37 | |
| mwhahaha | but thanks i'll try to use that instead | 19:37 |
| *** rkukura has quit IRC | 19:38 | |
| clarkb | mwhahaha: not sure I understand? its given to you in the src_dir var? | 19:38 |
| mordred | well, it's also predictable, so you don't _have_ to use that zuul.projects[] data | 19:38 |
| mordred | clarkb: src_dir only works for a single repo | 19:38 |
| clarkb | mordred: no its in there for each projects repo too | 19:38 |
| clarkb | so you can loop over it | 19:38 |
| mordred | good point | 19:38 |
| mwhahaha | zuul.projects['git.openstack.org/openstack/python-tripleoclient'].src_dir | 19:38 |
| AJaeger | clarkb: or use the above as I pointed out above | 19:39 |
| mwhahaha | that's one hell of a variable :D | 19:39 |
| mordred | yah. I guess it just depends on what you're wanting to do | 19:39 |
| mordred | if you find yourself writing zuul.projects['git.openstack.org/openstack/python-tripleoclient'].src_dir in a playbook - you're writing too many characters :) | 19:39 |
| mordred | but yes - it's totally there for the iterating | 19:39 |
| AJaeger | mwhahaha: feel free to show a change and ask for review here... | 19:40 |
| mordred | ++ | 19:40 |
| mwhahaha | i shall, just trying to figure out where the data is :) | 19:40 |
| AJaeger | mordred: time for a quick cleanup change review, please? https://review.openstack.org/609958 | 19:40 |
| * AJaeger waves good bye | 19:41 | |
| *** rkukura has joined #openstack-infra | 19:41 | |
| *** dhill_ has joined #openstack-infra | 19:41 | |
| mordred | AJaeger: done! have a good weekend | 19:42 |
| AJaeger | mordred: thanks - and thanks ;) | 19:43 |
| clarkb | loop: zuul.projects | dict2items # this doesn't exist until ansible 2.6 :/ then item.value.src_dir is your path? | 19:44 |
| clarkb | in 2.5 which zuul runs I guess you use with_something | 19:45 |
| *** apetrich has quit IRC | 19:48 | |
| *** apetrich has joined #openstack-infra | 19:49 | |
| smcginnis | If anyone other than Andreas has a moment to take a look at https://review.openstack.org/#/c/609827/ - I have some follow up work I want to do once I've been able to fully test that. | 19:49 |
| openstackgerrit | Merged openstack-infra/project-config master: Cleanup devstack-gate jobs https://review.openstack.org/609958 | 19:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add line to setup.cfg limiting to python >=3.5 https://review.openstack.org/610089 | 19:54 |
| *** rkukura has quit IRC | 19:55 | |
| clarkb | smcginnis: done. Keep in mind its hard to undo that with pbr once you've take the plunge | 19:55 |
| clarkb | (so maybe test with a repo where specific version numbers don't matter too much) | 19:56 |
| smcginnis | clarkb: Thanks! | 19:56 |
| smcginnis | Yeah, going to test with release-test before I attempt anything else there. | 19:56 |
| smcginnis | I suppose an attempt to revert the commit would just double the problem. | 19:56 |
| smcginnis | Assuming they don't remove the text from the commit. | 19:56 |
| clarkb | ya I want to say when zuul ran into an improper/unwated pbr version tag that we had to do a hard reset? | 19:57 |
| clarkb | corvus probably remembers | 19:57 |
| smcginnis | Seems like the only way to clear something like that out. | 19:57 |
| smcginnis | Kinda scary a little inconspicuous text in a commit message has the potential to cause so much trouble. :) | 19:58 |
| *** eernst has joined #openstack-infra | 20:02 | |
| corvus | yes. we took the extraordinary step of altering git history. | 20:10 |
| corvus | smcginnis, dhellmann: i can't recommend *against* using the Sem-Ver headers enough. in short, i think it's a really bad idea to encode actionable content like that in commit messages. the primary benefit of version control systems is that any change can be undone. but *that* change *can not* be undone. it's terrifying. | 20:12 |
| smcginnis | corvus: Any suggestions how we can increment version for those that want to be able to test upgrades to master after we cut stable branches? | 20:13 |
| clarkb | smcginnis: pip install -U? I'm not sure why the version matters for that | 20:14 |
| smcginnis | corvus: The idea with this is when we create stable branches, we have the auto generated patches that update reno and the stable .gitreview, etc. This would just be included with that. | 20:14 |
| smcginnis | clarkb: I don't think distro packagers can do that. | 20:14 |
| corvus | smcginnis: nope, i'm sorry i haven't engaged with the problem and i have to run now. i'll be happy to chat more next week, i just felt that since clarkb invoked my experience with the semver headers, i should share my 2-cents in case the lack of "undo" functionality had not occurred to you (it certainly hadn't to others, including me, until i saw the results) | 20:15 |
| smcginnis | corvus: No problem, thanks for that. | 20:15 |
| clarkb | smcginnis: ya its probably harder for distros since their versions have proper meaning typically. I'm not sure how you'd tackle that in a distro context | 20:16 |
| openstackgerrit | Merged openstack-infra/project-config master: update the branching script to include a semver bump https://review.openstack.org/609827 | 20:16 |
| smcginnis | clarkb: Doug detailed some of the challenges with this on the ML. Let me see if I can dig that up. | 20:16 |
| *** rkukura has joined #openstack-infra | 20:16 | |
| smcginnis | http://lists.openstack.org/pipermail/openstack-dev/2018-October/135706.html | 20:17 |
| *** ssbarnea_ has quit IRC | 20:17 | |
| smcginnis | I'm certainly open to other suggestions, but even with the inherent risks, this seemed like the best solution to the issue. | 20:17 |
| *** needscoffee is now known as kmalloc | 20:20 | |
| openstackgerrit | Merged openstack-infra/nodepool master: Add line to setup.cfg limiting to python >=3.5 https://review.openstack.org/610090 | 20:21 |
| openstackgerrit | Merged openstack-infra/zuul master: Quick-start: remove ansible Dockerfile https://review.openstack.org/609555 | 20:21 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool master: Run zuul-quick-start job https://review.openstack.org/610159 | 20:35 |
| *** ianychoi has joined #openstack-infra | 20:35 | |
| *** rkukura has quit IRC | 20:37 | |
| *** lujinluo has joined #openstack-infra | 20:37 | |
| *** ansmith has quit IRC | 20:40 | |
| *** eharney has quit IRC | 20:40 | |
| *** rkukura has joined #openstack-infra | 20:41 | |
| *** bobh has quit IRC | 20:48 | |
| openstackgerrit | Merged openstack-infra/zuul master: Build new container images for quick-start https://review.openstack.org/609763 | 20:51 |
| *** mriedem_away is now known as mriedem | 20:51 | |
| openstackgerrit | Merged openstack-infra/zuul master: Add line to setup.cfg limiting to python >=3.5 https://review.openstack.org/610089 | 20:52 |
| *** rkukura has quit IRC | 20:56 | |
| *** dave-mccowan has quit IRC | 20:57 | |
| *** rkukura has joined #openstack-infra | 21:00 | |
| *** trown is now known as trown|outtypewww | 21:00 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul-jobs master: Fix typo with ansible_galaxy_info.token https://review.openstack.org/610164 | 21:00 |
| openstackgerrit | Merged openstack-infra/zuul master: Quick-start: switch to polygerrit https://review.openstack.org/609828 | 21:08 |
| tonyb | dhellmann: Thanks. | 21:22 |
| *** jamesdenton has quit IRC | 21:24 | |
| openstackgerrit | Merged openstack-infra/zuul master: Remove unneeded cruft from bdist_wheel fix attempts https://review.openstack.org/610143 | 21:27 |
| openstackgerrit | Matt McEuen proposed openstack-infra/project-config master: New Airship project - Tempest Plugin https://review.openstack.org/610175 | 21:30 |
| *** PapaOurs is now known as bauzas | 21:32 | |
| smcginnis | gerritbot appears to be dead in #openstack-cinder | 21:39 |
| *** rkukura has quit IRC | 21:39 | |
| clarkb | smcginnis: it connects to channels in a least recently used rotation because there are too many channels for it to connect to directly anymore | 21:40 |
| clarkb | smcginnis: have there been missed events? or is it just gone? if no missed events and just gone it should join and emit the next event | 21:40 |
| clarkb | and I think the sasl auth is in place now so it shouldn't have the auth race issues with freenode +r rules anymore | 21:40 |
| smcginnis | Missed events. Or at least severely delayed. | 21:40 |
| smcginnis | Not a big deal, just noticed there wasn't an announcement for a patch I pushed up. | 21:41 |
| smcginnis | I'll check again on Monday. | 21:41 |
| *** lujinluo has quit IRC | 21:42 | |
| clarkb | smcginnis: 2018-10-12 21:38:32,626 INFO gerritbot: Sending "Sean McGinnis proposed openstack/cinder master: Allow setting volume readonly during creating https://review.openstack.org/610177" to #openstack-cinder | 21:42 |
| clarkb | it thinks it is sending things to the channel | 21:42 |
| smcginnis | Huh, nothing's shown up in channel. Unless irssi is blocking it somehow. | 21:42 |
| smcginnis | Anyway, gotta run. Thanks. | 21:43 |
| *** ansmith has joined #openstack-infra | 21:43 | |
| *** boden has quit IRC | 21:43 | |
| *** rkukura has joined #openstack-infra | 21:43 | |
| clarkb | ya eavesdrop doesn't see it either | 21:43 |
| clarkb | my current guess is we didn't restart it after the sasl change went in because that happened after the config updates | 21:47 |
| clarkb | but finding in the logs where it started is not easy | 21:47 |
| *** agopi_ has joined #openstack-infra | 21:47 | |
| *** lujinluo has joined #openstack-infra | 21:49 | |
| *** agopi has quit IRC | 21:49 | |
| clarkb | no ps says it started after the install should've happened | 21:50 |
| clarkb | fungi: ^ sasl may not have solved the problems. I wonder if there is something buggy about the channel joiing | 21:51 |
| *** rkukura has quit IRC | 21:52 | |
| prometheanfire | clarkb: is it ok if comments are lost from the resolved.conf file? | 21:53 |
| clarkb | prometheanfire: probably? usually the comments are "don't edit this by hand" | 21:54 |
| *** rlandy has quit IRC | 21:54 | |
| clarkb | grepping gerritbot logs I see it joining the cinder channel multiple times but never parting it (could be due to restarts) it should be connected to the channel? | 21:54 |
| *** jistr has quit IRC | 21:55 | |
| prometheanfire | clarkb: ok, because configparser drops all comments which sucks, and configobj isn't a python builtin | 21:56 |
| *** jistr has joined #openstack-infra | 21:56 | |
| *** tosky has quit IRC | 22:02 | |
| *** jistr has quit IRC | 22:05 | |
| *** agopi_ is now known as agopi | 22:05 | |
| *** mriedem is now known as mriedem_away | 22:05 | |
| *** jistr has joined #openstack-infra | 22:06 | |
| clarkb | I'm pretty stumped on this channel issue since the join happens long after the connection to the server but it would've been auth'd via sasl anyway | 22:12 |
| clarkb | the only thing I can figure is that freenode didn't get the message? | 22:12 |
| *** agopi is now known as agopi|brb | 22:13 | |
| openstackgerrit | Matthew Thode proposed openstack-infra/glean master: write dns info to networkd https://review.openstack.org/610107 | 22:16 |
| *** agopi|brb has quit IRC | 22:17 | |
| prometheanfire | clarkb, mordred, odyssey4me: ^ reviewable | 22:18 |
| prometheanfire | huh, no zuul dir for glean | 22:19 |
| *** pcaruana has quit IRC | 22:20 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Fix typo with ansible_galaxy_info.token https://review.openstack.org/610164 | 22:26 |
| *** eernst has quit IRC | 22:30 | |
| *** pbourke has quit IRC | 22:32 | |
| *** pbourke has joined #openstack-infra | 22:32 | |
| *** eernst has joined #openstack-infra | 22:33 | |
| *** eernst has quit IRC | 22:37 | |
| *** eernst has joined #openstack-infra | 22:37 | |
| *** eernst_ has joined #openstack-infra | 22:39 | |
| *** eernst has quit IRC | 22:39 | |
| *** eernst_ has quit IRC | 22:43 | |
| *** kjackal_v2 has quit IRC | 22:56 | |
| *** elbragstad has quit IRC | 22:56 | |
| *** sthussey has quit IRC | 23:03 | |
| *** elbragstad has joined #openstack-infra | 23:15 | |
| *** elbragstad has quit IRC | 23:15 | |
| *** owalsh_ has joined #openstack-infra | 23:18 | |
| *** kgiusti has left #openstack-infra | 23:18 | |
| *** owalsh has quit IRC | 23:21 | |
| *** agopi|brb has joined #openstack-infra | 23:25 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/project-config master: Disable packethost due to mirror outages https://review.openstack.org/610187 | 23:28 |
| clarkb | the mirror has been shutdown again. I think we should go ahead and do ^ | 23:28 |
| *** slaweq has quit IRC | 23:32 | |
| *** agopi|brb has quit IRC | 23:41 | |
| mwhahaha | speaking of mirrors, do we have the cloud images (like the centos one) available via a mirror? | 23:51 |
| *** gyee has quit IRC | 23:52 | |
| mwhahaha | i see the fedora one, but not centos | 23:52 |
| clarkb | mwhahaha: the fedora atomic images at http://mirror.dfw.rax.openstack.org/fedora/atomic/stable/ are the only ones I know of | 23:52 |
| prometheanfire | not sure who added gentoo to the glean gate, but, it's not passing | 23:52 |
| prometheanfire | http://logs.openstack.org/07/610107/2/check/nodepool-functional-py35-gentoo-src/ccdd739/job-output.txt.gz#_2018-10-12_22_42_09_829955 | 23:52 |
| prometheanfire | stuck on waitforimage | 23:53 |
| clarkb | prometheanfire: check the nodepool builder logs | 23:53 |
| mwhahaha | clarkb: yea the regular fedora one is there too http://mirror.bhs1.ovh.openstack.org/fedora/releases/28/Cloud/x86_64/images/ | 23:53 |
| prometheanfire | clarkb: images built successfully | 23:54 |
| prometheanfire | if that's what you are talking about | 23:54 |
| clarkb | waitforimage is waiting for the image to build and upload. | 23:55 |
| clarkb | http://logs.openstack.org/07/610107/2/check/nodepool-functional-py35-gentoo-src/ccdd739/controller/nodepool/nodepool-image-list.txt does not show the image being ready | 23:56 |
| clarkb | http://logs.openstack.org/07/610107/2/check/nodepool-functional-py35-gentoo-src/ccdd739/controller/logs/screen-nodepool-builder.txt.gz#_Oct_12_23_42_31_463377 that makes me think the build failed | 23:56 |
| *** EvilienM is now known as EmilienM | 23:57 | |
| clarkb | http://logs.openstack.org/07/610107/2/check/nodepool-functional-py35-gentoo-src/ccdd739/controller/logs/builds/gentoo-17-0-systemd-0000000003_log.txt.gz#_2018-10-12_23_42_24_086 is the error maybe? | 23:57 |
| prometheanfire | I think I have a review for that... | 23:58 |
| prometheanfire | somewhere, I added sudo to packages.yaml | 23:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!