| *** thorst has joined #openstack-ansible | 00:00 | |
| *** scarlisle has quit IRC | 00:02 | |
| *** arif-ali has joined #openstack-ansible | 00:05 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon: Standardise installation-related task files https://review.openstack.org/294095 | 00:06 |
|---|---|---|
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon: Use upper-constraints file in developer mode https://review.openstack.org/295153 | 00:06 |
| *** schwicht has joined #openstack-ansible | 00:11 | |
| -cloudnull- We still have a few patches in merge conflict. https://review.openstack.org/#/q/project:%255Eopenstack/openstack-ansible.*+-is:mergeable+is:open,n,z if you have time to review the patches and update them for the post IRR role out it'd be much appreciated. | 00:12 | |
| *** woodard has quit IRC | 00:22 | |
| *** woodard has joined #openstack-ansible | 00:23 | |
| *** fawadkhaliq has quit IRC | 00:23 | |
| *** wtcross has joined #openstack-ansible | 00:24 | |
| *** woodard has quit IRC | 00:24 | |
| *** izaakk has quit IRC | 00:27 | |
| *** wtcross has quit IRC | 00:28 | |
| *** fawadkhaliq has joined #openstack-ansible | 00:47 | |
| *** furlongm has quit IRC | 00:55 | |
| *** furlongm_ has joined #openstack-ansible | 00:55 | |
| *** ScarZy has quit IRC | 00:57 | |
| *** skamithi13 has quit IRC | 00:59 | |
| *** skamithi13 has joined #openstack-ansible | 01:01 | |
| *** schwicht_ has joined #openstack-ansible | 01:03 | |
| *** schwicht has quit IRC | 01:05 | |
| *** johnmilton has quit IRC | 01:05 | |
| *** fawadkhaliq has quit IRC | 01:05 | |
| *** fawadkhaliq has joined #openstack-ansible | 01:06 | |
| *** sdake has quit IRC | 01:06 | |
| *** sdake has joined #openstack-ansible | 01:08 | |
| *** ScarZy has joined #openstack-ansible | 01:09 | |
| *** sdake_ has joined #openstack-ansible | 01:11 | |
| *** tomoe_ has joined #openstack-ansible | 01:12 | |
| *** sdake has quit IRC | 01:14 | |
| *** neilus has joined #openstack-ansible | 01:17 | |
| *** neilus1 has joined #openstack-ansible | 01:19 | |
| *** neilus has quit IRC | 01:20 | |
| *** johnmilton has joined #openstack-ansible | 01:21 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 01:21 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:22 | |
| *** wtcross has joined #openstack-ansible | 01:25 | |
| *** wtcross has quit IRC | 01:29 | |
| *** schwicht_ has quit IRC | 01:32 | |
| *** schwicht has joined #openstack-ansible | 01:32 | |
| *** b3rnard0 is now known as b3rnard0_away | 01:35 | |
| *** neilus1 has quit IRC | 02:06 | |
| *** thorst has quit IRC | 02:10 | |
| *** thorst has joined #openstack-ansible | 02:11 | |
| *** iceyao has joined #openstack-ansible | 02:18 | |
| *** thorst has quit IRC | 02:19 | |
| *** neilus has joined #openstack-ansible | 02:22 | |
| *** fawadkhaliq has quit IRC | 02:24 | |
| *** wtcross has joined #openstack-ansible | 02:26 | |
| *** wtcross has quit IRC | 02:30 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-galera_client: Updated role using the Multi-Distro framework https://review.openstack.org/295599 | 02:41 |
| *** asettle has quit IRC | 02:44 | |
| *** iceyao has quit IRC | 02:45 | |
| *** iceyao has joined #openstack-ansible | 02:46 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-galera_server: Updated role using the Multi-Distro framework https://review.openstack.org/295601 | 02:54 |
| *** asettle has joined #openstack-ansible | 02:56 | |
| *** fawadkhaliq has joined #openstack-ansible | 03:00 | |
| *** iceyao has quit IRC | 03:06 | |
| *** iceyao has joined #openstack-ansible | 03:06 | |
| *** sdake_ is now known as sdake | 03:08 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-lxc_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295606 | 03:11 |
| *** thorst has joined #openstack-ansible | 03:17 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-memcached_server: [WIP] Multi-distro-ify role https://review.openstack.org/279608 | 03:19 |
| *** thorst has quit IRC | 03:24 | |
| *** wtcross has joined #openstack-ansible | 03:26 | |
| *** wtcross has quit IRC | 03:31 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-memcached_server: Updated role using the Multi-Distro framework https://review.openstack.org/279608 | 03:34 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-lxc_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295606 | 03:36 |
| *** SamYaple is now known as contributed_this | 03:47 | |
| *** contributed_this is now known as SamYaple | 03:48 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-openstack_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295613 | 03:51 |
| *** tomoe_ has quit IRC | 03:55 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-lxc_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295606 | 03:56 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-galera_server: Updated role using the Multi-Distro framework https://review.openstack.org/295601 | 03:57 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-openstack_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295613 | 04:01 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_keystone: Updated role using the Multi-Distro framework https://review.openstack.org/295615 | 04:05 |
| *** jmhale has quit IRC | 04:11 | |
| *** jmhale has joined #openstack-ansible | 04:16 | |
| *** jduhamel has quit IRC | 04:21 | |
| *** fawadkhaliq has quit IRC | 04:22 | |
| *** fawadkhaliq has joined #openstack-ansible | 04:22 | |
| *** thorst has joined #openstack-ansible | 04:24 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-openstack_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295613 | 04:26 |
| *** wtcross has joined #openstack-ansible | 04:27 | |
| *** thorst has quit IRC | 04:29 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-lxc_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295606 | 04:29 |
| *** wtcross has quit IRC | 04:32 | |
| *** furlongm_ is now known as furlongm | 04:36 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_aodh: Updated role using the Multi-Distro framework https://review.openstack.org/295620 | 04:38 |
| *** daneyon_ has quit IRC | 04:42 | |
| *** daneyon has joined #openstack-ansible | 04:43 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-openstack_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295613 | 04:52 |
| *** furlongm has quit IRC | 04:54 | |
| *** daneyon has quit IRC | 04:57 | |
| *** tomoe_ has joined #openstack-ansible | 05:11 | |
| *** furlongm has joined #openstack-ansible | 05:11 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-openstack_hosts: Updated role using the Multi-Distro framework https://review.openstack.org/295613 | 05:16 |
| *** thorst has joined #openstack-ansible | 05:26 | |
| *** wtcross has joined #openstack-ansible | 05:27 | |
| *** wtcross has quit IRC | 05:32 | |
| *** thorst has quit IRC | 05:34 | |
| *** neilus has quit IRC | 05:39 | |
| *** saneax_AFK is now known as saneax | 05:41 | |
| *** fawadkhaliq has quit IRC | 05:52 | |
| *** neilus has joined #openstack-ansible | 06:20 | |
| *** phiche has joined #openstack-ansible | 06:23 | |
| *** phiche has quit IRC | 06:24 | |
| *** phiche1 has joined #openstack-ansible | 06:24 | |
| *** tomoe_ has quit IRC | 06:30 | |
| *** thorst has joined #openstack-ansible | 06:31 | |
| *** neilus has quit IRC | 06:34 | |
| *** javeriak has joined #openstack-ansible | 06:38 | |
| *** thorst has quit IRC | 06:38 | |
| *** neilus has joined #openstack-ansible | 06:44 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Enable developer mode with upper-constraints https://review.openstack.org/295301 | 06:54 |
| *** phiche1 has quit IRC | 06:55 | |
| *** phiche has joined #openstack-ansible | 07:05 | |
| *** mikelk has joined #openstack-ansible | 07:11 | |
| *** neilus has quit IRC | 07:14 | |
| *** neilus has joined #openstack-ansible | 07:14 | |
| *** neilus has quit IRC | 07:23 | |
| *** wtcross has joined #openstack-ansible | 07:29 | |
| *** sdake has quit IRC | 07:29 | |
| *** McMurlock1 has joined #openstack-ansible | 07:33 | |
| *** wtcross has quit IRC | 07:34 | |
| *** thorst has joined #openstack-ansible | 07:36 | |
| *** karimb has joined #openstack-ansible | 07:42 | |
| *** thorst has quit IRC | 07:44 | |
| *** asettle has quit IRC | 07:45 | |
| *** admin0 has joined #openstack-ansible | 07:48 | |
| *** admin0 has quit IRC | 07:54 | |
| *** admin0 has joined #openstack-ansible | 07:54 | |
| *** asettle has joined #openstack-ansible | 07:57 | |
| *** pegmanm has joined #openstack-ansible | 08:00 | |
| *** asettle has quit IRC | 08:02 | |
| *** markvoelker has quit IRC | 08:16 | |
| admin0 | good morning | 08:18 |
| hughsaunders | morning admin0 | 08:22 |
| *** wtcross has joined #openstack-ansible | 08:30 | |
| *** permalac has quit IRC | 08:33 | |
| *** openstackgerrit has quit IRC | 08:33 | |
| *** openstackgerrit has joined #openstack-ansible | 08:34 | |
| *** wtcross has quit IRC | 08:35 | |
| *** pcaruana has joined #openstack-ansible | 08:40 | |
| *** thorst has joined #openstack-ansible | 08:42 | |
| *** thorst has quit IRC | 08:49 | |
| *** mgoddard has joined #openstack-ansible | 09:03 | |
| *** javeriak has quit IRC | 09:03 | |
| *** javeriak has joined #openstack-ansible | 09:03 | |
| *** markvoelker has joined #openstack-ansible | 09:17 | |
| odyssey4me | o/ | 09:26 |
| odyssey4me | hughsaunders mattt andymccr please review asap: https://review.openstack.org/#/q/project:openstack/openstack-ansible-ironic+status:open+label:Verified%252B1 | 09:26 |
| *** wtcross has joined #openstack-ansible | 09:31 | |
| *** wtcross has quit IRC | 09:35 | |
| *** neilus has joined #openstack-ansible | 09:41 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_swift: Limit local tasks to local hosts https://review.openstack.org/295693 | 09:44 |
| *** electrofelix has joined #openstack-ansible | 09:44 | |
| *** thorst has joined #openstack-ansible | 09:47 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Adjust swift plays to use unified os-swift role https://review.openstack.org/293911 | 09:50 |
| *** markvoelker has quit IRC | 09:52 | |
| *** thorst has quit IRC | 09:54 | |
| *** iceyao has quit IRC | 10:00 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: [WIP] cinder functional testing https://review.openstack.org/295265 | 10:04 |
| *** Andrew_jedi has joined #openstack-ansible | 10:11 | |
| Andrew_jedi | My system clock is UTC then why OSA containers have EDT timezone, can i fix this via any playbook config ? | 10:12 |
| odyssey4me | Andrew_jedi hmm, never noticed that before - it may be possible using the environment spec that's used for proxy vars | 10:15 |
| Andrew_jedi | odyssey4me: have you ever noticed the issue that 1 out of 10 cinder volumes fails on creation and 50% fail on deletion ? | 10:16 |
| odyssey4me | Andrew_jedi as you can see in https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/user_variables.yml#L121-L134 there is a way using the dict 'global_environment_variables' in /etc/openstack_deploy/user_variables.yml to set environment variables | 10:16 |
| odyssey4me | Andrew_jedi no, although that's almost certainly related to the capacity available to your controllers and the tuning you use for the db/rabbitmq | 10:17 |
| odyssey4me | it could also be a bad back-end | 10:17 |
| Andrew_jedi | odyssey4me: yes but i am not using any proxy | 10:19 |
| Andrew_jedi | odyssey4me: Yes, i am also inclined to think that the bad back-end is the problem here. | 10:19 |
| odyssey4me | Andrew_jedi the remark about global environment vars is typically used for proxies, but perhaps could also be used to set the TZ | 10:20 |
| odyssey4me | Andrew_jedi interesting though, my container's time zone is set to UTC | 10:21 |
| Andrew_jedi | odyssey4me: Ahhh okay | 10:22 |
| odyssey4me | it matches the time zone set on the host | 10:22 |
| Andrew_jedi | odyssey4me: I think what happened here is that during installation TZ was EDT | 10:22 |
| Andrew_jedi | and later it was changed to UTC | 10:22 |
| odyssey4me | it seems that the default behaviour is to match the host at creation time: https://github.com/lxc/lxc/issues/47 | 10:22 |
| Andrew_jedi | odyssey4me: Someone else changed the TZ :/ | 10:22 |
| odyssey4me | so the easiest way to change it would be something like this: cd /opt/openstack-ansible/playbooks; ansible all_containers -m shell -a 'echo Etc/UTC | tee /etc/timezone' | 10:25 |
| *** johnmilton has quit IRC | 10:25 | |
| odyssey4me | try that out on one non-critical container, before doing it to all | 10:25 |
| odyssey4me | either use --limit, or use a container name instead of 'all_containers' | 10:26 |
| *** wtcross has joined #openstack-ansible | 10:31 | |
| *** Andrew_jedi has quit IRC | 10:32 | |
| *** neilus has quit IRC | 10:33 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Enable developer mode with upper-constraints https://review.openstack.org/295301 | 10:37 |
| *** neilus has joined #openstack-ansible | 10:49 | |
| *** wtcross has quit IRC | 10:51 | |
| *** thorst has joined #openstack-ansible | 10:52 | |
| *** Andrew_jedi has joined #openstack-ansible | 10:54 | |
| *** asettle has joined #openstack-ansible | 10:56 | |
| Andrew_jedi | odyssey4me: Thanks :) | 10:57 |
| *** thorst has quit IRC | 10:59 | |
| *** wtcross has joined #openstack-ansible | 11:04 | |
| *** wtcross has quit IRC | 11:09 | |
| *** schwicht has quit IRC | 11:17 | |
| *** weshay has joined #openstack-ansible | 11:35 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: [WIP] cinder functional testing https://review.openstack.org/295265 | 11:38 |
| *** iceyao has joined #openstack-ansible | 11:42 | |
| *** thorst has joined #openstack-ansible | 11:47 | |
| *** markvoelker has joined #openstack-ansible | 11:48 | |
| *** johnmilton has joined #openstack-ansible | 11:59 | |
| mhayden | buenos dias | 12:04 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: [WIP] cinder functional testing https://review.openstack.org/295265 | 12:07 |
| *** mathlin has quit IRC | 12:09 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate: Simplify test inventory https://review.openstack.org/295748 | 12:12 |
| odyssey4me | o/ mhayden | 12:12 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Adjust swift plays to use unified os-swift role https://review.openstack.org/293911 | 12:14 |
| *** Andrew_jedi_ has joined #openstack-ansible | 12:17 | |
| *** Andrew_jedi has quit IRC | 12:18 | |
| *** Andrew_jedi_ is now known as Andrew_jedi | 12:18 | |
| *** markvoelker has quit IRC | 12:21 | |
| mhayden | howdy | 12:25 |
| mattt | howdy mhayden | 12:26 |
| *** skamithi13 has quit IRC | 12:26 | |
| *** skamithi13 has joined #openstack-ansible | 12:27 | |
| *** skamithi13 has quit IRC | 12:28 | |
| *** skamithi13 has joined #openstack-ansible | 12:28 | |
| *** wtcross has joined #openstack-ansible | 12:29 | |
| *** pjm_ has joined #openstack-ansible | 12:35 | |
| *** mmm has joined #openstack-ansible | 12:35 | |
| *** neilus has quit IRC | 12:41 | |
| *** retreved has joined #openstack-ansible | 12:43 | |
| *** schwicht has joined #openstack-ansible | 12:44 | |
| *** retreved has quit IRC | 12:46 | |
| pjm_ | Hello, when I'm trying to install the OpenStack AIO | 12:47 |
| pjm_ | it gives me a error (not specific) | 12:47 |
| *** retreved has joined #openstack-ansible | 12:47 | |
| pjm_ | - [ Error Info - 326 0 ] --------------------------------------------- | 12:47 |
| pjm_ | I run the commands that were in the page | 12:47 |
| pjm_ | is there a way to have a more verbose debug? | 12:47 |
| *** markvoelker has joined #openstack-ansible | 12:47 | |
| pjm_ | thanks for the attention :) | 12:48 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: [WIP] Cinder functional testing https://review.openstack.org/295265 | 12:52 |
| *** woodard has joined #openstack-ansible | 12:52 | |
| *** woodard has quit IRC | 12:52 | |
| *** woodard has joined #openstack-ansible | 12:53 | |
| *** neilus has joined #openstack-ansible | 12:54 | |
| *** woodard has quit IRC | 12:55 | |
| automagically | morning | 12:56 |
| *** woodard has joined #openstack-ansible | 12:56 | |
| odyssey4me | pjm_ there is a ton of information above that which would be more useful for debugging - can you pastebin the full log? | 12:56 |
| pjm_ | where the full log is stored? | 12:56 |
| pjm_ | Or I should be running the script saving the output to a file, like | 12:58 |
| pjm_ | ./bootstrap-aio.sh > output.log | 12:58 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate: Remove the database creation & messaging tasks from the role https://review.openstack.org/295778 | 12:59 |
| odyssey4me | pjm_ bootstrap-aio is only the first part - and the log outputs to the screen in stdout | 12:59 |
| odyssey4me | pjm did it fail in the first part? | 12:59 |
| pjm_ | sorry it was the last one, the run-playbook | 13:00 |
| odyssey4me | pjm_ can you see from the last few lines of the log which playbook it failed on? | 13:00 |
| pjm_ | first I did: scripts/bootstrap-ansible.sh, scripts/bootstrap-aio.sh and then | 13:00 |
| pjm_ | openstack-ansible os-keystone-install.yml | 13:00 |
| odyssey4me | pjm_ ok, did you checkout a tag, or a branch? which was it? | 13:00 |
| pjm_ | I did checkout to version 12.0.0 and 12.0.8 | 13:01 |
| odyssey4me | and which tasks failed? can you copy/paste the output on your screen into a pastebin | 13:01 |
| odyssey4me | ? | 13:01 |
| pjm_ | I was thinking to change to the 11.0.0 | 13:01 |
| pjm_ | my screen was cleared :\ but i didn't find any information usefull. I will run again and paste the output to a file, and then upload it to pastebin | 13:02 |
| pjm_ | it's a good idea? | 13:02 |
| odyssey4me | pjm_ you can't switch willy nilly - each major version relates to a major openstack version and they are incompatible if you've part implemented things | 13:02 |
| pjm_ | (but the last few lines that I saw the useful info about an error was that number) | 13:02 |
| odyssey4me | as we know there's an issue with 12.0.8, I'd recommend checking out the head of liberty: git checkout -b origin/liberty | 13:03 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Remove old MariaDB apt sources https://review.openstack.org/289977 | 13:03 |
| pjm_ | hmm, ok thanks. And I should do teardown.sh or more scripts for cleanup? | 13:03 |
| odyssey4me | then execute all these things: http://docs.openstack.org/developer/openstack-ansible/install-guide/app-minorupgrade.html | 13:03 |
| odyssey4me | steps 4-8 | 13:04 |
| pjm_ | thank you :) | 13:04 |
| pjm_ | when I put to run the script should I save the stdout to a file? | 13:04 |
| pjm_ | for debug in case of errors? | 13:04 |
| odyssey4me | pjm_ if you want to you can execute: export ANSIBLE_LOG_PATH=/foo/bar.log | 13:05 |
| odyssey4me | that'll tell ansible to log to a file | 13:05 |
| pjm_ | loool it's much so easy that way :) | 13:06 |
| pjm_ | thank you and sorry for being so noob | 13:06 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_designate: Simplify test inventory https://review.openstack.org/295748 | 13:07 |
| odyssey4me | pjm_ we all start somewhere :) | 13:07 |
| mhayden | Apsu: imma start hacking on that lbaasv2 backport for liberty | 13:08 |
| pjm_ | odyssey4me: you're right, i'm trying to learn ansible and deployment with OpenStack | 13:08 |
| pjm_ | didn't know if it's better to first deploy manually and then understand how do it with a configuration management, or if it's the same | 13:08 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: Cinder functional testing https://review.openstack.org/295265 | 13:09 |
| odyssey4me | o/ automagically | 13:09 |
| pjm_ | odyssey4me: what do you mean? | 13:10 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_client: Updated role using the Multi-Distro framework https://review.openstack.org/295599 | 13:10 |
| odyssey4me | jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr I worked on simplifying the inventory to make use of the standard Ansible inventory mechanism for tests. In my view this is a good thing as it serves the purpose for testing, but also shows how to do a non-dynamic inventory with out stuff. Comments please: https://review.openstack.org/295748 | 13:11 |
| * automagically looking | 13:11 | |
| odyssey4me | pjm_ I mean that we all (including those you may call experts) started with the basics. | 13:11 |
| pjm_ | sorry, i was thinking that the word "automagically" was for me. Yes I agree | 13:12 |
| *** b3rnard0_away is now known as b3rnard0 | 13:13 | |
| mhayden | i love it when things just work automagically like that | 13:14 |
| *** adreznec has joined #openstack-ansible | 13:15 | |
| mhayden | speaking of irc nicks, every time i see pabelanger, i read it as pabel-anger and think he is angry | 13:15 |
| mhayden | my brain is broken, but we already knew that | 13:15 |
| pabelanger | ANGER | 13:19 |
| odyssey4me | my brain tends to switch between that and something that sounds more like bollinger (said the way Bond says it) | 13:20 |
| mhayden | then again, every time i see hughsaunders, i read it as HUG-H-SAUNDERS | 13:23 |
| odyssey4me | automagically FYI, using that format of inventory you're able to execute parts of the playbook instead of having to go through everything as we do now | 13:23 |
| odyssey4me | right now you have to go through all the galera/keystone stuff in order to test a change in the role being tested... with a static inventory like that you won't have to - you can simply execute the role | 13:24 |
| mattt | odyssey4me: https://review.openstack.org/#/c/295265/10/tests/inventory | 13:25 |
| mattt | that is the key | 13:25 |
| mattt | i think i can actually clean that up so variables are only set under [all] right ? | 13:25 |
| mattt | odyssey4me: https://review.openstack.org/#/c/295265/10/tox.ini | 13:25 |
| *** neilus1 has joined #openstack-ansible | 13:26 | |
| *** scarlisle has joined #openstack-ansible | 13:28 | |
| *** furlongm_ has joined #openstack-ansible | 13:28 | |
| *** neilus1 has quit IRC | 13:29 | |
| *** neilus1 has joined #openstack-ansible | 13:30 | |
| *** furlongm has quit IRC | 13:30 | |
| *** neilus has quit IRC | 13:30 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_cinder: Cinder functional testing https://review.openstack.org/295265 | 13:31 |
| *** wtcross has quit IRC | 13:31 | |
| *** neilus1 has quit IRC | 13:32 | |
| *** javeriak has quit IRC | 13:32 | |
| *** markvoelker_ has joined #openstack-ansible | 13:34 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server: Updated role using the Multi-Distro framework https://review.openstack.org/295601 | 13:36 |
| *** markvoelker has quit IRC | 13:36 | |
| *** wtcross has joined #openstack-ansible | 13:37 | |
| automagically | odyssey4me: Given the reliance of most services on keystone and galera anyway for overall functionality, I’m not sure how much that saves us, but that is a benefit that was not immediately clear to me | 13:38 |
| *** markvoelker has joined #openstack-ansible | 13:38 | |
| odyssey4me | automagically when you're implementing changes in a role and testing those changes, it saves an awful lot of time :) | 13:38 |
| odyssey4me | of course this benefit requires the test playbook to be broken out, which many aren't | 13:39 |
| *** woodard has quit IRC | 13:39 | |
| odyssey4me | basically you do the setup of the infra once, then you run and reirun the play for the role execution on its own as much as you want | 13:39 |
| odyssey4me | for gating specifically this makes no difference | 13:39 |
| automagically | odyssey4me: Right, makes sense | 13:39 |
| *** markvoelker_ has quit IRC | 13:40 | |
| *** saneax is now known as saneax_AFK | 13:40 | |
| mattt | odyssey4me: https://review.openstack.org/#/c/292911/6/tests/test.yml | 13:43 |
| *** ametts has joined #openstack-ansible | 13:44 | |
| *** Andrew_jedi has quit IRC | 13:46 | |
| *** chhavi has joined #openstack-ansible | 13:50 | |
| *** javeriak has joined #openstack-ansible | 13:52 | |
| *** jthorne has joined #openstack-ansible | 13:54 | |
| openstackgerrit | git-harry proposed openstack/openstack-ansible-os_keystone: WIP: Use os-service-installer https://review.openstack.org/292303 | 13:55 |
| *** pjm_ has quit IRC | 13:57 | |
| *** mgoddard_ has joined #openstack-ansible | 13:59 | |
| *** b3rnard0 is now known as b3rnard0_away | 13:59 | |
| *** mgoddard has quit IRC | 14:03 | |
| *** asettle has quit IRC | 14:04 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:05 | |
| *** spotz_zzz is now known as spotz | 14:09 | |
| *** bryan_att has joined #openstack-ansible | 14:09 | |
| Apsu | mhayden: Sounds good. | 14:11 |
| *** Mudpuppy has joined #openstack-ansible | 14:17 | |
| *** neilus has joined #openstack-ansible | 14:23 | |
| openstackgerrit | Merged openstack/openstack-ansible-ironic: Enable developer mode with upper-constraints https://review.openstack.org/295301 | 14:24 |
| *** woodard has joined #openstack-ansible | 14:27 | |
| cloudnull | odyssey4me: one of the issues we've had with static inventory is a relience on dict type vars in inventory which static inventory cant support. If we can make that go (upstream ansible) we can greatly simplify our supported inventory formats. | 14:30 |
| *** b3rnard0_away is now known as b3rnard0 | 14:30 | |
| odyssey4me | cloudnull we have a dict in there, and it works - at least for the dict requires for role testing | 14:33 |
| odyssey4me | cloudnull we simply apply group/host vars as per a standard inventory | 14:33 |
| cloudnull | its read in as a string | 14:33 |
| cloudnull | a unicode string | 14:33 |
| *** openstackgerrit has quit IRC | 14:33 | |
| *** openstackgerrit has joined #openstack-ansible | 14:34 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: [WIP] Add LBaaS v2 support in Liberty https://review.openstack.org/295866 | 14:34 |
| hughsaunders | mhayden: ({) | 14:35 |
| *** asettle has joined #openstack-ansible | 14:36 | |
| *** wtcross has quit IRC | 14:37 | |
| cloudnull | i think the addition of host / group vars outside of inventory is more complex than using the dynamic inventory with _hosts defined within it. | 14:37 |
| odyssey4me | cloudnull I don't understand the problem - perhaps you can illustrate it with something specific? I may be being thick here. The inventory itself is static and only needs INI formatted things (host/group memberships) - the group_vars and host_vars are expressed in yaml and seem to work perfectly for the requirements we have for role testing. | 14:37 |
| odyssey4me | The intent is to provide something simple, standard ansible and something that fulfills the needs of role testing. The intent is not to replace the dynamic inventory at all - it's just to facilitate role testing. | 14:39 |
| cloudnull | yes, its thing like http://cdn.pasteraw.com/nztecu9b63r59hwmu9anvqszhkjwiaw would be impossible in static inventory without making changes to the checkedout source code | 14:40 |
| odyssey4me | yeah, but that doesn't matter for role testing - the static inventory is a fixed point to use for testing | 14:40 |
| odyssey4me | if a consumer consumes the role they can learn from the inventory in tests to define their own | 14:41 |
| *** asettle has quit IRC | 14:41 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Consolidate pip install tasks https://review.openstack.org/295383 | 14:43 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Re-deploy the venv if it mismatches the repo https://review.openstack.org/295390 | 14:44 |
| cloudnull | this is all fine testing if we thing its best however I'd caution us using it as an example for "non-dynamic inventory". IMO itll just create more confusion around inventory . | 14:44 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Switch from MySQL-python to PyMySQL https://review.openstack.org/295432 | 14:44 |
| cloudnull | which we already have enough of | 14:44 |
| odyssey4me | cloudnull right now the inventory in the tests is inconvenient and difficult to understand - https://review.openstack.org/295748 is meant to lower the barrier for people who're used to Ansible static inventories, and also make the testing layout easier to understand for the role in question | 14:47 |
| odyssey4me | it took me half a day to figure out how we were doing inventory when first looking into how to simplify the tests | 14:48 |
| openstackgerrit | Merged openstack/openstack-ansible-ironic: Consolidate pip install tasks https://review.openstack.org/295383 | 14:49 |
| cloudnull | my other question is, how is this simpler? -- i mean its essentially _not_ testing what we're actually using . | 14:50 |
| cloudnull | I get it, just trying to understand the motivation. | 14:50 |
| odyssey4me | it allows me to run any part of the play, without having to execute this task: https://github.com/openstack/openstack-ansible-os_keystone/blob/master/tests/test.yml#L97-L118 | 14:53 |
| cloudnull | odyssey4me: regarding https://review.openstack.org/#/q/status:open+branch:master+topic:multi_distro_framework why would we not finish the impl for that, we have it in several roles and not others? | 14:53 |
| cloudnull | fair enough | 14:53 |
| odyssey4me | and it demonstrates how to use the role with a static inventory, which was part of the intention when splitting out the roles | 14:53 |
| openstackgerrit | Merged openstack/openstack-ansible-ironic: Re-deploy the venv if it mismatches the repo https://review.openstack.org/295390 | 14:54 |
| odyssey4me | furthermore https://github.com/openstack/openstack-ansible-os_keystone/blob/master/tests/test.yml#L97-L118 isn't doing what we do in the integrated build either | 14:54 |
| *** mgoddard_ has quit IRC | 14:55 | |
| cloudnull | it is. just without the python executable | 14:55 |
| odyssey4me | the integrated build provides the opportunity for us to test with the dyanmic inventory - I thin that using the static inventory will help us understand assumptions we're making to better the independent use of the roles - just like decoupling the roles from each other has done | 14:55 |
| cloudnull | its a like for like data structure | 14:55 |
| *** mgoddard has joined #openstack-ansible | 14:55 | |
| odyssey4me | and the resulting inventory is also like for like with this static inventory | 14:55 |
| cloudnull | where the other is not. but it looks fine. | 14:55 |
| cloudnull | like i said was just trying to understand the motivation | 14:55 |
| cloudnull | odyssey4me: its not, if you dump the inventory they'l be different. | 14:56 |
| odyssey4me | see http://pastebin.com/8NpnjH4X | 14:57 |
| *** sdake has joined #openstack-ansible | 14:58 | |
| *** izaakk has joined #openstack-ansible | 15:00 | |
| openstackgerrit | Merged openstack/openstack-ansible-ironic: Switch from MySQL-python to PyMySQL https://review.openstack.org/295432 | 15:00 |
| odyssey4me | it looks to me pretty much like the same sort of inventory: http://pastebin.com/fwnE5qrP | 15:02 |
| *** iceyao has quit IRC | 15:07 | |
| odyssey4me | mhayden minor tweak on https://review.openstack.org/290526 | 15:07 |
| *** keedya has joined #openstack-ansible | 15:07 | |
| mhayden | odyssey4me: i'll check | 15:10 |
| *** weezS has joined #openstack-ansible | 15:12 | |
| cloudnull | odyssey4me: sounds good to me . | 15:15 |
| openstackgerrit | git-harry proposed openstack/openstack-ansible-os_keystone: WIP: Use os-service-installer https://review.openstack.org/292303 | 15:15 |
| *** galstrom_zzz is now known as galstrom | 15:16 | |
| cloudnull | ^ thats a lot of juno-isms being receated. | 15:19 |
| cloudnull | odyssey4me: have you seen this where a var is not marked as undefined when used in var file ? specifically -- https://review.openstack.org/#/c/295613/5/vars/ubuntu-14.04.yml -- which is throwing an error ``'ansible_kernel' is undefined`` | 15:34 |
| odyssey4me | cloudnull heh, that's a fun one | 15:37 |
| odyssey4me | your debug var a little higher shows it, yet it doesn't evaluate in the if statement | 15:37 |
| cloudnull | yea the fact is being set (added debug to it) but its not read in the var file | 15:37 |
| cloudnull | yup. | 15:37 |
| cloudnull | its odd | 15:37 |
| odyssey4me | oh hang on - this is when the var is moved from defaults to the var file? | 15:37 |
| odyssey4me | hmm | 15:38 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Include security role in setup-hosts.yml https://review.openstack.org/290526 | 15:38 |
| cloudnull | yup | 15:38 |
| odyssey4me | well, perhaps that should be switched to use a task instead | 15:38 |
| cloudnull | the fact is gathered before the var file is loaded | 15:38 |
| cloudnull | but its still not read | 15:38 |
| cloudnull | Im. poking aroung the ansible internals to see why, was curious if you (or others) had seen that before | 15:39 |
| odyssey4me | it may be that the var file cannot include system vars? | 15:39 |
| cloudnull | that may be | 15:39 |
| odyssey4me | I've only seen something similar with the ansible cpu count - which is why we default it everywhere we use it | 15:39 |
| *** mmm has quit IRC | 15:40 | |
| *** eil397 has joined #openstack-ansible | 15:40 | |
| cloudnull | yea, its an odd one. | 15:41 |
| odyssey4me | cloudnull to answer your earlier question around the multi-OS stuff - it's already agreed to be a focal point for next cycle, so I think it makes sense to rather get going on that once we've cut the stable/mitaka branch... | 15:41 |
| *** eil397 has quit IRC | 15:41 | |
| odyssey4me | right now we need to finalise functional testing in the roles, update the SHA's in the roles to validate that they work, then update the SHA's for the integrated tests | 15:41 |
| *** eil397 has joined #openstack-ansible | 15:41 | |
| odyssey4me | the primary focus is the integrated build roles | 15:42 |
| cloudnull | that fine. but by not implementing the basic framework we'll have roles that have it and not kind spread all about. | 15:42 |
| odyssey4me | sure, that's why I suggested that the line be drawn at the infra roles where it's low impact | 15:42 |
| cloudnull | which is ok. but i was looking at trying to make them all more uniform | 15:42 |
| cloudnull | along those lines I was looking through https://github.com/openstack/openstack-ansible-specs and we should move our specs forward to newton that wont be implemented in this cycle. | 15:44 |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate: Remove the database creation & messaging tasks from the role https://review.openstack.org/295778 | 15:44 |
| *** arif-ali has quit IRC | 15:44 | |
| odyssey4me | yep, that makes sense - would you mind putting together a review for that? | 15:44 |
| cloudnull | sure | 15:44 |
| *** KLevenstein has joined #openstack-ansible | 15:45 | |
| odyssey4me | cloudnull FYI mattt figured out how to make the tox environment run more normally in the gate: https://review.openstack.org/295265 | 15:47 |
| odyssey4me | it turns out that -infra do some funky things | 15:47 |
| *** arif-ali has joined #openstack-ansible | 15:50 | |
| *** h1nch has quit IRC | 15:50 | |
| *** neilus has quit IRC | 15:51 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: [WIP] Add LBaaS v2 support in Liberty https://review.openstack.org/295866 | 15:52 |
| cloudnull | ah. -- funky things strikes again | 15:53 |
| *** skamithi13 has quit IRC | 15:55 | |
| odyssey4me | bug triage here cloudnull, mattt, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, Sam-I-Am, odyssey4me, serverascode, rromans, erikmwilson, mancdaz, _shaps_, BjoernT, claco, echiu, dstanek, jwagner, ayoung, prometheanfire, evrardjp, arbrandes, mhayden, scarlisle, luckyinva, ntt, javeriak, automagically, spotz, vdo, jmccrory, alextricity25, jasondotstar, KLevenstein, admin0, michaelgugino, ametts, v1k0d3n, | 15:58 |
| odyssey4me | severion, bgmccollum | 15:58 |
| *** jmccrory_ has joined #openstack-ansible | 15:58 | |
| spotz | o/ | 15:58 |
| odyssey4me | https://bugs.launchpad.net/openstack-ansible/+bugs?search=Search&field.status=New | 15:59 |
| cloudnull | o/ | 15:59 |
| eil397 | o/ | 15:59 |
| v1k0d3n | o/ | 15:59 |
| *** chhavi has quit IRC | 15:59 | |
| automagically | o/ | 15:59 |
| *** skamithi13 has joined #openstack-ansible | 15:59 | |
| odyssey4me | first up - mhayden have you made any progress on confirming https://bugs.launchpad.net/openstack-ansible/+bug/1556204 ? | 15:59 |
| openstack | Launchpad bug 1556204 in openstack-ansible "radvd not starting with IPv6 subnets" [Medium,New] - Assigned to Major Hayden (rackerhacker) | 15:59 |
| *** skamithi13 has quit IRC | 15:59 | |
| *** jorge_munoz has joined #openstack-ansible | 16:00 | |
| mattt | o/ | 16:00 |
| prometheanfire | hi | 16:00 |
| *** jorge_munoz_ has joined #openstack-ansible | 16:03 | |
| odyssey4me | no update for now - will leave as-is | 16:03 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1557682 | 16:03 |
| openstack | Launchpad bug 1557682 in openstack-ansible "repo_build cannot refer to variables in Creating OSA requirement wheels " [Undecided,New] | 16:03 |
| odyssey4me | this looks like a feature request from stevelle with regards to the repo-build process? | 16:04 |
| hughsaunders | o/ | 16:04 |
| stevelle | yes | 16:04 |
| stevelle | seems like a feature request | 16:05 |
| *** keedya has quit IRC | 16:05 | |
| odyssey4me | this looks like it'll be useful in the next cycle as the gnocci role comes into play | 16:05 |
| odyssey4me | any volunteers to take a look into it? | 16:05 |
| *** jorge_munoz has quit IRC | 16:05 | |
| *** jorge_munoz_ is now known as jorge_munoz | 16:05 | |
| d34dh0r53 | o/ | 16:05 |
| mattt | sounds like stevelle has this one :P | 16:06 |
| stevelle | I really don't mattt :) | 16:06 |
| odyssey4me | cloudnull perhaps you can sync up with stevelle and take a crack at it? | 16:06 |
| cloudnull | sure | 16:07 |
| *** h1nch has joined #openstack-ansible | 16:07 | |
| odyssey4me | alright, that'd be great | 16:07 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1558440 | 16:07 |
| openstack | Launchpad bug 1558440 in openstack-ansible "Adding new compute node missing ceilometer instructions" [Undecided,New] | 16:07 |
| odyssey4me | ah, I see metering_compute_containers, but no hosts related membership | 16:08 |
| spotz | Is the nova_key comments related or no? | 16:09 |
| logan- | nova_key stuff was addressed by https://review.openstack.org/#/c/294379/ | 16:10 |
| spotz | ok I"ll grab the ceilometer edit then and put that in a comment | 16:10 |
| odyssey4me | thanks spotz | 16:10 |
| odyssey4me | and finallly: https://bugs.launchpad.net/openstack-ansible/+bug/1559619 | 16:11 |
| openstack | Launchpad bug 1559619 in openstack-ansible "Could not fetch URL https://rpc-repo.rackspace.com/os-releases/12.0.8" [Undecided,New] | 16:11 |
| odyssey4me | I'm not sure why this matters, unless someone was trying to clone from rpc-repo. | 16:12 |
| spotz | odyssey4me I though automagically removed the rpc-repo references last week | 16:12 |
| odyssey4me | didn't we discuss this recently and remove something related from docs | 16:12 |
| automagically | I removed them from the docs | 16:12 |
| automagically | But there is still a reference in the play | 16:12 |
| automagically | https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L41 | 16:13 |
| odyssey4me | yeah, that's fine - it doesn't require anything, it simply uses stuff from there if it's available | 16:13 |
| automagically | But _should it?_ | 16:14 |
| mattt | so is it that we don't want people cloning any more? | 16:14 |
| odyssey4me | We'll tackle that in Newton. I've uncovered some nastiness in there when attempting to remove it. | 16:14 |
| mattt | if so, can we not just update the bug to let the reporter know? | 16:15 |
| odyssey4me | mattt people can clone if they want to, but from their own repo | 16:15 |
| odyssey4me | ah, found the doc change: https://review.openstack.org/#/c/293024/ | 16:16 |
| mattt | playbooks/repo-clone-mirror.yml still points to rep-repo tho no? | 16:17 |
| odyssey4me | mattt yep - nothing uses that though. We should perhaps figure out what to do with it. | 16:18 |
| odyssey4me | Is it useful to anyone? Should it be kept? | 16:18 |
| mattt | well people use it because we've gone back and forth on this a few times it feels | 16:18 |
| odyssey4me | Maybe we should remove the source so that the clone hard fails indicating a missing var? | 16:19 |
| odyssey4me | Although I'm sort-of in favor of just removing the play altogether. | 16:19 |
| mattt | dunno, but not sure we should be changing this stuff mid-way through a cycle | 16:19 |
| *** admin0 has quit IRC | 16:19 | |
| cloudnull | +1 on removing the play. | 16:20 |
| cloudnull | while its nice to be able to clone a mirror but the mirror is RAX managed and not always up-to-date especially when tracking releases and branches outside of the RAX product. | 16:22 |
| *** fignew has joined #openstack-ansible | 16:22 | |
| cloudnull | if we could point that to an infra resource that'd be nice. | 16:22 |
| cloudnull | but idk thats possible. | 16:23 |
| mattt | or update the play so that it spits out an error if it's trying to use rpc-repo | 16:23 |
| odyssey4me | yeah, I already spoke to infra and they're not keen - they already have a wheel repo | 16:23 |
| mattt | removing playbooks midcycle feels a bit shoddy | 16:24 |
| cloudnull | then I'm +1 on seeing the play go the way of the dodo bird. | 16:24 |
| eil397 | it is not the same repo. as i Know | 16:24 |
| *** fignew has quit IRC | 16:24 | |
| *** fignew has joined #openstack-ansible | 16:24 | |
| odyssey4me | as far as I know, RPC is moving away from using the clone mechanism anyway - so the likelihood of rpc-repo staying up to date is going to be less and less | 16:24 |
| *** sdake has quit IRC | 16:25 | |
| odyssey4me | so yeah, cloudnull perhaps you can submit a review - we can discuss in review | 16:25 |
| odyssey4me | automagically I think it would be best to backport that doc fix to the liberty branch | 16:25 |
| mattt | cloudnull: want me to update the play to spit out an error (in liberty) | 16:26 |
| mattt | and then we can deprecate it in mitaka | 16:26 |
| *** daneyon has joined #openstack-ansible | 16:26 | |
| spotz | odyssey4me could the play be changed to pick up a user var with a repo? | 16:26 |
| mattt | s/deprecate/remove/ | 16:26 |
| odyssey4me | mattt yeah, I think that's a good idea - let's make it clear that the option is deprecated in liberty and removed in mitaka | 16:26 |
| mattt | odyssey4me: ok i can do that | 16:26 |
| odyssey4me | spotz yeah, it already can - it's just not clear that it can - so we need to resolve that | 16:27 |
| cloudnull | mattt: +1 | 16:27 |
| automagically | odyssey4me: I’ll propose the backport | 16:27 |
| odyssey4me | cool, thanks | 16:28 |
| spotz | automagically You going to bug the stronger wording too? Or is that what you're talking about?:) | 16:28 |
| automagically | spotz: I’ll backport the doc change that removes the suggestion | 16:28 |
| spotz | cool:) | 16:29 |
| *** sdake has joined #openstack-ansible | 16:29 | |
| *** raddaoui has joined #openstack-ansible | 16:29 | |
| odyssey4me | that backport should probably have a release note added to it automagically to ensure that anyone picking up a fresh tag is aware of the change | 16:29 |
| automagically | Got it | 16:30 |
| *** jorge_munoz has quit IRC | 16:30 | |
| odyssey4me | or maybe mattt's change should rather have it | 16:30 |
| mattt | yeah i'll relnote it | 16:30 |
| odyssey4me | actually, I think it makes more sense for the play change to have it | 16:30 |
| automagically | cool | 16:30 |
| odyssey4me | ok, are there any bugs or reviews that anyone wants to drawr attention to? | 16:30 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Removed the repo clone mirror play https://review.openstack.org/295941 | 16:30 |
| spotz | My only one up for review got moved to Not Current so I just have my bugs to work on | 16:33 |
| eil397 | I would like to remove https://bugs.launchpad.net/openstack-ansible/+bug/1518485 "lxcbr0 disappeared run-playbooks.sh" | 16:33 |
| openstack | Launchpad bug 1518485 in openstack-ansible " lxcbr0 disappeared run-playbooks.sh " [Undecided,Incomplete] - Assigned to Anton Haldin (anton-haldin) | 16:33 |
| eil397 | I'm not sure what is the status of it. I can check and remove after it. | 16:34 |
| *** mikelk has quit IRC | 16:35 | |
| odyssey4me | eil397 you're seeing this in a ci job? | 16:35 |
| odyssey4me | eil397 and in that job you're running run-aio-build.sh ? | 16:35 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Removed the repo clone mirror play https://review.openstack.org/295941 | 16:35 |
| eil397 | I've created internal ci job to install os on few hardware servers | 16:36 |
| eil397 | and at the time I've created this describtion. it was pretty stable for reproducing. on ubuntu machines. | 16:36 |
| odyssey4me | eil397 can you make the CI job do the steps in http://docs.openstack.org/developer/openstack-ansible/developer-docs/quickstart-aio.html instead of running run-aio-build.sh ? | 16:37 |
| odyssey4me | it's possible that it is a race condition, except that we run CI jobs all day and don't hit this - so it's a little weird | 16:37 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: Removing reference to internal RPC repo https://review.openstack.org/295943 | 16:37 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Removed the repo clone mirror play https://review.openstack.org/295941 | 16:37 |
| odyssey4me | eil397 FYI for our CI job we run https://github.com/openstack/openstack-ansible/blob/master/scripts/gate-check-commit.sh instead which is built specifically for gating | 16:38 |
| eil397 | odyssey4me: issue for me in my cases was in run-playbooks.sh | 16:38 |
| eil397 | this thing does not work stable for me https://github.com/openstack/openstack-ansible/blob/master/scripts/run-playbooks.sh#L58 | 16:38 |
| *** jguy_ has joined #openstack-ansible | 16:38 | |
| eil397 | (ifdown lxcbr0 || true); ifup lxcbr0 | 16:39 |
| eil397 | I can propose other way to do it. but. I completly agree that all this logic is for ci infra. and it is not important. | 16:39 |
| odyssey4me | eil397 considering that none of us can reproduce this, perhaps you should submit a patch for something that works for you and we can discuss the alternative in review? | 16:40 |
| eil397 | odyssey4me: thank you for link. | 16:40 |
| cloudnull | spotz: any chance you might be able to have another look at https://review.openstack.org/#/c/290873/ | 16:40 |
| jguy_ | Trying to setup an AIO, running into a problem on one of the repo containers: from M2Crypto import SSL | 16:40 |
| jguy_ | ImportError: No module named M2Crypto | 16:40 |
| jguy_ | ---------------------------------------- | 16:40 |
| jguy_ | Failed building wheel for pywbem | 16:40 |
| jguy_ | Running setup.py clean for pywbem | 16:40 |
| cloudnull | hughsaunders: had a commend regarding some of the changes | 16:40 |
| automagically | jguy_: See https://github.com/openstack/openstack-ansible/commit/4e89921b9d35fbb27b2cdd733c75d30b0e61e957 | 16:41 |
| cloudnull | jguy_: that was recently resolved in the branches. | 16:41 |
| cloudnull | i dont think the changes have made it to a tag yet | 16:41 |
| jguy_ | Ahhh...ok | 16:41 |
| odyssey4me | jguy_ yeah, update to the head of liberty, or apply that patch | 16:41 |
| jguy_ | ok. Will do. | 16:42 |
| hughsaunders | cloudnull: spotz yeah, minor comments, I think its a really useful addition overall | 16:42 |
| odyssey4me | (assuming you're using liberty) | 16:42 |
| jguy_ | I am trying to. ;) | 16:42 |
| *** eil397 has quit IRC | 16:46 | |
| spotz | cloudnull I just looked at it and it said Not Current, will make revisions | 16:47 |
| *** eil397 has joined #openstack-ansible | 16:47 | |
| spotz | I do think eventually we need to seperate all those docs into Ceph and NetApps | 16:47 |
| hughsaunders | spotz: that probably means you were viewing an old revision? | 16:47 |
| spotz | I thought I refreshed hughsaunders but I do keep tabs open for stuff I'm doing or reviewing so I can keep track and follow-up when needed | 16:48 |
| *** MCoLo__ has joined #openstack-ansible | 16:50 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_designate: Simplify test inventory https://review.openstack.org/295748 | 16:51 |
| *** pcaruana has quit IRC | 16:52 | |
| *** MCoLo has quit IRC | 16:52 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift: Limit local tasks to local hosts https://review.openstack.org/295693 | 16:53 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_keystone: Test Keystone Mitaka RC1 in our role https://review.openstack.org/295953 | 16:56 |
| *** neilus has joined #openstack-ansible | 16:56 | |
| *** keedya has joined #openstack-ansible | 16:57 | |
| daneyon | I am using Ansible to manage CoreOS systems. Unfortunately the Ansible fact 'ansible_distribution' does not work for CoreOS: https://gist.github.com/danehans/dc7081b4ac5dfe857cfe#file-coreos_ansible_facts-L99-L102 | 16:58 |
| daneyon | However, ansible_lsb does work: https://gist.github.com/danehans/dc7081b4ac5dfe857cfe#file-coreos_ansible_facts-L258-L263 | 16:58 |
| daneyon | I am trying to look through the Ansible code to see how the ansible_distribution fact is derived. Can someone help point me in the right direction? | 16:59 |
| cloudnull | hum . | 17:00 |
| daneyon | larsks if you're around, I could use your guidance for ^ | 17:00 |
| cloudnull | ive not looked into those internals | 17:00 |
| odyssey4me | o/ daneyon which version of Ansible? cloudnull's been digging through it today :p | 17:00 |
| automagically | https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py <- daneyon | 17:01 |
| daneyon | cloudnull odyssey4me: oops... I meant to send this to the ansible irc channel. I'm using 1.9.4 | 17:01 |
| eil397 | : - ) | 17:02 |
| automagically | daneyon: https://github.com/ansible/ansible/blob/stable-1.9/lib/ansible/module_utils/facts.py#L241 | 17:02 |
| larsks | daneyon: I am around, but automagically has you pointed in the right direction. Beware: messy python lies ahead! | 17:02 |
| daneyon | automagically I did a search for ansible_distribution against ^ and didn't find anything | 17:03 |
| cloudnull | daneyon: https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py#L282 | 17:03 |
| cloudnull | doesnt look like coreos is a defined | 17:03 |
| cloudnull | even in master | 17:03 |
| daneyon | automagically cloudnull let me review | 17:03 |
| cloudnull | *devel | 17:03 |
| larsks | daneyon: the 'ansible_' prefix is added later (just look for 'distribution'). | 17:04 |
| spotz | hughsaunders: Can I get the translation for SPOF?:) | 17:04 |
| odyssey4me | spotz Single Point of Failure | 17:04 |
| *** jthorne has quit IRC | 17:05 | |
| spotz | heheh, I've never called it that | 17:05 |
| cloudnull | daneyon: well it is defined if the distro is set as "NA" https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py#L507-L525 and the platform information is returned as "coreos" https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py#L335 | 17:05 |
| *** jthorne has joined #openstack-ansible | 17:06 | |
| cloudnull | daneyon: on coreos can you import plaform and run platform.dist() | 17:06 |
| cloudnull | it would seem to me that its return value is not some variant of coreos | 17:07 |
| daneyon | cloudnull let me did into that | 17:07 |
| daneyon | cloudnull this file does not exist: https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py#L520 | 17:08 |
| cloudnull | but it has to hit the conditional first and the distrobution is set here https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts.py#L335-L336 | 17:09 |
| daneyon | cloudnull I could add update.conf to manage releases, but I want ansible to support a stock system | 17:09 |
| cloudnull | seems like the fact module will need some love to properlly support coreos | 17:11 |
| daneyon | cloudnull I'm going to test a few things and will let you know how it goes. | 17:12 |
| cloudnull | yea please do. :) | 17:12 |
| *** phiche has quit IRC | 17:12 | |
| cloudnull | hows it going btw, long time no talk :) | 17:12 |
| SamYaple | o/ daneyon youre just everywhere arent you | 17:13 |
| *** McMurlock1 has quit IRC | 17:14 | |
| daneyon | SamYaple lol... Not easy making coreos and ansible BFF's, but easier than containerizing OpenStack :-o | 17:14 |
| *** neilus has quit IRC | 17:15 | |
| *** sdake_ has joined #openstack-ansible | 17:16 | |
| *** sdake has quit IRC | 17:16 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone: Use keystone-manage bootstrap command https://review.openstack.org/293816 | 17:17 |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible: Added configuration information for using Ceph with Cinder https://review.openstack.org/290873 | 17:18 |
| openstackgerrit | git-harry proposed openstack/openstack-ansible-os_keystone: WIP: Use os-service-installer https://review.openstack.org/292303 | 17:18 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_keystone: Test Keystone Mitaka RC1 in our role https://review.openstack.org/295953 | 17:21 |
| *** tiagogomes_ has joined #openstack-ansible | 17:21 | |
| *** tiagogomes has quit IRC | 17:24 | |
| *** doublek has joined #openstack-ansible | 17:33 | |
| *** jorge_munoz has joined #openstack-ansible | 17:40 | |
| *** jduhamel has joined #openstack-ansible | 17:42 | |
| *** krotscheck has quit IRC | 17:43 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: [WIP] Add LBaaS v2 support in Liberty https://review.openstack.org/295866 | 17:44 |
| *** pcaruana has joined #openstack-ansible | 17:44 | |
| spotz | NOOOO not mhayden and lbaas v2:) | 17:45 |
| *** retreved has quit IRC | 17:45 | |
| mhayden | haha, i need some coffee, obviously | 17:46 |
| mhayden | didn't realize my cherry picks pulled the tasks into a directory called 'tasks' a the root | 17:46 |
| mhayden | sigh | 17:46 |
| *** schwicht has quit IRC | 17:48 | |
| *** Mudpuppy has quit IRC | 17:50 | |
| spotz | mhayden get more coffee you'll feel better | 17:51 |
| *** Mudpuppy has joined #openstack-ansible | 17:51 | |
| *** jorge_munoz has quit IRC | 17:55 | |
| *** phiche has joined #openstack-ansible | 17:58 | |
| *** KLevenstein has quit IRC | 18:02 | |
| *** doublek has left #openstack-ansible | 18:09 | |
| mhayden | hah | 18:19 |
| *** sdake_ is now known as sdake | 18:25 | |
| *** retreved has joined #openstack-ansible | 18:27 | |
| *** retreved has quit IRC | 18:28 | |
| *** retreved has joined #openstack-ansible | 18:29 | |
| *** karimb has quit IRC | 18:33 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 18:33 | |
| *** fawadkhaliq has joined #openstack-ansible | 18:36 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_barbican: Implement base configuration for OpenStack-CI https://review.openstack.org/295996 | 18:38 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_barbican: Remove dependency on the Keystone admin auth token https://review.openstack.org/295999 | 18:40 |
| *** weezS has quit IRC | 18:48 | |
| *** KLevenstein has joined #openstack-ansible | 18:49 | |
| *** arif-ali has quit IRC | 18:50 | |
| *** schwicht has joined #openstack-ansible | 18:51 | |
| *** weezS has joined #openstack-ansible | 18:51 | |
| *** daledude has joined #openstack-ansible | 19:04 | |
| daledude | im going from 12.0.5 to 12.0.8...is building the updated repository step from the docs no longer needed? | 19:04 |
| cloudnull | daledude: you need to run the repo-build play to create the new deps | 19:05 |
| cloudnull | as well as the venvs | 19:05 |
| automagically | We really need a minor version upgrade doc | 19:06 |
| daledude | there is a minor ver upgrade oc | 19:06 |
| daledude | http://docs.openstack.org/developer/openstack-ansible/liberty/install-guide/app-minorupgrade.html | 19:06 |
| daledude | and it says to run openstack-ansible repo-install.yml | 19:06 |
| automagically | Ah, my bad | 19:06 |
| daledude | but that tries to connect to https://rpc-repo.rackspace.com | 19:07 |
| daledude | ill try repo-build.yml instead | 19:07 |
| cloudnull | repo-install will call repo-build | 19:07 |
| cloudnull | in kilo that would try and do a clone | 19:07 |
| cloudnull | but it was revised in liberty | 19:07 |
| daledude | k | 19:08 |
| *** electrofelix has quit IRC | 19:12 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_barbican: Implement base configuration for OpenStack-CI https://review.openstack.org/295996 | 19:13 |
| odyssey4me | ug, repo-install runs repo-build | 19:14 |
| odyssey4me | *uh | 19:14 |
| odyssey4me | cloudnull daledude ^ | 19:14 |
| cloudnull | yes? | 19:14 |
| odyssey4me | nvm, I see you mentioned that | 19:14 |
| odyssey4me | daledude the only situation which we know of where you may end up with an error is if you happen to have upgraded from a kilo environment | 19:15 |
| daledude | i never used kilo | 19:16 |
| odyssey4me | in that case you'll find that you have a locked down pip conf | 19:16 |
| daledude | i went from 12.0.3 to 12.0.5 and now "upgrading" to 12.0.8 | 19:16 |
| odyssey4me | hmm, and do you have anything related to the repo configured in your /etc/openstack_deploy/user.* files or /etc/openstack_deploy/openstac_user_config.yml file? | 19:16 |
| *** ametts has quit IRC | 19:17 | |
| *** arif-ali has joined #openstack-ansible | 19:17 | |
| daledude | i have a repo-infra_hosts section in openstack_user_config.yml | 19:19 |
| odyssey4me | daledude that's fine - just hosts in there, no host vars? | 19:20 |
| daledude | correct | 19:20 |
| mhayden | so i'm running into that "Failed to build pywbem" error now, but i have the patch applied to my git repo | 19:20 |
| mhayden | and i rebuilt the repo already | 19:21 |
| odyssey4me | mhayden are you building against the head of liberty? | 19:21 |
| mhayden | odyssey4me: yes sir | 19:22 |
| mhayden | should i not be? | 19:22 |
| odyssey4me | daledude ok, can you pastebin the resulting log from executing the repo-install playbook with '-vv' for verbose logging | 19:22 |
| odyssey4me | mhayden for your patches, yes - that includes the fixups for the pywbem thingamy | 19:22 |
| mhayden | odyssey4me: after applying the patch, should i do anything else besides the repo-build play? | 19:24 |
| odyssey4me | mhayden just repo-build, then whichever other plays you'd like | 19:24 |
| *** admin0 has joined #openstack-ansible | 19:24 | |
| mhayden | the requirements_local_filtered.txt still ends up with pywbem in it | 19:25 |
| openstackgerrit | Merged openstack/openstack-ansible-os_barbican: Implement base configuration for OpenStack-CI https://review.openstack.org/295996 | 19:25 |
| odyssey4me | mhayden hmm, that's odd | 19:25 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_barbican: Remove dependency on the Keystone admin auth token https://review.openstack.org/295999 | 19:25 |
| daledude | odyssey4me: by running the repo-build instead of the documented repo-install it doesnt try to connect to rpc server. but, getting error: Failed building wheel for pywbem | 19:27 |
| daledude | looking to do pstebin | 19:27 |
| odyssey4me | daledude yeah, that's a known issue - apply https://review.openstack.org/294934 | 19:27 |
| odyssey4me | daledude can you try running repo-server instead? | 19:28 |
| admin0 | do i need to create glance, cinder user beforehand in ceph ? or will it ssh and do it itself ? | 19:28 |
| daledude | trying | 19:28 |
| admin0 | and on the storage server example, even if i have 3 mons, i jsut need to point to one ? | 19:29 |
| odyssey4me | admin0 personally I have no idea, you may have to walk the code to find that out | 19:29 |
| odyssey4me | admin0 for ceph you always want to point to them all, in case one goes down | 19:29 |
| odyssey4me | the ceph client libraries will automatically connect to another is their primary goes down | 19:29 |
| *** fawadkhaliq has quit IRC | 19:30 | |
| odyssey4me | if you only have one mon, your clients will hang, and you will cry :p | 19:30 |
| *** fawadkhaliq has joined #openstack-ansible | 19:30 | |
| admin0 | odyssey4me: i mean on example on storage_hosts section, only 1 mon is used for example .. do i need to have 3 storage hosts with 3 mons, or 1 storage hosts with 3 mons | 19:31 |
| admin0 | that is where the confusion lies | 19:31 |
| admin0 | because if the storage is ceph, there are no storage hosts right .. just ceph | 19:32 |
| admin0 | storage_hosts: | 19:32 |
| odyssey4me | admin0 a link to what you're referring to? | 19:32 |
| admin0 | uno momento | 19:32 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_barbican: Remove dependency on the Keystone admin auth token https://review.openstack.org/295999 | 19:34 |
| odyssey4me | admin0 as per the examples in https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/user_variables.yml#L81-L99 you should add the mons, how ever many you have - typically ceph requires an odd number of mons for better fencing | 19:34 |
| mhayden | cloudnull: i have your patch applied ( https://review.openstack.org/#/c/294934/ ) but the repo server still fails when rebuilding | 19:35 |
| daledude | odyssey4me: repo-server runs clean | 19:35 |
| odyssey4me | admin0 then as per https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L562-L564 and https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L603-L620 you can setup your storage hosts | 19:35 |
| odyssey4me | mhayden pastebin or it never happened :p | 19:35 |
| mhayden | odyssey4me: haha, in progress | 19:36 |
| mhayden | odyssey4me / cloudnull: http://paste.openstack.org/show/491502/ | 19:36 |
| odyssey4me | daledude so repo-server ran clean after you executed repo-build? hmm, I think git-harry and I chatted about this the other day | 19:36 |
| admin0 | odyssey4me: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L562-L564 is the reason of my confusion .. what exactly is a storage host here :D if I already specify 3 mons above? | 19:36 |
| admin0 | are those the osds ? | 19:36 |
| odyssey4me | will have to try and figure that one out | 19:36 |
| mhayden | odyssey4me: seems like the filtering isn't working -- pywbem is sneaking into my requirements_local_filtered.txt on the repo server | 19:37 |
| odyssey4me | admin0 the mons are the mons, the storage hosts host the OSD's | 19:37 |
| odyssey4me | admin0 actually, I have no idea | 19:37 |
| odyssey4me | oh wait - I think I have it | 19:37 |
| odyssey4me | admin0 essentially the storage_hosts run cinder-volume | 19:37 |
| admin0 | no take a good U turn :) | 19:37 |
| admin0 | no => you | 19:37 |
| odyssey4me | admin0 so in this case, for ceph, you actually want cinder-volume to run in a container on the controllers - I think | 19:38 |
| odyssey4me | so you want to switch on_metal to false - unless you wish to waste a host | 19:38 |
| odyssey4me | for any networked cinder storage, you can run cinder-volume in a container | 19:38 |
| odyssey4me | then I think the storage_hosts which you're using will become your controllers, as they'll get containers put on them | 19:39 |
| odyssey4me | I may be wrong- but I know you can rebuild so if I screwed up it's ok :) | 19:39 |
| admin0 | :D | 19:39 |
| odyssey4me | mhayden it's coming from somewhere - can you figure out where from? | 19:40 |
| mhayden | odyssey4me: still looking | 19:40 |
| admin0 | so 1. have cinder volume run on containers by removing is_metal .. on storage_hosts, define all ceph osds ? and on user_variable, put all mons | 19:41 |
| odyssey4me | admin0 no, don't define ceph osd's at all - those are managed outside of OSA | 19:41 |
| odyssey4me | storage_hosts should include the hosts which should run the cinder-volume containers | 19:42 |
| mhayden | so pywbem doesn't appear anywhere in liberty trunk right now that i can find -- i wonder if something on the repo server is causing problems | 19:42 |
| mattt | odyssey4me: i think admin0 jerry-rigged the OSDs into osa like we do in rpc-openstack | 19:42 |
| odyssey4me | mhayden yeah, but it may be coming in from something you're adding? | 19:42 |
| spotz | Wasn't there a review just yesterday about pywbem? | 19:42 |
| admin0 | mattt: that one worked for deployment :D | 19:43 |
| admin0 | but i destroyed that env | 19:43 |
| admin0 | and now i have one blank env and 1 ceph cluster ready | 19:43 |
| admin0 | and looking to tie them up togehter | 19:43 |
| *** egonzalez has joined #openstack-ansible | 19:44 | |
| odyssey4me | admin0 yep, so in OSA we only instrument the client access to the cluster - we don't touch the cluster itserlf | 19:44 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 19:44 | |
| admin0 | yep. and that is what i am trying to work on next | 19:45 |
| *** johnmilton has quit IRC | 19:45 | |
| admin0 | so i have server c20,c21,c22 already running ceph cluster with the proper pools and auth added .. and i have a bunch of other servers where I want to run the osa now, and try to use glance and cinder with ceph | 19:46 |
| odyssey4me | admin0 we're looking forward to your next howto ;) | 19:47 |
| admin0 | :D | 19:47 |
| mattt | admin0: https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/ceph_client/defaults/main.yml | 19:48 |
| *** jmccrory_ has quit IRC | 19:48 | |
| mhayden | odyssey4me: it seems to be sneaking into this fact -> https://github.com/openstack/openstack-ansible/blob/liberty/playbooks/roles/repo_build/tasks/repo_set_facts.yml#L42-L47 | 19:50 |
| mattt | mhayden: stepping in late but make sure you don't have stale facts | 19:50 |
| mattt | those are a bugger | 19:50 |
| mhayden | mattt: OMG i hope that's not it | 19:50 |
| mhayden | :| | 19:50 |
| mattt | mhayden: the # of hours i've wasted dealing w/ that | 19:51 |
| * mattt points at the stars | 19:51 | |
| spotz | heheh | 19:51 |
| *** javeriak has quit IRC | 19:52 | |
| * mhayden clears some facts and crosses fingers | 19:52 | |
| odyssey4me | lol, that is a good one to remember | 19:53 |
| mhayden | i wonder if we ought to put a reminder about that at the top of the minor upgrade docs? | 19:53 |
| mattt | admin0: our docs are a bit scattered when it comes to ceph integration :( | 19:53 |
| mattt | admin0: let me know if i can help, but if you're going through the process it'd be great if you could update the openstack-ansible docs w/ what you find | 19:54 |
| admin0 | i went through rpcov11 and osa last time to tie in ceph from rpco to osa | 19:54 |
| admin0 | it worked :D | 19:54 |
| mattt | admin0: why did you redo it all then? | 19:54 |
| mattt | (serious question :) | 19:54 |
| admin0 | free time, 3-4 test clusters that i can re-do anytime, a big hunger to learn more :D | 19:55 |
| admin0 | well, that test was for ceph .. yesterday i tried to load vpnaas, broke neutron .. | 19:55 |
| admin0 | that integration is saved and i will document it too | 19:56 |
| mattt | admin0: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example#L603-L620 | 19:56 |
| mattt | so that is for your cinder-volume bit | 19:56 |
| admin0 | its mostly for documentaiton .. how to configure ceph with osa is done . now is how to configure existing ceph with osa | 19:56 |
| mattt | then iirc to get glance using ceph you need to set glance_default_store: rbd and for nova ephemeral disks in rbd you set nova_libvirt_images_rbd_pool: vms | 19:57 |
| *** scarlisle has quit IRC | 19:57 | |
| odyssey4me | oh, ansible released v1.9.5 - who knew | 19:58 |
| admin0 | odyssey4me: i am mostly confused in this part: http://docs.rackspace.com/rpc/api/v11/bk-rpc-installation/content/ch-ceph.html | 19:58 |
| *** jlvillal has quit IRC | 19:58 | |
| admin0 | why just 1 infra in storage_hosts section .. should it be controllers .. or osds | 19:59 |
| admin0 | sorry that was for mattt :D | 19:59 |
| mattt | controllers, OSDs does not know about or care about your osds | 19:59 |
| mattt | sorry, OSA does not know about or care about :) | 20:00 |
| odyssey4me | admin0 ignore that doc - as OSA doesn't setup the ceph osd's, OSA doesn't care about them | 20:00 |
| odyssey4me | what mattt said | 20:00 |
| admin0 | so that is about setting it up, and not consuming it | 20:01 |
| * odyssey4me goes back to ironic matters | 20:01 | |
| admin0 | :D | 20:01 |
| *** jlvillal has joined #openstack-ansible | 20:03 | |
| *** phiche has quit IRC | 20:09 | |
| *** scarlisle has joined #openstack-ansible | 20:11 | |
| openstackgerrit | Merged openstack/openstack-ansible-specs: Spec to introduce & integrate Gnocchi role https://review.openstack.org/295574 | 20:16 |
| *** egonzalez has quit IRC | 20:16 | |
| spotz | going through the review list. alot of -1 on verification | 20:23 |
| *** fawadkhaliq has quit IRC | 20:30 | |
| *** fawadkhaliq has joined #openstack-ansible | 20:30 | |
| *** jorge_munoz has joined #openstack-ansible | 20:34 | |
| *** fawadkhaliq has quit IRC | 20:40 | |
| *** fawadkhaliq has joined #openstack-ansible | 20:40 | |
| *** fawadkhaliq has quit IRC | 20:50 | |
| *** fawadkhaliq has joined #openstack-ansible | 20:51 | |
| *** jaypipes has quit IRC | 20:52 | |
| *** asettle has joined #openstack-ansible | 20:53 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_barbican: Remove dependency on the Keystone admin auth token https://review.openstack.org/295999 | 20:54 |
| *** weezS has quit IRC | 20:55 | |
| *** schwicht has quit IRC | 20:58 | |
| *** thorst has quit IRC | 21:00 | |
| *** thorst has joined #openstack-ansible | 21:00 | |
| *** schwicht has joined #openstack-ansible | 21:01 | |
| *** thorst has quit IRC | 21:05 | |
| *** woodard_ has joined #openstack-ansible | 21:06 | |
| *** weezS has joined #openstack-ansible | 21:06 | |
| *** woodard has quit IRC | 21:09 | |
| *** woodard_ has quit IRC | 21:10 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Update local_install script for non-tox testing https://review.openstack.org/296060 | 21:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Remove unnecessary apt package installation https://review.openstack.org/296063 | 21:13 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Remove dependency on the Keystone admin auth token https://review.openstack.org/296066 | 21:17 |
| *** Mudpuppy has quit IRC | 21:18 | |
| *** jorge_munoz has quit IRC | 21:19 | |
| *** thorst has joined #openstack-ansible | 21:22 | |
| pabelanger | Ya, something is wrong with ansible, it's been hung for about 45mins now | 21:23 |
| *** schwicht has quit IRC | 21:23 | |
| pabelanger | I wonder if I stopped the node when ansible was running at the same time | 21:23 |
| pabelanger | Doh, wrong channel | 21:24 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Remove the database creation & messaging tasks from the role https://review.openstack.org/296069 | 21:24 |
| pabelanger | sorry bout that | 21:24 |
| odyssey4me | lol pabelanger | 21:24 |
| odyssey4me | too many windows, not enough coffee :p | 21:24 |
| pabelanger | odyssey4me: indeed, up to 15 now | 21:25 |
| *** thorst has quit IRC | 21:26 | |
| *** keedya has quit IRC | 21:28 | |
| mhayden | odyssey4me: it's past your bedtime | 21:33 |
| *** phiche has joined #openstack-ansible | 21:33 | |
| *** weezS has quit IRC | 21:35 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ironic: Implement initial test inventory/plays https://review.openstack.org/296077 | 21:38 |
| *** pjm6 has joined #openstack-ansible | 21:39 | |
| pjm6 | Hi there | 21:40 |
| odyssey4me | mrda mcarden neillc That set of patches to the Ironic role bring it to the point where it converges for me. I may have missed something, but it should give you a starting point to work from and I'll revisit and test on a fresh instance tomorrow. Note that it seems to break keystone - not sure why yet. | 21:40 |
| pjm6 | I was here in afternoon because I had an error in deploying openstack ansible AIO | 21:40 |
| odyssey4me | afk for a bit | 21:41 |
| *** jguy_ has quit IRC | 21:41 | |
| cloudnull | ohai pjm6 | 21:43 |
| pjm6 | hi cloudnull | 21:47 |
| pjm6 | I collect some logs, but didn't understand well | 21:47 |
| pjm6 | the stdout that I collect was the following | 21:47 |
| pjm6 | http://pastebin.com/bW6hEVDh | 21:47 |
| * cloudnull looking | 21:47 | |
| *** openstackgerrit has quit IRC | 21:48 | |
| cloudnull | so is your deployment not building contianers? | 21:49 |
| cloudnull | what OS are you deploying into ? and what are the specs? | 21:49 |
| cloudnull | do you have the stdout for the error that was thrown ? | 21:49 |
| *** openstackgerrit has joined #openstack-ansible | 21:49 | |
| mcarden | odyssey4me: 296077? OK - I'll check it out. | 21:50 |
| pjm6 | I think it build containers because i can see them | 21:50 |
| pjm6 | I'm going to post the | 21:50 |
| pjm6 | second ansible_log | 21:50 |
| mrda | thanks odyssey4me - I'll take a look this morning | 21:54 |
| pjm6 | http://pastebin.com/dtBLfqmZ | 21:54 |
| pjm6 | that was the second time that I run | 21:55 |
| pjm6 | the first time it didn't give error in the file (when using export ANSIBLE_LOG_PATH) | 21:56 |
| sigmavirus24 | cloudnull: odyssey4me I tagged you into a mailing list thread that you might have missed otherwise | 21:57 |
| cloudnull | which one ? | 21:58 |
| odyssey4me | mcarden neillc mrda these: https://review.openstack.org/#/q/project:+openstack/openstack-ansible-ironic+owner:jesse.pretorius%2540rackspace.co.uk+status:open+label:Verified%252B1 | 22:01 |
| sigmavirus24 | cloudnull: I have to run but it was about pbr semver tags | 22:02 |
| mrda | odyssey4me: thanks | 22:02 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:02 | |
| coolj | pjm6: looks like two of your rabbit containers and a heat_apis container canot be reached via ssh | 22:02 |
| cloudnull | ill go look. | 22:02 |
| cloudnull | thanks sigmavirus24_awa | 22:02 |
| pjm6 | coolj: it seems like that, but what I do was just running the commands that were in the documentation. | 22:04 |
| pjm6 | If I use teardown.sh and run over again, it will be a clean install? | 22:04 |
| coolj | pjm6: not sure tbh, i've never used it! i would investigate why those containers are unreachable and go from there. | 22:07 |
| pjm6 | hmm, ok thank you :) | 22:07 |
| *** KLevenstein has quit IRC | 22:08 | |
| pjm6 | but the AIO with those three commands was supposed to be working on demand, right ? | 22:08 |
| *** retreved has quit IRC | 22:10 | |
| coolj | pjm6: i have not used the aio personally (i am just a deployer, i don't directly work on osa), sorry i can't provide more info! | 22:11 |
| pjm6 | ah ok, no problem. thank you for help :) | 22:12 |
| pjm6 | but you use OSA to deploy, right? | 22:12 |
| odyssey4me | pjm, it works for us for the most part in most situations - we test and have commit tests using it - so it's pretty robust | 22:13 |
| odyssey4me | (the AIO I mean) | 22:13 |
| odyssey4me | pjm6 ^ | 22:13 |
| *** darrenc is now known as darrenc_afk | 22:14 | |
| pjm6 | yes i believe, I will try using teardown.sh and use the "Rebuild an AIO" guide | 22:14 |
| spotz | Hrm the Van Zandt hotel allows any pet you can fit through the door..... mhayden should I being the pony? | 22:14 |
| pjm6 | and try to deploy again | 22:14 |
| neillc | odyssey4me: I have some working tox based tests for ironic. will now try and reconcile your new changes with what I have | 22:15 |
| odyssey4me | pjm6 if you teardown too often, it may be best to restart fresh - a minimum ubuntu install on a VM with a single NIC and internet access | 22:15 |
| pjm6 | hm i think iu | 22:17 |
| pjm6 | will do that | 22:17 |
| pjm6 | i was getting xenstore-read error in a VM (in XenServer hypervisor) | 22:17 |
| pjm6 | but i was using the tag 12.0.0 or 12.0.9 | 22:17 |
| spotz | pjm6 There was a bug for a day or so I think where I had that error too, but the fix went out quickly. Can't remember which version though | 22:19 |
| *** galstrom is now known as galstrom_zzz | 22:19 | |
| pjm6 | the error of xenstore-read ? | 22:19 |
| pjm6 | I'm trying since yesterday spotz | 22:19 |
| spotz | pjm6 Yeah I think so, it was a couple months ago | 22:20 |
| pjm6 | but I think the branch liberty it's currently fine | 22:20 |
| pjm6 | hmm, I will try with branch liberty instead of tags, thanks =) | 22:20 |
| spotz | pjm6 Yeah Liberty was fine, we were doing some testing on master | 22:20 |
| odyssey4me | pjm6 yeah, until the next liberty tag I'd suggest using the head of liberty for now | 22:21 |
| pjm6 | thanks odyssey4me and spotz | 22:21 |
| pjm6 | odyssey4me i'm using it since you told me in this morning, but I didn't try in the VM | 22:21 |
| odyssey4me | pjm6 if you've torn down and will be rebuilding an AIO and have customised nothing - it may be good to remove /etc/openstack_deploy/ after the teardown | 22:22 |
| pjm6 | btw: it is easy to make a distribute deployment with the following structure: 2 master nodes (with RabbitMQ, Horizon, and common openstack services), 2 network nodes, 2 controller nodes and 2 compute nodes ? | 22:23 |
| pjm6 | odyssey4me thanks, i will do that in my physical machine and try in the Virtual Machine to see if the error goes out :) | 22:23 |
| odyssey4me | pjm6 VM is better to make it easier to destroy and recreate, but yeah sometimes it may not work | 22:24 |
| odyssey4me | it all depends on the platform and the cpu support it has | 22:24 |
| pjm6 | i'm using XenServer with 2 CPUs Xeon (if I remember it has 40vCores) | 22:25 |
| *** coolj has quit IRC | 22:27 | |
| *** coolj has joined #openstack-ansible | 22:28 | |
| *** scarlisle has quit IRC | 22:34 | |
| *** darrenc_afk is now known as darrenc | 22:35 | |
| openstackgerrit | Merged openstack/openstack-ansible: Added memcache flush to upgrade process https://review.openstack.org/294115 | 22:41 |
| *** spotz is now known as spotz_zzz | 22:42 | |
| odyssey4me | neillc Hopefully I'll be able to stand up a working standalone build tomorrow. For now at least this converges. I just need to do a fresh build to confirm that before patching tox to see if the gate agrees. We're busy wrestling with some stuff to hopefully make tox behave more consistently between infra and a build on a cloud instance see mattt's work and some notes in https://review.openstack.org/295265 | 22:42 |
| odyssey4me | the python env is set properly to the tox python executable in the gate, but not when we do it in a cloud instance - that's where it stands now | 22:43 |
| odyssey4me | the problem trying to get solved is that we test on a cloud instance and it's fine, but it doesn't work in the gate | 22:44 |
| odyssey4me | anyway, I'm out for the night - have a great day y'all | 22:44 |
| *** schwicht has joined #openstack-ansible | 22:45 | |
| *** phiche1 has joined #openstack-ansible | 22:48 | |
| pjm6 | thanks, for you to odyssey4me | 22:48 |
| *** phiche has quit IRC | 22:50 | |
| *** networkninja has joined #openstack-ansible | 22:55 | |
| admin0 | odyssey4me: https://www.openstackfaq.com/openstack-ansible-ceph/ — ceph added for deployment :D | 22:58 |
| admin0 | but this is just install .. integration — will check tomororw | 22:58 |
| admin0 | *tomrorow | 22:58 |
| admin0 | *tomorrow | 22:58 |
| admin0 | :D | 22:58 |
| *** admin0 has quit IRC | 23:04 | |
| *** saneax_AFK is now known as saneax | 23:05 | |
| *** schwicht has quit IRC | 23:06 | |
| *** izaakk has quit IRC | 23:17 | |
| *** phiche1 has quit IRC | 23:21 | |
| *** schwicht has joined #openstack-ansible | 23:25 | |
| *** schwicht has quit IRC | 23:30 | |
| openstackgerrit | Charles Farquhar proposed openstack/openstack-ansible: Add pip_get_pip_options to fall back Install PIP https://review.openstack.org/296112 | 23:32 |
| *** jguy_ has joined #openstack-ansible | 23:34 | |
| openstackgerrit | Charles Farquhar proposed openstack/openstack-ansible-pip_install: Add pip_get_pip_options to fall back Install PIP https://review.openstack.org/296114 | 23:36 |
| *** daledude has left #openstack-ansible | 23:39 | |
| *** markvoelker has quit IRC | 23:44 | |
| *** schwicht has joined #openstack-ansible | 23:46 | |
| *** markvoelker has joined #openstack-ansible | 23:47 | |
| *** fawadkhaliq has quit IRC | 23:48 | |
| *** fawadkhaliq has joined #openstack-ansible | 23:49 | |
| *** johnmilton has joined #openstack-ansible | 23:51 | |
| *** schwicht has quit IRC | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!