| *** johnmilton has joined #openstack-ansible | 00:01 | |
| *** busterswt has quit IRC | 00:02 | |
| *** busterswt has joined #openstack-ansible | 00:05 | |
| *** sdake has joined #openstack-ansible | 00:06 | |
| *** ChrisBenson has quit IRC | 00:09 | |
| *** busterswt has quit IRC | 00:09 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: WIP Initial convergence testing https://review.openstack.org/304887 | 00:11 |
|---|---|---|
| *** thorst_ has joined #openstack-ansible | 00:14 | |
| *** asettle has joined #openstack-ansible | 00:18 | |
| *** fawadkhaliq has quit IRC | 00:21 | |
| *** fawadkhaliq has joined #openstack-ansible | 00:21 | |
| *** sdake_ has joined #openstack-ansible | 00:22 | |
| *** sdake has quit IRC | 00:23 | |
| *** asettle has quit IRC | 00:24 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: WIP Initial convergence testing https://review.openstack.org/304887 | 00:25 |
| *** nhadzter has quit IRC | 00:29 | |
| *** arif-ali has quit IRC | 00:43 | |
| *** arif-ali has joined #openstack-ansible | 00:50 | |
| *** fishcried has joined #openstack-ansible | 00:53 | |
| *** sdake_ has quit IRC | 00:58 | |
| *** darrenc is now known as darrenc_afk | 01:00 | |
| *** johnmilton has quit IRC | 01:01 | |
| *** fishcried has quit IRC | 01:02 | |
| *** fishcried has joined #openstack-ansible | 01:02 | |
| *** johnmilton has joined #openstack-ansible | 01:08 | |
| *** thorst_ has quit IRC | 01:09 | |
| *** eric_lopez has joined #openstack-ansible | 01:09 | |
| *** elo has joined #openstack-ansible | 01:09 | |
| *** eric_lopez has left #openstack-ansible | 01:10 | |
| *** elo has quit IRC | 01:11 | |
| *** elo has joined #openstack-ansible | 01:14 | |
| *** fishcried has quit IRC | 01:20 | |
| *** lihg has quit IRC | 01:26 | |
| *** darrenc_afk is now known as darrenc | 01:34 | |
| *** jthorne_ has joined #openstack-ansible | 01:40 | |
| *** jthorne has quit IRC | 01:40 | |
| *** ChrisBenson has joined #openstack-ansible | 01:40 | |
| *** busterswt has joined #openstack-ansible | 01:40 | |
| *** fawadkhaliq has quit IRC | 01:43 | |
| *** fawadkhaliq has joined #openstack-ansible | 01:44 | |
| *** thorst_ has joined #openstack-ansible | 01:53 | |
| *** busterswt has quit IRC | 01:54 | |
| *** busterswt has joined #openstack-ansible | 01:57 | |
| *** thorst_ has quit IRC | 01:57 | |
| *** busterswt has quit IRC | 02:08 | |
| *** busterswt has joined #openstack-ansible | 02:12 | |
| *** busterswt has quit IRC | 02:29 | |
| *** sdake has joined #openstack-ansible | 02:32 | |
| *** sdake has quit IRC | 02:32 | |
| *** sdake has joined #openstack-ansible | 02:32 | |
| *** nhadzter has joined #openstack-ansible | 02:34 | |
| *** thorst_ has joined #openstack-ansible | 02:34 | |
| *** busterswt has joined #openstack-ansible | 02:36 | |
| *** asettle has joined #openstack-ansible | 02:37 | |
| *** thorst_ has quit IRC | 02:40 | |
| *** sdake_ has joined #openstack-ansible | 02:42 | |
| *** busterswt has quit IRC | 02:43 | |
| *** sdake has quit IRC | 02:45 | |
| *** tomoe_ has joined #openstack-ansible | 02:53 | |
| *** busterswt has joined #openstack-ansible | 02:54 | |
| *** asettle has quit IRC | 02:57 | |
| *** asettle has joined #openstack-ansible | 02:58 | |
| *** b3rnard0_away is now known as b3rnard0 | 02:58 | |
| *** sdake_ is now known as sdkae | 03:05 | |
| *** shausy has joined #openstack-ansible | 03:12 | |
| *** shausy has quit IRC | 03:21 | |
| *** fawadkhaliq has quit IRC | 03:22 | |
| *** fawadkhaliq has joined #openstack-ansible | 03:23 | |
| *** fishcried has joined #openstack-ansible | 03:39 | |
| *** ChrisBenson has quit IRC | 03:39 | |
| *** ChrisBenson has joined #openstack-ansible | 03:39 | |
| *** severion has quit IRC | 03:40 | |
| *** v1k0d3n has joined #openstack-ansible | 03:40 | |
| *** tomoe_ has quit IRC | 03:44 | |
| *** b3rnard0 is now known as b3rnard0_away | 03:49 | |
| *** elgertam has quit IRC | 03:52 | |
| *** alex______ has joined #openstack-ansible | 03:56 | |
| alex______ | has anyone seen this before on latest liberty branch? repo_build is failing with greenlet dependency problem as greenlet wheel for some reason is built as: | 03:56 |
| alex______ | lrwxrwxrwx 1 nginx www-data 77 Apr 13 04:37 greenlet-0.4.9-cp27-cp27mu-manylinux1_x86_64.whl -> /var/www/repo/pools/greenlet/greenlet-0.4.9-cp27-cp27mu-manylinux1_x86_64.whl | 03:57 |
| alex______ | i.e. weird arch | 03:57 |
| alex______ | this is normal x64 ubuntu 14.04 latest | 03:57 |
| *** jamielennox is now known as jamielennox|away | 03:58 | |
| *** ChrisBenson1 has joined #openstack-ansible | 04:00 | |
| *** ChrisBenson has quit IRC | 04:00 | |
| *** jamielennox|away is now known as jamielennox | 04:06 | |
| *** jayc has joined #openstack-ansible | 04:09 | |
| *** yarkot_ has joined #openstack-ansible | 04:10 | |
| *** weezS has joined #openstack-ansible | 04:16 | |
| *** busterswt has quit IRC | 04:24 | |
| *** yatin has joined #openstack-ansible | 04:33 | |
| *** asettle has quit IRC | 04:37 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ironic: Update ironic.conf for swift and keystone compat https://review.openstack.org/301712 | 04:38 |
| *** shausy has joined #openstack-ansible | 04:58 | |
| *** czunker has joined #openstack-ansible | 05:00 | |
| *** fawadkhaliq has quit IRC | 05:11 | |
| *** fawadkhaliq has joined #openstack-ansible | 05:11 | |
| *** javeriak has joined #openstack-ansible | 05:18 | |
| *** fawadkhaliq has quit IRC | 05:20 | |
| *** fawadkhaliq has joined #openstack-ansible | 05:20 | |
| *** fishcried has quit IRC | 05:21 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 05:32 |
| *** chhavi has joined #openstack-ansible | 05:32 | |
| *** asettle has joined #openstack-ansible | 05:35 | |
| *** javeriak has quit IRC | 05:37 | |
| *** weezS has quit IRC | 05:46 | |
| *** javeriak has joined #openstack-ansible | 05:52 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 05:54 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 06:00 |
| *** jayc has quit IRC | 06:08 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server: Add executable bit to run_tests.sh https://review.openstack.org/301007 | 06:11 |
| *** sdkae is now known as sdake | 06:12 | |
| *** fishcried has joined #openstack-ansible | 06:12 | |
| *** sanjay__u has joined #openstack-ansible | 06:14 | |
| *** v1k0d3n has quit IRC | 06:25 | |
| *** v1k0d3n has joined #openstack-ansible | 06:32 | |
| *** chhavi has quit IRC | 06:41 | |
| *** sanjay__u has quit IRC | 06:48 | |
| *** mikelk has joined #openstack-ansible | 06:49 | |
| *** sanjay__u has joined #openstack-ansible | 06:49 | |
| *** chhavi has joined #openstack-ansible | 06:53 | |
| *** sanjayu___ has joined #openstack-ansible | 06:58 | |
| *** sanjay__u has quit IRC | 07:01 | |
| *** fawadkhaliq has quit IRC | 07:10 | |
| *** admin0 has joined #openstack-ansible | 07:11 | |
| *** asettle has quit IRC | 07:12 | |
| *** yatin has quit IRC | 07:20 | |
| *** asettle has joined #openstack-ansible | 07:24 | |
| *** gparaskevas has joined #openstack-ansible | 07:25 | |
| *** asettle has quit IRC | 07:32 | |
| *** neilus has joined #openstack-ansible | 07:34 | |
| *** sanjayu___ has quit IRC | 07:37 | |
| *** sanjay__u has joined #openstack-ansible | 07:37 | |
| *** sanjayu___ has joined #openstack-ansible | 07:41 | |
| *** yarkot_ has quit IRC | 07:42 | |
| *** sanjay__u has quit IRC | 07:44 | |
| *** sdake_ has joined #openstack-ansible | 07:46 | |
| *** Oku_OS-away is now known as Oku_OS | 07:46 | |
| *** sdake has quit IRC | 07:49 | |
| *** sdake_ is now known as sdake | 07:50 | |
| evrardjp | good morning everyone | 08:08 |
| *** ChrisBenson1 has quit IRC | 08:10 | |
| *** evrardjp has quit IRC | 08:11 | |
| *** fishcried has quit IRC | 08:11 | |
| *** evrardjp has joined #openstack-ansible | 08:11 | |
| *** yatin has joined #openstack-ansible | 08:16 | |
| *** chhavi has quit IRC | 08:18 | |
| *** yatin has quit IRC | 08:20 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-security: Add ability to enable unattended upgrades https://review.openstack.org/304096 | 08:21 |
| *** sanjayu___ has quit IRC | 08:21 | |
| *** yatin has joined #openstack-ansible | 08:22 | |
| *** keedya has quit IRC | 08:28 | |
| *** subscope has joined #openstack-ansible | 08:34 | |
| *** asettle has joined #openstack-ansible | 08:37 | |
| *** chhavi has joined #openstack-ansible | 08:40 | |
| *** javeriak has quit IRC | 08:41 | |
| *** asettle has quit IRC | 08:41 | |
| *** javeriak has joined #openstack-ansible | 08:46 | |
| *** javeriak has quit IRC | 08:53 | |
| *** RTSL has joined #openstack-ansible | 09:01 | |
| admin0 | morning all | 09:02 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: updated ironic playbook inclusion for tftp network https://review.openstack.org/303659 | 09:03 |
| evrardjp | morning admin0 | 09:04 |
| *** persia has quit IRC | 09:04 | |
| *** persia has joined #openstack-ansible | 09:06 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: updated ironic playbook inclusion for tftp network https://review.openstack.org/303659 | 09:08 |
| *** nhadzter has quit IRC | 09:09 | |
| odyssey4me | mrda cloudnull andymccr o/ | 09:10 |
| odyssey4me | lol | 09:10 |
| odyssey4me | whoops | 09:10 |
| odyssey4me | good morning all | 09:10 |
| * odyssey4me goes to get coffee | 09:10 | |
| odyssey4me | alex______ no, we haven't seen that - is that the head of the branch or a tag? | 09:11 |
| odyssey4me | admin0 are you still trying to figure out how to SSL? | 09:11 |
| admin0 | yes sir :) | 09:11 |
| admin0 | odyssey4me: this is the last attempt log: https://gist.github.com/a1git/cb178caea150e49e7e31ba848cd7f39e | 09:11 |
| admin0 | on what cloudnull suggested | 09:11 |
| admin0 | which still fails | 09:12 |
| odyssey4me | admin0 are you building this fresh, or applying it as a patch to an existing environment? | 09:14 |
| alex______ | odyssey4me it's head of liberty branch | 09:14 |
| odyssey4me | alex______ hmm, we last merged a patch into the liberty branch on Apr 8 | 09:15 |
| alex______ | i could use system pip to install greenlet-0.4.9-cp27-cp27mu-manylinux1_x86_64.whl, but pip 7.1 complained.. didn't like package type.. i did this hack: | 09:16 |
| alex______ | diff --git a/playbooks/roles/repo_build/tasks/repo_post_build.yml b/playbooks/roles/repo_build/tasks/repo_post_build.yml index 7dc2cce..c1d4888 100644 --- a/playbooks/roles/repo_build/tasks/repo_post_build.yml +++ b/playbooks/roles/repo_build/tasks/repo_post_build.yml @@ -63,6 +63,15 @@ tags: - repo-create-release-links +- name: Hack to fix up greenlet + file: + dest: "{{ repo_build_release_path }}/{{ repo | 09:16 |
| alex______ | oops | 09:16 |
| odyssey4me | alex______ but we ran some patch tests yesterday - for example https://review.openstack.org/304456 - and it tested just fine | 09:16 |
| alex______ | basically changed the symlink and it worked fine. | 09:16 |
| odyssey4me | alex______ ah, how is it that you ended up with an old version of pip? | 09:16 |
| odyssey4me | the new version should be forcing that to be upgraded | 09:16 |
| admin0 | odyssey4me: its a new environment | 09:17 |
| odyssey4me | was this built previously and you've updated to a newer tag? | 09:17 |
| admin0 | completely new setup | 09:17 |
| alex______ | odyssey4me: /tmp/openstack-venv-builder/venvs/nova/bin/pip was 7.1.. are you saying should have been 8? | 09:18 |
| odyssey4me | alex______ yes, most definitely | 09:18 |
| odyssey4me | alex______ see https://github.com/openstack/openstack-ansible/blob/master/requirements.txt#L11-L13 | 09:18 |
| odyssey4me | or rather, for liberty, https://github.com/openstack/openstack-ansible/blob/liberty/requirements.txt#L11-L13 | 09:19 |
| odyssey4me | if you've just updated from a previous tag, then did you go through all the steps here after you checked out the new tag? http://docs.openstack.org/developer/openstack-ansible/install-guide/app-minorupgrade.html | 09:19 |
| odyssey4me | if it's a fresh environment then I'd love to know how you ended up with the wrong version of pip | 09:20 |
| admin0 | odyssey4me: that was for me or alex with a long tail :D | 09:20 |
| alex______ | odyssey4me: hmm.. let me see if i can reproduce.. now i'm very confused. =) | 09:20 |
| odyssey4me | admin0 are you aware that https://gist.github.com/a1git/cb178caea150e49e7e31ba848cd7f39e#file-gistfile1-txt-L24 sets up LBaaSv1, from which there is no migration of load balancers possible to LBaaSv2 ? | 09:21 |
| admin0 | i will remove that line now | 09:21 |
| admin0 | all i want is to do a fresh start with SSLs :D | 09:21 |
| odyssey4me | I think we did a pretty decent job of noting that in http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-network-services.html#load-balancing-service-optional | 09:21 |
| odyssey4me | ok, admin0 at no time are you actually informing the services that they should have SSL endpoints | 09:22 |
| odyssey4me | you're informing the LB that it needs to be setup for SSL, but that's it | 09:22 |
| alex______ | does anyone know of a way to recover an orphaned cinder volume on nfs (i.e. the volume file exists on an nfs mount, but no meta data or anything on it in the db). | 09:22 |
| admin0 | so how do I do that ? | 09:22 |
| odyssey4me | admin0 do you only want the public endpoint to be SSL? | 09:23 |
| admin0 | if applicable, all .. else only public is good enough | 09:23 |
| admin0 | its a wildcard cert and everything if ssl will alsow rok | 09:23 |
| admin0 | odyssey4me: the most desirable condition is to have all endpoints in SSL | 09:28 |
| odyssey4me | ok, gimme a moment to just verify what's changed in that patch | 09:31 |
| *** javeriak has joined #openstack-ansible | 09:32 | |
| *** yatin has quit IRC | 09:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_barbican: Enable functional convergence testing https://review.openstack.org/301422 | 09:34 |
| odyssey4me | admin0 the protocols to be used for the endpoints are all keyed from these vars: https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/hosts.yml#L97-L100 | 09:34 |
| odyssey4me | it looks like it needs a little work to ensure that *all* endpoints actually use those vars, but many of them do | 09:34 |
| admin0 | so for each service, i need to do this step | 09:34 |
| odyssey4me | admin0 no, just add those vars to your user_variables and make the value https | 09:38 |
| admin0 | aha | 09:39 |
| admin0 | got it | 09:39 |
| admin0 | :D | 09:39 |
| admin0 | will try again | 09:39 |
| admin0 | so this treats horizon differntly | 09:39 |
| *** subscope has quit IRC | 09:39 | |
| *** andymccr_ has joined #openstack-ansible | 09:40 | |
| odyssey4me | yes, all SSL termination is set to be at the load balancer | 09:40 |
| odyssey4me | the patch isn't complete, so some of the endpoints won't be correctly set - if you can pastebin the endpoint list once the build is complete then that'd be helpful | 09:40 |
| *** fishcried has joined #openstack-ansible | 09:41 | |
| *** fishcried has quit IRC | 09:42 | |
| *** yatin has joined #openstack-ansible | 09:42 | |
| *** javeriak_ has joined #openstack-ansible | 09:45 | |
| *** dweaver` has joined #openstack-ansible | 09:45 | |
| *** pjm6 has joined #openstack-ansible | 09:47 | |
| *** jmccrory_ has joined #openstack-ansible | 09:47 | |
| *** javeriak has quit IRC | 09:47 | |
| *** RTSL has quit IRC | 09:47 | |
| *** jthorne_ has quit IRC | 09:47 | |
| *** jmccrory has quit IRC | 09:47 | |
| *** dalees has quit IRC | 09:47 | |
| *** andymccr has quit IRC | 09:47 | |
| *** mancdaz has quit IRC | 09:47 | |
| *** dweaver has quit IRC | 09:47 | |
| *** h1nch has quit IRC | 09:47 | |
| *** jthorne has joined #openstack-ansible | 09:48 | |
| *** jmccrory_ is now known as jmccrory | 09:48 | |
| *** h1nch has joined #openstack-ansible | 09:53 | |
| *** fishcried has joined #openstack-ansible | 10:00 | |
| admin0 | odyssey4me: for now i started with just openstack-ansible haproxy-install.yml — to see if haproxy listens to it properly | 10:01 |
| admin0 | even with those settings, i see ssl only on keystone and horizon | 10:01 |
| admin0 | and on nothing else | 10:01 |
| admin0 | let me gist the latest config, run and what i see | 10:02 |
| admin0 | odyssey4me: https://gist.github.com/a1git/b463c1f1ae20bc039eb31bff6c1e1a19 | 10:03 |
| *** fishcried has quit IRC | 10:04 | |
| *** pjm6 has quit IRC | 10:15 | |
| *** pjm6 has joined #openstack-ansible | 10:16 | |
| *** mancdaz has joined #openstack-ansible | 10:16 | |
| *** asettle has joined #openstack-ansible | 10:18 | |
| *** javeriak_ has quit IRC | 10:19 | |
| *** sdake_ has joined #openstack-ansible | 10:20 | |
| *** sdake has quit IRC | 10:22 | |
| *** asettle has quit IRC | 10:23 | |
| *** deadnull has joined #openstack-ansible | 10:25 | |
| *** sanjay__u has joined #openstack-ansible | 10:26 | |
| admin0 | odyssey4me: this openstack_service_internaluri_proto: https — those are handled by the setup openstack/keystone thing i think .. .. but if haproxy is not even setup @ start listing to https, then endpoints will be there, but non-functional | 10:27 |
| *** chhavi has quit IRC | 10:27 | |
| admin0 | and that patch was about using haproxy wtih ssl for everything | 10:27 |
| *** pjm6 has quit IRC | 10:31 | |
| *** sanjay__u has quit IRC | 10:31 | |
| *** openstackstatus has quit IRC | 10:32 | |
| *** openstack has joined #openstack-ansible | 11:33 | |
| *** pjm6_ has joined #openstack-ansible | 11:33 | |
| *** jwitk0 has joined #openstack-ansible | 11:34 | |
| *** openstackgerrit has quit IRC | 11:35 | |
| *** pjm6 has quit IRC | 11:36 | |
| *** bapalm has quit IRC | 11:36 | |
| *** tiagogomes has quit IRC | 11:36 | |
| *** tiagogomes has joined #openstack-ansible | 11:36 | |
| *** jwitko has quit IRC | 11:37 | |
| yatin | chhavi: i guess...it takes care of db and rabbitmq | 11:37 |
| *** johnmilton has joined #openstack-ansible | 11:37 | |
| yatin | chhavi: look at os_keystone role | 11:37 |
| *** clickboom has joined #openstack-ansible | 11:38 | |
| hughsaunders | evrardjp: reviewed | 11:38 |
| javeriak | hey odyssey4me; so our setup went back to the sluggish state the next morning; after working fine for a few hours; nothing was done on it meanwhile but seems like some kind of buildup happened meanwhile :( | 11:39 |
| *** bapalm has joined #openstack-ansible | 11:41 | |
| odyssey4me | chhavi yes, the implementation of the db and rabbitmq details is done as they are required for keystone to work | 11:42 |
| odyssey4me | chhavi also, other dependencies (such as pip, mariadb client, etc) are also installed because they're required | 11:42 |
| odyssey4me | javeriak hmm, that's weird - perhaps there's some sort of memory allocation limit that's been reached? | 11:43 |
| chhavi | ok, then i need to check if i am missing something | 11:43 |
| javeriak | yes that seems likely | 11:43 |
| odyssey4me | javeriak I've engaged with the Performance Team and agreed that we'll setup a test plan to try to find some sensible defaults and isolate performance limits, etc - but that will only happen next cycle so it doesn't really help right now | 11:43 |
| *** Oku_OS is now known as Oku_OS-away | 11:44 | |
| javeriak | odyssey4me or some kind os queue's got filled up | 11:44 |
| odyssey4me | javeriak do you perhaps have some sort of monitoring in place (or could you put something in place) to monitor memory usage, cpu usage, mariadb performance metrics and rabbitmq performance metrics? | 11:44 |
| javeriak | odyssey4me no i didnt setup something yet, but you're right, let me do that now | 11:45 |
| javeriak | ive been trying to cleanup for more than a day now, and even just deleting all the stuff in openstack is taking ages | 11:45 |
| odyssey4me | mancdaz is this failure happening in the initial build of mariadb/galera with v5.5 (not 10)? | 11:46 |
| mancdaz | odyssey4me the non-upgrade part of the test is done with 10, and that's what I'm running | 11:46 |
| mancdaz | so I've just commented out the upgrade play, essentially, to test this problem | 11:47 |
| odyssey4me | ok, so this is a 10 only issue then - I thought that perhaps the issue may be related to us using the same tasks/templates for both versions and that there may be an issue with doing that | 11:48 |
| odyssey4me | javeriak yeah, if we can have those monitored and see trends then that would really help to isolate the issues | 11:48 |
| javeriak | odyssey4me im afraid we only have two more days with it :(; not much time to look into this further | 11:49 |
| *** nhadzter has quit IRC | 11:49 | |
| *** retreved has joined #openstack-ansible | 11:49 | |
| odyssey4me | mancdaz would this perhaps relate to the version of percona-xtrabackup we're using? isn't that what handles the sst? | 11:49 |
| odyssey4me | the mariadb version will be the latest, and may have an issue using that old version | 11:50 |
| mancdaz | odyssey4me I'm currnetly testing with the latest version of xtrabackup-v2. Not convinced that's the issue but hopefully it will be | 11:50 |
| *** openstackgerrit has joined #openstack-ansible | 11:50 | |
| odyssey4me | javeriak if you look at the logs when you're processing deletions/creations, are you seeing any disconnects/reconnects or anything else that would indicate that something's having to wait up | 11:51 |
| odyssey4me | it may just be that due to the tests you're running those values need to be a lot higher? | 11:51 |
| *** sdake is now known as sdake_busy_webin | 11:51 | |
| odyssey4me | mancdaz ah ok, as I recall there needed to be some config changes to make that work - which is why we ended up pinning it | 11:52 |
| mancdaz | odyssey4me do you happen to remember what they were? | 11:52 |
| odyssey4me | realistically I'd like us to move to using MariaDB 10.1 for master as it includes a lot of the extra packages by default... but the current issue needs to be resolved without that as it'll apply to the Liberty & Mitaka branches | 11:52 |
| odyssey4me | mancdaz we did the pin because there wasn't time to figure that out :/ | 11:53 |
| mancdaz | ha | 11:53 |
| *** thorst_ has joined #openstack-ansible | 11:53 | |
| *** gfa is now known as gfa_ | 11:53 | |
| *** gfa_ is now known as gfa | 11:53 | |
| odyssey4me | I think a new breaking change version was released just before we were scheduled to release liberty, so we pinned. | 11:53 |
| odyssey4me | unfortunately we didn't circle back afterwards to unpin and resolve that | 11:54 |
| admin0 | stable/mitaka failing @ TASK: [openstack-ansible-security | V-38496 - Default operating system accounts (other than root) must be locked] *** | 11:54 |
| admin0 | -vvvv fails to show that acocun tin question | 11:55 |
| *** weshay has joined #openstack-ansible | 11:55 | |
| odyssey4me | admin0 you can disable the hardening if you like to move it along, I suspect that issue has been resolved already - I'm working on a tool to help with the SHA bumping of roles today and the bump for the 13.0.1 release will include the fix | 11:55 |
| *** mouchon has quit IRC | 11:56 | |
| *** neilus1 has joined #openstack-ansible | 11:56 | |
| admin0 | doing that | 11:56 |
| *** sanjay__u has quit IRC | 11:58 | |
| *** neilus has quit IRC | 11:59 | |
| *** yatin has quit IRC | 12:00 | |
| pjm6_ | if I can use openstack-ansible invetory dynamic file generated, its possible to use without openstack-ansible script? | 12:00 |
| hughsaunders | pjm6_: you want to use openstack_inventory.json without the OSA dynamic_inventory.py script? | 12:02 |
| evrardjp | hughsaunders thanks | 12:02 |
| pjm6_ | hughsaunders: I think it is. I'm developing a task in ansible to deploy my custom dashboard to the current horizon_containers | 12:03 |
| pjm6_ | but I have to use the inventory.json file | 12:03 |
| odyssey4me | pjm6_ it is possible to use alternative inventories if you want to - ansible expects the inventory in an ini format, or a json format - we use the ini format in the role tests - what're you actually hoping to achieve? | 12:04 |
| pjm6_ | odyssey4me: is to use the info of the horizon containers (that are in that file) for uploading my custom dashboard and install the libs in pip | 12:04 |
| pjm6_ | in manner fact, I don't wanna to change the inventory | 12:04 |
| pjm6_ | only fetch the hosts info (in this case horizons host) | 12:05 |
| *** pjm6_ is now known as pjm6 | 12:05 | |
| odyssey4me | pjm6 ok, are you trying to copy a set of files over to the horizon hosts? and do these files need to contain some sort of info from the inventory? | 12:06 |
| pjm6 | Yes | 12:07 |
| pjm6 | well the only info that need is the IP | 12:07 |
| pjm6 | for making the SSH connection | 12:07 |
| odyssey4me | pjm6 ok, is your code having to make an ssh connection? | 12:07 |
| pjm6 | well i think it needs for installing the PIP requirements | 12:08 |
| odyssey4me | this seems like a pretty bad thing to be doing in website, FYI | 12:08 |
| pjm6 | really ? | 12:08 |
| odyssey4me | pjm6 ah, no need to do that - are those requirements in pypi, or in a private repo? | 12:08 |
| pjm6 | so what would be the best approach to copy the custom dashboard? | 12:08 |
| odyssey4me | and are they accessible via http/https? | 12:08 |
| pjm6 | in a private repo | 12:08 |
| pjm6 | Yes | 12:08 |
| pjm6 | I put in that way (because I was seeing how you do) and seems more automatic | 12:09 |
| odyssey4me | alright, and is that private repo setup so that you could use pip install against it? | 12:09 |
| pjm6 | Yes | 12:09 |
| *** thorst_ has quit IRC | 12:09 | |
| pjm6 | basically what I wanna to do is | 12:09 |
| *** chhavi has quit IRC | 12:09 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: Fixing keepalived bug when 2+ backup nodes have the same priority https://review.openstack.org/279730 | 12:09 |
| *** Oku_OS-away is now known as Oku_OS | 12:09 | |
| odyssey4me | alright, then I expect that you want the pip files installed into the horizon venv | 12:09 |
| pjm6 | Sure :) | 12:09 |
| pjm6 | Ensure that database is installed with empty data (run only once), install pip files inside horizon venv, copy dashboard and enabled file to the horizon hosts, and then restart apache2 | 12:10 |
| odyssey4me | are the customisations for horizon built in the manner that custom modules are done? https://github.com/openstack/openstack-ansible-os_horizon/blob/master/tasks/horizon_post_install.yml#L96-L117 | 12:10 |
| pjm6 | but do you think this is a bad approach? | 12:10 |
| pjm6 | well I read that but i'm not sure if is the same thing | 12:11 |
| pjm6 | a module is different from a dashboard? | 12:11 |
| pjm6 | https://github.com/openstack/horizon/tree/master/openstack_dashboard/dashboards => i created an extra folder here | 12:12 |
| odyssey4me | ok, putting the files into a horizon venv is easy enough - put this list var into your user_variables.yml file, and add the other packages to it, then you need to give the repo server access to your private repo, let me find the correct var to set - gimme a minute | 12:12 |
| odyssey4me | pjm6 which release/tag/branch are you installing? | 12:12 |
| *** clickboom has quit IRC | 12:12 | |
| pjm6 | thanks odyssey4me. | 12:13 |
| pjm6 | I'm using stable/liberty | 12:13 |
| odyssey4me | pjm, hmm - this will be harder with liberty than it is with mitaka | 12:15 |
| pjm6 | odyssey4me: but its hard to use another playbook (or task) in ansible | 12:16 |
| odyssey4me | in mitaka we allow the repo build process to make use of a list of extra indexes to find packages: https://github.com/openstack/openstack-ansible-repo_build/blob/stable/mitaka/defaults/main.yml#L55-L57 | 12:16 |
| pjm6 | that use the | 12:16 |
| *** automagically has joined #openstack-ansible | 12:16 | |
| pjm6 | hosts: horizon_all from that file | 12:16 |
| odyssey4me | we don't have that in liberty | 12:16 |
| pjm6 | hmm nice :) i'm seeing, that is more extensible | 12:16 |
| odyssey4me | so yeah - as way you can do it is just to create a playbook that does whatever you need to do, then ensure that you have an ansible.cfg in the same folder as the playbook which tells ansible to make use of the dynamic inventory | 12:17 |
| odyssey4me | so, for instance, you can have your playbooks in /opt/my_custom_stuff/ | 12:17 |
| pjm6 | ahh now I see what i need that file :D | 12:17 |
| pjm6 | but that would be a bad practice/choice, right? | 12:18 |
| odyssey4me | then you copy https://github.com/openstack/openstack-ansible/blob/liberty/playbooks/ansible.cfg over to that folder, change the paths to be explicit, for example lookup_plugins should have the value /opt/openstack-ansible/playbooks/plugins/lookups | 12:18 |
| odyssey4me | the inventory should point to /opt/openstack-ansible/playbooks/inventory | 12:19 |
| odyssey4me | that way you can have any playbooks doing whatever you want, and they can still use the whole inventory | 12:19 |
| pjm6 | Thanks odyssey4me :D I will try that | 12:19 |
| pjm6 | but why that will be a bad thing? will exists another alternative? | 12:20 |
| odyssey4me | it's not bad practise at all - in fact this is far easier to do that trying to fork the repo to add your own things | 12:20 |
| mhayden | buenos dias | 12:20 |
| pjm6 | I was thinking in using ansible because doing it manually accross all would be "massive" | 12:20 |
| pjm6 | yeah odyssey4me, with this i'm trying not to break the openstack-ansible scripts | 12:21 |
| pjm6 | and be able to add my costumizations :) | 12:21 |
| pjm6 | hi mhayden | 12:21 |
| *** automagically has quit IRC | 12:21 | |
| odyssey4me | pjm6 great - ideally to simplify doing this in the future you can consider how the existing playbooks/roles could reduce your need for custom playbooks, then you can submit a patch to the horizon role which would work for you and others who want to do similar things | 12:23 |
| odyssey4me | generally if you have to do your own things, then idea is to try and work with the community to find a solution in a future version so that once it's released you can reduce the amount of custom stuff you have to carry alone and we can all improve things together as a community | 12:23 |
| pjm6 | yes odyssey4me :) i was thinking in that | 12:24 |
| pjm6 | but because my ansible acknowledgement is not too much | 12:25 |
| pjm6 | i didn't see that, but as far as I learn more and more, I would like to do that yes .D | 12:25 |
| joker_ | I got ERROR: Inventory script https://gist.github.com/celikmustafa89/c708065bc566714336c8f60fb43e936f | 12:26 |
| joker_ | can someone help me? thanks | 12:26 |
| odyssey4me | joker_ do you perhaps have hosts in openstack_user_config which are named 'infra' or 'haproxy', or the same as any of the group names? | 12:28 |
| odyssey4me | otherwise it'd be best to add your openstack_user_config to the gist so that it can be inspected | 12:28 |
| *** chhavi has joined #openstack-ansible | 12:30 | |
| *** b3rnard0_away is now known as b3rnard0 | 12:37 | |
| joker_ | odyssey4me: https://gist.github.com/celikmustafa89/3f50fea4690938867716a7adaaf01474 | 12:37 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 12:38 |
| joker_ | odyssey4me : https://gist.github.com/celikmustafa89/3f50fea4690938867716a7adaaf01474 | 12:39 |
| joker_ | odyssey4me I think there is no such a conflict | 12:40 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: Improved logging for memcached https://review.openstack.org/305165 | 12:41 |
| *** thorst_ has joined #openstack-ansible | 12:42 | |
| *** thorst_ has quit IRC | 12:42 | |
| admin0 | setup-infra failed because it could it restart haproxy .. haproxy already running .. i had to do apt-get autoremove --purge haproxy to get it going again .. i normally run the haproxy playbook first before I run setup-hosts .. so something changed here as well | 12:43 |
| admin0 | it cound not* | 12:43 |
| odyssey4me | joker_ hmm, ok - that seems to look fine - do you have an inventory json file in /etc/openstack_deploy yet? | 12:44 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 12:44 |
| odyssey4me | joker_ also, is this a new environment? or is it an environment which you had setup before and have made inventory changes to? | 12:44 |
| odyssey4me | admin0 interesting - the haproxy playbook is now included in the setup-infrastructure playbook, so you don't have to run it seperately | 12:45 |
| odyssey4me | but it is weird that a re-run of the playbook caused a problem | 12:45 |
| joker_ | odyssey4me it's new | 12:45 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: Improved logging for memcached https://review.openstack.org/305165 | 12:45 |
| joker_ | odyssey4me there is openstack_inventory.json file | 12:46 |
| odyssey4me | joker_ ok, so I expect that if there is an existing fact cache or inventory then it's expendable? if so, check for the inventory json file and the fact cache and remove them | 12:46 |
| admin0 | it was not the case with liberty .. and also, i had to run it because it failed @ some repo stuff where it was lookign for a local repo endpoint provided by that step .. so i had that habit to run it beforehand | 12:46 |
| odyssey4me | it might be that you had something previously, then an edit changed something and it got muddled | 12:46 |
| odyssey4me | admin0 yep, although I think we backported that inclusion into liberty recently - can't recall | 12:47 |
| odyssey4me | but anyway, it's now included - the issue with the repo/packages will need looking into to isolate the issue | 12:47 |
| admin0 | as long as it works( either way) i am fine with | 12:47 |
| odyssey4me | the playbook should be possible to run over and over again | 12:47 |
| *** klamath has joined #openstack-ansible | 12:49 | |
| joker_ | odyssey4me : should I remove "openstack_inventory.json" ?? | 12:49 |
| *** klamath has quit IRC | 12:49 | |
| *** klamath has joined #openstack-ansible | 12:49 | |
| admin0 | oops :D | 12:50 |
| admin0 | when you re-run the playbooks, then your containers will be re-created with new ips | 12:50 |
| admin0 | if you do that | 12:50 |
| admin0 | i mean brand new containers .. | 12:52 |
| admin0 | crashing the controllers ( if not enough ram ) | 12:52 |
| odyssey4me | joker_ it may be safer to move it somewhere, but yes | 12:52 |
| odyssey4me | and clear the fact cache too, if there is one | 12:52 |
| odyssey4me | joker_ I'm assuming that you don't have any containers or anything setup yet | 12:53 |
| joker_ | odyssey4me : ı did not run yet | 12:53 |
| joker_ | no container nothing | 12:53 |
| joker_ | by fact, you mean ansible_facts directory, right? | 12:53 |
| joker_ | there is no file in that directory | 12:53 |
| *** schwicht has joined #openstack-ansible | 12:54 | |
| javeriak | odyssey4me re-running the nova plays has fixed it again, i didnt even change any config this time... this is off | 12:55 |
| javeriak | odd* | 12:55 |
| joker_ | odyssey4me here is ls output under /etc/openstack_deploy directory https://gist.github.com/celikmustafa89/9bb2834027b956edd7c5f30cd6509ecc/revisions | 12:55 |
| odyssey4me | javeriak odd, did any of the services restart in the play execution? | 12:56 |
| javeriak | i think it has a handler to all the nova ones | 12:56 |
| odyssey4me | javeriak sure, but those shouldn't fire if no configs were changed - unless we happen to have an issue with properly keeping things idempotent, which is possible | 12:57 |
| odyssey4me | joker_ so clear the contents of the ansible_facts directory | 12:58 |
| javeriak | odyssey4me but why see this is only a large install | 12:58 |
| odyssey4me | and move the inventory json file out somewhere as a backup | 12:58 |
| odyssey4me | joker_ another thing you can actually do is execute the dynamic inventory directory - it'll spit out a resulting json file to stdout if the config is valid | 12:58 |
| joker_ | odyssey4me ok | 12:59 |
| joker_ | it worked | 12:59 |
| joker_ | thank you | 12:59 |
| odyssey4me | right, so I expect that the old inventory was broken in some way based on config fiddles | 12:59 |
| odyssey4me | you could compare the new and old one to see where if you're keen to understand, but it seems rather academic at this point :p | 12:59 |
| joker_ | I destroyed the old one :D | 13:00 |
| *** jaypipes has quit IRC | 13:07 | |
| *** sdake has joined #openstack-ansible | 13:10 | |
| *** sdake_busy_webin has quit IRC | 13:12 | |
| *** elgertam has joined #openstack-ansible | 13:12 | |
| joker_ | odyssey4me because of name conflict. I gave haproxy name as haproxy at the beginning. As you said before it is related with that name conflict. thank you again ... | 13:12 |
| *** sdake_ has joined #openstack-ansible | 13:13 | |
| odyssey4me | joker_ aha, that makes sense - we have a bug for that, but haven't yet settled on a solution I don't think | 13:13 |
| *** sdake has quit IRC | 13:15 | |
| joker_ | odyssey4me : can you share the bug link? | 13:16 |
| odyssey4me | joker_ sure - https://bugs.launchpad.net/openstack-ansible/+bug/1512883 | 13:17 |
| openstack | Launchpad bug 1512883 in openstack-ansible "Host name of haproxy in openstack_user_config.yml causes recursion error" [Medium,In progress] - Assigned to Ala Raddaoui (raddaoui-ala) | 13:17 |
| *** sdake_ is now known as sdake | 13:18 | |
| *** sdake is now known as sdake_busy | 13:18 | |
| *** busterswt has joined #openstack-ansible | 13:21 | |
| *** Mudpuppy has joined #openstack-ansible | 13:24 | |
| *** Bjoern_ has joined #openstack-ansible | 13:25 | |
| *** Bjoern_ is now known as Bjoern_zZzZzZzZ | 13:25 | |
| *** automagically has joined #openstack-ansible | 13:28 | |
| *** jthorne has quit IRC | 13:30 | |
| *** shausy has quit IRC | 13:30 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 13:30 |
| joker_ | odyssey4me : I am getting this error over and over again. I guess, it upgrades pip to 8.1.1. But in the next steps, installation needs pip 7.2.1 ---- Is that normal? ----https://gist.github.com/celikmustafa89/b461461201a014f6a7fed68127d2eae1 | 13:31 |
| *** jthorne has joined #openstack-ansible | 13:31 | |
| odyssey4me | joker_ no, that's not normal - looking | 13:31 |
| odyssey4me | joker_ this is liberty, right? | 13:32 |
| joker_ | yes | 13:32 |
| odyssey4me | can you include the part of the log where it ends up choosing the older pip version? | 13:32 |
| *** pjm6_ has joined #openstack-ansible | 13:34 | |
| *** jthorne_ has joined #openstack-ansible | 13:34 | |
| joker_ | I am not there yet. I ll share it when I come that step. That was the error I got my previous ansible installation which was crushed. | 13:35 |
| *** kysse_ has joined #openstack-ansible | 13:36 | |
| *** retreved_ has joined #openstack-ansible | 13:37 | |
| evrardjp | thanks for the quick review... I didn't expect one so soon ;) | 13:37 |
| odyssey4me | joker_ that first set should fail and the failures be ignored, it should then fall back to installing in isolated mode and get the right version from pypi | 13:37 |
| odyssey4me | joker_ FYI we've cleaned that up quite substantially in Mitaka | 13:38 |
| joker_ | Ok | 13:38 |
| evrardjp | my last comment was for automagically | 13:39 |
| joker_ | thank you very much | 13:39 |
| *** keedya has joined #openstack-ansible | 13:39 | |
| joker_ | odyssey4me : ı want to ask another thing about haproxy. under the global_overrides: heading. I am confusing at internal_lb_vip_addresss and external_lb_vip_address ? what am I have to write there? here is my config file -> https://gist.github.com/celikmustafa89/3f50fea4690938867716a7adaaf01474 | 13:40 |
| odyssey4me | joker_ the internal address should match the haproxy host's address, unless you have more than one in which case you need to make it the keepalived vip address | 13:41 |
| odyssey4me | joker_ the external address is only used for the service catalog and keystone will show it as the address used for the public endpoints | 13:41 |
| odyssey4me | joker_ I'm not sure, but I think that the external address can simply be a DNS name if you want | 13:42 |
| *** xar-- has joined #openstack-ansible | 13:43 | |
| joker_ | odyssey4me : I got the following error when ı run setup-hosts.yml | 13:43 |
| joker_ | fatal: [b3lab_haproxy] => SSH Error: data could not be sent to the remote host. Make sure this host can be reached over ssh | 13:43 |
| *** jthorne has quit IRC | 13:43 | |
| *** retreved has quit IRC | 13:43 | |
| *** pjm6 has quit IRC | 13:43 | |
| *** dmellado has quit IRC | 13:43 | |
| *** arbrandes1 has quit IRC | 13:43 | |
| *** xar- has quit IRC | 13:43 | |
| *** kysse has quit IRC | 13:43 | |
| *** admiralboom has quit IRC | 13:43 | |
| *** dmellado_ has joined #openstack-ansible | 13:43 | |
| *** mgoddard_ has joined #openstack-ansible | 13:45 | |
| *** clickboom has joined #openstack-ansible | 13:48 | |
| *** mgoddard has quit IRC | 13:49 | |
| *** admiralboom has joined #openstack-ansible | 13:50 | |
| *** arbrandes1 has joined #openstack-ansible | 13:50 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:51 | |
| *** KLevenstein has joined #openstack-ansible | 13:53 | |
| admin0 | i am also hitting the “msg: the python mysqldb module is required” | 13:53 |
| *** thorst_ has joined #openstack-ansible | 13:55 | |
| odyssey4me | admin0 hmm, that's with mitaka, right? | 13:56 |
| admin0 | stable/mitaka :) | 13:57 |
| odyssey4me | admin0 ah, we fixed that but I haven't updated the SHA's yet - lemme dig up that fix quickly | 13:57 |
| admin0 | you can just tell me the tag/branch that works :D | 13:57 |
| odyssey4me | admin0 https://github.com/openstack/openstack-ansible-repo_build/commit/0ce640328e484fcb442fdae66bb7edee9ce4d232 | 13:57 |
| *** schwicht has quit IRC | 13:57 | |
| odyssey4me | admin0 I'll be submitting a patch shortly which will most likely become 13.0.1 and that will include all the updates we've made since 13.0.0 to fix issues like these | 13:58 |
| admin0 | ok | 13:58 |
| admin0 | i will wait for that then | 13:58 |
| javeriak | odyssey4me seen this before: Request failed: <urlopen error [Errno -2] Name or service not known for 'Get Modern PIP using fallback URL' ? | 13:58 |
| admin0 | play “league of legends” to kill time :D | 13:58 |
| odyssey4me | mitaka is still a bit experimental as it's a fresh release, but that's why we've been asking for more testers :) | 13:58 |
| Mudpuppy | league of ragers :) | 13:59 |
| odyssey4me | joker_ have you replicated your public ssh key (/root/.ssh/id_rsa.pub) from the deployment host to all the other hosts? | 13:59 |
| joker_ | ok | 14:00 |
| *** mgoddard_ has quit IRC | 14:00 | |
| *** mgoddard has joined #openstack-ansible | 14:00 | |
| *** jayc has joined #openstack-ansible | 14:00 | |
| odyssey4me | joker_ if you don't do that, then the containers don't get the key and everything goes sideways | 14:00 |
| admin0 | Mudpuppy: its to experience all human emotions: rage, helplessness, betrayal, accomplishment, anger — where else can you get all of that in 30 mins :D | 14:00 |
| odyssey4me | if the containers are already built, you'll have to destroy and rebuild them because the key gets added at build time | 14:00 |
| joker_ | ok | 14:01 |
| joker_ | I got it | 14:01 |
| *** czunker has quit IRC | 14:02 | |
| *** persia has quit IRC | 14:02 | |
| *** itsuugo has quit IRC | 14:02 | |
| automagically | I’ve been doing some testing with Mitaka lately as well FWIW | 14:03 |
| *** ametts has joined #openstack-ansible | 14:03 | |
| odyssey4me | automagically excellent, good to know | 14:04 |
| *** gfa has quit IRC | 14:04 | |
| *** gfa has joined #openstack-ansible | 14:04 | |
| odyssey4me | admin0 a way that you can progress is to change all the SHA's set in ansible-role-requirements.yml to be 'stable/mitaka' - that'll include all the most recent fixes | 14:04 |
| admin0 | odyssey4me: is there a one-liner command that i can copy/paste :D | 14:05 |
| admin0 | and be done with it ? | 14:05 |
| odyssey4me | I'm just trying to figure out how to make python's yaml.dump to output the ansible-role-requirements file in the way that I want it... I'm stuck on trying to get the dicts to be sorted in the output :/ | 14:05 |
| odyssey4me | admin0 gimme a sec to get that for you | 14:05 |
| *** itsuugo has joined #openstack-ansible | 14:08 | |
| *** clickboom has quit IRC | 14:09 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: Improved logging for memcached https://review.openstack.org/305165 | 14:09 |
| *** thorst_ has quit IRC | 14:10 | |
| *** clickboom has joined #openstack-ansible | 14:11 | |
| *** gfa is now known as gfa_ | 14:11 | |
| odyssey4me | admin0 in /opt/openstack-ansible execute this: sed 's|version: [[:alnum:]].*|version: stable/mitaka|' ansible-role-requirements.yml | 14:12 |
| odyssey4me | automagically logan- your thoughts on merging https://review.openstack.org/302385 would be appreciated - liberty is stable so we'd like diverse input before simply adding a new 'feature' this late in the game - please vote and add comments in the review | 14:13 |
| * automagically looking | 14:13 | |
| odyssey4me | Anyone else who has a liberty environment running should also please add a vote | 14:14 |
| odyssey4me | preferably with a comment indicating why you voted what you did | 14:14 |
| admin0 | if you have a liberty in production already, would people feel comfortable to run this playbook .. because it locks down stuff … it might break what you have ( at least that is the feeling ) | 14:17 |
| admin0 | or to be in production, they might have already hardened some parts themselves, which this might break | 14:17 |
| odyssey4me | admin0 by default it will not change anything because the hardening is disabled by default | 14:18 |
| odyssey4me | so basically the whole role will be skipped, unless someone enables hardening | 14:18 |
| automagically | admin0: I provided commentary on my vote, but effectively operators running Liberty will make two choices, both to upgrade to the latest Liberty tag/release as well as to enable the role explicitly | 14:19 |
| odyssey4me | so a deployer would have to specifically enable this | 14:19 |
| automagically | In both cases, I’d expect operators to use best practices and test these changes in a non-production environment prior to applying in production | 14:19 |
| odyssey4me | ++ | 14:19 |
| odyssey4me | ok, I need some help from anyone who's better at python than I am (which is probably just about everyone) - I've built a CLI tool to do small edits of the ansible-role-requirements file which has a very well known yaml structure | 14:20 |
| odyssey4me | it's a list of dicts | 14:20 |
| admin0 | yes but again, your test env will be like 10 nodes and prodiuction like 50+ with 100s of customers or 1000s of vms .. and with security hardening like these ( which appear post deployments ) its uncomfortable because you cannot test everything that you do in production | 14:20 |
| odyssey4me | the trouble is that whenever I write the changed file out, the dicts are never ordered because python dicts aren't ordered | 14:20 |
| odyssey4me | so the dict structure order changes every time I run the tool | 14:21 |
| odyssey4me | I'd like to have the yaml.dump output the dicts in a sorted order. | 14:22 |
| *** asettle has joined #openstack-ansible | 14:22 | |
| odyssey4me | I'm stuck on figuring out how to do this without doing something stupid and basically writing stuff to the file without using the dumper. | 14:22 |
| odyssey4me | it looks like maybe I can adjust how the dumper handles the output using a constructor? I'm not sure if that makes sense though - I'd like some direction from anyone who can comprehend http://pyyaml.org/wiki/PyYAMLDocumentation#DumpingYAML better than I can | 14:23 |
| odyssey4me | hughsaunders jmccrory automagically ^ perhaps you guys can provide advice? | 14:23 |
| *** elgertam has quit IRC | 14:23 | |
| odyssey4me | git-harry cloudnull ^ | 14:24 |
| odyssey4me | admin0 if that's the case, then don't enable it :p | 14:24 |
| * mhayden high fives automagically :) | 14:24 | |
| admin0 | my saying is, set it to false if it was false before, and then give the option to test and set it to true :) | 14:25 |
| automagically | admin0: And that is how it will ship in Liberty if approved for merge | 14:26 |
| odyssey4me | admin0 that's exactly what the patch does - it only turns it on for the AIO and therefore gating | 14:26 |
| automagically | odyssey4me: Taking a look at your YAML dumping problem | 14:26 |
| admin0 | ok | 14:26 |
| *** asettle has quit IRC | 14:27 | |
| automagically | odyssey4me: Maybe give this a whirl: http://www.grantjenks.com/docs/sortedcontainers/sorteddict.html | 14:27 |
| automagically | Or, better yet: https://docs.python.org/2/library/collections.html#collections.OrderedDict | 14:28 |
| odyssey4me | automagically so the trouble is not sorting the dict - that's easy - the trouble is that I have a yaml object with the data in it, and the dumper doesn't give me control over the output of the structures | 14:28 |
| automagically | Ah, so dumping a sorted dict results in it “unsorting" | 14:29 |
| odyssey4me | yep | 14:29 |
| odyssey4me | hmm, it seems that this looks like it could do it | 14:30 |
| odyssey4me | https://sourceforge.net/p/yaml/mailman/message/23596068/ | 14:30 |
| odyssey4me | lemme give it a bash | 14:30 |
| *** pjm6 has joined #openstack-ansible | 14:31 | |
| automagically | Yep, that looks promising | 14:31 |
| *** spotz_zzz is now known as spotz | 14:33 | |
| *** clickboom has quit IRC | 14:33 | |
| *** pjm6_ has quit IRC | 14:33 | |
| odyssey4me | hmm, on quick inspection that appears to have worked :) | 14:33 |
| * odyssey4me does a happy dance | 14:34 | |
| admin0 | haha :) [Set requirement names >> Set stripped global requirement names >> Set stripped local requirement names >>Set global normalized requirement names >> Set local normalized requirement names >> Set filtered requirement names ] — as a deployer i am like huh ? why not “fixing names” | 14:34 |
| *** elo has quit IRC | 14:34 | |
| *** pjm6 has quit IRC | 14:35 | |
| *** elo has joined #openstack-ansible | 14:35 | |
| *** pjm6 has joined #openstack-ansible | 14:35 | |
| cloudnull | morning | 14:35 |
| admin0 | morning | 14:35 |
| cloudnull | hows it ? | 14:36 |
| admin0 | raining outside, no ssl inside | 14:36 |
| admin0 | :D | 14:36 |
| cloudnull | hahaa | 14:36 |
| cloudnull | raining here too | 14:36 |
| *** thorst_ has joined #openstack-ansible | 14:37 | |
| cloudnull | hailing last night | 14:37 |
| admin0 | i was enjoing the playbooks .. seeing [Set requirement names >> Set stripped global requirement names >> Set stripped local requirement names >>Set global normalized requirement names >> Set local normalized requirement names >> Set filtered requirement names ] and thinking .. why not just write “fixing names” and be done with it | 14:37 |
| odyssey4me | cloudnull yeah, took a look through the ssl enablement patch earlier and it still needs much work | 14:37 |
| cloudnull | they are, in the end, all used for different purposes | 14:37 |
| odyssey4me | check the review comments :) | 14:38 |
| *** thorst_ has quit IRC | 14:38 | |
| cloudnull | will do in a bit... | 14:38 |
| *** keedya has quit IRC | 14:38 | |
| cloudnull | still consuming coffee | 14:38 |
| hughsaunders | odyssey4me: happy to help with python script, if you have a link to the script and a description of your aim.. | 14:38 |
| *** keedya has joined #openstack-ansible | 14:39 | |
| admin0 | odyssey4me: grep version /opt/openstack-ansible/ansible-role-requirements.yml — gives me 29 lines of mitaka and nothing else .. .. but when i run the playbooks again and again, still fails at “the python mysqldb module is required” | 14:39 |
| admin0 | i will delete all containers and retry | 14:40 |
| admin0 | just to see if its something exsisting | 14:40 |
| *** pjm6 has quit IRC | 14:41 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: Move inventory environment loading to function https://review.openstack.org/290740 | 14:42 |
| odyssey4me | admin0 can you provide a verbose log of the playbook execution that results in that error? I'd like to see where it hits the issue | 14:44 |
| javeriak | guys needs some help, a rerun of playbooks is failing of 'Get Modern PIP using fallback URL' ; pip is already there on these nodes so dont knw why it wants to try and reinstall | 14:44 |
| *** yarkot_ has joined #openstack-ansible | 14:45 | |
| cloudnull | if cores have a moment to review https://review.openstack.org/#/c/304385/ -- the ceph ansible people are waiting on that to go in for their upcoming release -- not that it's our top priority but it'd be good to be neighborly if at all possible. | 14:45 |
| cloudnull | javeriak: is the repo server down or being rebuilt ? if you remove /root/pip/pip.conf and itll install from upstream. | 14:46 |
| cloudnull | admin0 odyssey4me looking at the ssl bits now. | 14:47 |
| cloudnull | im rekicking my dev lab so will be a min | 14:47 |
| evrardjp | I got a weird issue, I don't know if anybody got that issue already: I do openstack-ansible -e debug=False, and in my rendered template I see False, but when I put debug: False in a file named user_jph.yml with the standard openstack-ansible CLI I see true rendered | 14:47 |
| automagically | Ah yes, I saw the ceph-ansible Github issue around that. Reviewing now | 14:47 |
| cloudnull | thanks automagically | 14:48 |
| *** toddnni has quit IRC | 14:48 | |
| javeriak | cloudnull repo servers are up, dont know why it wants to install pip everything, let me just remove the conf | 14:48 |
| evrardjp | the file is well listed when I see the first line of the openstack-ansible line | 14:48 |
| cloudnull | ansible -m shell -a 'rm /root/pip/pip.conf' $NAMEOFGROUP | 14:48 |
| evrardjp | like it's overwritten by something | 14:48 |
| odyssey4me | evrardjp user_jph is read before user_variables, so a value in user_variables is likely overriding it | 14:49 |
| evrardjp | I thought nothing should override the user_* variables | 14:49 |
| evrardjp | no way | 14:49 |
| cloudnull | evrardjp: yes | 14:49 |
| evrardjp | :) | 14:49 |
| evrardjp | I meant no way I did't look at that | 14:49 |
| odyssey4me | evrardjp the user_* files are read in alphanumeric order | 14:49 |
| cloudnull | the filenames are alpha sorted. | 14:49 |
| cloudnull | ha. what odyssey4me said :) | 14:50 |
| evrardjp | dammit | 14:50 |
| evrardjp | thanks guys | 14:50 |
| evrardjp | I remember why I used user_variables_jph in the past now | 14:50 |
| odyssey4me | they're read in alphanumeric order by bash, then ansible merges them over the top of each other in the order it's given the files - so the last key's value wins | 14:50 |
| evrardjp | yeah yeah understood | 14:51 |
| evrardjp | I now have my user wariables | 14:51 |
| *** clickboom has joined #openstack-ansible | 14:52 | |
| evrardjp | and I feel dumb | 14:52 |
| hughsaunders | evrardjp: sounds a little german | 14:52 |
| *** phalmos has joined #openstack-ansible | 14:52 | |
| odyssey4me | evrardjp you're rusty, we'll forgive you | 14:52 |
| mgariepy | automagically, can yoy +2 this change please :) https://review.openstack.org/#/c/304157/ | 14:55 |
| automagically | mgariepy: I’ll certainly review it | 14:55 |
| *** toddnni has joined #openstack-ansible | 14:55 | |
| automagically | or re-review in this case | 14:55 |
| mgariepy | hehe ;) thanks a lot. | 14:55 |
| odyssey4me | haha | 14:56 |
| mgariepy | since the base changed will it need rebase ? or if it's applying correctly it's ok ? | 14:56 |
| mgariepy | parent* | 14:56 |
| odyssey4me | mgariepy no need to rebase - it's done automatically by zuul\ | 14:56 |
| mgariepy | ok | 14:56 |
| admin0 | automagically: on building the wheel, i see lots of lines like Skipping virtualenv, due to already being wheel .. but we changed versions .. so i think it just does not have it | 14:57 |
| odyssey4me | the only time a rebase is actually required is when there's a merge conflict | 14:57 |
| odyssey4me | admin0 nope, that's a normal part of the process | 14:57 |
| odyssey4me | if it's already built the wheel before, it doesn't need to build it again | 14:57 |
| *** mgoddard_ has joined #openstack-ansible | 14:58 | |
| *** mgoddard has quit IRC | 14:58 | |
| openstackgerrit | Merged openstack/openstack-ansible-plugins: Update the config_template plugin https://review.openstack.org/304385 | 14:59 |
| odyssey4me | cloudnull we should probably backport that ^ at least to mitaka | 15:00 |
| cloudnull | yes | 15:04 |
| *** mgoddard__ has joined #openstack-ansible | 15:05 | |
| *** javeriak_ has joined #openstack-ansible | 15:05 | |
| cloudnull | https://review.openstack.org/#/c/305337/ | 15:05 |
| admin0 | stuck there TASK: [repo_build | Check apt last update file] ****** .not even the full lines of **** like others | 15:05 |
| *** neilus1 has quit IRC | 15:06 | |
| *** jaypipes has joined #openstack-ansible | 15:06 | |
| admin0 | while true do ; redo ; done :D | 15:06 |
| *** javeriak has quit IRC | 15:07 | |
| *** mgoddard_ has quit IRC | 15:08 | |
| *** sdake_busy has quit IRC | 15:08 | |
| *** sanjay__u has joined #openstack-ansible | 15:11 | |
| *** schwicht has joined #openstack-ansible | 15:13 | |
| automagically | Would appreciate if folks interested in the dynamic inventory work session in Austin, take a look at https://etherpad.openstack.org/p/openstack-ansible-newton-dynamic-inventory and add/edit/comment | 15:14 |
| *** gfa_ is now known as gfa | 15:14 | |
| automagically | ^ palendae, cloudnull, odyssey4me | 15:14 |
| * cloudnull looking | 15:15 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 15:15 |
| palendae | automagically: Ah yeah, was going to do that today | 15:15 |
| *** clickboom has quit IRC | 15:16 | |
| *** sdake has joined #openstack-ansible | 15:17 | |
| spotz | automagically: That's a good idea putting an etherpad out ahead of time | 15:18 |
| odyssey4me | evrardjp would it be sensible for us to fix the keepalived role to a SHA, and update that SHA for each stable branch tag in OSA, or should we leave it to be a tag? | 15:21 |
| odyssey4me | spotz yes, that would be ideal for all sessions | 15:21 |
| evrardjp | odyssey4me the goal of having a tag is to use ansible-galaxy | 15:22 |
| spotz | odyssey4me Ok I'll try to get something up today | 15:22 |
| *** clickboom has joined #openstack-ansible | 15:22 | |
| admin0 | odyssey4me: i deleted the containers, ran the playbooks again, stuck @ the same: https://gist.github.com/a1git/ed8c6622402dab8cb6042f22c037b75f | 15:23 |
| evrardjp | ansible-galaxy website works fine with tags, and used to not work with SHAs | 15:23 |
| admin0 | need to run .. brb | 15:23 |
| *** clickboom has quit IRC | 15:23 | |
| odyssey4me | evrardjp yeah, we're switching to using the git method if we use a SHA | 15:23 |
| odyssey4me | so it'll bypass the galaxy API query | 15:24 |
| evrardjp | if we switch everything to git there is no need to use ansible-galaxy cli at all | 15:24 |
| *** admin0 has quit IRC | 15:24 | |
| evrardjp | nor use the format they use | 15:24 |
| evrardjp | they require* | 15:24 |
| evrardjp | just don't expect any publicity of the work you've done inside the ansible community, or lobbying to roles you could eventually consume | 15:25 |
| odyssey4me | admin0 can you provide the complete verbose output of the whole galera_install playbook | 15:25 |
| cloudnull | evrardjp: i think the roles will be published to galaxy eventually | 15:25 |
| odyssey4me | evrardjp right now none of our roles are published in galaxy, so that's a moot point | 15:26 |
| cloudnull | so we should be able to get the publicity out of having the roles in the broader ansible galaxy community | 15:26 |
| evrardjp | cloudnull I think that's a good idea to bring more ppl contributing to them | 15:26 |
| evrardjp | especially to easy independant roles | 15:26 |
| *** clickboom has joined #openstack-ansible | 15:26 | |
| odyssey4me | I'm specifically looking at whether it makes sense for us to us ea SHA for the keepalived role, or whether we should use a tag as we do now. | 15:26 |
| evrardjp | odyssey4me you could use keepalived sha if you prefer | 15:26 |
| cloudnull | we can continue using the role as they are and using the git method however i'd caution us getting away from the galaxy requirements | 15:27 |
| evrardjp | if it makes your life easier | 15:27 |
| evrardjp | I don't personally care | 15:27 |
| palendae | Yeah, we should keep the galaxy requirements | 15:27 |
| *** galstrom_zzz is now known as galstrom | 15:27 | |
| *** pjm6 has joined #openstack-ansible | 15:27 | |
| palendae | I do think we should talk to the Ansible team to see if we can get fallback/secondary URLs for galaxy requirements | 15:27 |
| palendae | In case galaxy breaks again | 15:27 |
| palendae | Or even if it doesn't | 15:27 |
| evrardjp | I just think that if we move away from ansible-galaxy requiremets, we'll have work to do to come back | 15:27 |
| palendae | ^ yep | 15:28 |
| evrardjp | palendae it's not really hard to write a fallback url to be honest | 15:28 |
| palendae | evrardjp: No, but it has to be done in their code | 15:28 |
| evrardjp | indeed | 15:28 |
| evrardjp | the code is really not meant for that at the beginning, but it's perfectly doable, I tried to hack it quickly | 15:29 |
| evrardjp | I don't have the code anymore though | 15:29 |
| evrardjp | anyway | 15:29 |
| evrardjp | up to you odyssey4me | 15:29 |
| evrardjp | generally, when ppl release something, they prefer using tag numbers instead of shas :D | 15:30 |
| odyssey4me | yes, we do want to publish the role at some point but that will not necessarily be for our consumption until we have them published and we decide to consume it in that way - so for now I'm just trying to determine how we keep things up to date | 15:30 |
| *** weezS has joined #openstack-ansible | 15:30 | |
| evrardjp | some companies don't count the commit as a single piece of work | 15:30 |
| odyssey4me | for the keepalived role in particular, we have no current way of regularly checking for a new tag and consuming it | 15:30 |
| *** phalmos has quit IRC | 15:30 | |
| evrardjp | me ? | 15:30 |
| pjm6 | guys for adding a new database and user permissions, i just need to add in one of the galera gluster and it will propagate to the others, or exists something that I must be aware? | 15:30 |
| odyssey4me | so I'm asking whether we should just SHA bump as we do for other things, or whether it makes better sense to find the latest tag and go with that? | 15:30 |
| odyssey4me | pjm6 if you add it to galera, it'll replicate | 15:31 |
| evrardjp | if you don't put version, ansible-galaxy cli will automatically fetch the latest version | 15:31 |
| pjm6 | Ok odyssey4me, tks :) and i just need to access galera container and add it, right? | 15:31 |
| odyssey4me | pjm6 just point it at the same address used for all other galera transactions | 15:31 |
| evrardjp | latest tagged version to be clear | 15:31 |
| odyssey4me | pjm6 look at the keystone playbook as a sample - it includes the creation of a db for keystone | 15:31 |
| pjm6 | Nice :D thanks odyssey4me | 15:31 |
| evrardjp | or to be even more precise: latest imported tagged version | 15:32 |
| odyssey4me | evrardjp yep, but for repeatable builds we must fix it to something - we can't have today's OSA tag build something different today than it does tomorrow | 15:32 |
| *** jayc has quit IRC | 15:32 | |
| evrardjp | for stable branch I'd keep a tag | 15:32 |
| odyssey4me | ok, so fixed to a tag then - thanks | 15:32 |
| pjm6 | btw: odyssey4me, evrardjp besides that function (def allowed) of horizon (that I talk yesterday) works for a custom policy rules, I will use the keystone role associated with a project :) | 15:32 |
| openstackgerrit | Merged openstack/openstack-ansible-security: Fix flake8 violation in conf.py https://review.openstack.org/304815 | 15:36 |
| evrardjp | pjm6 that's a more standard way of doing | 15:37 |
| *** jayc has joined #openstack-ansible | 15:37 | |
| pjm6 | evrardjp: yes, but for the positive side, discovered an other way to do it :D | 15:37 |
| evrardjp | feel free to write it on a blog and share me the url :) | 15:37 |
| evrardjp | and to share me* ? | 15:38 |
| openstackgerrit | Merged openstack/openstack-ansible-specs: PowerVM Virt Driver Support https://review.openstack.org/302941 | 15:38 |
| pjm6 | about what evrardjp? the code def allowed() ? | 15:38 |
| *** yarkot_ has quit IRC | 15:38 | |
| evrardjp | pjm6 yup | 15:38 |
| evrardjp | this way I don't lose it :) | 15:38 |
| *** openstack has joined #openstack-ansible | 16:10 | |
| evrardjp | not sure to understand your question | 16:10 |
| *** Xu_Chao has quit IRC | 16:10 | |
| pjm6 | source /openstack/venvs/neutron-12.0.10/bin/activate | 16:10 |
| pjm6 | Was to use the python environment in the bash :) is that command | 16:11 |
| *** sdake has quit IRC | 16:11 | |
| *** schwicht has joined #openstack-ansible | 16:13 | |
| *** sdake has joined #openstack-ansible | 16:15 | |
| *** wangqwsh has joined #openstack-ansible | 16:15 | |
| *** ChrisBenson has joined #openstack-ansible | 16:17 | |
| *** sdake_ has quit IRC | 16:17 | |
| *** automagically has quit IRC | 16:20 | |
| *** phalmos has quit IRC | 16:21 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Fix flake8 violation in conf.py https://review.openstack.org/305389 | 16:21 |
| *** Bjoern_zZzZzZzZ has quit IRC | 16:21 | |
| cloudnull | pjm6: all of the executables are in the venv and can be found here : /openstack/venvs/$SERVICE-$VERSION/bin/$PROGRAM | 16:22 |
| cloudnull | if you dont want to source the venv | 16:22 |
| cloudnull | you dont have to | 16:22 |
| pjm6 | ohh thanks :D | 16:22 |
| cloudnull | you can just run the bin | 16:22 |
| pjm6 | cloudnull: do you have a testing machine with OSA? | 16:22 |
| cloudnull | IE /openstack/venvs/neutron-12.0.10/bin/neutron net-list will work just fine | 16:22 |
| pjm6 | thanks :D that is more easy of course | 16:22 |
| cloudnull | yesi have a test lab | 16:23 |
| pjm6 | can you try use the vpnaas please? | 16:23 |
| pjm6 | I use it and found that | 16:23 |
| pjm6 | neutron_vpnaas python module aren't being loaded to the contianer | 16:23 |
| pjm6 | or i'm doing it wrong :( | 16:23 |
| pjm6 | what i was trying to do is go to machine copy the git of neutron_vpnaas and the error disappear but didn't solve the problem | 16:24 |
| cloudnull | let me go pull that in to this lab. | 16:24 |
| pjm6 | well i have to run come back later | 16:24 |
| cloudnull | it may be a min | 16:24 |
| cloudnull | ok | 16:24 |
| *** pjm6 has quit IRC | 16:25 | |
| mhayden | evrardjp / pjm6: i haven't tested vpnaas quite yet | 16:26 |
| evrardjp | good to know | 16:26 |
| *** schwicht has quit IRC | 16:26 | |
| *** wangqwsh has quit IRC | 16:31 | |
| *** chhavi has quit IRC | 16:31 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Update source-branch-updater to work with IRR's https://review.openstack.org/297083 | 16:32 |
| odyssey4me | cloudnull ^ I'd appreciate a review to ensure that I'm not missing something obvious | 16:32 |
| *** admin0 has joined #openstack-ansible | 16:32 | |
| * cloudnull looking now | 16:33 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_ceilometer: Update paste, policy and rootwrap configurations 2016-04-13 https://review.openstack.org/305397 | 16:35 |
| *** thorst_ has joined #openstack-ansible | 16:35 | |
| *** phalmos has joined #openstack-ansible | 16:36 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron: Update paste, policy and rootwrap configurations 2016-04-13 https://review.openstack.org/305398 | 16:36 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova: Update paste, policy and rootwrap configurations 2016-04-13 https://review.openstack.org/305399 | 16:37 |
| *** javeriak_ has quit IRC | 16:41 | |
| *** thorst_ has quit IRC | 16:44 | |
| *** schwicht has joined #openstack-ansible | 16:45 | |
| *** Oku_OS is now known as Oku_OS-away | 16:47 | |
| *** sdake has quit IRC | 16:49 | |
| *** admin0 has quit IRC | 16:49 | |
| *** chhavi has joined #openstack-ansible | 16:49 | |
| *** jiteka has quit IRC | 16:52 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: WIP Initial convergence testing https://review.openstack.org/304887 | 16:53 |
| odyssey4me | evrardjp did you mean to duplicate a patch? https://review.openstack.org/305165 / https://review.openstack.org/305161 ? | 16:55 |
| evrardjp | one is targetting osa the other the role itself | 16:55 |
| evrardjp | want me to update the commit txt? | 16:56 |
| odyssey4me | ah, silly me - never mind | 16:56 |
| *** sanjay__u has quit IRC | 16:56 | |
| evrardjp | it's not done anyway it doesn't pass the gate while it works on aio | 16:56 |
| evrardjp | you have time :D | 16:56 |
| *** pjm6 has joined #openstack-ansible | 16:58 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 16:58 |
| odyssey4me | cloudnull ^ FYI there was a bug registered ages ago to make that happen, so well done for closing a bug :) | 16:59 |
| odyssey4me | I just added the reference to the commit msg & topic | 16:59 |
| *** clickboom has joined #openstack-ansible | 17:02 | |
| *** javeriak has joined #openstack-ansible | 17:03 | |
| *** gfa is now known as gfa_ | 17:04 | |
| *** v1k0d3n has quit IRC | 17:05 | |
| *** javeriak_ has joined #openstack-ansible | 17:06 | |
| *** mgoddard has joined #openstack-ansible | 17:07 | |
| *** mgoddard__ has quit IRC | 17:07 | |
| *** javeriak has quit IRC | 17:10 | |
| *** severion has joined #openstack-ansible | 17:10 | |
| cloudnull | thanks odyssey4me --- | 17:11 |
| cloudnull | seems im having connection issues. | 17:11 |
| cloudnull | if miss message / dont respond please resend | 17:11 |
| *** javeriak_ has quit IRC | 17:17 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 17:17 |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible: Improved logging for memcached (OSA calling part) https://review.openstack.org/305165 | 17:19 |
| evrardjp | ^ no mistake anymore between the two commits now | 17:20 |
| jauyeung | hi there, has anyone seen this error 'LinuxBridgeNeutronAgentRPC' object has no attribute 'state_rpc' with the linux bridge agent? thanks | 17:23 |
| odyssey4me | jauyeung not that I've seen | 17:24 |
| *** Bjoern_ has joined #openstack-ansible | 17:25 | |
| jauyeung | all my network:dhcp ports are showing down... thought this might be the reason | 17:26 |
| jauyeung | no other errors in my logs | 17:26 |
| jauyeung | strange | 17:26 |
| *** v1k0d3n has joined #openstack-ansible | 17:27 | |
| v1k0d3n | hey guys...have a question | 17:27 |
| v1k0d3n | so i'm starting to get into the whole operational aspect of monitoring osa, and i am monitoring the nova_console container backend. | 17:28 |
| v1k0d3n | on all the clusters i have built, this one seems to always be in a down state in haproxy web-ui. | 17:28 |
| v1k0d3n | oh sorry...this is for nova_console_novnc, and you know what...i guess that makes sense! | 17:29 |
| *** Bjoern_ has quit IRC | 17:29 | |
| odyssey4me | alright, I'm done for today - going offline - chat tomorrow | 17:30 |
| evrardjp | bye! | 17:31 |
| evrardjp | leaving too | 17:31 |
| *** javeriak has joined #openstack-ansible | 17:31 | |
| *** ChrisBenson has quit IRC | 17:32 | |
| *** ChrisBenson has joined #openstack-ansible | 17:32 | |
| cloudnull | have a good one | 17:33 |
| pjm6 | cya odyssey4me | 17:34 |
| *** admin0 has joined #openstack-ansible | 17:36 | |
| *** chhavi has quit IRC | 17:36 | |
| *** ChrisBenson has quit IRC | 17:41 | |
| *** ChrisBenson has joined #openstack-ansible | 17:42 | |
| spotz | night odyssey4me and evrardjp | 17:43 |
| *** javeriak_ has joined #openstack-ansible | 17:45 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 17:46 |
| *** javeriak has quit IRC | 17:46 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_horizon: Add option to enable from VPNaaS in Horizon https://review.openstack.org/305421 | 17:49 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_horizon: Add option to enable from VPNaaS in Horizon https://review.openstack.org/305421 | 17:49 |
| pjm6 | bye evrardjp | 17:49 |
| *** sdake has joined #openstack-ansible | 17:55 | |
| *** eil397 has joined #openstack-ansible | 18:06 | |
| *** persia has joined #openstack-ansible | 18:07 | |
| *** pjm6 has quit IRC | 18:13 | |
| *** briancubed has joined #openstack-ansible | 18:13 | |
| *** alex______ has joined #openstack-ansible | 18:15 | |
| *** alex______ has quit IRC | 18:18 | |
| *** alex______ has joined #openstack-ansible | 18:18 | |
| *** woodard has joined #openstack-ansible | 18:20 | |
| alex______ | would this be considered a bug? if you use cinder + nfs driver, the volumes on the storage nodes end up being owned by dnsmasq:116 in our case, as that maps to libvirt-qemu:kvm on the compute nodes. should their be a libvert-qemu:kvm user/group on storage with matching uid/gid so perms are correct? | 18:21 |
| *** woodard has quit IRC | 18:22 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_neutron: Add options to auto enable VPNaaS https://review.openstack.org/305432 | 18:23 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Add option to auto enable from VPNaaS in Horizon https://review.openstack.org/305433 | 18:24 |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 18:24 |
| cloudnull | alex______: thats something we've talked about in the past | 18:25 |
| cloudnull | idk if its a bug | 18:25 |
| cloudnull | however i'd like to talk about trying to come up with an scheme that will help us make sure the uid and gid are configurable for cases of shared storage | 18:25 |
| cloudnull | i think a little for to that effect has gone into nova via mhayden? | 18:26 |
| alex______ | cloudnull: gotcha. yes, standardizing uid/gid would be useful.. i typo'd one of our storage nodes in the ansible config, and so only provisioned 3 out of 4 of them for cinder, but 4 out of 4 for swift. when I went to add cinder onto the missing node, the uid/gid of swift and cinder were reversed, so files were getting created as swift user on one node. =) | 18:29 |
| cloudnull | sounds like a great time :) | 18:29 |
| * cloudnull is kidding | 18:29 | |
| alex______ | bunch of find's and chowns. =) | 18:30 |
| *** woodard has joined #openstack-ansible | 18:31 | |
| mhayden | cloudnull: oops, what did i do? | 18:31 |
| cloudnull | it would be good to coordinate the uid/gid for all services we support. its just not been done yet | 18:32 |
| cloudnull | mhayden: didnt you work on nova uid/gid things ? | 18:32 |
| cloudnull | alex______: https://github.com/openstack/openstack-ansible-os_nova/blob/master/defaults/main.yml#L57-L67 | 18:32 |
| alex______ | can't imagine the upgrade script for that change. =) | 18:32 |
| cloudnull | yup | 18:33 |
| alex______ | oh, interesting, thanks! | 18:33 |
| cloudnull | i think thats why its not been done by default | 18:33 |
| cloudnull | https://github.com/openstack/openstack-ansible-os_nova/commit/f4821e47b1cdafbb00d1949b292a40aaa687f3f1 | 18:33 |
| cloudnull | mhayden: it is all your fault =) | 18:33 |
| cloudnull | j/k | 18:33 |
| cloudnull | alex______: its not implemented in cinder or swfit it seems | 18:34 |
| *** jwitko has joined #openstack-ansible | 18:36 | |
| *** pjm6 has joined #openstack-ansible | 18:36 | |
| *** woodard has quit IRC | 18:36 | |
| mhayden | ah, i did do those | 18:37 |
| *** woodard has joined #openstack-ansible | 18:37 | |
| mhayden | did we find a bug? | 18:37 |
| mhayden | sorry, i've got a lotta windows open | 18:37 |
| *** jwitk0 has quit IRC | 18:37 | |
| *** Tebro has quit IRC | 18:38 | |
| alex______ | no, no bug, just can add to it to have cinder / swift also have set uid's | 18:38 |
| *** charz has quit IRC | 18:38 | |
| cloudnull | just a continuation of the shared storage convo and being able to set the UID/GID | 18:39 |
| *** bapalm has quit IRC | 18:39 | |
| *** fignew has quit IRC | 18:39 | |
| *** hlkv6 has quit IRC | 18:39 | |
| *** marcoamorales has quit IRC | 18:39 | |
| cloudnull | alex______: are you going to be at the summit? | 18:39 |
| *** gfa_ has quit IRC | 18:39 | |
| *** weshay has quit IRC | 18:39 | |
| *** openstackgerrit has quit IRC | 18:39 | |
| *** sshen has quit IRC | 18:39 | |
| *** galstrom has quit IRC | 18:39 | |
| alex______ | sadly, no, and i missed it last year when it was in my own backdoor. =) | 18:40 |
| cloudnull | :( next one :) | 18:40 |
| *** sshen has joined #openstack-ansible | 18:41 | |
| cloudnull | maybe we can get that convo within the working sessions https://etherpad.openstack.org/p/openstack-ansible-newton-roadmap-brainstorm? | 18:41 |
| *** gfa_ has joined #openstack-ansible | 18:41 | |
| *** kong has quit IRC | 18:41 | |
| *** meteorfox has quit IRC | 18:41 | |
| pjm6 | in fresh installing mitaka (after removing liberty) for some reason in my bare metal (compute and storage node, which is the same) failed the instalation of pip requirements | 18:41 |
| pjm6 | it seems that the repo server (port 8181) is not available | 18:42 |
| pjm6 | <html><body><h1>503 Service Unavailable</h1> | 18:42 |
| pjm6 | anyone had this problem? | 18:42 |
| *** phiche has joined #openstack-ansible | 18:43 | |
| mongo2 | pjm6: not on this release but I did need to make sure to up nfile limits in /etc/security in some older versions. | 18:43 |
| *** Tebro has joined #openstack-ansible | 18:44 | |
| *** bapalm has joined #openstack-ansible | 18:44 | |
| pjm6 | hmm in the repo container? | 18:44 |
| *** galstrom has joined #openstack-ansible | 18:44 | |
| mongo2 | but that may not be the issue, also remember that * no longer covers root | 18:45 |
| *** charz has joined #openstack-ansible | 18:45 | |
| pjm6 | the log is this: http://pastebin.com/sxycTznC | 18:46 |
| pjm6 | in the machien with that IP (infra host) | 18:47 |
| *** hlkv6 has joined #openstack-ansible | 18:47 | |
| pjm6 | same error | 18:47 |
| jmccrory | pjm6: have you run the haproxy-install playbook? | 18:47 |
| mongo2 | pjm6: this was in the bare metal host, but it passes through at the time the container starts. | 18:47 |
| *** automagically has joined #openstack-ansible | 18:48 | |
| mongo2 | and you have check that the container does have the port open with netstat -tnl right? | 18:48 |
| pjm6 | jmccrory, i was doing the setup-hosts.yml (in the docs that process comes after) but i did yet | 18:48 |
| pjm6 | for strange in the infra host i had no containers | 18:48 |
| *** michaelgugino has joined #openstack-ansible | 18:48 | |
| pjm6 | but in the host i have port 8181 open | 18:49 |
| jmccrory | oh, you may have an existing pip.conf from your previous install | 18:49 |
| jmccrory | ~/.pip/pip.conf | 18:49 |
| pjm6 | yes that could be possible jmccrory | 18:49 |
| pjm6 | 12.0.10 | 18:49 |
| mongo2 | ya, no continers would cause a problem. | 18:49 |
| pjm6 | yeah, mitaka is 13.0.0. | 18:49 |
| pjm6 | so its better remove all | 18:50 |
| *** johnmilton has quit IRC | 18:50 | |
| pjm6 | folder /root/.pip/* | 18:50 |
| pjm6 | ? | 18:50 |
| jmccrory | yep, it'll be recreated | 18:52 |
| *** weshay has joined #openstack-ansible | 18:52 | |
| *** fignew has joined #openstack-ansible | 18:53 | |
| pjm6 | thanks jmccrory, mongo2 :) | 18:53 |
| pjm6 | well it seems that had passed =) | 18:54 |
| pjm6 | that task, tks | 18:54 |
| *** deadnull has quit IRC | 18:54 | |
| *** openstackgerrit has joined #openstack-ansible | 18:55 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 18:55 |
| briancubed | Like pjm6, I'm doing a fresh install, only I'm using liberty. I'm also running into a problem with repo server when running the repo-build playbook. A log snippet is here: http://pastebin.com/JhWKw7iW. | 18:56 |
| *** admin0 has quit IRC | 18:56 | |
| *** thorst_ has joined #openstack-ansible | 18:56 | |
| pjm6 | briancubed, are you using branch stable/liberty? | 18:58 |
| pjm6 | i'm saying this because my previous instalation was in the branch stable/libety | 18:59 |
| *** jayc has quit IRC | 18:59 | |
| briancubed | I am using a fork created by a co-worker. I believe it's from tag 12.0.8???? | 18:59 |
| pjm6 | and had 12.0.10 | 18:59 |
| pjm6 | Yes, that why i was saying | 18:59 |
| *** weezS has joined #openstack-ansible | 18:59 | |
| pjm6 | i used the branch stable/liberty | 18:59 |
| *** Rodrigo_BR has joined #openstack-ansible | 19:00 | |
| briancubed | So, pjm6, you think I should use stable/liberty instead? | 19:00 |
| pjm6 | in my opinion i think yes | 19:00 |
| pjm6 | (if I understood well) | 19:01 |
| *** thorst_ has quit IRC | 19:01 | |
| pjm6 | the stable/liberty is the branch that don't have nothing experimental (anyone correct me if i'm wrong) | 19:01 |
| briancubed | I can try stable/liberty without our integration changes. If it works, I can make the integration changes by hand and retry. | 19:01 |
| *** sdake_ has joined #openstack-ansible | 19:01 | |
| pjm6 | probably that tag are using an commit that could not be working | 19:01 |
| briancubed | I am willing to give stable/liberty a try | 19:02 |
| *** alex______ has quit IRC | 19:02 | |
| pjm6 | well I used it a two weeks ago and everything works fine =D | 19:03 |
| pjm6 | i'm trying to use now the mitaka because i need the vpnaas | 19:03 |
| *** sdake has quit IRC | 19:03 | |
| *** meteorfox has joined #openstack-ansible | 19:06 | |
| *** kong has joined #openstack-ansible | 19:09 | |
| *** admin0 has joined #openstack-ansible | 19:10 | |
| *** jayc has joined #openstack-ansible | 19:11 | |
| *** hybridpollo has joined #openstack-ansible | 19:11 | |
| *** javeriak_ has quit IRC | 19:14 | |
| odyssey4me | anyone testing out Mitaka may wish to add this patch, as this includes the final Mitaka release of the OpenStack projects: https://review.openstack.org/302359 | 19:16 |
| odyssey4me | 13.0.0 was based on RC1 | 19:16 |
| *** KLevenstein has quit IRC | 19:18 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server: Add executable bit to run_tests.sh https://review.openstack.org/301007 | 19:20 |
| *** thorst_ has joined #openstack-ansible | 19:21 | |
| *** thorst_ has quit IRC | 19:21 | |
| *** thorst_ has joined #openstack-ansible | 19:22 | |
| odyssey4me | jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr an quick review of https://review.openstack.org/302359 to prep for the next tag would be appreciated | 19:22 |
| odyssey4me | it also brings in all the fixes we identified in the RC period, so it's pretty essential | 19:23 |
| cloudnull | bin done did it | 19:23 |
| odyssey4me | cloudnull thanks, I think stevelle could use another vote on https://review.openstack.org/304328 | 19:23 |
| odyssey4me | and https://review.openstack.org/304328 for that matter | 19:24 |
| pjm6 | odyssey4me, for adding that patch I must change git branch to that one? | 19:24 |
| stevelle | odyssey4me: are you trying to topple my tower of babel? | 19:24 |
| odyssey4me | lol, I meant https://review.openstack.org/304336 | 19:24 |
| odyssey4me | stevelle if you wanna build the largest dep chain known to gerrit, we can leave it be if you like :p | 19:25 |
| stevelle | :D | 19:25 |
| odyssey4me | alright, I'm going offline for realsies this time | 19:26 |
| stevelle | I seriously doubt I would beat the glance artifacts chain :) | 19:26 |
| pjm6 | in this lines | 19:31 |
| pjm6 | https://github.com/openstack/openstack-ansible-rsyslog_client/blob/2246bd2beec155f55496861a82df50d0cc4418a0/tasks/rsyslog_client_install.yml#L39-L46 | 19:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_gnocchi: Enable docs task https://review.openstack.org/304336 | 19:31 |
| pjm6 | that should be sudo mode? | 19:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_gnocchi: Enabling bashate and pep8 lint checks https://review.openstack.org/304328 | 19:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_gnocchi: Enable ansible lint and syntax tests https://review.openstack.org/304343 | 19:32 |
| pjm6 | i'm asking because i'm getting error saying that its not possible to open apt of perconna | 19:32 |
| pjm6 | but the file exists | 19:32 |
| pjm6 | msg: Failed to lock apt for exclusive operation | 19:32 |
| pjm6 | or because we put user as root, we have privileges as a root user? | 19:33 |
| *** marcoamorales has joined #openstack-ansible | 19:35 | |
| openstackgerrit | Merged openstack/openstack-ansible: Fix configuration string for haproxy https://review.openstack.org/304157 | 19:38 |
| *** Tridde has joined #openstack-ansible | 19:42 | |
| *** sdake_ has quit IRC | 19:48 | |
| *** jwitko has quit IRC | 19:48 | |
| *** schwicht has quit IRC | 19:48 | |
| *** elo has quit IRC | 19:48 | |
| *** tiagogomes has quit IRC | 19:48 | |
| *** markvoelker has quit IRC | 19:48 | |
| *** clsacramento has quit IRC | 19:48 | |
| *** mongo2 has quit IRC | 19:48 | |
| *** LanceHaig has quit IRC | 19:48 | |
| *** rromans has quit IRC | 19:48 | |
| *** Trident has quit IRC | 19:48 | |
| *** homerp has quit IRC | 19:48 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: Move inventory environment loading to function https://review.openstack.org/290740 | 19:48 |
| automagically | Would appreciate it if I could get some more reviews on https://review.openstack.org/#/c/298916/ that’s been out there for awhile | 19:50 |
| *** jwitk0 has joined #openstack-ansible | 19:50 | |
| automagically | Nothing urgent about, but I like to draw attention to stuff once it surpasses 2 weeks after initial submission | 19:51 |
| *** rromans_ has joined #openstack-ansible | 19:54 | |
| *** schwicht_ has joined #openstack-ansible | 19:54 | |
| spotz | looking automagically I tend to squirrel from the channel so if I'm not on when submitted I don't always see it | 19:55 |
| automagically | thx spotz | 19:55 |
| *** Rodrigo_BR has quit IRC | 19:55 | |
| *** LanceHai- has joined #openstack-ansible | 19:56 | |
| *** clsacramento_ has joined #openstack-ansible | 19:57 | |
| *** homerp_ has joined #openstack-ansible | 20:02 | |
| *** sdake_ has joined #openstack-ansible | 20:03 | |
| *** jwitko has joined #openstack-ansible | 20:03 | |
| *** schwicht has joined #openstack-ansible | 20:03 | |
| *** elo has joined #openstack-ansible | 20:03 | |
| *** tiagogomes has joined #openstack-ansible | 20:03 | |
| *** markvoelker has joined #openstack-ansible | 20:03 | |
| *** clsacramento has joined #openstack-ansible | 20:03 | |
| *** mongo2 has joined #openstack-ansible | 20:03 | |
| *** LanceHaig has joined #openstack-ansible | 20:03 | |
| *** rromans has joined #openstack-ansible | 20:03 | |
| *** homerp has joined #openstack-ansible | 20:03 | |
| *** dmsimard1 has joined #openstack-ansible | 20:03 | |
| pjm6 | in mitaka anyone had problems in setup-hosts.yml regarding rsyslog_client | 20:04 |
| *** sdake_ has quit IRC | 20:04 | |
| *** jwitko has quit IRC | 20:04 | |
| *** schwicht has quit IRC | 20:04 | |
| *** elo has quit IRC | 20:04 | |
| *** tiagogomes has quit IRC | 20:04 | |
| *** markvoelker has quit IRC | 20:04 | |
| *** clsacramento has quit IRC | 20:04 | |
| *** mongo2 has quit IRC | 20:04 | |
| *** LanceHaig has quit IRC | 20:04 | |
| *** rromans has quit IRC | 20:04 | |
| *** homerp has quit IRC | 20:04 | |
| pjm6 | in the "Update apt if needed" | 20:04 |
| *** dmsimard1 is now known as dmsimard | 20:05 | |
| *** LanceHai- is now known as LanceHaig | 20:05 | |
| *** spotz has quit IRC | 20:05 | |
| *** spotz has joined #openstack-ansible | 20:08 | |
| openstackgerrit | Merged openstack/openstack-ansible-security: Fix flake8 violation in conf.py https://review.openstack.org/305389 | 20:10 |
| *** mongo2_ has joined #openstack-ansible | 20:11 | |
| *** sdake_ has joined #openstack-ansible | 20:11 | |
| *** elo has joined #openstack-ansible | 20:11 | |
| *** tiagogomes has joined #openstack-ansible | 20:11 | |
| *** markvoelker has joined #openstack-ansible | 20:11 | |
| *** mongo2 has joined #openstack-ansible | 20:11 | |
| *** mongo2 has quit IRC | 20:11 | |
| briancubed | hey, pjm6! Just wanted to let you know that I switched to stable/liberty and that repo-build error I couldn't figure out went away... | 20:12 |
| *** eric_lopez has joined #openstack-ansible | 20:12 | |
| *** sdake has joined #openstack-ansible | 20:12 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts: conditionally include the scsi_dh kernel module https://review.openstack.org/300621 | 20:12 |
| briancubed | I have a bit of work to do to move my integration code, but that's less frustrating than chasing failures! :-) | 20:13 |
| *** michaelgugino has quit IRC | 20:14 | |
| *** markvoelker_ has joined #openstack-ansible | 20:14 | |
| pjm6 | the same error ? :\ | 20:14 |
| *** tiagogomes has quit IRC | 20:15 | |
| *** markvoelker has quit IRC | 20:15 | |
| *** sdake_ has quit IRC | 20:16 | |
| *** elo has quit IRC | 20:16 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 20:18 |
| briancubed | Spoke too soon. Hit a failure on the next playbook, galera-install: http://pastebin.com/kyngtv1n | 20:20 |
| pjm6 | so it fix the first? :D | 20:21 |
| *** Nepoc has quit IRC | 20:21 | |
| pjm6 | do you have generated the user_secrets.yml? | 20:21 |
| briancubed | yes, using stable/liberty fixed the first error. Now I see this new one. Let me check the user-secrets.yml... | 20:22 |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon: Update SSL key / cert distribution for Horizon https://review.openstack.org/304113 | 20:22 |
| pjm6 | briancubed, btw: when you switch branch, did you run bootstrap-ansible script? | 20:23 |
| *** automagically has quit IRC | 20:23 | |
| briancubed | pjm6: Oops. No. | 20:23 |
| pjm6 | I think you must do that | 20:23 |
| *** jayc has quit IRC | 20:23 | |
| pjm6 | to fetch the new git repo checkouts :) | 20:23 |
| briancubed | okay. let me do that. | 20:24 |
| *** jayc has joined #openstack-ansible | 20:24 | |
| mhayden | i'm deploying OSA mitaka but the deployment fails when it tries to create galera users: msg: the python mysqldb module is required | 20:25 |
| briancubed | FWIW, /etc/openstack_deploy/user_secrets.yml has no passwords in it. Should I have added passwords there? | 20:25 |
| mhayden | but it seems that the repo server is building mysql-python | 20:25 |
| *** phiche has quit IRC | 20:27 | |
| *** tiagogomes has joined #openstack-ansible | 20:28 | |
| pjm6 | yes briancubed | 20:32 |
| pjm6 | wait | 20:32 |
| briancubed | Looks like a step I missed in the docs: python pw-token-gen.py --file /etc/openstack_deploy/user_secrets.yml | 20:33 |
| briancubed | From here: http://docs.openstack.org/developer/openstack-ansible/liberty/install-guide/configure-creds.html?highlight=user_secrets | 20:34 |
| pjm6 | yes i was finding that | 20:34 |
| pjm6 | to give you | 20:35 |
| briancubed | My guess is that I should run the python command, above, destroy containers, create containers, then resume... Cool? | 20:35 |
| *** weshay has quit IRC | 20:35 | |
| pjm6 | that's probably our error | 20:35 |
| pjm6 | well if you're in the | 20:36 |
| pjm6 | openstack-playbook | 20:36 |
| pjm6 | or infraestructure | 20:36 |
| pjm6 | I think its better | 20:36 |
| *** krotscheck is now known as krotscheck_dcm | 20:36 | |
| *** klamath_ has joined #openstack-ansible | 20:37 | |
| *** Mudpuppy_ has joined #openstack-ansible | 20:37 | |
| *** odyssey4me_ has joined #openstack-ansible | 20:38 | |
| *** jwitko has joined #openstack-ansible | 20:39 | |
| briancubed | pjm6: Better to destroy/create containers? | 20:39 |
| pjm6 | Yes if you can | 20:39 |
| briancubed | yes, I can. thank you. | 20:39 |
| pjm6 | it will be a clean install (i'm not sure if in the step you are) | 20:39 |
| pjm6 | had already need of the secretsw | 20:39 |
| *** lykinsbd has joined #openstack-ansible | 20:39 | |
| pjm6 | at least you guarantee that are no miss steps :D | 20:39 |
| briancubed | I'm working through failures in running setup-infrastructure.yml | 20:40 |
| briancubed | so I can destroy/create containers, then restart setup_infrastrcture | 20:40 |
| *** sshen_ has joined #openstack-ansible | 20:41 | |
| *** promethe1nfire has joined #openstack-ansible | 20:42 | |
| *** markvoelker has joined #openstack-ansible | 20:42 | |
| *** kysse_ has joined #openstack-ansible | 20:42 | |
| *** raginbaj- has joined #openstack-ansible | 20:43 | |
| *** lbragstad_ has joined #openstack-ansible | 20:44 | |
| *** nb1 has joined #openstack-ansible | 20:45 | |
| *** erikwilson has joined #openstack-ansible | 20:45 | |
| *** ajo_ has joined #openstack-ansible | 20:45 | |
| *** brad[]` has joined #openstack-ansible | 20:45 | |
| *** markvoelker_ has quit IRC | 20:47 | |
| *** jwitk0 has quit IRC | 20:47 | |
| *** sshen has quit IRC | 20:47 | |
| *** kysse has quit IRC | 20:47 | |
| *** admiralboom has quit IRC | 20:47 | |
| *** Mudpuppy has quit IRC | 20:47 | |
| *** klamath has quit IRC | 20:47 | |
| *** xek has quit IRC | 20:47 | |
| *** brad[] has quit IRC | 20:47 | |
| *** lykinsbd_ has quit IRC | 20:47 | |
| *** mgariepy has quit IRC | 20:47 | |
| *** prometheanfire has quit IRC | 20:47 | |
| *** ajo has quit IRC | 20:47 | |
| *** raginbajin has quit IRC | 20:47 | |
| *** odyssey4me has quit IRC | 20:47 | |
| *** erikmwilson has quit IRC | 20:47 | |
| *** lbragstad has quit IRC | 20:47 | |
| *** pabelanger has quit IRC | 20:47 | |
| *** larsks has quit IRC | 20:47 | |
| *** raginbaj- is now known as raginbajin | 20:47 | |
| *** larsks has joined #openstack-ansible | 20:47 | |
| pjm6 | destroy and then run setup-hosts (if i'm not mistaken) and then infraestructure | 20:47 |
| *** promethe1nfire is now known as prometheanfire | 20:49 | |
| *** pabelanger has joined #openstack-ansible | 20:49 | |
| *** nb1 is now known as admiralboom | 20:50 | |
| *** mgariepy has joined #openstack-ansible | 20:53 | |
| briancubed | pjm6: yes. I'll do that. thanks. | 20:54 |
| *** pabelanger has quit IRC | 20:58 | |
| *** pabelanger has joined #openstack-ansible | 20:58 | |
| pjm6 | no problem, hope it works briancubed :) | 20:58 |
| mhayden | i think the repo server builds mysql-python against an older version of galera | 20:58 |
| mhayden | and then the library is nowhere to be found on the galera nodes themselves | 20:58 |
| mhayden | either i found a bug or i'm an idiot | 20:58 |
| *** clickboom has quit IRC | 21:01 | |
| mhayden | yeah, that seems to be the issue | 21:01 |
| mhayden | created a bug: https://bugs.launchpad.net/openstack-ansible/+bug/1570095 | 21:04 |
| openstack | Launchpad bug 1570095 in openstack-ansible "mysql-python builds against old version of libmariadbclient" [Undecided,New] | 21:04 |
| *** johnmilton has joined #openstack-ansible | 21:05 | |
| *** briancubed has quit IRC | 21:05 | |
| *** sdake has quit IRC | 21:08 | |
| *** retreved_ has quit IRC | 21:10 | |
| jmccrory | mhayden: mitaka branch might not have included the most recent repo_build role. https://review.openstack.org/#/c/301103/ should have fixed that | 21:10 |
| jmccrory | https://bugs.launchpad.net/openstack-ansible/+bug/1565128 | 21:11 |
| openstack | Launchpad bug 1565128 in openstack-ansible mitaka "mariadb10 libs are needed when building the repo" [Critical,Fix committed] - Assigned to Jesse Pretorius (jesse-pretorius) | 21:11 |
| mhayden | jmccrory: darnit -- and i remember looking at that one! | 21:11 |
| mhayden | thanks, jmccrory | 21:11 |
| jmccrory | sure np | 21:11 |
| mhayden | i stumbled into an openstack-ansible-security bug that was already fixed too | 21:11 |
| mhayden | working with these independent roles is ... interesting | 21:12 |
| mhayden | night, folks | 21:12 |
| pjm6 | jmccrory, that was already merged in git repo (i think) | 21:13 |
| pjm6 | jmccrory, do you have a clue why in setup-hosts i'm failing in the Update apt repo? | 21:13 |
| jmccrory | later, mhayden | 21:13 |
| pjm6 | cya mhayden | 21:13 |
| jmccrory | pjm6: not sure. have a paste of the error you're getting? | 21:14 |
| pjm6 | yes, just waiting to the command ends for pasting the text | 21:14 |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 21:16 |
| *** asettle has joined #openstack-ansible | 21:17 | |
| *** phiche has joined #openstack-ansible | 21:19 | |
| pjm6 | jmccrory, here http://pastebin.com/VrXyhyG9 | 21:19 |
| pjm6 | i have read that error could be because not running sudo (in ansible) but as we are the root user , shouldn't be that | 21:19 |
| *** schwicht_ has quit IRC | 21:19 | |
| *** admin0 has quit IRC | 21:21 | |
| jmccrory | are infra1, haproxy1, and compute1 all the same host? | 21:21 |
| *** lbragstad_ is now known as lbragstad | 21:21 | |
| *** phiche1 has joined #openstack-ansible | 21:22 | |
| *** phiche has quit IRC | 21:23 | |
| pjm6 | no | 21:24 |
| pjm6 | infra1 and haproxy1 same | 21:24 |
| pjm6 | compute1 different | 21:24 |
| jmccrory | oh ok, the host name doesn't have to match the group, it'll get assigned to the right groups in the inventory. think that explains the first couple 'failed to lock apt' errors. | 21:26 |
| pjm6 | sorry didn't understand. | 21:26 |
| pjm6 | my config maybe is wrong? | 21:27 |
| pjm6 | (i'm using the same as used in the liberty) | 21:27 |
| jmccrory | oh ok, yeah so in your openstack_user_config. the same hostname should be used consistently for each individual host | 21:28 |
| pjm6 | hmm | 21:29 |
| pjm6 | for example | 21:29 |
| pjm6 | (maybe i'm wrong this) | 21:30 |
| jmccrory | when the inventory json file is generated it'll add the hosts to the correct groups based on your user_config. | 21:30 |
| *** thorst_ has quit IRC | 21:30 | |
| pjm6 | http://pastebin.com/cicvhtX6 | 21:30 |
| pjm6 | in this case | 21:30 |
| pjm6 | all should be infra1 ? | 21:30 |
| jmccrory | yep | 21:31 |
| pjm6 | hmm ok that could explain things | 21:31 |
| pjm6 | thanks jmccrory :D | 21:31 |
| pjm6 | then I should run again | 21:31 |
| pjm6 | the same playbook setup-hosts.yml | 21:31 |
| pjm6 | or should delete | 21:31 |
| pjm6 | a file first? | 21:31 |
| jmccrory | have you created any containers yet? | 21:32 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 21:32 | |
| pjm6 | i think the playbook didn't create yet | 21:32 |
| pjm6 | at least i do | 21:32 |
| pjm6 | lxc-ls | 21:32 |
| pjm6 | and nothing appears in the infra1 | 21:32 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 21:32 | |
| jmccrory | oh ok, should be fine to delete openstack_inventory.json then | 21:33 |
| jmccrory | it'll get recreated on your next playbook run | 21:33 |
| pjm6 | well i deleted this two openstack_inventory.json and openstack_hostnames_ips.yml | 21:34 |
| pjm6 | and now don't appear the log1 and haproxy1 | 21:34 |
| pjm6 | :D | 21:34 |
| jmccrory | ok cool | 21:34 |
| pjm6 | lets see what happens thanks | 21:34 |
| pjm6 | but this was working on liberty, so I discarted that file | 21:35 |
| jmccrory | hmm, apt update task might have been skipped or worked differently in liberty. i'll check | 21:36 |
| *** thorst has joined #openstack-ansible | 21:36 | |
| pjm6 | well it seems that rsyslog_client 12.0.0 don't exist :\ | 21:38 |
| pjm6 | at least in the | 21:38 |
| pjm6 | https://github.com/openstack/openstack-ansible-rsyslog_client | 21:38 |
| *** Mudpuppy has joined #openstack-ansible | 21:39 | |
| pjm6 | well it seems it passes that step jmccrory :D | 21:40 |
| pjm6 | thanks very much | 21:40 |
| jmccrory | ah ok good. sure, no problem | 21:40 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 21:40 |
| *** thorst has quit IRC | 21:40 | |
| *** Mudpuppy_ has quit IRC | 21:41 | |
| *** fignew has quit IRC | 21:41 | |
| pjm6 | hope that now could make an deployment functional and stable as liberty =D | 21:41 |
| *** admiralboom has quit IRC | 21:41 | |
| *** clsacramento_ has quit IRC | 21:41 | |
| *** clsacramento has joined #openstack-ansible | 21:43 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 21:45 |
| *** galstrom is now known as galstrom_zzz | 21:45 | |
| *** jauyeung has quit IRC | 21:48 | |
| *** keedya has quit IRC | 21:54 | |
| *** thorst has joined #openstack-ansible | 21:54 | |
| *** fignew has joined #openstack-ansible | 21:55 | |
| *** phiche1 has quit IRC | 21:57 | |
| *** thorst has quit IRC | 21:58 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 22:00 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:02 | |
| *** sdake has joined #openstack-ansible | 22:02 | |
| *** phalmos has quit IRC | 22:06 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible: Isolate Ansible from the deployment host https://review.openstack.org/304840 | 22:06 |
| *** b3rnard0 is now known as b3rnard0_away | 22:08 | |
| *** busterswt has quit IRC | 22:09 | |
| *** _hanhart has quit IRC | 22:11 | |
| *** jayc has quit IRC | 22:18 | |
| *** clickboom has joined #openstack-ansible | 22:18 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 22:20 |
| *** _hanhart has joined #openstack-ansible | 22:21 | |
| *** clickboom has quit IRC | 22:26 | |
| *** spotz is now known as spotz_zzz | 22:28 | |
| *** jayc has joined #openstack-ansible | 22:29 | |
| *** klamath_ has quit IRC | 22:30 | |
| *** nb1 has joined #openstack-ansible | 22:34 | |
| *** ametts has quit IRC | 22:39 | |
| *** jayc has quit IRC | 22:44 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 22:45 |
| *** ChrisBenson has quit IRC | 22:48 | |
| *** ChrisBenson has joined #openstack-ansible | 22:48 | |
| *** sdake has quit IRC | 22:48 | |
| *** schwicht has joined #openstack-ansible | 22:52 | |
| *** weezS has quit IRC | 22:53 | |
| *** schwicht_ has joined #openstack-ansible | 22:55 | |
| *** schwicht has quit IRC | 22:56 | |
| *** sdake has joined #openstack-ansible | 23:00 | |
| *** andymccr_ is now known as andymccr | 23:02 | |
| *** woodard has quit IRC | 23:09 | |
| *** woodard has joined #openstack-ansible | 23:10 | |
| openstackgerrit | Steve Lewis (stevelle) proposed openstack/openstack-ansible-os_gnocchi: Initial convergence testing https://review.openstack.org/304887 | 23:17 |
| *** ChrisBenson has quit IRC | 23:18 | |
| *** johnmilton has quit IRC | 23:18 | |
| openstackgerrit | Merged openstack/openstack-ansible-memcached_server: Improved logging for memcached https://review.openstack.org/305161 | 23:20 |
| *** ChrisBenson has joined #openstack-ansible | 23:20 | |
| jwitko | hey guys after a fresh install of some additional compute nodes I'm noticing live migration on those new comp nodes is failing due to the /usr/bin/kvm binary missing. is this not installed as part of the OSA install? | 23:38 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Resolve issues with swift tests https://review.openstack.org/304350 | 23:38 |
| *** schwicht_ has quit IRC | 23:41 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron: Fix missing 'qos' in extension drivers https://review.openstack.org/304305 | 23:55 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!