| *** thorst has joined #openstack-ansible | 00:16 | |
| *** thorst has quit IRC | 00:21 | |
| *** Oku_OS-away has joined #openstack-ansible | 00:22 | |
| *** mancdaz has joined #openstack-ansible | 00:22 | |
| *** yifei has joined #openstack-ansible | 00:23 | |
| *** woodard_ has quit IRC | 00:48 | |
| *** woodard has joined #openstack-ansible | 00:48 | |
| *** sboyron has quit IRC | 00:50 | |
| *** thorst has joined #openstack-ansible | 00:52 | |
| *** thorst has quit IRC | 00:57 | |
| *** thorst has joined #openstack-ansible | 01:27 | |
| *** thorst has quit IRC | 01:28 | |
| *** thorst has joined #openstack-ansible | 01:29 | |
| *** thorst has quit IRC | 01:30 | |
| *** openstackgerrit has joined #openstack-ansible | 01:32 | |
| openstackgerrit | Kaio Kassiano Moura Oliveira proposed openstack/openstack-ansible-os_monasca-agent master: Fix default role missing pip packages https://review.openstack.org/525022 | 01:32 |
|---|---|---|
| *** thorst has joined #openstack-ansible | 02:01 | |
| *** hw_wutianwei has joined #openstack-ansible | 02:04 | |
| *** thorst has quit IRC | 02:06 | |
| *** thorst has joined #openstack-ansible | 02:10 | |
| *** thorst has quit IRC | 02:14 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_designate master: Fix ansible-lint test failures https://review.openstack.org/525031 | 02:47 |
| *** thorst has joined #openstack-ansible | 03:17 | |
| *** thorst has quit IRC | 03:22 | |
| *** chhavi has joined #openstack-ansible | 03:40 | |
| *** thorst has joined #openstack-ansible | 03:57 | |
| *** cuongnv has joined #openstack-ansible | 03:59 | |
| *** thorst has quit IRC | 04:02 | |
| *** weezS has joined #openstack-ansible | 04:13 | |
| *** weezS has quit IRC | 04:19 | |
| *** udesale has joined #openstack-ansible | 04:25 | |
| *** udesale has quit IRC | 04:25 | |
| *** udesale has joined #openstack-ansible | 04:25 | |
| *** bhujay has joined #openstack-ansible | 04:29 | |
| *** dasTor has quit IRC | 04:35 | |
| *** dasTor has joined #openstack-ansible | 04:35 | |
| *** thorst has joined #openstack-ansible | 04:37 | |
| *** thorst has quit IRC | 04:41 | |
| *** m|y|k has joined #openstack-ansible | 04:42 | |
| *** m|y|k has quit IRC | 04:42 | |
| *** thorst has joined #openstack-ansible | 05:10 | |
| *** thorst has quit IRC | 05:15 | |
| *** andymccr has quit IRC | 05:32 | |
| *** andymccr has joined #openstack-ansible | 05:33 | |
| *** cloudnull has quit IRC | 05:33 | |
| *** thorst has joined #openstack-ansible | 05:37 | |
| *** cloudnull has joined #openstack-ansible | 05:41 | |
| *** thorst has quit IRC | 05:42 | |
| *** thorst has joined #openstack-ansible | 05:45 | |
| *** thorst has quit IRC | 05:50 | |
| *** hybridpolio has quit IRC | 06:00 | |
| *** thorst has joined #openstack-ansible | 06:24 | |
| *** thorst has quit IRC | 06:29 | |
| *** jbadiapa has joined #openstack-ansible | 06:33 | |
| *** gkadam_ has quit IRC | 06:47 | |
| *** thorst has joined #openstack-ansible | 06:57 | |
| *** thorst has quit IRC | 07:01 | |
| *** mardim has quit IRC | 07:03 | |
| jafeha | good morning | 07:07 |
| *** jvidal has joined #openstack-ansible | 07:14 | |
| *** dayou has quit IRC | 07:14 | |
| *** cshen has joined #openstack-ansible | 07:23 | |
| *** askb_ has joined #openstack-ansible | 07:25 | |
| *** askb_ has quit IRC | 07:29 | |
| *** michelv has joined #openstack-ansible | 07:31 | |
| *** jafeha has quit IRC | 07:33 | |
| *** indistylo has joined #openstack-ansible | 07:33 | |
| *** askb_ has joined #openstack-ansible | 07:34 | |
| *** thorst has joined #openstack-ansible | 07:36 | |
| *** thorst has quit IRC | 07:41 | |
| *** mardim has joined #openstack-ansible | 07:44 | |
| *** rgogunskiy has joined #openstack-ansible | 07:45 | |
| evrardjp | good morning | 07:46 |
| *** mardim has quit IRC | 07:47 | |
| *** shardy has joined #openstack-ansible | 07:48 | |
| *** pmannidi has quit IRC | 07:50 | |
| evrardjp | hwoarang: are you there? I have some proposition for gate issues due to the packages/lxc for centos/suse: what if we remove the suse/centos jobs from gates, but keep them in the check state, as voting. We'd still need to pass there before going further. But we don't double the risks. | 07:55 |
| evrardjp | also, cores: we need to merge this: https://review.openstack.org/#/c/524961/ | 07:56 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-openstack_hosts master: Run pip once and for all https://review.openstack.org/524962 | 07:59 |
| *** jafeha has joined #openstack-ansible | 08:01 | |
| jafeha | good morning again | 08:01 |
| evrardjp | good morning jafeha :) | 08:01 |
| evrardjp | I am clueless about this issue: http://logs.openstack.org/24/504224/23/check/openstack-ansible-deploy-all_metal-ubuntu-xenial/72f8220/job-output.txt.gz#_2017-12-03_16_55_11_162483 | 08:01 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_keystone master: [DNM] debug why on metal fails on gates and not on my machine. https://review.openstack.org/524954 | 08:05 |
| *** mardim has joined #openstack-ansible | 08:08 | |
| *** epalper has joined #openstack-ansible | 08:11 | |
| *** gkadam_ has joined #openstack-ansible | 08:15 | |
| *** mbuil has joined #openstack-ansible | 08:17 | |
| *** sxc731 has joined #openstack-ansible | 08:21 | |
| *** pcaruana has joined #openstack-ansible | 08:23 | |
| openstackgerrit | Periyasamy Palanisamy proposed openstack/openstack-ansible-specs master: blueprint spec for opendaylight bgpvpn support https://review.openstack.org/523171 | 08:27 |
| hwoarang | evrardjp: hello. What gate issues with lxc? What broke again? | 08:33 |
| evrardjp | oh no it's the usual download timeouts things | 08:41 |
| evrardjp | hwoarang: but now that you're there, could you give this https://review.openstack.org/#/c/524961/ a nudge? | 08:42 |
| evrardjp | same for this: https://review.openstack.org/#/c/515405/ | 08:43 |
| evrardjp | both would unblock gates | 08:43 |
| hwoarang | evrardjp: well what timeouts? :( | 08:44 |
| hwoarang | the lxc downloads from images.linuxcountainers.org? or the cache preparation? because the cache preparation was due to mirror changes from pip_install to openstack_hosts that we fixed on friday | 08:44 |
| evrardjp | yes | 08:45 |
| evrardjp | the images | 08:45 |
| evrardjp | but that happens once in a while | 08:45 |
| evrardjp | that's not really a problem, just a minor annoyance | 08:45 |
| evrardjp | I thought this could be a different way of thinking | 08:45 |
| evrardjp | still require a pass, but not two. | 08:46 |
| *** sboyron has joined #openstack-ansible | 08:46 | |
| evrardjp | I think the more urgent is those 2 I gave you, because they are SURE blocking some of my patches :p | 08:46 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-openstack_hosts master: Run pip once and for all https://review.openstack.org/524962 | 08:47 |
| hwoarang | iirc we have some harcoded dns entiries for linuxcontainers.org | 08:48 |
| hwoarang | so we may be missing some server side load balancing :/ | 08:48 |
| evrardjp | well let me paste you a link when this comes again | 08:49 |
| evrardjp | also fun stuff, I think the rabbitmq version bump kinda destroyed some things for suse. | 08:51 |
| evrardjp | http://logs.openstack.org/66/524166/1/check/openstack-ansible-functional-opensuse-423/0cc191a/job-output.txt.gz#_2017-12-03_13_44_48_460254 | 08:52 |
| evrardjp | but right now the priority is to the other two. | 08:52 |
| *** ThomasS has quit IRC | 08:56 | |
| *** vnogin has joined #openstack-ansible | 08:57 | |
| *** armaan has joined #openstack-ansible | 08:59 | |
| *** hamza21 has joined #openstack-ansible | 08:59 | |
| *** vnogin has quit IRC | 09:01 | |
| hwoarang | hmm | 09:05 |
| evrardjp | odyssey4me: thanks for the reviews! | 09:07 |
| evrardjp | and good morning :) | 09:07 |
| evrardjp | hwoarang: I will work on the rabbitmq issue for suse now | 09:08 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate master: Use group_names to check a host belongs to group https://review.openstack.org/523275 | 09:10 |
| hwoarang | let me know if you need help | 09:11 |
| evrardjp | right now I will build my first opensuse on my machine :p | 09:11 |
| *** pbandark has joined #openstack-ansible | 09:12 | |
| *** SerenaFeng has joined #openstack-ansible | 09:13 | |
| *** SerenaFeng has quit IRC | 09:15 | |
| hwoarang | the rabbitmq failure is strange because we install the package from the distro repos | 09:16 |
| hwoarang | not from upstream | 09:16 |
| hwoarang | lets see | 09:16 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Run ARA only if enabled https://review.openstack.org/512216 | 09:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Only build ARA report when test fails https://review.openstack.org/512217 | 09:18 |
| evrardjp | I thought that was previous, and we upgrade to something with packages | 09:19 |
| evrardjp | oh you mean the defaults | 09:19 |
| evrardjp | let me check | 09:19 |
| *** sboyron has quit IRC | 09:21 | |
| *** vnogin has joined #openstack-ansible | 09:24 | |
| hwoarang | yeah | 09:24 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Follow symlinks in the lookup https://review.openstack.org/524961 | 09:25 |
| *** sboyron has joined #openstack-ansible | 09:29 | |
| *** SmearedBeard has joined #openstack-ansible | 09:29 | |
| *** armaan has quit IRC | 09:31 | |
| *** rodolof has joined #openstack-ansible | 09:32 | |
| *** armaan has joined #openstack-ansible | 09:33 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Fix gate overrides https://review.openstack.org/525108 | 09:36 |
| evrardjp | hwoarang: ^ | 09:36 |
| hwoarang | hmm | 09:37 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Use one play for container creation https://review.openstack.org/525109 | 09:37 |
| *** sboyron has quit IRC | 09:38 | |
| hwoarang | evrardjp: it worked before so can you explain a situation where the result is bad? | 09:38 |
| evrardjp | let me rephrase that english thing | 09:38 |
| hwoarang | im curous | 09:38 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins stable/pike: Follow symlinks in the lookup https://review.openstack.org/525111 | 09:40 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins stable/ocata: Follow symlinks in the lookup https://review.openstack.org/525112 | 09:41 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins stable/newton: Follow symlinks in the lookup https://review.openstack.org/525113 | 09:41 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Fix SUSE gate overrides outside gates https://review.openstack.org/525108 | 09:43 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Fix SUSE gate overrides outside gates https://review.openstack.org/525108 | 09:44 |
| evrardjp | hwoarang: better english ^ | 09:44 |
| evrardjp | like what odyssey4me could have said recently, that's less chewbacca-ey | 09:46 |
| evrardjp | I don't remember what he said, but it was very imaged. | 09:46 |
| hwoarang | ah that's better | 09:46 |
| evrardjp | haha. Knew it! | 09:46 |
| openstackgerrit | Periyasamy Palanisamy proposed openstack/openstack-ansible-os_neutron master: add bgpvpn ml2 config support https://review.openstack.org/522598 | 09:47 |
| evrardjp | it all makes sense to me when I see the code in front of me, but I still have trouble to translate in english and project to other ppl minds. It's an effort :p | 09:47 |
| hwoarang | hahah | 09:48 |
| evrardjp | hwoarang: what I have noticed is that the change started to appear between the 1 and 2 of december | 09:49 |
| odyssey4me | evrardjp you'll need to add something like https://review.openstack.org/#/c/515405/ to https://github.com/openstack/openstack-ansible-tests/blob/79e11d35d2749ef99616495cd11942ce8a43a5e6/test-install-openstack-hosts.yml | 09:49 |
| evrardjp | (for rabbitmq, not for my language skills) | 09:49 |
| evrardjp | odyssey4me: true, we are not reusing the integrated for all of this. | 09:50 |
| evrardjp | thanks! | 09:50 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/pike: Remove jinja2 delimiters from when clause https://review.openstack.org/525117 | 09:53 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 09:53 |
| openstackgerrit | Periyasamy Palanisamy proposed openstack/openstack-ansible-os_neutron master: add bgpvpn ml2 config support https://review.openstack.org/522598 | 09:56 |
| *** Smeared_Beard has joined #openstack-ansible | 09:58 | |
| *** SmearedBeard has quit IRC | 09:59 | |
| odyssey4me | evrardjp some feedback in https://review.openstack.org/525118 | 10:00 |
| evrardjp | odyssey4me: do we need the serial 1? | 10:04 |
| evrardjp | in the Playbook for creating containers | 10:04 |
| evrardjp | if we do not, I remove it and it's faster when doing the openstack_hosts | 10:04 |
| odyssey4me | https://review.openstack.org/521598 | 10:05 |
| evrardjp | mmm | 10:06 |
| evrardjp | maybe we should still keep the 2 plays for now, and consolidate on this patch 521598? | 10:06 |
| evrardjp | there doesn't seem to have consensus | 10:06 |
| odyssey4me | which is why you shouldn't mck with that piece | 10:07 |
| odyssey4me | the consensus is that we're happy to drop serial 1, but only if it's explained | 10:07 |
| evrardjp | ok let me adapt my patch real quick | 10:07 |
| odyssey4me | right now it's just being dropped with no reference to the original reason it was put there | 10:07 |
| odyssey4me | most tests don't have that many containers, so series doesn't matter that much | 10:08 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins stable/ocata: Follow symlinks in the lookup https://review.openstack.org/525112 | 10:08 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins stable/newton: Follow symlinks in the lookup https://review.openstack.org/525113 | 10:08 |
| odyssey4me | serial: 1 I mean | 10:08 |
| *** cuongnv has quit IRC | 10:10 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 10:10 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 10:10 |
| *** vnogin has quit IRC | 10:10 | |
| evrardjp | updated. | 10:10 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 10:12 |
| evrardjp | good catch! | 10:12 |
| *** vnogin has joined #openstack-ansible | 10:12 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 10:13 |
| evrardjp | hwoarang: https://review.openstack.org/#/c/525118/5 should be enough to make the rabbitmq work. I've tried an early patchset on my machine and it went further than rabbit. | 10:17 |
| hwoarang | hmm but why... | 10:18 |
| evrardjp | odyssey4me: I don't know if you've seen but we've finally bumped pike sha's. It seems the sweet spot is around 2h15. | 10:19 |
| evrardjp | of run time | 10:19 |
| evrardjp | hwoarang: because this way we ensure we apply repo management to containers too? | 10:19 |
| evrardjp | I will have patches to improve the lxc_container_create and lxc_hosts | 10:19 |
| evrardjp | https://review.openstack.org/#/c/524865/ is for uca file management, so we can use the uca file when copying data from hosts to containers | 10:20 |
| hwoarang | evrardjp: yeah but i mean i dont understand why that patch fixes rabbitmq | 10:20 |
| evrardjp | for suse, I don't know what we do in the lxc_hosts copy steps I will have a look | 10:20 |
| hwoarang | for suse | 10:20 |
| evrardjp | hwoarang: probably because of the version available | 10:21 |
| odyssey4me | hwoarang it's because the pip_install role doesn't have the repo management any more | 10:21 |
| odyssey4me | evrardjp broke everything ;) | 10:21 |
| hwoarang | but we don't use external repo for rabbitmq | 10:21 |
| hwoarang | we use the one from the default repos so we don't care if external repos have been applied or not | 10:21 |
| evrardjp | we use the openstack ones Pike right? | 10:21 |
| hwoarang | and default repos are always in the zuul images | 10:21 |
| hwoarang | hmm | 10:21 |
| evrardjp | I thought 42.3 was like very old rabbitmq | 10:22 |
| hwoarang | that may makes sense. rabbitmq may be comming from the cloud repo | 10:22 |
| hwoarang | :/ | 10:22 |
| hwoarang | ok then i get it now | 10:22 |
| evrardjp | 3.5.8 from 42.3 | 10:22 |
| evrardjp | yeah | 10:22 |
| hwoarang | good point | 10:22 |
| evrardjp | but yes I broke everything. I think it's for the best in the long run | 10:22 |
| evrardjp | but it is quite a ride. | 10:23 |
| evrardjp | :) | 10:23 |
| *** vnogin has quit IRC | 10:23 | |
| openstackgerrit | Merged openstack/openstack-ansible-plugins stable/pike: Follow symlinks in the lookup https://review.openstack.org/525111 | 10:23 |
| evrardjp | the lookup merge is fast! | 10:23 |
| *** ianychoi_ is now known as ianychoi | 10:24 | |
| *** electrofelix has joined #openstack-ansible | 10:24 | |
| *** vnogin has joined #openstack-ansible | 10:27 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_client master: Remove pip_install dependency https://review.openstack.org/524963 | 10:29 |
| *** zkynet has joined #openstack-ansible | 10:30 | |
| *** hamza21 has quit IRC | 10:32 | |
| evrardjp | odyssey4me: I'd like your opinion on this: http://logs.openstack.org/24/504224/23/check/openstack-ansible-deploy-all_metal-ubuntu-xenial/094621a/job-output.txt.gz#_2017-12-04_09_08_11_594884 | 10:33 |
| evrardjp | I thought of doing like what we do in rpc, replace aio1 with $(hostname) in all the /etc/openstack_deploy files. | 10:33 |
| evrardjp | just a well placed sed. | 10:33 |
| evrardjp | would you do that, or go for 1) a different templating in the bootstrap-host role, for the generation of the o_u_c and the conf.d/* | 10:34 |
| evrardjp | 2) adapt the hosts file | 10:35 |
| odyssey4me | a simple sed should do that, and would cause less confusion | 10:35 |
| evrardjp | ok. | 10:35 |
| odyssey4me | templating out the files could be done later, if at all | 10:35 |
| evrardjp | I don't like the idea of touching this bash thingy for /etc/hosts. | 10:35 |
| evrardjp | that's too big of a job. | 10:36 |
| *** taseer1 has joined #openstack-ansible | 10:38 | |
| *** Taseer has quit IRC | 10:39 | |
| *** taseer2 has joined #openstack-ansible | 10:39 | |
| *** ThomasS has joined #openstack-ansible | 10:40 | |
| ThomasS | hi | 10:41 |
| ThomasS | does anyone know how to fix the object/object bug when uploading an image in horizon? | 10:41 |
| ThomasS | can it be fixed in pike with a configuration change? | 10:41 |
| *** stuartgr has joined #openstack-ansible | 10:42 | |
| *** taseer1 has quit IRC | 10:43 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Build everything on metal https://review.openstack.org/504224 | 10:44 |
| ThomasS | it looks like i am hitting this bug | 10:44 |
| ThomasS | https://bugs.launchpad.net/horizon/+bug/1703708 | 10:44 |
| openstack | Launchpad bug 1703708 in OpenStack Dashboard (Horizon) "Horizon image upload with bad Glance CORS config fails with "[object Object]"" [High,Fix released] - Assigned to Pascal Boutin (pboutin) | 10:44 |
| *** armaan has quit IRC | 10:45 | |
| *** rodolof has quit IRC | 10:45 | |
| ThomasS | this looks like it is presenting a better error message but not a solution for the wrong cors config :-) | 10:45 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_keystone master: [DNM] debug why on metal fails on gates and not on my machine. https://review.openstack.org/524954 | 10:46 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_neutron master: Fix swift port conflict https://review.openstack.org/524874 | 10:49 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Change opendaylight default port https://review.openstack.org/524873 | 10:49 |
| evrardjp | mardim: mbuil could you re-review: https://review.openstack.org/#/c/524874/ ? | 10:50 |
| *** vnogin has quit IRC | 10:50 | |
| *** armaan has joined #openstack-ansible | 10:54 | |
| *** armaan has quit IRC | 10:57 | |
| *** gun1x has left #openstack-ansible | 11:00 | |
| odyssey4me | evrardjp FYI found the issue with the docs publishing: https://review.openstack.org/525154 | 11:03 |
| mardim | evrardjp, I will after lunch :) | 11:04 |
| *** armaan has joined #openstack-ansible | 11:06 | |
| *** vnogin has joined #openstack-ansible | 11:11 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server master: Allow the upstream pypi mirror to be changed https://review.openstack.org/525155 | 11:12 |
| *** taseer2 is now known as Taseer | 11:14 | |
| openstackgerrit | Merged openstack/openstack-ansible-specs master: Simplify repo management https://review.openstack.org/521078 | 11:20 |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Remove old fact cleanup tasks https://review.openstack.org/523481 | 11:21 |
| evrardjp | odyssey4me: good catch. | 11:22 |
| *** hamza21 has joined #openstack-ansible | 11:41 | |
| *** shardy has quit IRC | 11:44 | |
| *** ivveh has joined #openstack-ansible | 11:45 | |
| *** shardy has joined #openstack-ansible | 11:45 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the default pip index to use the pypi cache when possible https://review.openstack.org/522802 | 11:51 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the default pip index to use the pypi cache when possible https://review.openstack.org/522802 | 11:53 |
| *** smatzek has joined #openstack-ansible | 11:56 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the default pip index to use the pypi cache when possible https://review.openstack.org/522802 | 12:02 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the default pip index to use the pypi cache when possible https://review.openstack.org/522802 | 12:04 |
| *** savvas_ has quit IRC | 12:08 | |
| *** savvas has joined #openstack-ansible | 12:08 | |
| *** michelv has quit IRC | 12:11 | |
| *** hamza21 has quit IRC | 12:12 | |
| *** michelv has joined #openstack-ansible | 12:14 | |
| *** hamza21 has joined #openstack-ansible | 12:17 | |
| *** sxc731 has quit IRC | 12:17 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Prepare containers with openstack-hosts role https://review.openstack.org/525118 | 12:19 |
| ThomasS | does nobody know how to configure the cors corretly in OSA to fix the error on image upload in horizon? | 12:19 |
| ThomasS | https://bugs.launchpad.net/horizon/+bug/1703708 | 12:19 |
| openstack | Launchpad bug 1703708 in OpenStack Dashboard (Horizon) "Horizon image upload with bad Glance CORS config fails with "[object Object]"" [High,Fix released] - Assigned to Pascal Boutin (pboutin) | 12:19 |
| *** armaan has quit IRC | 12:20 | |
| *** armaan has joined #openstack-ansible | 12:20 | |
| *** chhavi has quit IRC | 12:23 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Use Ansible 2.4 https://review.openstack.org/522778 | 12:26 |
| *** hamza21 has quit IRC | 12:31 | |
| *** zkynet has quit IRC | 12:31 | |
| *** vnogin has quit IRC | 12:32 | |
| *** vnogin has joined #openstack-ansible | 12:32 | |
| *** sep has joined #openstack-ansible | 12:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Force the uca filename https://review.openstack.org/524865 | 12:37 |
| *** vnogin has quit IRC | 12:42 | |
| *** hw_wutianwei has quit IRC | 12:43 | |
| *** woodard_ has joined #openstack-ansible | 12:49 | |
| *** woodard has quit IRC | 12:49 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-os_neutron master: Add OVS-NSH support for opensuse https://review.openstack.org/524635 | 12:51 |
| *** Mr_Smurf has joined #openstack-ansible | 12:54 | |
| *** wspthr has quit IRC | 12:55 | |
| Mr_Smurf | Do I have have to upgrade every version step when upgrading minor versions or can I just go the the latest in one jump? | 12:55 |
| *** wspthr has joined #openstack-ansible | 12:57 | |
| *** vnogin has joined #openstack-ansible | 12:58 | |
| odyssey4me | Mr_Smurf you should be able to upgrade to the latest minor rev in one go | 12:58 |
| *** sboyron has joined #openstack-ansible | 12:59 | |
| Mr_Smurf | odyssey4me: I was trying to upgrade only one of the mgmt nodes first to see if stuff works but setup-infrastructure.yml on my repo container does not fetch everything i need | 13:00 |
| odyssey4me | Mr_Smurf did you follow this process? https://docs.openstack.org/openstack-ansible/pike/user/minor-upgrade.html | 13:01 |
| Mr_Smurf | I'm upgrading newton from 14.0.7 to 14.2.12 (https://docs.openstack.org/openstack-ansible/newton/upgrade-guide/minor-upgrade.html) | 13:02 |
| Mr_Smurf | I get the empty directory /var/www/repo/os-releases/14.2.12 but it does not fetch anyting from git and I don't get anything new under /var/www/repo/venvs/ | 13:03 |
| odyssey4me | Mr_Smurf ok, did you do bootstrap-ansible after doing the new checkout? also, if you run repo-build.yml does it fail the build tasks, or skip them? | 13:05 |
| Mr_Smurf | I did repo build with --limit to one of the repo containers but I will try to run without limits | 13:07 |
| *** sboyron has quit IRC | 13:07 | |
| odyssey4me | Mr_Smurf yeah, the repo build does not destroy any data so it's safe to do so - also it decides which repo container to do the build on so I don't think limit will work there | 13:08 |
| odyssey4me | if you used limit previously, then it's probably better to use repo-install.yml so that repo_server runs first | 13:08 |
| Mr_Smurf | ok, I'll try that :) | 13:09 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Use one play for container creation https://review.openstack.org/525109 | 13:10 |
| *** vnogin has quit IRC | 13:10 | |
| Mr_Smurf | It seems to be doing stuff :) | 13:10 |
| Mr_Smurf | well, it exploded in the end though... (item=need_wheel_build) => {"failed": true, "item": "need_wheel_build", "msg": "Destination file /etc/ansible/facts.d/openstack_ansible.fact not writable"} | 13:11 |
| Mr_Smurf | I do not have any /etc/ansible directory in the repo container so I can understand if it is not writeable | 13:11 |
| odyssey4me | Mr_Smurf ah, you may need to run lxc-containers-create first to get that created | 13:13 |
| Mr_Smurf | so how much will explode when I run that? | 13:15 |
| odyssey4me | if you're concerned about it affecting other services, run with --limit lxc_hosts,repo_all | 13:15 |
| odyssey4me | lxc_hosts is included just to get facts from them - no changes will be made there | 13:16 |
| Mr_Smurf | ok, good :) | 13:16 |
| Mr_Smurf | I was planing to upgrade the services on one of the mgmt-nodes first to see that evertying is working | 13:18 |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate master: Fix ansible-lint test failures https://review.openstack.org/525031 | 13:18 |
| *** zkynet has joined #openstack-ansible | 13:18 | |
| mhayden | morning folks | 13:20 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-galera_client master: Remove pip_install dependency https://review.openstack.org/524963 | 13:22 |
| openstackgerrit | Merged openstack/openstack-ansible-repo_server master: Allow the upstream pypi mirror to be changed https://review.openstack.org/525155 | 13:24 |
| evrardjp | ThomasS: https://gist.github.com/evrardjp/5f1ae8d1afdf7b23d3224bc55388adab ? | 13:24 |
| mgariepy | morning | 13:28 |
| *** armaan has quit IRC | 13:29 | |
| ThomasS | let me check | 13:29 |
| *** armaan has joined #openstack-ansible | 13:30 | |
| ThomasS | i will try this one thanks :-) | 13:31 |
| *** armaan has quit IRC | 13:32 | |
| *** armaan has joined #openstack-ansible | 13:33 | |
| *** portante has quit IRC | 13:35 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate master: Use group_names to check a host belongs to group https://review.openstack.org/523275 | 13:35 |
| *** lihi has quit IRC | 13:35 | |
| *** oanson has quit IRC | 13:35 | |
| *** portante has joined #openstack-ansible | 13:36 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/pike: Gather more logs when running tests https://review.openstack.org/524863 | 13:37 |
| ThomasS | @evrardjp upload works now flawless, on completion it shows [object Object] again | 13:37 |
| *** lihi has joined #openstack-ansible | 13:37 | |
| *** oanson has joined #openstack-ansible | 13:37 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate stable/pike: Fix ansible-lint test failures https://review.openstack.org/525180 | 13:40 |
| ThomasS | uploading a larger iso fails with same error. maybe only the first one worked because i used the small cirros image | 13:41 |
| openstackgerrit | Merged openstack/openstack-ansible-os_glance stable/pike: Remove jinja2 delimiters from when https://review.openstack.org/524867 | 13:41 |
| ThomasS | no the object object error pops up immediately | 13:41 |
| *** chhavi has joined #openstack-ansible | 13:43 | |
| *** woodard_ has quit IRC | 13:43 | |
| evrardjp | mmm | 13:43 |
| evrardjp | I am no expert in all of that | 13:44 |
| *** woodard has joined #openstack-ansible | 13:44 | |
| evrardjp | I think this could be a different thing | 13:44 |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron stable/pike: Remove jinja2 delimiters from when clause https://review.openstack.org/525117 | 13:45 |
| *** bhujay has quit IRC | 13:46 | |
| ThomasS | ok, but thanks for helping with that anyway | 13:47 |
| evrardjp | so it's helping for small uploads but not large uploads? | 13:47 |
| ThomasS | i guess everyone has this problem when deploying pike with OSA | 13:47 |
| ThomasS | i will try again | 13:47 |
| evrardjp | I don't know | 13:47 |
| evrardjp | I can try later. | 13:47 |
| *** jamesdenton has joined #openstack-ansible | 13:48 | |
| odyssey4me | ThomasS are you using ceph or swift or file system for the glance back-end? | 13:48 |
| ThomasS | ceph | 13:48 |
| openstackgerrit | Manuel Buil proposed openstack/openstack-ansible-os_neutron master: Provide support for SFC deployments https://review.openstack.org/510909 | 13:50 |
| odyssey4me | ThomasS hmm, perhaps logan- will be online later and can confirm the issue or suggest changes of config | 13:51 |
| odyssey4me | I'v eonly ever hear dissues like that when using ceph and I suspect that there's a horizon config that needs changing, assuming you're experiencing this in horizon? | 13:51 |
| *** michelv has quit IRC | 13:51 | |
| odyssey4me | I wonder if this might not relate? https://review.openstack.org/#/c/516751/ | 13:53 |
| *** thorst has joined #openstack-ansible | 13:59 | |
| ThomasS | yes it it only in horizon | 14:02 |
| ThomasS | works from the utility container | 14:02 |
| *** vnogin has joined #openstack-ansible | 14:02 | |
| *** hamza21 has joined #openstack-ansible | 14:03 | |
| ThomasS | so it looks like the mentioned horizon bug | 14:04 |
| ThomasS | https://bugs.launchpad.net/horizon/+bug/1703708 | 14:04 |
| openstack | Launchpad bug 1703708 in OpenStack Dashboard (Horizon) "Horizon image upload with bad Glance CORS config fails with "[object Object]"" [High,Fix released] - Assigned to Pascal Boutin (pboutin) | 14:04 |
| odyssey4me | ThomasS are the settings shown in the bug set in your horizon container? | 14:05 |
| ThomasS | no | 14:07 |
| ThomasS | i was using the defaults from OSA | 14:07 |
| odyssey4me | is there a cors section in your glance-api.conf? | 14:07 |
| odyssey4me | try adding https://review.openstack.org/#/c/516751/ to your horizon role and re-run the horizon playbook and see if it helps? | 14:08 |
| ThomasS | i added this https://gist.github.com/evrardjp/5f1ae8d1afdf7b23d3224bc55388adab | 14:08 |
| ThomasS | and done a glance-config run | 14:08 |
| *** mbuil has quit IRC | 14:09 | |
| odyssey4me | oh ok, personally I've never seen how the cors stuff should work so I can't really advise there, but I do suspect that the patch I mentioned may help | 14:10 |
| odyssey4me | it would seem that glance itself is working fine, given that it works from the utility container | 14:10 |
| ThomasS | i kept the default | 14:11 |
| ThomasS | horizon_images_allow_location: False | 14:11 |
| ThomasS | yeah i guess it is really a horizon problem | 14:11 |
| ThomasS | setting this to true will not help i guess | 14:12 |
| odyssey4me | ah, I see that's the default and it seems you're right | 14:12 |
| odyssey4me | I'm out of ideas for now then - hopefully someone else can help a bit later when the US crew comes online. | 14:14 |
| evrardjp | yeah, maybe robcresswell can help too, I remember we discussed CORS in the past with him. | 14:15 |
| logan- | reading up.. horizon uploads to glance failing? | 14:16 |
| logan- | backing store shouldn't really matter on that. glance api interaction is the same either way. usually its a CORS issue | 14:16 |
| logan- | the best place to troubleshoot that is in the browser console | 14:17 |
| logan- | you'll probably see the browser blocking the upload request to the glance api due to it being on a different hostname/protocol/whatever than horizon | 14:18 |
| logan- | last time this came up i put up an example cors config: https://gist.github.com/Logan2211/32fa5fb0c6e0c2775b14ede5174d6c07 | 14:18 |
| logan- | another option is setting horizon's glance upload mode to legacy | 14:18 |
| evrardjp | logan-: I think I pasted something based on your paste | 14:18 |
| evrardjp | I remember we had that conversation | 14:19 |
| logan- | we should improve docs around this issue | 14:19 |
| evrardjp | but now I can't say why it would work on cirros vs a large image | 14:20 |
| logan- | yeah true | 14:20 |
| evrardjp | I agree, or have better defaults | 14:20 |
| evrardjp | I mean this could probably be wired in group vars | 14:20 |
| logan- | i'm not sure how you would do that | 14:20 |
| logan- | it is domain specific | 14:20 |
| *** goldenfri has joined #openstack-ansible | 14:21 | |
| *** sxc731 has joined #openstack-ansible | 14:21 | |
| logan- | suppose maybe you could have the wiring ready to consume openstack_cors_allowed_origin and leave openstack_cors_allowed_origin undefined | 14:22 |
| evrardjp | logan-: well if move the use of external_lb_vip_address to using external_lb_fqdn, we can then use | 14:22 |
| evrardjp | that value to use it in | 14:22 |
| evrardjp | allowed_origin by default | 14:22 |
| logan- | it wouldn't help much for my use | 14:22 |
| logan- | my api endpoints are on api.region.cloud.domain.net and my horizon is on cloud.domain.net | 14:22 |
| logan- | the external_lb_vip_address isnt an issue | 14:23 |
| logan- | the same host is already allowed | 14:23 |
| logan- | the issue could arise from having IP based vip_address and access the horizon thru a domain name | 14:23 |
| logan- | (because horizon would use the IP based catalog endpoint to reach glance) | 14:23 |
| evrardjp | yeah I understand, a difference between horizon url and endpoint basically | 14:24 |
| logan- | right | 14:24 |
| evrardjp | that's simple to fix too | 14:24 |
| evrardjp | we can, while we run the horizon play, use that to compare with the service catalog | 14:24 |
| ThomasS | ok | 14:24 |
| ThomasS | i will try your settings and do a horizon confifg | 14:25 |
| ThomasS | will report back | 14:25 |
| evrardjp | then if we have something different, enable CORS for this different thing | 14:25 |
| evrardjp | ThomasS: we are talking about future state here :) | 14:25 |
| logan- | osa doesn't know about my alias | 14:25 |
| jamesdenton | What is the best method for defining a secondary group file - to perform tasks against groups that are aren't defined in the dynamic inventory? Is that possible? | 14:25 |
| logan- | depends what version of osa jamesdenton | 14:25 |
| jamesdenton | This is.. 14 | 14:25 |
| evrardjp | yeah and depends for what too :) | 14:25 |
| evrardjp | N | 14:26 |
| evrardjp | mmm | 14:26 |
| evrardjp | what do you want to achieve? | 14:26 |
| evrardjp | why not adding your own group? | 14:26 |
| evrardjp | into the dynamic inventory? | 14:26 |
| *** mbuil has joined #openstack-ansible | 14:26 | |
| jamesdenton | Basically just want to perform a maintenance against all hosts, but in waves. So wave1 would include these 15 hosts, wave 2 another 15, etc | 14:26 |
| evrardjp | logan-: what do you mean by "doesn't know about my alias" ? | 14:26 |
| evrardjp | the service catalog must still know the api endpoints, right? | 14:27 |
| logan- | evrardjp: the url used to access horizon is not configured in osa anywhere except for my cors configuration | 14:27 |
| logan- | horizon isn't an api endpoint | 14:27 |
| evrardjp | ok so the external_lb_vip_address isn't what's used for horizon | 14:27 |
| logan- | providing a list of cors hosts is not something you can accurately detect | 14:27 |
| logan- | correct | 14:27 |
| logan- | it needs to be deployer provided | 14:27 |
| evrardjp | yeah in that case, that wouldn't work. | 14:27 |
| evrardjp | but you know stuff :p | 14:28 |
| evrardjp | you're now in pro-league | 14:28 |
| logan- | you would not need to configure cors for external_lb_vip_address if that is the address you access horizon on | 14:28 |
| evrardjp | (not Jupiler) | 14:28 |
| logan- | because horizon on external_lb_vip_address == glance on external_lb_vip_address | 14:28 |
| *** sboyron has joined #openstack-ansible | 14:28 | |
| evrardjp | logan-: exactly | 14:28 |
| evrardjp | so using it to see if there is a difference between endpoints seem fair. | 14:28 |
| evrardjp | But when you think about use cases | 14:29 |
| evrardjp | most ppl will probably use external_lb_vip_address for all the openstack bits including endpoints, and if they need to put horizon on a different domain, they would configure JUST horizon differently | 14:29 |
| evrardjp | so there is probably little to no value of that detection system. | 14:30 |
| *** smatzek has quit IRC | 14:31 | |
| *** smatzek has joined #openstack-ansible | 14:31 | |
| evrardjp | and I don't think doing a CORS on allowed_origin: * is a good default. | 14:31 |
| logan- | yeah | 14:31 |
| evrardjp | so yeah, docs is probably our best bet. When ThomasS will have figured out why the image upload fails in his case, we should probably start writing a use case document | 14:32 |
| evrardjp | in the docs | 14:32 |
| evrardjp | jamesdenton: I see, you could use serial for example | 14:34 |
| Mr_Smurf | should the /etc/ansible/facts.d/openstack_ansible.fact file in the containers have more info than [lxc] variant = default in them? | 14:34 |
| evrardjp | if it's for compute nodes, you can probably configure different files in conf.d | 14:35 |
| evrardjp | to split in different things | 14:35 |
| evrardjp | it would still be the same group, but you can remove files before running, and restore after. Ugly but works. | 14:36 |
| *** smatzek has quit IRC | 14:36 | |
| Mr_Smurf | I get the error "msg": "'dict object' has no attribute 'ansible_processor_vcpus'" but ansible_processor_vcpus exists in the file for the container in /etc/openstack_deploy/ansible_facts/ | 14:36 |
| logan- | yeah would be good to get some bugs opened when this comes up with a dump of catalog, browser console error, and url horizon url | 14:36 |
| logan- | horizon url* | 14:36 |
| jamesdenton | thank you, evrardjp. I will check that out! | 14:36 |
| evrardjp | Mr_Smurf: /etc/ansible/facts.d/openstack_ansible.fact are what's named "local facts", there are facts that are computed on the remote node. | 14:37 |
| evrardjp | /etc/openstack_deploy/ansible_facts/ are on the deploy node | 14:37 |
| evrardjp | what's missing in your case is a gather_facts: True in a playbook | 14:37 |
| evrardjp | jamesdenton: backup everything beforehand | 14:38 |
| evrardjp | :) | 14:38 |
| evrardjp | because messing with the inventory is NEVER good :p | 14:38 |
| evrardjp | but IIRC, that's what cloudnull was doing in OSIC | 14:38 |
| jamesdenton | Yeah for now I will stick to a long --limit string, but i will look into something else in the future | 14:38 |
| evrardjp | logan-: duely noted. Will remember to ask that. | 14:38 |
| evrardjp | jamesdenton: in ansible 2.4 that will be SUPER easy. | 14:39 |
| evrardjp | I can promise you that. | 14:39 |
| evrardjp | But it won't be backported. | 14:39 |
| jamesdenton | no worries. Everyone will just upgrade to the latest release. | 14:44 |
| jamesdenton | :D | 14:44 |
| evrardjp | :) | 14:45 |
| evrardjp | never heard that before. | 14:45 |
| evrardjp | :D | 14:45 |
| *** sboyron has quit IRC | 14:46 | |
| *** thorst has quit IRC | 14:46 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Remove keystone_token_driver variable https://review.openstack.org/523495 | 14:47 |
| openstackgerrit | Merged openstack/openstack-ansible master: Adapt to openstack_hosts/pip_install changes https://review.openstack.org/515405 | 14:48 |
| *** thorst has joined #openstack-ansible | 14:48 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible master: Reduce console output in gate jobs https://review.openstack.org/524721 | 14:49 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible master: Avoid using 'roles' reserved variable https://review.openstack.org/524660 | 14:50 |
| *** sboyron has joined #openstack-ansible | 14:51 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_swift stable/pike: Remove sleep option from service handlers https://review.openstack.org/525221 | 14:51 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_swift stable/ocata: Remove sleep option from service handlers https://review.openstack.org/525222 | 14:52 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_swift stable/pike: Remove jinja2 delimiters from when https://review.openstack.org/525223 | 14:52 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_neutron stable/pike: Remove send_arp_for_ha option https://review.openstack.org/525224 | 14:52 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_horizon stable/pike: Remove enable_{firewall,vpn} options https://review.openstack.org/525225 | 14:52 |
| *** bhujay has joined #openstack-ansible | 14:53 | |
| *** gouthamr has joined #openstack-ansible | 14:53 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Use Ansible 2.4 https://review.openstack.org/522778 | 14:54 |
| *** vnogin has quit IRC | 14:56 | |
| * mhayden is getting his backport on | 14:56 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_keystone stable/pike: Remove keystone_token_driver variable https://review.openstack.org/525227 | 14:57 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible master: Fix crontab errors on CentOS/SUSE https://review.openstack.org/520682 | 14:57 |
| *** rgogunskiy has quit IRC | 14:57 | |
| *** esberglu has joined #openstack-ansible | 14:58 | |
| mhayden | cloudnull / evrardjp: if y'all get a moment for this quick one -> https://review.openstack.org/523941 | 14:58 |
| *** weezS has joined #openstack-ansible | 14:58 | |
| *** smatzek has joined #openstack-ansible | 14:59 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible master: Build everything on metal https://review.openstack.org/504224 | 15:00 |
| evrardjp | mhayden: sure | 15:01 |
| evrardjp | I am stuck anyway, can't figure out things rn. | 15:01 |
| evrardjp | mhayden: do you know you're on schedule? | 15:02 |
| *** indistylo has quit IRC | 15:02 | |
| evrardjp | this week it's the deprecatations week. | 15:02 |
| evrardjp | :) | 15:02 |
| mhayden | ooh fancy | 15:02 |
| * mhayden didn't know he was on schedule | 15:02 | |
| *** vnogin has joined #openstack-ansible | 15:03 | |
| *** sboyron has quit IRC | 15:03 | |
| *** thorst has quit IRC | 15:03 | |
| *** woodard has quit IRC | 15:04 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Remove the curly braces in when https://review.openstack.org/524866 | 15:06 |
| evrardjp | yeah this week we should go over projects and see what's deprecated, and adopt our templates | 15:07 |
| evrardjp | so I planned to check all the pike deprecations this week | 15:08 |
| evrardjp | mhayden: https://etherpad.openstack.org/p/osa-queens-m2-deprecations | 15:11 |
| *** phalmos has joined #openstack-ansible | 15:12 | |
| *** bhujay has quit IRC | 15:13 | |
| *** bhujay has joined #openstack-ansible | 15:19 | |
| cloudnull | mornings | 15:20 |
| evrardjp | moring cloudnull ! | 15:21 |
| hwoarang | good morning! | 15:21 |
| mhayden | evrardjp: woot | 15:22 |
| ThomasS | hi | 15:22 |
| mhayden | ohai ThomasS | 15:22 |
| ThomasS | so i done this config https://gist.github.com/Logan2211/32fa5fb0c6e0c2775b14ede5174d6c07 | 15:22 |
| evrardjp | yup? | 15:22 |
| ThomasS | and also done horizon-config | 15:22 |
| *** phalmos has quit IRC | 15:22 | |
| ThomasS | fiest upload works again with any image size and shows object/object after completion | 15:23 |
| ThomasS | next upload does not work :-( | 15:23 |
| cloudnull | hows it on this fine day ? | 15:23 |
| *** thorst has joined #openstack-ansible | 15:24 | |
| *** hamza21 has quit IRC | 15:24 | |
| *** sxc731 has quit IRC | 15:26 | |
| *** esberglu_ has joined #openstack-ansible | 15:27 | |
| *** thorst has quit IRC | 15:29 | |
| *** esberglu has quit IRC | 15:29 | |
| *** vnogin has quit IRC | 15:34 | |
| *** vnogin has joined #openstack-ansible | 15:40 | |
| *** hw_wutianwei has joined #openstack-ansible | 15:41 | |
| *** phalmos has joined #openstack-ansible | 15:43 | |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Add basic plugin support for nspawn type containers https://review.openstack.org/490689 | 15:44 |
| *** sboyron has joined #openstack-ansible | 15:51 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Fix SUSE gate overrides outside gates https://review.openstack.org/525108 | 15:51 |
| *** sxc731 has joined #openstack-ansible | 15:52 | |
| *** jvidal has quit IRC | 15:53 | |
| *** esberglu_ has quit IRC | 15:56 | |
| *** esberglu has joined #openstack-ansible | 15:56 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Add OVS-NSH support for opensuse https://review.openstack.org/524635 | 16:00 |
| *** michelv has joined #openstack-ansible | 16:02 | |
| *** armaan has quit IRC | 16:02 | |
| Adri2000 | hi | 16:03 |
| *** armaan has joined #openstack-ansible | 16:03 | |
| Adri2000 | I'm a bit concerned about https://git.openstack.org/cgit/openstack/openstack-ansible-lxc_container_create/commit/?h=stable/pike&id=05b84528d100ab73b91e119dd379a8ed0726db7d | 16:03 |
| Adri2000 | just deployed this on a prod environment with a 3-node galera | 16:04 |
| Adri2000 | and some of the changes triggered a restart of *all* the lxc containers, including the galera ones, at the same time | 16:04 |
| Adri2000 | galera doesn't like it ... | 16:04 |
| evrardjp | WOW. | 16:06 |
| evrardjp | I didn't see it | 16:06 |
| evrardjp | oh yeah I was on holiday. | 16:06 |
| *** zkynet has quit IRC | 16:07 | |
| evrardjp | I see what it was done there. | 16:07 |
| evrardjp | cloudnull: spotz mhayden ^ check at this. This is not good. | 16:08 |
| *** armaan has quit IRC | 16:09 | |
| evrardjp | I bumped the role, as I usually do, so I am also responsible. | 16:09 |
| evrardjp | What should we do to detect this | 16:10 |
| mgariepy | that would be really nice to be able to detect this. | 16:12 |
| mgariepy | can we add a reno notes at least to warn users that it can happen ? | 16:13 |
| odyssey4me | well, is this in a released version - or is it at the head? | 16:13 |
| evrardjp | it probably is. | 16:13 |
| odyssey4me | if at the head, then we have detected it ;) | 16:13 |
| evrardjp | it is already tagged I think | 16:13 |
| evrardjp | let me double check | 16:13 |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/pike: Disable offloading in test by default https://review.openstack.org/524860 | 16:15 |
| *** hw_wutianwei has quit IRC | 16:15 | |
| evrardjp | it is not tagged! | 16:15 |
| *** sboyron has quit IRC | 16:15 | |
| evrardjp | Anyway, we should still be careful. | 16:15 |
| evrardjp | it's stable... | 16:16 |
| evrardjp | I mean stable is in the title. | 16:16 |
| odyssey4me | I'll bet the change to add create=dir was the cause: https://github.com/openstack/openstack-ansible-lxc_container_create/blob/master/tasks/lxc_container_config.yml#L44 | 16:17 |
| Adri2000 | odyssey4me: this, and the task "Write default container config" as well I think | 16:18 |
| Adri2000 | possibly | 16:18 |
| Adri2000 | because it is changed in that commit, and it has a notify to restart containers | 16:19 |
| evrardjp | odyssey4me: I am doubting of the need of the backport, tbh. This looks more like a feature than a bug. | 16:19 |
| odyssey4me | this may have something to do with the same issue: https://review.openstack.org/#/c/524735/1 | 16:21 |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate stable/pike: Fix ansible-lint test failures https://review.openstack.org/525180 | 16:22 |
| *** vnogin has quit IRC | 16:23 | |
| odyssey4me | we can implement the thing to prevent the restart from happening on a new environment, so it'll prevent that until the next major upgrade | 16:23 |
| mgariepy | would it be possible to have the galera and rabbitmq cluster to restart container only 1 at the time and wait for the cluster et be back to a healthy state before doing the next one ? | 16:24 |
| odyssey4me | mgariepy right now this is happening in lxc-containers-create, which touches all containers at once | 16:24 |
| odyssey4me | we either need to allow the restart not to happen, then make it happen later in serial - or we need to switch to a JIT container-create execution which only runs against the container in the playbook that runs in serial anyway | 16:25 |
| evrardjp | I am fine with that, but this is code that can break, I was more thinking of something that wouldn't pass gates if we are changing too many things. | 16:26 |
| evrardjp | I don't know how we can do that | 16:26 |
| mgariepy | serial restart would be nice but also restarting like neutron agent container can do some damage. | 16:27 |
| evrardjp | the only thing I see is the container uptime, which we won't be able to use in our gates | 16:27 |
| evrardjp | yeah and things like that. | 16:27 |
| evrardjp | we should maybe have minor update testing in periodics that shows downtime? | 16:28 |
| spotz | I like odyssey4me's idea | 16:28 |
| odyssey4me | well, unfortunately the JIT thing cannot and should not be backported - it's a big change | 16:28 |
| evrardjp | yeah | 16:28 |
| *** thorst has joined #openstack-ansible | 16:28 | |
| evrardjp | minor update testing something we can do. From tag x to head of branch. | 16:28 |
| evrardjp | if we see downtime, we can use that before tagging next release. | 16:29 |
| odyssey4me | sure | 16:29 |
| evrardjp | we still release every 2 weeks, so it's good enough for deployers I guess? | 16:29 |
| evrardjp | it's a lot of work though | 16:29 |
| evrardjp | but I think it could be a good thing | 16:30 |
| evrardjp | what do you think odyssey4me ? | 16:30 |
| odyssey4me | sure, if people are actually going to watch it | 16:31 |
| evrardjp | I am supposed to watch periodic logs every day, but I admit, I do that every 2 days. we have so many things pending right now... | 16:32 |
| odyssey4me | for now better to try to resolve this issue before the next release | 16:33 |
| *** thorst has quit IRC | 16:33 | |
| evrardjp | And that would catch issues like the one Adri2000 just mentioned. In any case, without Adri2000 we wouldn't have known for this. | 16:33 |
| evrardjp | periodics or not. | 16:33 |
| evrardjp | agreed odyssey4me . | 16:33 |
| *** Cronix has joined #openstack-ansible | 16:35 | |
| evrardjp | any idea there? | 16:36 |
| mgariepy | can we have a special test for commits that are tagging release ? | 16:36 |
| evrardjp | to solve the issue? Backporting the patch you pasted odyssey4me ? | 16:36 |
| evrardjp | mgariepy: we could technically, but these would be long tests. | 16:36 |
| evrardjp | So I am sure these are best left as periodics. | 16:37 |
| mgariepy | how long does it take ? 2h ? | 16:37 |
| evrardjp | if you have to deploy version x and do a minor update to latest, I suspect more than that. | 16:37 |
| mgariepy | have it run automatically when you bump release would be better IMO. | 16:38 |
| evrardjp | probably max 4h. | 16:38 |
| mgariepy | wow. that's long. | 16:38 |
| odyssey4me | oh bugger - there's a regression in there too | 16:38 |
| evrardjp | odyssey4me: in there? | 16:38 |
| odyssey4me | lemme push up some patches | 16:38 |
| evrardjp | should we revert? | 16:38 |
| evrardjp | you are all scaring me now. | 16:39 |
| odyssey4me | no, it adds more value than the pain it creates - and I think the other backends are being used already by some | 16:39 |
| openstackgerrit | Manuel Buil proposed openstack/openstack-ansible-os_neutron master: Provide support for SFC deployments https://review.openstack.org/510909 | 16:39 |
| savvas | hey guys, any support for Murano in the pipeline? Or recommended deployment for use with OpenStack Ansible? | 16:40 |
| *** bhujay has quit IRC | 16:40 | |
| *** chyka has joined #openstack-ansible | 16:41 | |
| evrardjp | savvas: murano isn't in the pipeline. I heard people interested by it, but I haven't seen commits. | 16:41 |
| odyssey4me | is the murano project still alive? | 16:41 |
| openstackgerrit | Manuel Buil proposed openstack/openstack-ansible-os_neutron master: Provide support for SFC deployments https://review.openstack.org/510909 | 16:42 |
| savvas | it caught my interest when watching this https://www.youtube.com/watch?time_continue=1680&v=Su4X2w-X-IE | 16:43 |
| evrardjp | odyssey4me: I don't think murano is still alive. | 16:44 |
| savvas | mm | 16:46 |
| odyssey4me | oh, never mind - turns out the regression I thought I spotted is not there | 16:47 |
| evrardjp | that's the good news. | 16:47 |
| evrardjp | don't tell me there is a bad news that go with it. | 16:47 |
| evrardjp | :) | 16:47 |
| odyssey4me | well, I suspect that https://github.com/openstack/openstack-ansible-lxc_container_create/commit/c82c7a60e9df2931ab0b456deb26b615d9ccd4e6 is the cause of the restarts - but that patch is actually good to have... we just did not cater for preventing the restart | 16:48 |
| savvas | https://docs.openstack.org/murano/latest/ does seem to be maintained still | 16:48 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift stable/pike: Remove jinja2 delimiters from when https://review.openstack.org/525223 | 16:48 |
| odyssey4me | well, as it turns out we already have a var preventing the restart in the role: https://github.com/openstack/openstack-ansible-lxc_container_create/blob/master/handlers/main.yml#L34 | 16:55 |
| *** woodard has joined #openstack-ansible | 16:56 | |
| *** esberglu has quit IRC | 16:56 | |
| *** esberglu has joined #openstack-ansible | 16:58 | |
| *** thorst has joined #openstack-ansible | 17:02 | |
| *** woodard has quit IRC | 17:08 | |
| *** woodard has joined #openstack-ansible | 17:08 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-pip_install master: Add find-links, even when not being locked down https://review.openstack.org/525257 | 17:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Set the default pip index to use the pypi cache when possible https://review.openstack.org/522802 | 17:11 |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone stable/pike: Remove keystone_token_driver variable https://review.openstack.org/525227 | 17:12 |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/pike: Gather more logs when running tests https://review.openstack.org/524863 | 17:13 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift stable/ocata: Remove sleep option from service handlers https://review.openstack.org/525222 | 17:14 |
| *** udesale has quit IRC | 17:17 | |
| *** chyka has quit IRC | 17:20 | |
| *** chyka has joined #openstack-ansible | 17:21 | |
| odyssey4me | Adri2000 can you give us your source version and current version from that minor upgrade? | 17:23 |
| *** michelv has quit IRC | 17:27 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Change deprecated nova_metadata_ip option https://review.openstack.org/523941 | 17:32 |
| *** SerenaFeng has joined #openstack-ansible | 17:37 | |
| *** SerenaFeng has quit IRC | 17:37 | |
| *** galstrom_zzz is now known as galstrom | 17:38 | |
| *** galstrom is now known as galstrom_zzz | 17:38 | |
| *** gkadam_ has quit IRC | 17:38 | |
| openstackgerrit | Manuel Buil proposed openstack/openstack-ansible master: Add networking-sfc repo to repo_packages https://review.openstack.org/525264 | 17:39 |
| *** thorst_ has joined #openstack-ansible | 17:41 | |
| *** thorst has quit IRC | 17:44 | |
| *** mbuil has quit IRC | 17:45 | |
| *** thorst_ has quit IRC | 17:51 | |
| mhayden | is it possible to specify an extra pip wheel to be built with repo_build? | 17:57 |
| * mhayden hunts | 17:57 | |
| *** indistylo has joined #openstack-ansible | 17:58 | |
| stuartgr | is there a way to assign compute nodes into nova availability zones during deployment, or does it have to be done after the deployment is complete? | 17:58 |
| *** epalper has quit IRC | 17:59 | |
| *** wspthr has quit IRC | 18:01 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Fix swift port conflict https://review.openstack.org/524874 | 18:01 |
| *** thorst has joined #openstack-ansible | 18:02 | |
| *** wspthr has joined #openstack-ansible | 18:04 | |
| *** chhavi has quit IRC | 18:05 | |
| *** thorst has quit IRC | 18:07 | |
| *** Cronix has quit IRC | 18:08 | |
| *** cshen_ has joined #openstack-ansible | 18:12 | |
| *** thorst has joined #openstack-ansible | 18:15 | |
| ArchiFleKs | odyssey4me: it was from 16.0.4 to head of stable/pike, I have to admit we are not using tags but the stable/pike, but we will switch to tags from now on | 18:18 |
| *** jwitko_ has joined #openstack-ansible | 18:20 | |
| *** thorst has quit IRC | 18:20 | |
| ArchiFleKs | evrardjp: I think it is possible to prevent at least galera or some container group (like neutron) to restart even when the handler is called, with the lxc_container_allow_restarts var. If we use the group_var / host_var mecanism in /etc/openstack_deploy that you pushed a while ago. Right ? | 18:21 |
| *** cshen_ has quit IRC | 18:22 | |
| *** thorst has joined #openstack-ansible | 18:24 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_neutron master: Ensure LBaaSv2 deployment is tested https://review.openstack.org/525272 | 18:35 |
| *** RandomTech has joined #openstack-ansible | 18:35 | |
| *** hamza21 has joined #openstack-ansible | 18:36 | |
| *** thorst has quit IRC | 18:38 | |
| mhayden | fedora's COPR is down | 18:40 |
| mhayden | so that will break CentOS 7's LXC pkgs for a bit :/ | 18:40 |
| *** sxc731 has quit IRC | 18:40 | |
| *** thorst has joined #openstack-ansible | 18:41 | |
| *** thorst has quit IRC | 18:43 | |
| * mhayden needs to figure out if we can get that mirrored :/ | 18:43 | |
| mgariepy | mhayden, do you know why it's down that often ? | 18:43 |
| mhayden | i have no idea | 18:43 |
| mhayden | bah https://status.fedoraproject.org/ | 18:45 |
| * mhayden will look around in infra land to see if we can mirror it | 18:45 | |
| evrardjp | mhayden: could you submit a patch to move the jobs to -nv temporarily? | 18:45 |
| mhayden | During the week of 2017-12-04 - 2017-12-08, all Fedora Infrastructure servers get moved | 18:46 |
| mhayden | https://status.fedoraproject.org/q4maint.html | 18:46 |
| mhayden | EGADS | 18:46 |
| *** gkadam_ has joined #openstack-ansible | 18:47 | |
| RandomTech | hey anyone here familar with rabbitmq? | 18:52 |
| RandomTech | we are having rabbit mq error out/fail during our setup | 18:52 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_glance master: Fix systemd init template for program override https://review.openstack.org/524778 | 18:59 |
| *** weezS has quit IRC | 19:04 | |
| *** stuartgr has quit IRC | 19:04 | |
| *** juliovp01 has joined #openstack-ansible | 19:06 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_glance master: Fix systemd init template for program override https://review.openstack.org/524778 | 19:07 |
| *** weezS has joined #openstack-ansible | 19:08 | |
| *** gouthamr has quit IRC | 19:20 | |
| *** vnogin has joined #openstack-ansible | 19:24 | |
| *** savvas_ has joined #openstack-ansible | 19:26 | |
| *** savvas_ has quit IRC | 19:27 | |
| *** vnogin has quit IRC | 19:28 | |
| *** savvas_ has joined #openstack-ansible | 19:28 | |
| RandomTech | mhayden: you availible a second. im having a probem with my setup | 19:28 |
| *** sxc731 has joined #openstack-ansible | 19:36 | |
| *** gouthamr has joined #openstack-ansible | 19:37 | |
| mhayden | RandomTech: just got back from eating -- what's the issue? | 19:39 |
| mhayden | also, which branch/tag are you deploying? | 19:39 |
| RandomTech | we used '# git clone -b 16.0.1 https://git.openstack.org/openstack/openstack-ansible \ /opt/openstack-ansible' for the pull as per the documention here: https://docs.openstack.org/project-deploy-guide/openstack-ansible/pike/deploymenthost.html | 19:41 |
| RandomTech | one second while i get the error messages compiled in a paste | 19:41 |
| Adri2000 | odyssey4me: yeah as ArchiFleKs said, so went openstack-ansible 27fa064a0eb2c2b17c361e176abe06375f8dc55f (16.0.4) to 16c69046bfd90d1b984de43bc6267fece6b75f1c (head) | 19:41 |
| *** juliovp0_ has joined #openstack-ansible | 19:41 | |
| *** thorst has joined #openstack-ansible | 19:42 | |
| Adri2000 | btw, regarding the lxc_container_allow_restarts mechanism I noticed it's used in scripts/run-upgrade.sh. so at least it means a major upgrade wouldn't be hit by this issue | 19:42 |
| *** sxc731 has quit IRC | 19:43 | |
| *** juliovp0_ has quit IRC | 19:43 | |
| *** juliovp0_ has joined #openstack-ansible | 19:44 | |
| RandomTech | mhayden: here it is: http://paste.openstack.org/show/628118/ | 19:44 |
| *** chyka_ has joined #openstack-ansible | 19:44 | |
| RandomTech | we tried rerunning it and i split up the error messages accordingly | 19:44 |
| *** gouthamr has quit IRC | 19:45 | |
| *** juliovp01 has quit IRC | 19:45 | |
| *** dave-mccowan has joined #openstack-ansible | 19:45 | |
| *** cshen_ has joined #openstack-ansible | 19:46 | |
| mhayden | hmm, looks like the cluster somehow got interrupted | 19:46 |
| mhayden | is only one of the nodes misbehaving? | 19:46 |
| RandomTech | it didnt get to the point in which it could try to join the others from what i could tell | 19:47 |
| *** chyka has quit IRC | 19:47 | |
| RandomTech | we can send you a full run log if it would help but it would take a moment to grab | 19:47 |
| RandomTech | the palybook decided to stop after the errors listed there instead of completing | 19:48 |
| mgariepy | cloudnull, are you around ? | 19:48 |
| *** chyka_ has quit IRC | 19:49 | |
| mhayden | RandomTech: well, one option is to go into your third node there and stop rabbitmq completely and restart it | 19:49 |
| RandomTech | okay it does seem that it may be only that container | 19:49 |
| mhayden | just to see if something needs to get cleared | 19:49 |
| mhayden | when it's gotten really out of hand before, i've thrown out the container and rebuilt it | 19:50 |
| RandomTech | so doing the teardown and rerunning setup-infrastructure? | 19:50 |
| *** thorst has quit IRC | 19:51 | |
| *** chyka has joined #openstack-ansible | 19:54 | |
| mhayden | you could tear down that third rabbitmq container, then rerun lxc-container-create.yml, then rerun the rabbitmq playbook | 19:58 |
| mhayden | but i wonder if you can just run start_app inside that container and make it happy | 19:58 |
| *** chyka has quit IRC | 19:59 | |
| *** hybridpollo has joined #openstack-ansible | 20:01 | |
| *** hamza21 has quit IRC | 20:01 | |
| *** savvas_ has quit IRC | 20:02 | |
| *** armaan has joined #openstack-ansible | 20:03 | |
| *** thorst has joined #openstack-ansible | 20:04 | |
| mhayden | ugh, python-ldap wants pyasn1-modules>0.1.5 but upper constraints limits to 0.1.0 | 20:05 |
| mhayden | in pike | 20:05 |
| * mhayden digs | 20:05 | |
| *** cshen_ has quit IRC | 20:06 | |
| *** hybridpollo has quit IRC | 20:09 | |
| *** indistylo has quit IRC | 20:12 | |
| *** hybridpollo has joined #openstack-ansible | 20:14 | |
| RandomTech | mhayden: is the version the documentation says to clone outdated. i noticed it doesnt match up with the version the documentation is named after | 20:18 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_keystone master: Add pyldap to Keystone packages https://review.openstack.org/525297 | 20:20 |
| *** cshen_ has joined #openstack-ansible | 20:21 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_keystone stable/pike: Remove python-ldap and add pyldap https://review.openstack.org/525299 | 20:22 |
| mhayden | RandomTech: not sure what you're asking | 20:22 |
| *** juliovp0_ has quit IRC | 20:22 | |
| *** juliovp01 has joined #openstack-ansible | 20:23 | |
| RandomTech | if you go to: "https://docs.openstack.org/openstack-ansible-haproxy_server/pike/" its labeled 16.0.2 but it has you run a command to clone 16.0.1 | 20:23 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_keystone stable/pike: Remove python-ldap and add pyldap https://review.openstack.org/525299 | 20:23 |
| mhayden | the 16.0.2 docs should be very close | 20:23 |
| *** juliovp0_ has joined #openstack-ansible | 20:25 | |
| *** juliovp0_ has quit IRC | 20:25 | |
| *** juliovp01 has quit IRC | 20:27 | |
| *** cshen_ has quit IRC | 20:34 | |
| *** armaan has quit IRC | 20:34 | |
| *** cshen_ has joined #openstack-ansible | 20:34 | |
| *** gouthamr has joined #openstack-ansible | 20:36 | |
| *** vnogin has joined #openstack-ansible | 20:38 | |
| *** savvas_ has joined #openstack-ansible | 20:38 | |
| RandomTech | ya, mhayden i just wasnt sure if that page needed updating | 20:42 |
| *** armaan has joined #openstack-ansible | 20:45 | |
| *** vnogin has quit IRC | 20:50 | |
| *** gkadam_ has quit IRC | 20:51 | |
| *** thorst has quit IRC | 20:56 | |
| *** chyka has joined #openstack-ansible | 21:02 | |
| *** thorst has joined #openstack-ansible | 21:02 | |
| *** chyka has quit IRC | 21:02 | |
| *** armaan has quit IRC | 21:04 | |
| *** savvas_ has quit IRC | 21:05 | |
| *** chyka has joined #openstack-ansible | 21:05 | |
| *** shardy has quit IRC | 21:05 | |
| *** thorst has quit IRC | 21:07 | |
| *** cshen_ has quit IRC | 21:08 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_neutron master: Ensure LBaaSv2 deployment is tested https://review.openstack.org/525272 | 21:08 |
| *** woodard has quit IRC | 21:13 | |
| *** thorst has joined #openstack-ansible | 21:17 | |
| *** vnogin has joined #openstack-ansible | 21:19 | |
| *** thorst has quit IRC | 21:20 | |
| *** thorst has joined #openstack-ansible | 21:21 | |
| *** smatzek has quit IRC | 21:23 | |
| *** cshen_ has joined #openstack-ansible | 21:26 | |
| *** vnogin has quit IRC | 21:29 | |
| *** vnogin has joined #openstack-ansible | 21:32 | |
| *** thorst has quit IRC | 21:36 | |
| *** pcaruana has quit IRC | 21:39 | |
| *** cshen_ has quit IRC | 21:44 | |
| *** cmart has joined #openstack-ansible | 21:45 | |
| *** SerenaFeng has joined #openstack-ansible | 21:45 | |
| *** SerenaFeng has quit IRC | 21:46 | |
| *** SerenaFeng has joined #openstack-ansible | 21:47 | |
| *** vnogin has quit IRC | 21:48 | |
| *** cshen_ has joined #openstack-ansible | 21:50 | |
| *** thorst has joined #openstack-ansible | 21:50 | |
| *** thorst has quit IRC | 21:54 | |
| *** vnogin has joined #openstack-ansible | 21:56 | |
| *** cshen_ has quit IRC | 21:58 | |
| *** gouthamr has quit IRC | 22:04 | |
| *** RandomTech has quit IRC | 22:06 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-os_neutron stable/pike: Change deprecated nova_metadata_ip option https://review.openstack.org/525333 | 22:07 |
| *** thorst has joined #openstack-ansible | 22:10 | |
| *** thorst has quit IRC | 22:14 | |
| cmart | Hi folks -- my dynamic inventory script is generating inventory with ansible_host of "null" for all the containers. This is causing a template in the HAProxy role to fail. Anyone else run into this recently? | 22:18 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_glance master: Fix systemd init template for program override https://review.openstack.org/524778 | 22:19 |
| *** phalmos has quit IRC | 22:22 | |
| *** pmannidi has joined #openstack-ansible | 22:43 | |
| openstackgerrit | Frank Zhang proposed openstack/openstack-ansible-ops master: User marker to skip completed plays https://review.openstack.org/525339 | 22:44 |
| openstackgerrit | Frank Zhang proposed openstack/openstack-ansible-ops master: Use marker to skip completed plays https://review.openstack.org/525339 | 22:45 |
| *** thorst has joined #openstack-ansible | 22:53 | |
| *** weezS has quit IRC | 22:56 | |
| *** thorst has quit IRC | 22:58 | |
| *** savvas has quit IRC | 23:03 | |
| *** gouthamr has joined #openstack-ansible | 23:08 | |
| *** jose-phillips has joined #openstack-ansible | 23:18 | |
| jose-phillips | Hi | 23:18 |
| jose-phillips | one question | 23:18 |
| jose-phillips | if i use openstack-ansbile | 23:18 |
| jose-phillips | im able to modify some configurations by hand | 23:18 |
| jose-phillips | ? | 23:18 |
| *** esberglu has quit IRC | 23:18 | |
| *** esberglu has joined #openstack-ansible | 23:19 | |
| *** esberglu has quit IRC | 23:23 | |
| cmart | jose-phillips, the short answer is yes you can, but which configurations are you talking about? | 23:24 |
| jose-phillips | cinder.conf | 23:24 |
| jose-phillips | im using juju but juju when i perform some changes to cinder.conf inside of a container | 23:24 |
| jose-phillips | after reboot the confiuration lost | 23:24 |
| jose-phillips | the question is if i switch to openstack-ansible i will able to modify cinder.conf | 23:24 |
| jose-phillips | and the changes will be persistent | 23:25 |
| cmart | Yes, persistent until the next time you run OSA and the config is re-generated | 23:25 |
| cmart | what changes do you want to make to cinder.conf? | 23:25 |
| jose-phillips | i need to add a new section called [NetApp] | 23:25 |
| jose-phillips | and the netapp storage filler settings | 23:26 |
| jose-phillips | also change enable_backends = NetApp | 23:26 |
| jose-phillips | on default section | 23:26 |
| cmart | I don't use it, but a quick search that OSA claims to support NetApp backend for Cinder: https://docs.openstack.org/openstack-ansible-os_cinder/latest/configure-cinder.html#configuring-cinder-to-use-a-netapp-appliance | 23:26 |
| cmart | so if that ends up working for your needs, then you won't need to edit cinder.conf by hand at all :) | 23:27 |
| jose-phillips | :) | 23:27 |
| jose-phillips | great | 23:27 |
| jose-phillips | another question | 23:27 |
| jose-phillips | im new on OSA | 23:27 |
| jose-phillips | i never used before | 23:27 |
| jose-phillips | is possible with osa install openstack without ceph and swift | 23:28 |
| cmart | it's preferable to just have OSA deploy whatever config you want. but you can certainly modify cinder.conf afterward, and that'll persist until it's overwritten by another OSA deployment. | 23:28 |
| jose-phillips | that i really dont need it | 23:28 |
| cmart | yes. I have a production cluster running right now without either Ceph or Swift. they are optional. | 23:28 |
| jose-phillips | Cool | 23:28 |
| jose-phillips | well time to left behind fuel and juju | 23:28 |
| jose-phillips | :) | 23:28 |
| jose-phillips | what about here | 23:29 |
| jose-phillips | is easy? | 23:29 |
| cmart | I don't understand | 23:30 |
| cmart | your question | 23:30 |
| jose-phillips | sorry i forgot the question lol | 23:30 |
| jose-phillips | hahah | 23:30 |
| jose-phillips | i mean | 23:30 |
| jose-phillips | is easy here upgrade openstack | 23:30 |
| jose-phillips | with osa? | 23:30 |
| jose-phillips | right now on my fuel deployment im complety stuck with mitaka | 23:30 |
| jose-phillips | and juju it's imposible configure netapp | 23:30 |
| cmart | yes. It's straightforward to upgrade an OSA-deployed cluster to the next version of OpenStack. Done it more than once (albeit only in test environments!) | 23:31 |
| *** esberglu has joined #openstack-ansible | 23:31 | |
| cmart | I think that's a design goal of the project, to always provide an upgrade path. | 23:31 |
| cmart | (obviously you need to have an existing cluster already deployed with OpenStack-Ansible for the upgrade to be smooth) | 23:32 |
| jose-phillips | lol | 23:32 |
| jose-phillips | i know | 23:32 |
| jose-phillips | well i guess is time to trying it | 23:32 |
| jose-phillips | btw who support really ansible? | 23:32 |
| cmart | enjoy. OSA is kind of a beast, but so much nicer than the alternatives, I think | 23:33 |
| jose-phillips | for example fuel is supported by Mirantis | 23:33 |
| jose-phillips | juju by cannocial | 23:33 |
| cmart | Ansible was acquired by Red Hat a few years ago | 23:33 |
| *** thorst has joined #openstack-ansible | 23:33 | |
| *** thorst has quit IRC | 23:33 | |
| jose-phillips | another question ansible can be added to MAAS to control the physical hardware? | 23:34 |
| cmart | (and OpenStack-Ansible originated with Rackspace, I think most of the core contributors work there. Rackspace uses OSA to support their private cloud offering) | 23:34 |
| cmart | jose-phillips I deploy my OpenStack hosts with MaaS, so I think the answer is "yes" | 23:34 |
| jose-phillips | and ansible? | 23:35 |
| cmart | oh wait, now I'm not sure what you're asking | 23:35 |
| jose-phillips | juju integrates with maas to control the hardware | 23:35 |
| jose-phillips | ansible is able to control the hardware thru MAAS api? | 23:35 |
| cmart | ah I see. OSA has no particular integration with MaaS that I'm aware of. I use MaaS to deploy my hosts, but I do it "by hand" | 23:36 |
| jose-phillips | cool | 23:36 |
| jose-phillips | i guess i will try it | 23:36 |
| jose-phillips | also ansible perform containers right | 23:36 |
| jose-phillips | ? | 23:36 |
| cmart | the trickiest thing is probably setting up the network requirements on your hosts (interfaces and bonds).. once you figure that out the rest is smoother | 23:36 |
| cmart | if by "Ansible perform containers" you mean that OSA uses Ansible to deploy OpenStack services inside containers, then the answer is yes | 23:37 |
| jose-phillips | cool | 23:38 |
| jose-phillips | well im destroying my juju environment to start over again | 23:38 |
| *** smatzek has joined #openstack-ansible | 23:38 | |
| *** esberglu has quit IRC | 23:39 | |
| jose-phillips | i already have configured the network with maas but i guess i have to perform the bridges inside of each node | 23:39 |
| *** smatzek has quit IRC | 23:43 | |
| *** thorst has joined #openstack-ansible | 23:45 | |
| cmart | yes, it's not too bad. Here we made this https://github.com/cyverse/openstack-ansible-host-prep to make the process easier, some of it's a little old but I'm still using it now to build a Pike cloud | 23:47 |
| *** woodard has joined #openstack-ansible | 23:47 | |
| cmart | prob best to just follow the OSA docs :) | 23:48 |
| *** thorst has quit IRC | 23:49 | |
| *** thorst has joined #openstack-ansible | 23:51 | |
| *** thorst has quit IRC | 23:55 | |
| *** woodard has quit IRC | 23:56 | |
| *** gouthamr has quit IRC | 23:56 | |
| *** gouthamr has joined #openstack-ansible | 23:58 | |
| *** cmart has quit IRC | 23:58 | |
| *** ianychoi has quit IRC | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!