| *** thorst has quit IRC | 00:01 | |
| *** gouthamr has quit IRC | 00:03 | |
| Drago | I have a question. How do I get a different version of a service installed? I'm installing Magnum from master on a cloud where everything else is stable/newton | 00:03 |
|---|---|---|
| dxiri | right now I am hitting a weird issue where the ssh lxc_container_create : Wait for ssh to be available is hanging | 00:04 |
| Drago | I tried editing https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml and https://github.com/openstack/openstack-ansible-os_magnum/blob/master/defaults/main.yml, but running repo_build.yml hasn't changed the venv | 00:05 |
| *** Mudpuppy has joined #openstack-ansible | 00:07 | |
| *** Mudpuppy has quit IRC | 00:11 | |
| dxiri | ahh I see why my ssh is hanging | 00:12 |
| dxiri | I have no dhcp on the mgmt network | 00:12 |
| dxiri | do I really need to setup a dhcp server? | 00:13 |
| dxiri | or is there a way for openstack-ansible to take care of that for me? | 00:13 |
| *** fops has joined #openstack-ansible | 00:14 | |
| *** Mudpuppy has joined #openstack-ansible | 00:15 | |
| *** fops_ has quit IRC | 00:17 | |
| *** LiYuenan has joined #openstack-ansible | 00:21 | |
| *** hj-hpe has quit IRC | 00:24 | |
| *** Mudpuppy has quit IRC | 00:25 | |
| stevelle | Drago: rather than editing files in roles or group_vars, you should generally override values in your user_variables.yml -- but setting overrides for these should do what you want https://github.com/openstack/openstack-ansible-os_magnum/blob/master/defaults/main.yml#L57-L58 | 00:31 |
| *** dxiri has quit IRC | 00:35 | |
| Drago | stevelle: General question: are all OSA variables flat and do user_variables always take precedent? As in, if I see "magnum_git_repo", can I be confident that I can put magnum_git_repo in my user_variables and it will override any other instance of magnum_git_repo? | 00:36 |
| stevelle | Drago: yes. | 00:37 |
| stevelle | there is some nuance to variables but user_variables and user_secrets should take precedence. | 00:37 |
| Drago | stevelle: I updated defaults/main's magnum_git_repo to be master a few tries ago, and after another run, I'm still pulling down the old version of magnum | 00:39 |
| Drago | Here's what I'm doing | 00:39 |
| Drago | run openstack-ansible repo_build.yml | 00:39 |
| Drago | curl -O http://162.209.99.181:8181/venvs/14.0.1/ubuntu/magnum-14.0.1-x86_64.tgz | 00:40 |
| Drago | Unzip it and check to see if the file has changed | 00:40 |
| Drago | Am I looking at the wrong tarball, or missing some other configuration? | 00:40 |
| Drago | I attached to the repo container and it actually cloned the correct commit | 00:40 |
| Drago | (I changed defaults/repo_packages/openstack_services.yml to have magnum at 57e74664d30dbf3cadaa247ca678e0d6d636536c, which is the latest commit on master) | 00:41 |
| stevelle | I'm guessing that it shouldn't be a magnum-14.0.1-x86_64 tarball | 00:42 |
| stevelle | https://github.com/openstack/openstack-ansible-os_magnum/blob/14.0.0/defaults/main.yml#L52 | 00:42 |
| stevelle | I have to dig a bit more to be sure but it seems like 14.0.1 isn't what you're trying to achieve anyway | 00:44 |
| Drago | stevelle: In os-magnum-install, theres `magnum_venv_tag: "{{ openstack_release }}"` | 00:45 |
| stevelle | and that would used in the os_magnum role to fetch the stable/newton venv tarball | 00:46 |
| stevelle | you might be building the right tarball in repo_build but fetching the wrong one because you haven't overridden the magnum_venv_tag | 00:47 |
| Drago | stevelle: /opt/openstack-ansible/playbooks/inventory/group_vars/all.yml:17:openstack_release: 14.0.1 | 00:48 |
| Drago | stevelle: So because it defaults to openstack_release, I need to override it so that it's not | 00:48 |
| *** c-mart has quit IRC | 00:48 | |
| stevelle | let me cross check a couple things to be sure before I confirm :) | 00:49 |
| Drago | stevelle: Okay, thanks a lot for your help by the way | 00:49 |
| stevelle | yes, Drago. It does look like just overriding magnum_venv_tag with the right value should be the fix. | 00:50 |
| stevelle | happy to help when I can | 00:50 |
| Drago | stevelle: What would the right value be? The same thing I used for magnum_git_install_branch? | 00:51 |
| stevelle | I'd probably check in your repo container to be sure you pick the right value | 00:51 |
| stevelle | poking at repo build to see if I can confirm what that value would be, but would be faster to check in the webroot | 00:51 |
| Drago | Running repo-build again with magnum_requirements_git_install_branch, magnum_venv_tag, and magnum_requirements_git_install_branch set to master | 00:54 |
| Drago | stevelle: Is there any way to cut down the amount of work repo-build is doing? It takes 15 minutes every time | 00:54 |
| stevelle | Drago: the role is already fairly minimal for your environment. We haven't added features to allow limiting of scope to a specific project | 00:56 |
| stevelle | but the fewer services you deploy (per conf.d/ and openstack_user_config.yml) the fewer get built in the repo | 00:57 |
| *** thorst has joined #openstack-ansible | 00:57 | |
| Drago | Makes sense | 00:57 |
| stevelle | it looks like repo_build_release_tag is being used in the naming of the tarball in the repo_build role which may be making things behave oddly for you in this case | 00:59 |
| *** javeriak has quit IRC | 01:00 | |
| Drago | stevelle: It is? | 01:00 |
| stevelle | https://github.com/openstack/openstack-ansible-repo_build/blob/master/templates/op-venv-script.sh.j2#L74 | 01:01 |
| Drago | stevelle: So if it has a different name at least, it shouldn't be skipped | 01:02 |
| Drago | but it doesn't look like it will be different if only magnum_venv_tag is different | 01:03 |
| stevelle | https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L110 | 01:03 |
| stevelle | that is suggesting that I led you astray, and that it would build a magnum-14.0.1-x86_64 tarball regardless of what magnum_venv_tag says | 01:04 |
| stevelle | there might be something else overriding that but I'd have to make a test env to tinker in | 01:04 |
| Drago | stevelle: I've heard of an —isolated option when installing a service. Can I use that to install a totally custom version? | 01:04 |
| *** thorst has quit IRC | 01:04 | |
| *** cathrichardson has quit IRC | 01:05 | |
| Drago | From what you're finding it seems like OSA expects everything to be at the same release, which probably makes sense for most production environments | 01:05 |
| *** cathrichardson has joined #openstack-ansible | 01:05 | |
| stevelle | truth | 01:05 |
| stevelle | we didn't make it easy to meet your expectation. What I might recommend for you is to set https://github.com/openstack/openstack-ansible-os_magnum/blob/master/defaults/main.yml#L61 true | 01:06 |
| stevelle | When that is set true we won't use a venv from the repo_build stage | 01:06 |
| Drago | stevelle: Fine by me! | 01:06 |
| Drago | heh | 01:06 |
| stevelle | you might have the best luck by destroying the magnum containers first and recreating them | 01:07 |
| Drago | stevelle: I'll try that next if this doesn't fully work | 01:07 |
| stevelle | setting developer mode and pointing your repo to the branch you want should be the easiest solution and is what we're using in glance to test an upgrade to Ocata | 01:08 |
| stevelle | I'm guessing you're not doing production work if you're installing magnum from master | 01:09 |
| Drago | stevelle: Nope, doing a workshop here in Barcelona, and have been having more difficulty than I expected | 01:10 |
| stevelle | egads. it's late there | 01:10 |
| Drago | stevelle: Yes | 01:10 |
| stevelle | condolences | 01:11 |
| Drago | stevelle: Unless you count keynotes, nothing's really happening until after noon, so it's okay | 01:11 |
| *** alij has joined #openstack-ansible | 01:13 | |
| *** v1k0d3n has joined #openstack-ansible | 01:17 | |
| *** alij has quit IRC | 01:17 | |
| *** javeriak has joined #openstack-ansible | 01:18 | |
| *** javeriak has quit IRC | 01:19 | |
| *** javeriak has joined #openstack-ansible | 01:19 | |
| *** javeriak has quit IRC | 01:20 | |
| Drago | stevelle: It's saying it can't find PyMySQL===0.7.9 or something like that | 01:20 |
| Drago | stevelle: I think it's because pip is trying to look in the venv for it. I added "pip_install_options: —isolated" into the group_vars for magnum | 01:20 |
| stevelle | if you set development_mode you don't have to much with pip_install_options | 01:21 |
| stevelle | muck* | 01:21 |
| stevelle | though there would be a pip config that locks down sources to point at the repo, if you didn't destroy the containers | 01:21 |
| Drago | As I understand it, that'll cause pip to ignore the venv… or something | 01:22 |
| *** v1k0d3n has quit IRC | 01:22 | |
| Drago | I ran it again and it succeeded | 01:22 |
| Drago | hopefully it installed the right things | 01:22 |
| Drago | Sweet, got the new file in the magnum container | 01:23 |
| Drago | Magnum commands at least work | 01:24 |
| Drago | hopefully even if the service didn't get restarted or something silly, it'll work because it just has to pick up a new Heat template file | 01:25 |
| *** hughmFLEXin has joined #openstack-ansible | 01:30 | |
| *** alij has joined #openstack-ansible | 01:32 | |
| *** javeriak has joined #openstack-ansible | 01:34 | |
| *** hughmFLEXin has quit IRC | 01:35 | |
| *** javeriak has quit IRC | 01:38 | |
| *** drifterza has quit IRC | 02:00 | |
| *** thorst has joined #openstack-ansible | 02:02 | |
| *** alij has quit IRC | 02:05 | |
| *** alij has joined #openstack-ansible | 02:08 | |
| *** thorst has quit IRC | 02:10 | |
| *** cfarquhar has joined #openstack-ansible | 02:15 | |
| *** Trident has quit IRC | 02:20 | |
| *** john51 has joined #openstack-ansible | 02:28 | |
| *** john51 has quit IRC | 02:30 | |
| *** john51 has joined #openstack-ansible | 02:33 | |
| *** john51 has quit IRC | 02:35 | |
| *** allanice001 has joined #openstack-ansible | 02:35 | |
| *** john51 has joined #openstack-ansible | 02:38 | |
| *** john51 has quit IRC | 02:40 | |
| *** john51 has joined #openstack-ansible | 02:43 | |
| *** agrebennikov has joined #openstack-ansible | 02:59 | |
| *** alij has quit IRC | 02:59 | |
| *** cathrichardson has quit IRC | 03:03 | |
| *** agrebennikov has quit IRC | 03:04 | |
| *** thorst has joined #openstack-ansible | 03:07 | |
| *** thorst has quit IRC | 03:15 | |
| *** allanice001 has quit IRC | 03:20 | |
| *** adrian_otto has quit IRC | 03:20 | |
| *** javeriak has joined #openstack-ansible | 03:23 | |
| *** allanice001 has joined #openstack-ansible | 03:30 | |
| *** alij has joined #openstack-ansible | 03:30 | |
| *** alij has quit IRC | 03:35 | |
| *** fops has quit IRC | 03:52 | |
| *** javeriak has quit IRC | 03:56 | |
| *** thorst has joined #openstack-ansible | 03:57 | |
| *** poopcat has quit IRC | 04:04 | |
| *** thorst has quit IRC | 04:05 | |
| *** alij has joined #openstack-ansible | 04:10 | |
| *** kysse_ has quit IRC | 04:23 | |
| *** kysse has joined #openstack-ansible | 04:25 | |
| *** thorst has joined #openstack-ansible | 04:28 | |
| *** allanice001 has quit IRC | 04:36 | |
| *** thorst has quit IRC | 04:37 | |
| *** LiYuenan has quit IRC | 04:41 | |
| *** javeriak has joined #openstack-ansible | 04:52 | |
| *** allanice001 has joined #openstack-ansible | 04:54 | |
| *** weezS has joined #openstack-ansible | 05:05 | |
| *** Drago has quit IRC | 05:14 | |
| *** thorst has joined #openstack-ansible | 05:22 | |
| *** thorst has quit IRC | 05:35 | |
| *** thorst has joined #openstack-ansible | 05:36 | |
| *** gouthamr has joined #openstack-ansible | 05:36 | |
| *** alij has quit IRC | 05:43 | |
| *** thorst_ has joined #openstack-ansible | 05:44 | |
| *** thorst has quit IRC | 05:44 | |
| *** thorst_ has quit IRC | 05:48 | |
| *** Jeffrey4l has quit IRC | 05:50 | |
| *** h5t4 has joined #openstack-ansible | 05:51 | |
| *** Jack_Iv has joined #openstack-ansible | 05:58 | |
| *** alij has joined #openstack-ansible | 05:59 | |
| *** sc68cal has quit IRC | 06:10 | |
| *** sc68cal has joined #openstack-ansible | 06:13 | |
| *** allanice001 has quit IRC | 06:13 | |
| *** woodard has joined #openstack-ansible | 06:16 | |
| *** drifterza has joined #openstack-ansible | 06:19 | |
| *** askb has quit IRC | 06:26 | |
| *** woodard has quit IRC | 06:39 | |
| *** charz_ has joined #openstack-ansible | 06:40 | |
| *** rickflare2 has joined #openstack-ansible | 06:42 | |
| *** charz has quit IRC | 06:42 | |
| *** galstrom_zzz has quit IRC | 06:42 | |
| *** rickflare has quit IRC | 06:42 | |
| *** galstrom_zzz has joined #openstack-ansible | 06:42 | |
| *** rickflare2 is now known as rickflare | 06:43 | |
| *** m4rx has joined #openstack-ansible | 06:46 | |
| *** gouthamr has quit IRC | 06:48 | |
| openstackgerrit | chandanc proposed openstack/openstack-ansible-os_neutron: Add support for Neutron FWaaS v2 https://review.openstack.org/388548 | 06:48 |
| *** johnmilton has quit IRC | 06:48 | |
| *** gtrxcb has quit IRC | 06:49 | |
| *** jwitko has quit IRC | 06:50 | |
| *** jwitko has joined #openstack-ansible | 06:50 | |
| *** johnmilton has joined #openstack-ansible | 06:51 | |
| *** weezS has quit IRC | 06:54 | |
| *** david-lyle has joined #openstack-ansible | 07:03 | |
| *** fxpester has joined #openstack-ansible | 07:03 | |
| *** thorst has joined #openstack-ansible | 07:09 | |
| *** vishwanathj has joined #openstack-ansible | 07:15 | |
| *** gouthamr has joined #openstack-ansible | 07:18 | |
| *** prometheanfire has quit IRC | 07:20 | |
| *** alij has quit IRC | 07:21 | |
| *** alij has joined #openstack-ansible | 07:24 | |
| *** asettle has joined #openstack-ansible | 07:24 | |
| *** prometheanfire has joined #openstack-ansible | 07:27 | |
| *** alij has quit IRC | 07:28 | |
| *** allanice001 has joined #openstack-ansible | 07:28 | |
| *** javeriak has quit IRC | 07:31 | |
| *** berendt has joined #openstack-ansible | 07:33 | |
| *** berendt has quit IRC | 07:41 | |
| *** smatzek has joined #openstack-ansible | 07:43 | |
| *** alij has joined #openstack-ansible | 07:45 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-galera_server: Fix MYSQLD_STARTUP_TIMEOUT for systemd https://review.openstack.org/389066 | 07:50 |
| *** haad1 has joined #openstack-ansible | 07:50 | |
| *** m4rx_ has joined #openstack-ansible | 07:53 | |
| *** m4rx has quit IRC | 07:56 | |
| *** asettle has quit IRC | 08:01 | |
| *** asettle has joined #openstack-ansible | 08:01 | |
| *** smatzek has quit IRC | 08:01 | |
| *** andymccr_ is now known as andymccr | 08:02 | |
| *** alij has quit IRC | 08:04 | |
| *** asettle has quit IRC | 08:05 | |
| *** thorst has quit IRC | 08:09 | |
| *** thorst has joined #openstack-ansible | 08:17 | |
| *** asettle has joined #openstack-ansible | 08:19 | |
| *** thorst has quit IRC | 08:21 | |
| *** m4rx_ has quit IRC | 08:21 | |
| *** m4rx has joined #openstack-ansible | 08:23 | |
| *** m4rx is now known as m4rx_ | 08:23 | |
| *** alij has joined #openstack-ansible | 08:26 | |
| *** asettle has quit IRC | 08:30 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_magnum: [WIP] Move to use on metal for gate timing improvements https://review.openstack.org/389708 | 08:34 |
| *** m4rx has joined #openstack-ansible | 08:34 | |
| *** m4rx_ has quit IRC | 08:35 | |
| *** smatzek has joined #openstack-ansible | 08:36 | |
| *** thorst has joined #openstack-ansible | 08:38 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_magnum: [WIP] Move to use on metal for gate timing improvements https://review.openstack.org/389708 | 08:42 |
| *** thorst has quit IRC | 08:42 | |
| *** thorst has joined #openstack-ansible | 08:44 | |
| openstackgerrit | Adrien Cunin proposed openstack/openstack-ansible: Protect CLI parameters passed through osa wrapper https://review.openstack.org/390480 | 08:47 |
| *** thorst has quit IRC | 08:48 | |
| *** smatzek has quit IRC | 08:48 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_magnum: [WIP] Move to use on metal for gate timing improvements https://review.openstack.org/389708 | 08:51 |
| *** gouthamr has quit IRC | 08:52 | |
| *** smatzek has joined #openstack-ansible | 08:53 | |
| *** thorst has joined #openstack-ansible | 08:54 | |
| *** david-lyle has quit IRC | 08:55 | |
| *** alij has quit IRC | 08:56 | |
| *** thorst has quit IRC | 08:58 | |
| *** fops has joined #openstack-ansible | 09:02 | |
| *** smatzek has quit IRC | 09:04 | |
| *** smatzek has joined #openstack-ansible | 09:07 | |
| *** alij has joined #openstack-ansible | 09:09 | |
| *** Drago1 has joined #openstack-ansible | 09:12 | |
| *** Drago1 has quit IRC | 09:13 | |
| *** Drago1 has joined #openstack-ansible | 09:13 | |
| *** ybabenko has joined #openstack-ansible | 09:17 | |
| *** thorst has joined #openstack-ansible | 09:17 | |
| *** adrian_otto has joined #openstack-ansible | 09:17 | |
| *** smatzek has quit IRC | 09:18 | |
| *** thorst_ has joined #openstack-ansible | 09:21 | |
| *** thorst has quit IRC | 09:22 | |
| *** adrian_otto has quit IRC | 09:24 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Identify virt type of PowerVM and KVM on Power https://review.openstack.org/389981 | 09:24 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Fix api_paste_config option https://review.openstack.org/386290 | 09:24 |
| *** thorst has joined #openstack-ansible | 09:25 | |
| *** berendt has joined #openstack-ansible | 09:25 | |
| openstackgerrit | Merged openstack/openstack-ansible-haproxy_server: [DOCS] Add note about extra VIP addresses https://review.openstack.org/389506 | 09:26 |
| *** thorst_ has quit IRC | 09:26 | |
| *** thorst has quit IRC | 09:29 | |
| *** thorst has joined #openstack-ansible | 09:29 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-haproxy_server: [DOCS] Add note about extra VIP addresses https://review.openstack.org/390496 | 09:29 |
| *** smatzek has joined #openstack-ansible | 09:31 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova: Fix api_paste_config option https://review.openstack.org/390497 | 09:31 |
| *** woodard has joined #openstack-ansible | 09:31 | |
| *** m4rx has quit IRC | 09:31 | |
| *** m4rx has joined #openstack-ansible | 09:33 | |
| *** Michaelg_barcelo has joined #openstack-ansible | 09:33 | |
| *** electrofelix has joined #openstack-ansible | 09:42 | |
| *** Jeffrey4l has joined #openstack-ansible | 09:43 | |
| *** smatzek has quit IRC | 09:45 | |
| *** pester has joined #openstack-ansible | 09:46 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Fix Ansible bootstrap upper constraint use https://review.openstack.org/390501 | 09:46 |
| *** smatzek has joined #openstack-ansible | 09:48 | |
| *** fxpester has quit IRC | 09:49 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Implement upper constraints for Ansible bootstrap https://review.openstack.org/389775 | 09:50 |
| openstackgerrit | Merged openstack/openstack-ansible: Using assertIsNone() instead of assertEqual(None) https://review.openstack.org/389992 | 09:55 |
| *** Drago1 has quit IRC | 09:58 | |
| openstackgerrit | Merged openstack/openstack-ansible-rsyslog_client: Fix errors due to repo_service_user_name setting https://review.openstack.org/390066 | 10:00 |
| *** Jeffrey4l has quit IRC | 10:03 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-rsyslog_client: Fix errors due to repo_service_user_name setting https://review.openstack.org/390506 | 10:04 |
| *** berendt has quit IRC | 10:05 | |
| *** adrian_otto has joined #openstack-ansible | 10:07 | |
| *** david-lyle has joined #openstack-ansible | 10:08 | |
| openstackgerrit | Merged openstack/openstack-ansible-repo_server: Fix errors due to repo_service_user_name setting https://review.openstack.org/390063 | 10:08 |
| *** Michaelg_barcelo has quit IRC | 10:08 | |
| *** Drago1 has joined #openstack-ansible | 10:09 | |
| *** Drago1 has quit IRC | 10:10 | |
| *** Drago1 has joined #openstack-ansible | 10:10 | |
| *** smatzek has quit IRC | 10:10 | |
| *** m4rx has quit IRC | 10:15 | |
| *** d3n14l has joined #openstack-ansible | 10:15 | |
| *** adrian_otto has quit IRC | 10:15 | |
| *** thorst has quit IRC | 10:15 | |
| *** gouthamr has joined #openstack-ansible | 10:16 | |
| *** smatzek has joined #openstack-ansible | 10:17 | |
| *** berendt has joined #openstack-ansible | 10:18 | |
| *** m4rx has joined #openstack-ansible | 10:18 | |
| *** thorst has joined #openstack-ansible | 10:19 | |
| *** d3n14l has quit IRC | 10:21 | |
| *** david-lyle has quit IRC | 10:22 | |
| *** thorst_ has joined #openstack-ansible | 10:23 | |
| *** thorst has quit IRC | 10:24 | |
| *** m4rx has quit IRC | 10:24 | |
| *** d3n14l has joined #openstack-ansible | 10:28 | |
| *** smatzek has quit IRC | 10:31 | |
| *** d3n14l has quit IRC | 10:33 | |
| *** m4rx has joined #openstack-ansible | 10:36 | |
| *** janki has joined #openstack-ansible | 10:36 | |
| *** d3n14l has joined #openstack-ansible | 10:36 | |
| *** allanice001 has quit IRC | 10:37 | |
| *** allanice001 has joined #openstack-ansible | 10:38 | |
| *** Jeffrey4l has joined #openstack-ansible | 10:38 | |
| *** d3n14l has quit IRC | 10:39 | |
| *** Drago1 has quit IRC | 10:39 | |
| *** d3n14l has joined #openstack-ansible | 10:42 | |
| *** janki has quit IRC | 10:43 | |
| *** Drago1 has joined #openstack-ansible | 10:45 | |
| *** Drago1 has quit IRC | 10:45 | |
| *** Drago1 has joined #openstack-ansible | 10:45 | |
| *** d3n14l has quit IRC | 10:50 | |
| *** d3n14l has joined #openstack-ansible | 10:51 | |
| *** d3n14l has quit IRC | 10:52 | |
| *** woodard has quit IRC | 10:52 | |
| *** berendt has quit IRC | 10:53 | |
| *** ybabenko has quit IRC | 10:53 | |
| *** gouthamr has quit IRC | 10:57 | |
| *** smatzek has joined #openstack-ansible | 10:57 | |
| *** thorst_ has quit IRC | 10:58 | |
| *** thorst has joined #openstack-ansible | 11:01 | |
| *** Jeffrey4l has quit IRC | 11:03 | |
| openstackgerrit | Merged openstack/openstack-ansible: Fix errors due to repo_service_user_name setting https://review.openstack.org/390065 | 11:03 |
| *** thorst has quit IRC | 11:05 | |
| *** alij has quit IRC | 11:05 | |
| *** alij has joined #openstack-ansible | 11:06 | |
| *** alij has quit IRC | 11:11 | |
| *** thorst has joined #openstack-ansible | 11:18 | |
| *** thorst has quit IRC | 11:23 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Fix errors due to repo_service_user_name setting https://review.openstack.org/390522 | 11:24 |
| *** admin0 has joined #openstack-ansible | 11:25 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Fix errors due to repo_service_user_name setting https://review.openstack.org/390523 | 11:25 |
| *** smatzek has quit IRC | 11:26 | |
| *** esberglu__ has joined #openstack-ansible | 11:28 | |
| *** esberglu__ has quit IRC | 11:29 | |
| *** admin0 has quit IRC | 11:30 | |
| *** alij has joined #openstack-ansible | 11:30 | |
| *** alij has quit IRC | 11:31 | |
| *** alij has joined #openstack-ansible | 11:31 | |
| *** johnmilton has quit IRC | 11:32 | |
| *** alij has quit IRC | 11:32 | |
| *** alij has joined #openstack-ansible | 11:32 | |
| *** retreved has joined #openstack-ansible | 11:35 | |
| *** alij has quit IRC | 11:37 | |
| *** alij has joined #openstack-ansible | 11:38 | |
| *** alij has quit IRC | 11:39 | |
| *** alij has joined #openstack-ansible | 11:39 | |
| *** Drago1 has quit IRC | 11:44 | |
| *** thorst has joined #openstack-ansible | 11:44 | |
| *** thorst has quit IRC | 11:48 | |
| *** thorst has joined #openstack-ansible | 11:50 | |
| *** smatzek has joined #openstack-ansible | 11:51 | |
| *** thorst has quit IRC | 11:55 | |
| *** jwitko has quit IRC | 11:55 | |
| *** schwicht has joined #openstack-ansible | 11:56 | |
| *** johnmilton has joined #openstack-ansible | 11:59 | |
| *** asettle has joined #openstack-ansible | 12:04 | |
| *** johnmilton has quit IRC | 12:04 | |
| *** johnmilton has joined #openstack-ansible | 12:05 | |
| *** asettle has quit IRC | 12:05 | |
| *** adrian_otto has joined #openstack-ansible | 12:10 | |
| *** Hosam has joined #openstack-ansible | 12:10 | |
| *** thorst has joined #openstack-ansible | 12:12 | |
| *** adrian_otto has quit IRC | 12:12 | |
| *** Hosam has quit IRC | 12:15 | |
| *** adrian_otto has joined #openstack-ansible | 12:15 | |
| *** adrian_otto has quit IRC | 12:15 | |
| *** Drago has joined #openstack-ansible | 12:15 | |
| *** ybabenko has joined #openstack-ansible | 12:15 | |
| mgariepy | good morning everyone | 12:15 |
| *** Drago has quit IRC | 12:15 | |
| *** thorst has quit IRC | 12:16 | |
| *** alij has quit IRC | 12:16 | |
| *** Drago has joined #openstack-ansible | 12:17 | |
| *** thorst has joined #openstack-ansible | 12:18 | |
| *** adrian_otto has joined #openstack-ansible | 12:18 | |
| *** v1k0d3n has joined #openstack-ansible | 12:21 | |
| *** berendt has joined #openstack-ansible | 12:24 | |
| *** gouthamr has joined #openstack-ansible | 12:25 | |
| *** admin0 has joined #openstack-ansible | 12:27 | |
| *** pcaruana has joined #openstack-ansible | 12:29 | |
| *** adrian_otto has quit IRC | 12:30 | |
| *** schwicht has quit IRC | 12:31 | |
| *** Drago has quit IRC | 12:32 | |
| *** Drago has joined #openstack-ansible | 12:33 | |
| *** gouthamr has quit IRC | 12:40 | |
| *** timrc_ is now known as timrc | 12:41 | |
| *** admin0 has quit IRC | 12:42 | |
| *** pcaruana has quit IRC | 12:44 | |
| *** alij has joined #openstack-ansible | 12:47 | |
| *** ybabenko has quit IRC | 12:51 | |
| *** rgogunskiy has joined #openstack-ansible | 12:52 | |
| *** d3n14l has joined #openstack-ansible | 12:54 | |
| *** thorst has quit IRC | 12:55 | |
| *** Hosam has joined #openstack-ansible | 12:58 | |
| *** Hosam has quit IRC | 12:58 | |
| *** berendt has quit IRC | 12:58 | |
| *** d3n14l has quit IRC | 12:58 | |
| *** Hosam has joined #openstack-ansible | 12:58 | |
| *** thorst_ has joined #openstack-ansible | 12:58 | |
| *** dxiri has joined #openstack-ansible | 12:59 | |
| *** v1k0d3n has quit IRC | 12:59 | |
| *** adrian_otto has joined #openstack-ansible | 13:00 | |
| dxiri | good day everyone | 13:00 |
| dxiri | have a simple question that I can't find the answer for, how does the containers get the IP they will use? | 13:00 |
| dxiri | do I need a dhcp server? or is it set statically from a pool? | 13:01 |
| *** janki has joined #openstack-ansible | 13:03 | |
| *** thorst_ has quit IRC | 13:03 | |
| *** gouthamr has joined #openstack-ansible | 13:03 | |
| *** d3n14l has joined #openstack-ansible | 13:04 | |
| *** admin0 has joined #openstack-ansible | 13:04 | |
| *** schwicht has joined #openstack-ansible | 13:05 | |
| *** alij has quit IRC | 13:05 | |
| *** thorst_ has joined #openstack-ansible | 13:09 | |
| *** ybabenko has joined #openstack-ansible | 13:09 | |
| *** schwicht has quit IRC | 13:09 | |
| *** smatzek has quit IRC | 13:11 | |
| *** klamath has joined #openstack-ansible | 13:13 | |
| *** klamath has quit IRC | 13:13 | |
| *** thorst_ has quit IRC | 13:13 | |
| *** ybabenko has quit IRC | 13:14 | |
| *** klamath has joined #openstack-ansible | 13:14 | |
| *** adrian_otto has quit IRC | 13:14 | |
| *** d3n14l has quit IRC | 13:15 | |
| *** jcannava has joined #openstack-ansible | 13:16 | |
| *** Jeffrey4l has joined #openstack-ansible | 13:18 | |
| *** cathrichardson has joined #openstack-ansible | 13:18 | |
| *** schwicht has joined #openstack-ansible | 13:19 | |
| *** Hosam has quit IRC | 13:20 | |
| *** cathrichardson has quit IRC | 13:21 | |
| *** cathrichardson has joined #openstack-ansible | 13:21 | |
| *** allanice001 has quit IRC | 13:22 | |
| *** Hosam has joined #openstack-ansible | 13:23 | |
| *** cathrichardson has quit IRC | 13:24 | |
| *** thorst_ has joined #openstack-ansible | 13:26 | |
| *** david-lyle has joined #openstack-ansible | 13:26 | |
| *** thorst_ has quit IRC | 13:27 | |
| *** cathrichardson has joined #openstack-ansible | 13:28 | |
| mgariepy | dxiri, it's generated from the pool https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L2-L11 | 13:28 |
| mgariepy | dxiri, with the network config for the containers like https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L25 | 13:30 |
| *** d3n14l has joined #openstack-ansible | 13:31 | |
| dxiri | mgariepy: thank you! the problem I am having is that I moved the management to a new vlan/network, but the containers are still trying to set an IP from the old pool | 13:33 |
| dxiri | I have destroyed the containers about the times now | 13:33 |
| dxiri | and that behavior keeps on happening | 13:33 |
| *** thorst_ has joined #openstack-ansible | 13:33 | |
| mgariepy | they are stored in : /etc/openstack_deploy/openstack_inventory.json | 13:34 |
| *** d3n14l has quit IRC | 13:34 | |
| dxiri | ah, can I nuke that file? | 13:34 |
| dxiri | so it gets generated again? | 13:34 |
| mgariepy | if you nuke it it will be regenerated with new info, but you will loose your current inventory. | 13:35 |
| dxiri | that's the idea :) | 13:35 |
| dxiri | want to start from scratch | 13:36 |
| mgariepy | if you want to start from scratch then yes you can nuke it ;) | 13:36 |
| mgariepy | don't ever do that in production tho. | 13:36 |
| mgariepy | all the container name will change, the ips will also etc.. | 13:37 |
| mgariepy | dxiri, which version of openstack are you installing ? | 13:37 |
| dxiri | right now I am trying to get this to run on a test env, so we should be good :) | 13:38 |
| dxiri | Newton | 13:38 |
| *** alij has joined #openstack-ansible | 13:38 | |
| mgariepy | ok | 13:38 |
| *** alij has quit IRC | 13:38 | |
| dxiri | tag 14.0.0 | 13:38 |
| *** alij has joined #openstack-ansible | 13:38 | |
| *** thorst_ has quit IRC | 13:38 | |
| dxiri | we have 12 nodes to play with here :D | 13:38 |
| dxiri | I'm starting with 3 and then will start to expand once I get this 3 running good | 13:39 |
| *** david-lyle has quit IRC | 13:39 | |
| mgariepy | great :) | 13:39 |
| mgariepy | I wish I would have a 12 nodes cluster to play with which isn't 10 years old ;) | 13:40 |
| dxiri | mgariepy: and supposing I need to do this on prod eventually, the best way would be to edit the json file with the correct IPs and then re-run the playbooks? | 13:40 |
| mgariepy | dxiri, i'm not sure, you will have to tests it. | 13:41 |
| mgariepy | maybe someone else would know better than me. | 13:41 |
| dxiri | well with 12 nodes I can afford to have staging and prod :D | 13:41 |
| dxiri | lol | 13:41 |
| *** schwicht has quit IRC | 13:42 | |
| mgariepy | I have 3 good nodes, plus a bunch of old servers (mostly for controllers) for staging. | 13:42 |
| admin0 | mgariepy: all you need is 1 good node to make it 15 :D | 13:43 |
| *** agrebennikov has joined #openstack-ansible | 13:43 | |
| admin0 | evrardjp: nice meeting you (in person) | 13:43 |
| mgariepy | my prod is 3 ctrl, 36 compute + ceph cluster | 13:43 |
| admin0 | not abel to meet anyone else :) | 13:43 |
| admin0 | ywr | 13:44 |
| admin0 | yet | 13:44 |
| *** jwitko has joined #openstack-ansible | 13:44 | |
| mgariepy | any core around ? https://review.openstack.org/#/c/390353/ ? | 13:44 |
| palendae | I think a lot of them are in sessions at the summit right now | 13:45 |
| *** Jeffrey4l has quit IRC | 13:45 | |
| mgariepy | yeah probably.. | 13:45 |
| *** openstackstatus has joined #openstack-ansible | 13:48 | |
| *** ChanServ sets mode: +v openstackstatus | 13:48 | |
| *** admin0 has quit IRC | 13:48 | |
| *** Drago has quit IRC | 13:49 | |
| *** schwicht has joined #openstack-ansible | 13:54 | |
| *** d3n14l has joined #openstack-ansible | 13:56 | |
| palendae | alextricity25, Did you fix your problem with the 'no such option --constraint' error? | 13:57 |
| palendae | Looks like I may need to upgrade virtualenv on the machine | 13:58 |
| alextricity25 | palendae: Honestly I haven't looked at it since I last brought it up. Sorry :/ | 13:58 |
| *** gouthamr has quit IRC | 13:58 | |
| palendae | alextricity25, np. Found https://bugs.launchpad.net/nova/+bug/1548712 which said 13.03 was too old...and I have 11 something | 13:59 |
| openstack | Launchpad bug 1548712 in OpenStack Compute (nova) "pip: no such option: -c in current Nova master testing setup" [Undecided,Invalid] | 13:59 |
| *** thorst_ has joined #openstack-ansible | 13:59 | |
| *** v1k0d3n has joined #openstack-ansible | 13:59 | |
| *** thorst_ has quit IRC | 13:59 | |
| *** admin0 has joined #openstack-ansible | 13:59 | |
| *** thorst_ has joined #openstack-ansible | 14:00 | |
| alextricity25 | Oh Interesting. Did upgrading do the trick for you? | 14:00 |
| *** d3n14l has quit IRC | 14:00 | |
| cloudnull | mornings all | 14:01 |
| palendae | Gotten past that, at least | 14:02 |
| palendae | alextricity25, Yep looks like that was it. I bet the images for the gate have the latest version pre-installed so that's why they don't see it | 14:03 |
| palendae | But a stock 14.04 image wouldn't | 14:03 |
| mgariepy | good morning cloudnull | 14:05 |
| cloudnull | o/ mgariepy | 14:05 |
| *** d3n14l has joined #openstack-ansible | 14:05 | |
| *** greghaynes has joined #openstack-ansible | 14:07 | |
| *** thorst_ has quit IRC | 14:08 | |
| *** retreved_ has joined #openstack-ansible | 14:09 | |
| *** d3n14l has quit IRC | 14:09 | |
| *** michaelg_barcelo has joined #openstack-ansible | 14:09 | |
| *** Mudpuppy has joined #openstack-ansible | 14:09 | |
| *** Mudpuppy has quit IRC | 14:10 | |
| *** thorst_ has joined #openstack-ansible | 14:11 | |
| *** retreved_ has quit IRC | 14:11 | |
| *** retreved_ has joined #openstack-ansible | 14:12 | |
| *** Mudpuppy has joined #openstack-ansible | 14:12 | |
| *** retreved has quit IRC | 14:12 | |
| *** arif-ali has quit IRC | 14:12 | |
| cloudnull | if folks are around to give these a review it'd be appreciated: https://review.openstack.org/#/q/project:%255Eopenstack/openstack-ansible.*+branch:stable/mitaka+status:open,n,z | 14:13 |
| cloudnull | ^ mitaka [old and boring i know :)] | 14:13 |
| openstackgerrit | Merged openstack/openstack-ansible-rsyslog_client: Fix errors due to repo_service_user_name setting https://review.openstack.org/390506 | 14:13 |
| cloudnull | https://review.openstack.org/#/q/project:%255Eopenstack/openstack-ansible.*+branch:stable/newton+status:open,n,z | 14:13 |
| cloudnull | ^ Newton [new and exciting] | 14:14 |
| cloudnull | :p | 14:14 |
| openstackgerrit | Merged openstack/openstack-ansible-haproxy_server: [DOCS] Add note about extra VIP addresses https://review.openstack.org/390496 | 14:14 |
| cloudnull | castulo: so I've now done three upgrades and have had 1 failure which was due to ceilometer apt packages not being available. | 14:15 |
| *** esberglu has joined #openstack-ansible | 14:15 | |
| *** arif-ali has joined #openstack-ansible | 14:15 | |
| cloudnull | going from mitaka to newton | 14:15 |
| cloudnull | if you have a chance today to give some more insight into the errors you've been seeing I'd love to see if we can get bugs created for them and get them taken care. | 14:15 |
| lbragstad | morning! curious if anyone has seen this in the os_keystone role recently http://cdn.pasteraw.com/37pg9l2acsfoark6u98it0r86nxcq4l | 14:16 |
| stevemar | lbragstad: morning?! | 14:17 |
| lbragstad | stevelle yep - morning | 14:17 |
| cloudnull | lbragstad: are you not in BCN ? | 14:17 |
| lbragstad | cloudnull i'm in Fargo, ND | 14:17 |
| palendae | cloudnull, Real small RAX crew in BCN afaik | 14:18 |
| palendae | Real, real small | 14:18 |
| cloudnull | lbragstad: ah. | 14:18 |
| lbragstad | our osic performance bot caught that ^ error while running os_keystone on master in the Ensure service tenant step | 14:18 |
| *** allanice001 has joined #openstack-ansible | 14:19 | |
| lbragstad | wondering if that is something anyone else has seen (it looks like a dependency issue) or it happens to be something I messed up | 14:19 |
| *** chris_hultin|AWA is now known as chris_hultin | 14:19 | |
| cloudnull | ive not seen that. | 14:19 |
| *** allanice001 has quit IRC | 14:20 | |
| lbragstad | ContextualVersionConflict: (pysaml2 4.3.0 (/openstack/venvs/keystone-untagged/lib/python2.7/site-packages), Requirement.parse('pysaml2<4.0.3,>=2.4.0'), set(['keystone'])) | 14:20 |
| *** berendt has joined #openstack-ansible | 14:20 | |
| *** v1k0d3n has quit IRC | 14:20 | |
| cloudnull | but whaat check is that? | 14:20 |
| cloudnull | master? | 14:20 |
| lbragstad | cloudnull yep - i run master for our performance job | 14:20 |
| lbragstad | keystone master + os_keystone master | 14:20 |
| *** thorst_ has quit IRC | 14:20 | |
| * lbragstad lives dangerously | 14:21 | |
| lbragstad | I have extra verbose output from ansible, too | 14:21 |
| lbragstad | http://cdn.pasteraw.com/ehy3v4qjv2jecbznyfd4y4m06t5wc0r | 14:21 |
| lbragstad | it appears that keystoneclient receives a 500 from keystone while trying to verify the service tenant exists | 14:22 |
| *** thorst_ has joined #openstack-ansible | 14:22 | |
| lbragstad | err - s/keystoneclient/keystoneauth/ | 14:24 |
| *** david-lyle has joined #openstack-ansible | 14:24 | |
| *** allanice001 has joined #openstack-ansible | 14:25 | |
| cloudnull | lbragstad: do you have access to the node where that ran ? | 14:25 |
| lbragstad | cloudnull I do | 14:26 |
| lbragstad | cloudnull it's a bare metal host but the ansible bits run on-metal within a container | 14:26 |
| cloudnull | can you cat out /opt/developer-pip-constraints.txt | 14:26 |
| lbragstad | cloudnull yep | 14:26 |
| *** thorst_ has quit IRC | 14:27 | |
| lbragstad | cloudnull | 14:27 |
| lbragstad | # cat /opt/developer-pip-constraints.txt | 14:27 |
| lbragstad | git+https://git.openstack.org/openstack/keystone@master#egg=keystone | 14:27 |
| cloudnull | ok. | 14:27 |
| dxiri | ok I found my problem....for some reason eth1 on the containers doesn't want to come up automatically | 14:27 |
| dxiri | but If I do ifup eth1 inside the container it works | 14:28 |
| dxiri | any idea why? | 14:28 |
| cloudnull | lbragstad: do you have the verbose output from the task "Install required pip packages" ? | 14:28 |
| lbragstad | cloudnull I can - let me grab it | 14:28 |
| cloudnull | dxiri: if you stop and start the container does it come up ? | 14:28 |
| *** Jeffrey4l has joined #openstack-ansible | 14:28 | |
| dxiri | cloudnull: checking | 14:28 |
| cloudnull | lxc-stop -n $NAME; lxc-start -dn $NAME | 14:28 |
| cloudnull | the lxc-container-create play is supposed to bring up all of the interfaces but maybe something went wrong? | 14:29 |
| cloudnull | or wasn't in place when it ran? | 14:29 |
| *** alij has quit IRC | 14:30 | |
| lbragstad | cloudnull http://cdn.pasteraw.com/h3xw3i3x8b92hijmj5hwl9xcuwst8e0 | 14:30 |
| dxiri | mmmm, I rebooted the host itself (kernel update) and I see they are running, but eth1 has no ip | 14:31 |
| dxiri | and ifup works | 14:31 |
| lbragstad | cloudnull and http://cdn.pasteraw.com/qahjip3vw47tba28q13kwc49dpt9vvq | 14:31 |
| *** berendt has quit IRC | 14:31 | |
| dxiri | my play is hanging (of course) on the "wait for container ssh" task | 14:31 |
| cloudnull | lbragstad: looks like the play installed pysaml2 here "/openstack/venvs/keystone-untagged/bin/pip install -U --constraint /opt/developer-pip-constraints.txt argparse keystone keystonemiddleware ldappool lxml PyMySQL oslo.log oslo.middleware pbr pycrypto pysaml2 python-keystoneclient python-ldap python-memcached python-openstackclient repoze.lru uWSGI" but in the output it's saying that the package is already up to | 14:35 |
| cloudnull | date? | 14:35 |
| cloudnull | meaning it was already in the venv? | 14:35 |
| lbragstad | cloudnull must have been - but i'm not exactly sure how | 14:36 |
| *** Hosam has quit IRC | 14:36 | |
| lbragstad | cloudnull this is how i'm running everything - https://github.com/lbragstad/keystone-performance/blob/master/run_everything.sh | 14:37 |
| cloudnull | maybe the venv was already there? | 14:37 |
| castulo | cloudnull: ok, I'm installing a new system from scratch and will retry the upgrade. I will keep you posted. | 14:38 |
| cloudnull | ok. | 14:38 |
| cloudnull | thanks | 14:38 |
| lbragstad | cloudnull hmm - this is a brand new container, so unless something about installing the database installs pysaml2? | 14:38 |
| cloudnull | dxiri: did it get past that part? | 14:38 |
| *** admin0 has quit IRC | 14:39 | |
| palendae | If people have time - I'm curious if I should keep going with this to be just deleting IPs, or deleting then regenerating. Is there a case for deleting and not regenerating? https://review.openstack.org/#/c/390375/ | 14:39 |
| *** janki has quit IRC | 14:40 | |
| *** david-lyle has quit IRC | 14:40 | |
| *** d3n14l has joined #openstack-ansible | 14:40 | |
| cloudnull | lbragstad: i'm trying to reproduce it nnow | 14:41 |
| *** d3n14l has quit IRC | 14:42 | |
| lbragstad | cloudnull thanks - i'm digging through the galera_server, galera_client roles because those are dependencies but I'm not seeing anything pull in pysaml2 yet | 14:42 |
| dxiri | cloudnull: ran it again (for the 4th time) and it worked now | 14:45 |
| cloudnull | lbragstad: so i too am getting the 4.3.0 version. | 14:45 |
| lbragstad | hmm | 14:45 |
| cloudnull | and I'm running it manualy in a new vevn | 14:46 |
| cloudnull | *venv | 14:46 |
| lbragstad | huh - so something is pulling in 4.3.0 | 14:46 |
| lbragstad | but we appear to be capped at 4.0.3 | 14:46 |
| cloudnull | interesting. | 14:47 |
| cloudnull | so if i remove the package on the cli it pulls in the correct requirement. | 14:47 |
| cloudnull | otherwise it takes from head. | 14:47 |
| cloudnull | pysaml2-4.0.2 | 14:47 |
| lbragstad | cloudnull so constraints aren't being applied? https://github.com/openstack/keystone/blob/master/requirements.txt#L36 | 14:47 |
| cloudnull | it would seem they're being applied but items on the CLI are not inheriting them . | 14:48 |
| cloudnull | IE if i remove that one package it works as expected. | 14:48 |
| lbragstad | weird - | 14:49 |
| neith | I keep getting this error http://pastebin.com/LE7SzLvZ | 14:49 |
| *** Drago has joined #openstack-ansible | 14:49 | |
| lbragstad | so removing pysaml2 (at version 4.3.0) | 14:49 |
| neith | either I had or remove the executable bit | 14:49 |
| *** Drago has quit IRC | 14:49 | |
| neith | *add | 14:49 |
| *** Drago has joined #openstack-ansible | 14:49 | |
| cloudnull | neith: can you execute the file and see if it throws an error? | 14:50 |
| neith | cloudnull: so obvious | 14:50 |
| neith | sorry | 14:50 |
| cloudnull | lbragstad: http://cdn.pasteraw.com/4ud8fqwp0owvjy041d9tj0djvhcchh3 | 14:50 |
| *** cathrich_ has joined #openstack-ansible | 14:51 | |
| cloudnull | neith: in ansible 1.9 it used to pass a stacktrace. | 14:51 |
| cloudnull | while in 2.x it masks it . | 14:51 |
| cloudnull | sorry. | 14:51 |
| lbragstad | cloudnull weird... | 14:51 |
| *** drifterza has quit IRC | 14:51 | |
| lbragstad | it should be getting constrained | 14:52 |
| *** cathrichardson has quit IRC | 14:52 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova: Fix api_paste_config option https://review.openstack.org/390497 | 14:53 |
| *** weezS has joined #openstack-ansible | 14:54 | |
| lbragstad | cloudnull did you test by removing https://github.com/openstack/openstack-ansible-os_keystone/blob/master/defaults/main.yml#L429 ? | 14:55 |
| *** alij has joined #openstack-ansible | 14:55 | |
| cloudnull | yes | 14:56 |
| lbragstad | cloudnull huh - yeah commenting out that line in my os_keystone role locally worked | 14:57 |
| *** allanice001 has quit IRC | 14:59 | |
| dxiri | cloudnull: happened again but on setup-infrastructure now :( | 14:59 |
| *** Hosam has joined #openstack-ansible | 14:59 | |
| *** pester has quit IRC | 15:00 | |
| *** markvoelker has joined #openstack-ansible | 15:01 | |
| cloudnull | dxiri: something has to be causing the interface not to come up. | 15:01 |
| cloudnull | maybe duplicate IPs? | 15:01 |
| cloudnull | also is this 16.04? | 15:02 |
| dxiri | 16.04 yes sir | 15:02 |
| dxiri | duplicate IPs? how? they are taken from the pool and I am sure about the excluded ranges :) | 15:02 |
| lbragstad | cloudnull do you want me to file a bug? | 15:03 |
| dxiri | does it matter that I have eth1 inside the container and eth1 on the host itself? | 15:03 |
| *** markvoelker_ has joined #openstack-ansible | 15:03 | |
| *** d3n14l has joined #openstack-ansible | 15:04 | |
| *** berendt has joined #openstack-ansible | 15:06 | |
| *** thorst_ has joined #openstack-ansible | 15:06 | |
| *** cathrich_ has quit IRC | 15:07 | |
| dxiri | for example...on the utility container, I have eth1@if192 on the container and eth1 on the host | 15:07 |
| *** markvoelker has quit IRC | 15:07 | |
| *** markvoel_ has joined #openstack-ansible | 15:07 | |
| cloudnull | dxiri: it should not matter at all. | 15:08 |
| cloudnull | dxiri: if you manually stop the container and start it does the device come up ? | 15:09 |
| *** d3n14l has quit IRC | 15:10 | |
| cloudnull | anything in the logs that would indicate there is an issue? | 15:10 |
| cloudnull | lbragstad: idk what the bug would be . | 15:10 |
| *** gouthamr has joined #openstack-ansible | 15:10 | |
| cloudnull | lbragstad: maybe file it in LP and we can raise other issues if neede.d | 15:10 |
| *** thorst_ has quit IRC | 15:11 | |
| lbragstad | cloudnull cool - done https://bugs.launchpad.net/openstack-ansible/+bug/1636547 | 15:11 |
| openstack | Launchpad bug 1636547 in openstack-ansible "Constraints aren't applied to pysaml2 when using os_keystone" [Undecided,New] | 15:11 |
| *** markvoelker_ has quit IRC | 15:11 | |
| *** david-lyle has joined #openstack-ansible | 15:11 | |
| *** h5t4 has quit IRC | 15:11 | |
| *** markvoel_ has quit IRC | 15:11 | |
| *** cathrichardson has joined #openstack-ansible | 15:11 | |
| lbragstad | cloudnull as a workaround should I override keystone_pip_packages? | 15:12 |
| lbragstad | in my performance project? | 15:12 |
| openstackgerrit | Lance Bragstad proposed openstack/openstack-ansible-os_keystone: Remove pysaml2 from keystone_pip_packages https://review.openstack.org/390587 | 15:13 |
| cloudnull | lbragstad: mind adding pip_install_upper_constraints: "http://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt?h=stable/newton" | 15:13 |
| *** markvoelker has joined #openstack-ansible | 15:13 | |
| cloudnull | or master in your case. | 15:13 |
| cloudnull | to the vars and rerunning ? | 15:13 |
| cloudnull | pip_install_upper_constraints: "http://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt?h=master" | 15:14 |
| lbragstad | cloudnull to os_keystone/defaults/main.yml ? | 15:14 |
| lbragstad | and adding PySAML2 back in? | 15:15 |
| *** cathrich_ has joined #openstack-ansible | 15:15 | |
| cloudnull | yes | 15:15 |
| cloudnull | and yes :) | 15:15 |
| lbragstad | cloudnull ok - running it now | 15:15 |
| *** thorst_ has joined #openstack-ansible | 15:15 | |
| lbragstad | cloudnull that seems to work, too | 15:16 |
| *** cathrich_ is now known as cathrichardson_ | 15:17 | |
| *** cathrichardson has quit IRC | 15:17 | |
| *** cathrichardson_ is now known as cathrichardson | 15:17 | |
| dxiri | cloudnull: restarting the container doesn't seem to help | 15:17 |
| dxiri | cloudnull: found this on dmesg inside one of the containers: | 15:17 |
| cloudnull | lbragstad: https://github.com/openstack/openstack-ansible-os_keystone/commit/c5a0cce3c695ccd6ecb594fc8b3148e085e9a5d2 | 15:17 |
| dxiri | [ 2715.103568] br-mgmt: port 39(2f1c5fe7_eth1) entered forwarding state | 15:17 |
| dxiri | [ 2715.103589] br-mgmt: port 39(2f1c5fe7_eth1) entered forwarding state | 15:17 |
| dxiri | [ 2715.104292] br-mgmt: port 39(2f1c5fe7_eth1) entered disabled state | 15:17 |
| dxiri | [ 2715.248374] eth1: renamed from veth24Y1D4 | 15:17 |
| lbragstad | cloudnull bah! | 15:18 |
| cloudnull | you've been odyssey4me'd | 15:18 |
| cloudnull | :p | 15:18 |
| lbragstad | cloudnull so - this change should go into os_kesytone though, right? | 15:19 |
| *** berendt has quit IRC | 15:19 | |
| *** thorst_ has quit IRC | 15:20 | |
| openstackgerrit | Lance Bragstad proposed openstack/openstack-ansible-os_keystone: Use pip_install_upper_constraints https://review.openstack.org/390587 | 15:20 |
| palendae | Does every role need that ^? | 15:20 |
| cloudnull | lbragstad: it looks like it's part of of the tests repo https://github.com/openstack/openstack-ansible-tests/blob/master/test-vars.yml#L415 | 15:20 |
| palendae | Oof | 15:21 |
| cloudnull | and for newton https://github.com/openstack/openstack-ansible-tests/blob/stable/newton/test-vars.yml#L415 | 15:21 |
| lbragstad | cloudnull ah - so the tests add it... but it's up to the individual projects to define it if they consume os_keystone? | 15:21 |
| cloudnull | so the tests repo adds in the common bits shared acorss many roles. | 15:21 |
| cloudnull | yes. it would seem that way . | 15:22 |
| palendae | lbragstad, Sounds like. The integrated part uses --constraints itself | 15:22 |
| lbragstad | alrighty - i can patch my project | 15:22 |
| cloudnull | we should have a note about that in the docs http://docs.openstack.org/developer/openstack-ansible-os_keystone/ | 15:23 |
| cloudnull | ^ that's kinda a big usability issue if folks don't know to set that in dev mode. | 15:23 |
| cloudnull | odyssey4me: | 15:23 |
| palendae | Yeah | 15:23 |
| *** rgogunskiy has quit IRC | 15:23 | |
| palendae | That's part of the problem with multiple repos like this | 15:24 |
| lbragstad | patched - https://github.com/lbragstad/keystone-performance/commit/30709a85a974b3e7416a5d2ceb63028a308173eb | 15:24 |
| *** esberglu has left #openstack-ansible | 15:24 | |
| palendae | Not sure how to get around that very well; need docs for all the roles, and/or the roles constrain themselves | 15:24 |
| cloudnull | all of the roles have docs. | 15:25 |
| palendae | need docs for implied dependencies* | 15:25 |
| cloudnull | yea. | 15:25 |
| lbragstad | doing a clean run on a new container | 15:25 |
| palendae | I'm aware they have docs files, but probably missing that verbage | 15:25 |
| cloudnull | palendae: indeed it is . | 15:25 |
| palendae | Thanks for trying this stuff in isolation, lbragstad | 15:26 |
| cloudnull | its noted in an example here http://docs.openstack.org/developer/openstack-ansible-pip_install/ | 15:26 |
| lbragstad | palendae no problem | 15:26 |
| cloudnull | but that's super obscure. | 15:26 |
| *** Jeffrey4l has quit IRC | 15:27 | |
| palendae | A guess, but pip_install's unlikely to be used much outside the integrated project | 15:27 |
| palendae | roles like keystone or swift have a better chance, I'd think | 15:27 |
| cloudnull | ++ | 15:27 |
| *** tschacara has joined #openstack-ansible | 15:28 | |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts: LXC version to 2.0.5 on CentOS https://review.openstack.org/390353 | 15:29 |
| lbragstad | cloudnull palendae awesome - thanks for the help. I've updated my bug and abandon my patch | 15:29 |
| cloudnull | thanks again for pounding on that | 15:30 |
| cloudnull | dxiri: from within the container, when eth1 is not up, is there anything in those logs that indicates there was an issue ? | 15:30 |
| *** asettle has joined #openstack-ansible | 15:31 | |
| neith | I know its not related to OSA but I cant ping between host on br-mgmt interfaces | 15:31 |
| neith | I don't know how to debug | 15:31 |
| cloudnull | neith: you have two hosts on br-mgmt and you're not seeing traffic between them ? | 15:32 |
| neith | cloudnull: right | 15:32 |
| cloudnull | can you ping one host and then do a tcpdump on the other to see if you're able to see anything ? | 15:32 |
| cloudnull | also are the two interfaces on the same subnet? | 15:33 |
| neith | cloudnull: I can try, yes they are | 15:33 |
| cloudnull | are you able to send traffic over other networks to your various hosts? | 15:33 |
| neith | cloudnull: thats what I am checking | 15:34 |
| *** c-mart has joined #openstack-ansible | 15:34 | |
| *** michaelg_barcelo has quit IRC | 15:35 | |
| *** admin0 has joined #openstack-ansible | 15:38 | |
| neith | cloudnull: without the bridge I can ping between hosts | 15:39 |
| neith | using the same /24 | 15:39 |
| *** weezS has quit IRC | 15:41 | |
| *** admin0 has left #openstack-ansible | 15:41 | |
| *** thorst_ has joined #openstack-ansible | 15:42 | |
| *** alij has quit IRC | 15:42 | |
| cloudnull | is the interface within the bridge. | 15:42 |
| *** david-lyle has quit IRC | 15:42 | |
| cloudnull | do you have something like "bridge_port ethX" in the interface config? | 15:43 |
| *** thorst__ has joined #openstack-ansible | 15:43 | |
| neith | cloudnull: bridge_ports ens4.10 | 15:46 |
| neith | 15:46 | |
| *** thorst___ has joined #openstack-ansible | 15:46 | |
| *** thorst_ has quit IRC | 15:47 | |
| *** gouthamr has quit IRC | 15:47 | |
| *** thorst__ has quit IRC | 15:47 | |
| *** markvoelker has quit IRC | 15:47 | |
| cloudnull | so this is running on vlan 10? and when you're able to ping between the hosts is that also running on ens4.10 ? | 15:47 |
| *** allanice001 has joined #openstack-ansible | 15:48 | |
| neith | cloudnull: I should try | 15:48 |
| *** markvoelker has joined #openstack-ansible | 15:50 | |
| -cloudnull- there will be no community bug triage today as most folks are away at the summit. If there are bugs that we need to take note of and or should be talking about at please raise in the channel however a formal bug triage process will not be happening today. | 15:52 | |
| *** d3n14l has joined #openstack-ansible | 15:53 | |
| *** admin0 has joined #openstack-ansible | 15:53 | |
| neith | cloudnull: I think my swith does not support vlan | 15:53 |
| *** allanice001 has quit IRC | 15:54 | |
| cloudnull | ah this may be | 15:54 |
| cloudnull | is it a managed switch ? | 15:54 |
| neith | cloudnull: no | 15:55 |
| neith | cloudnull: of course | 15:55 |
| cloudnull | bummer. | 15:55 |
| neith | cloudnull: it was a PoC | 15:55 |
| neith | I'll try do test without vlans | 15:55 |
| *** gouthamr has joined #openstack-ansible | 15:55 | |
| cloudnull | you can do it all without vlan. | 15:55 |
| *** d3n14l has quit IRC | 15:56 | |
| *** c-mart has quit IRC | 15:56 | |
| cloudnull | you may need to use fewer networks (depending on your config and the number of nics you have) . | 15:56 |
| *** d3n14l has joined #openstack-ansible | 15:56 | |
| neith | cloudnull: I really dont get why in the doc they use vlan because they uses different subnets | 15:56 |
| neith | and only one nic | 15:56 |
| *** markvoelker has quit IRC | 15:57 | |
| cloudnull | neith: the doc, even in test, is trying to build the most prod ready env possible. That said maybe we should produce a non-vlan example doc. | 15:59 |
| *** allanice001 has joined #openstack-ansible | 15:59 | |
| neith | cloudnull: well I need to reconfigure to use only one nic and no vlans then | 16:00 |
| cloudnull | 1 nic no vlans right ? | 16:01 |
| neith | cloudnull: exaclty | 16:01 |
| *** m4rx has quit IRC | 16:02 | |
| *** d3n14l has quit IRC | 16:03 | |
| neith | cloudnull: any recommandation? | 16:06 |
| *** hughmFLEXin has joined #openstack-ansible | 16:07 | |
| cloudnull | neith: i'd do something like so http://cdn.pasteraw.com/hctl4wzlpsdfufja5rbx86wazvlr629 | 16:09 |
| cloudnull | sorry -- one correction - http://cdn.pasteraw.com/6lq1h4ebyau6c1crwrxq488td49bap3 | 16:10 |
| *** alij has joined #openstack-ansible | 16:10 | |
| *** asettle has quit IRC | 16:10 | |
| cloudnull | idk what your interface config looks like nor the general topology but that should give you an idea on how to solve it | 16:11 |
| neith | cloudnull: great | 16:11 |
| neith | thks | 16:11 |
| cloudnull | anytime | 16:11 |
| neith | I'll try tomorrow | 16:11 |
| cloudnull | cool | 16:12 |
| *** h5t4 has joined #openstack-ansible | 16:12 | |
| *** weezS has joined #openstack-ansible | 16:13 | |
| cloudnull | that should give you a prod like setup which can be moved into something using vlans later, if you need vlan that is. | 16:14 |
| *** c-mart has joined #openstack-ansible | 16:14 | |
| *** admin0 has quit IRC | 16:15 | |
| *** gouthamr has quit IRC | 16:15 | |
| *** cathrichardson has quit IRC | 16:16 | |
| neith | cloudnull: so in this configuration, I nerver configure static IPs for br-vlan br-storage etc... ? | 16:16 |
| * neith is trying to understand too much new things simultaneously | 16:17 | |
| cloudnull | you likely will need to i just was copying the example from the docs and modifying it for a control bridge | 16:17 |
| neith | ok | 16:17 |
| cloudnull | the control bridge in that example is creating a bunch of veth pairs which are then plugged into the other bridges. in this way we have the same bridge layout that you'd use in prod but using a single interface. | 16:18 |
| *** cathrichardson has joined #openstack-ansible | 16:26 | |
| *** Hosam has quit IRC | 16:31 | |
| *** jheroux has joined #openstack-ansible | 16:36 | |
| *** tschacara has quit IRC | 16:36 | |
| *** haad1 has quit IRC | 16:37 | |
| *** hj-hpe has joined #openstack-ansible | 16:37 | |
| *** thorst___ has quit IRC | 16:40 | |
| *** Drago has quit IRC | 16:40 | |
| *** thorst_ has joined #openstack-ansible | 16:44 | |
| *** berendt has joined #openstack-ansible | 16:47 | |
| *** Drago has joined #openstack-ansible | 16:47 | |
| *** thorst_ has quit IRC | 16:49 | |
| *** Drago has quit IRC | 16:51 | |
| *** c-mart has quit IRC | 16:51 | |
| *** tschacara has joined #openstack-ansible | 16:53 | |
| *** weezS has quit IRC | 16:53 | |
| *** BjoernT has joined #openstack-ansible | 16:54 | |
| *** berendt has quit IRC | 16:58 | |
| *** markvoelker has joined #openstack-ansible | 16:58 | |
| *** javeriak has joined #openstack-ansible | 17:02 | |
| dxiri | cloudnull: do containers need internet access? | 17:05 |
| dxiri | my management bridge does not have internet access, thats via another gateway | 17:05 |
| *** c-mart has joined #openstack-ansible | 17:06 | |
| *** javeriak has quit IRC | 17:06 | |
| *** javeriak has joined #openstack-ansible | 17:06 | |
| *** Drago has joined #openstack-ansible | 17:07 | |
| *** Drago has joined #openstack-ansible | 17:07 | |
| *** gouthamr has joined #openstack-ansible | 17:08 | |
| *** fops has quit IRC | 17:10 | |
| dxiri | - network: | 17:11 |
| dxiri | container_bridge: "br-mgmt" | 17:11 |
| dxiri | container_type: "veth" | 17:11 |
| dxiri | container_interface: "eth1" | 17:11 |
| dxiri | ip_from_q: "container" | 17:11 |
| dxiri | type: "raw" | 17:11 |
| dxiri | group_binds: | 17:11 |
| dxiri | - all_containers | 17:11 |
| dxiri | - hosts | 17:11 |
| dxiri | should "container_interface" be the raw device the br-mgmt bridge is on? | 17:12 |
| dxiri | I see all "container_interface" variables be stuff like "eth12" or "eth11", which are non existent devices, only this br-mgmt bridge is set to an interface that actually exists | 17:13 |
| cloudnull | dxiri: container_interface will be created within the container when the container is built. | 17:20 |
| cloudnull | container bridge is the host device that ties them together. | 17:20 |
| Drago | dxiri: I have an AIO install, seems to be working, and my setting is eth1 | 17:21 |
| dxiri | ah...so maybe that's the problem? the container is getting confused since I also have an eth1 on the host itself | 17:21 |
| cloudnull | that should not be a problem | 17:21 |
| cloudnull | my lab is trusty and still using eth style devices. | 17:21 |
| dxiri | I think I see what's going on...just to test I changed eth1 to eth100 on the container interface | 17:24 |
| dxiri | and now that its creating the containers again | 17:24 |
| dxiri | it still trying to use eth1 | 17:24 |
| dxiri | so there is some leftover config that may be causing connectivity problems? | 17:24 |
| dxiri | is there anything else that I need to do besides lxc-containers-destroy? | 17:25 |
| *** gouthamr has quit IRC | 17:26 | |
| cloudnull | dxiri: no. that will destroy it all . | 17:28 |
| cloudnull | in terms of the containers and all of the container data | 17:28 |
| *** kvcobb2 has joined #openstack-ansible | 17:36 | |
| Matias | hey guys, did someone else had this problem with stable/newton? http://paste.openstack.org/show/586964/ | 17:37 |
| Matias | apparently it was a regression from RC3 | 17:37 |
| Matias | chown nova:nova /var/lock/nova fixes the issue, although I have no idea if the directory was supposed to be owned by nova | 17:38 |
| stevelle | I think it should have been writable by nova at least | 17:38 |
| Matias | it was owned by root:root before the chown | 17:40 |
| Matias | permissions are 755 | 17:41 |
| Matias | does somebody has a working installation to check the owner/perms, please? | 17:41 |
| cloudnull | Matias: I have a running newton deployment , just a moment. | 17:51 |
| Matias | thanks cloudnull :) | 17:53 |
| cloudnull | http://cdn.pasteraw.com/nhp4crxqpup0g22xxd5p3ambw8xgoft | 17:53 |
| cloudnull | mine is nova:nova | 17:53 |
| Matias | /etc/tmpfiles.d/nova.conf only specifies /var/lock/nova-compute | 17:54 |
| Matias | cloudnull: what about the directory? | 17:54 |
| cloudnull | my newton is on 14.04 | 17:54 |
| cloudnull | it's likely an issue with 16.04 + systemd | 17:54 |
| Matias | hmmm | 17:55 |
| cloudnull | which means we need to likely cater for that here https://github.com/openstack/openstack-ansible-os_nova/blob/master/templates/nova-systemd-tempfiles.j2 or somewhere in the systemd init setup here: https://github.com/openstack/openstack-ansible-os_nova/blob/master/tasks/nova_init_systemd.yml | 17:56 |
| *** d3n14l has joined #openstack-ansible | 17:57 | |
| cloudnull | here -- https://github.com/openstack/openstack-ansible-os_nova/blob/master/defaults/main.yml#L362-L389 | 17:57 |
| cloudnull | the service name is being used for the tempfiles dir | 17:58 |
| *** poopcat has joined #openstack-ansible | 18:00 | |
| Drago | How can I set up OSA on a node with an ephemeral disk? | 18:01 |
| cloudnull | Drago: is this for a new installation of OpenStack-Ansible ? | 18:01 |
| Drago | cloudnull: Yes | 18:01 |
| cloudnull | like an AIO ? | 18:01 |
| Drago | cloudnull: AIO, actually | 18:01 |
| Drago | Yep! | 18:02 |
| cloudnull | ok. | 18:02 |
| Matias | another important issue.. but I think someone is already working on this? https://bugs.launchpad.net/horizon/+bug/1632383 | 18:03 |
| openstack | Launchpad bug 1632383 in OpenStack Dashboard (Horizon) "The current Horizon settings indicate no valid image creation methods are available" [High,In progress] - Assigned to Richard Jones (r1chardj0n3s) | 18:03 |
| cloudnull | I'd partition the ephemeral disk into two and then mount 50% in /openstack 25% /var/lib/nova 25% /var/lib/lxc | 18:03 |
| cloudnull | **I'd partition the ephemeral disk into three... | 18:03 |
| Drago | cloudnull: To be explicit, I'm using the io1-15 flavor on RS public cloud | 18:03 |
| Drago | It reports 161 GB in fdisk -l | 18:04 |
| Drago | cloudnull: I haven't worked with them much before. What command should I use to do the mounting? Is it like CBS? | 18:04 |
| cloudnull | Matias: that was merged into the branch | 18:04 |
| cloudnull | https://review.openstack.org/#/c/389761/ | 18:04 |
| cloudnull | i don't think there's a tag with that in it yet. | 18:04 |
| Matias | ah, nice.. i was using the tag | 18:05 |
| *** adrian_otto has joined #openstack-ansible | 18:06 | |
| cloudnull | Drago: it like an extra disk. -- http://cdn.pasteraw.com/l0ao4z95dv7k3l6tp7nsnlaqyzoe74s -- | 18:09 |
| cloudnull | that should do the trick | 18:09 |
| cloudnull | ^ Many assumptions there. proceed at your own risk ^ | 18:10 |
| cloudnull | :) | 18:10 |
| cloudnull | opps typo on the last one, but you get the point. | 18:12 |
| Drago | cloudnull: I see it | 18:13 |
| Drago | Trying it now | 18:13 |
| cloudnull | http://cdn.pasteraw.com/8e4vi5354s5rsyc1o48wcd8xm1lg0dk | 18:13 |
| cloudnull | typo corrected and mkfs added ^ | 18:13 |
| Drago | cloudnull: Okay, all that worked fine and I see them when I run df -h | 18:16 |
| Drago | cloudnull: /dev/xvde actually | 18:16 |
| Drago | just fyi | 18:16 |
| Drago | Running bootstrapping now | 18:16 |
| cloudnull | cool | 18:16 |
| *** allanice001 has quit IRC | 18:17 | |
| Drago | cloudnull: It failed because it thinks there's not enough space available | 18:18 |
| cloudnull | on which partition ? | 18:19 |
| Drago | cloudnull: / | 18:19 |
| cloudnull | how much is available on / | 18:20 |
| Drago | cloudnull: 36 GB, which is why I wanted to mount ephemeral disk on it :) | 18:20 |
| cloudnull | in the gate from rax we use general1-8 | 18:20 |
| cloudnull | I think the aio scripts assume >=60gb on / | 18:21 |
| Matias | https://github.com/openstack/openstack-ansible-os_nova/blob/master/templates/nova.conf.j2#L245 | 18:22 |
| *** electrofelix has quit IRC | 18:22 | |
| Matias | I don't understand why this didn't fail before.. maybe the lock didn't exist before the final release? | 18:22 |
| Drago | cloudnull: Looking to see if there's a way to skip the check | 18:23 |
| Drago | cloudnull: Because it should have enough space, just not report it correctly, right? | 18:23 |
| cloudnull | Drago: right | 18:23 |
| cloudnull | the check is in the bootstrap role | 18:23 |
| cloudnull | Matias: idk tbh. that seems suspect. | 18:24 |
| *** schwicht has quit IRC | 18:25 | |
| cloudnull | http://cdn.pasteraw.com/nrcvol99xbzvoh4f8omse2pbqehh2rc | 18:25 |
| cloudnull | a quick look through the roles and it looks like we define it a a few places. | 18:25 |
| *** d3n14l has quit IRC | 18:27 | |
| cloudnull | Matias: this looks like a bug in the way the init bits are being layed down | 18:28 |
| dxiri | where do I set DNS settings for the containers? | 18:28 |
| Drago | cloudnull: Instead of partitioning myself, can I just use bootstrap_host_data_disk_device? | 18:28 |
| Matias | cloudnull: do you have /var/lock/nova-compute in trusty? | 18:28 |
| Drago | It has parted commands in it | 18:29 |
| Matias | my /var/lock/nova-compute is empty, idk if it was supposed to exist at all | 18:29 |
| *** adrian_otto has quit IRC | 18:30 | |
| cloudnull | Drago: yea, that may work. | 18:30 |
| *** fops has joined #openstack-ansible | 18:30 | |
| cloudnull | Matias: i do have that dir however my env is an upgraded one so maybe existing from before which is why i dont see that issue. | 18:30 |
| cloudnull | I had a greenfield newton deploy yesterday but I began working on upgrade things so i dont have a vanilla one to test with | 18:31 |
| Matias | cloudnull: http://paste.openstack.org/show/586973/ the only reference to it is in tmpfiles.d, so maybe you are right and it shouldn't be constructed using the service name in this case | 18:33 |
| Matias | or lock_path should be changed in nova.conf | 18:33 |
| *** fops has quit IRC | 18:33 | |
| cloudnull | i'm thinking that the lock path should be it's own var. | 18:34 |
| cloudnull | based on what was changed here https://github.com/openstack/openstack-ansible-os_nova/commit/9a17ca682db8d9f90ae89004ae44c51024236ac6 | 18:38 |
| cloudnull | it looks like we can just change "/var/lock/{{ item.value.service_name }}" to "/var/lock/{{ SOMEVAR }}" | 18:39 |
| *** weezS has joined #openstack-ansible | 18:39 | |
| *** fops has joined #openstack-ansible | 18:39 | |
| *** fops_ has joined #openstack-ansible | 18:40 | |
| *** d3n14l has joined #openstack-ansible | 18:44 | |
| *** fops has quit IRC | 18:44 | |
| *** Matias has quit IRC | 18:45 | |
| *** mpjetta has joined #openstack-ansible | 18:47 | |
| *** Matias has joined #openstack-ansible | 18:48 | |
| *** fops has joined #openstack-ansible | 18:55 | |
| *** schwicht has joined #openstack-ansible | 18:56 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Resolve distro package index issue https://review.openstack.org/390635 | 18:56 |
| *** fops_ has quit IRC | 18:58 | |
| palendae | cloudnull, Ocata will have no Ansible 1.9 support, right? | 19:06 |
| cloudnull | it was dropped in newton | 19:07 |
| palendae | Completely? | 19:07 |
| cloudnull | completly | 19:07 |
| cloudnull | **completely | 19:07 |
| palendae | Ok, we'll probably want to remove the 'ansible_ssh_host' entry from inventory then | 19:07 |
| palendae | http://docs.ansible.com/ansible/intro_inventory.html - " | 19:07 |
| palendae | Ansible 2.0 has deprecated the “ssh” from ansible_ssh_user, ansible_ssh_host, and ansible_ssh_port to become ansible_user, ansible_host, and ansible_port. If you are using a version of Ansible prior to 2.0, you should continue using the older style variables (ansible_ssh_*). These shorter variables are ignored, without warning, in older versions of Ansible." | 19:07 |
| palendae | My guess is that's used in a bunch of places though | 19:08 |
| *** alij has quit IRC | 19:08 | |
| cloudnull | yes. to my knowledge it still works in 2.1 (maybe 2.2) | 19:09 |
| cloudnull | and we need to remove it from the following too http://cdn.pasteraw.com/epfi2etl1s0lv9mkza1lo9tva7a6h6r | 19:09 |
| palendae | Yeah, deprecated != gone | 19:09 |
| palendae | But we should probably clean house just to avoid confusion | 19:09 |
| cloudnull | agreed | 19:10 |
| palendae | And yes, that was my concern with deleting it outright | 19:10 |
| palendae | That it would be referenced in other playbooks or roles | 19:10 |
| palendae | Since Newton was the cut over, IMO it should retain both | 19:10 |
| palendae | But Ocata+ can drop | 19:10 |
| palendae | I'll try to start on that if I can get past some issues with my current patch | 19:11 |
| *** Jack_Iv has quit IRC | 19:11 | |
| cloudnull | sounds good | 19:11 |
| Matias | cloudnull: https://github.com/openstack/openstack-ansible-os_nova/blob/0d075aad354b4fd707810ac4de1b3e734a2ebb8a/tasks/nova_pre_install.yml#L73 | 19:14 |
| *** weezS has quit IRC | 19:15 | |
| Matias | cloudnull: the issue would be present only after reboot | 19:15 |
| Matias | maybe in 14.04 /var/lock is not a ramdisk, I don't remember | 19:15 |
| *** Drago has quit IRC | 19:18 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Updated for depricated "ansible_ssh_user" https://review.openstack.org/390636 | 19:19 |
| cloudnull | Matias: it's there after reboot. | 19:19 |
| cloudnull | this is something we're going to need to address. | 19:19 |
| *** esberglu has joined #openstack-ansible | 19:21 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add command to remove IPs from inventory https://review.openstack.org/390375 | 19:23 |
| stevelle | palendae: file a bug for the removal? | 19:27 |
| palendae | stevelle, That's a good idea :) | 19:28 |
| Matias | https://bugs.launchpad.net/openstack-ansible/+bug/1636604 | 19:28 |
| openstack | Launchpad bug 1636604 in openstack-ansible "Nova fails to launch any instances after the compute host is rebooted" [Undecided,New] | 19:28 |
| cloudnull | looks like most of the vars have been removed from the roles. | 19:28 |
| dxiri | how can I set a dns server and default gateway from within the container? | 19:31 |
| palendae | stevelle, https://bugs.launchpad.net/openstack-ansible/+bug/1636606 | 19:31 |
| openstack | Launchpad bug 1636606 in openstack-ansible "`ansible_ssh_*` variables deprecated in Ansible 2.x" [Low,New] | 19:31 |
| dxiri | isn't that done automatically? | 19:31 |
| *** tschacara has quit IRC | 19:31 | |
| *** Mudpuppy_ has joined #openstack-ansible | 19:32 | |
| dxiri | I thought the management bridge was only for management inside openstack and that it didn't require internet access :( | 19:32 |
| *** Mudpuppy has quit IRC | 19:35 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server: Remove depricated ansible_ssh_host variable https://review.openstack.org/390641 | 19:35 |
| esberglu | Hi. I am trying to get an openstack-ansible aio setup working. Right now my /etc/resolv.conf file is being overwritten when running the run-playbook.sh script with 127.0.0.1. Anyone aware of what might be causing this? | 19:35 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-haproxy_server: Remove depricated ansible_ssh_host variable https://review.openstack.org/390642 | 19:35 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops: Remove depricated ansible_ssh_host variable https://review.openstack.org/390643 | 19:36 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_aodh: Remove depricated ansible_ssh_host variable https://review.openstack.org/390644 | 19:36 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_barbican: Remove depricated ansible_ssh_host variable https://review.openstack.org/390645 | 19:36 |
| *** Mudpuppy_ has quit IRC | 19:36 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_ceilometer: Remove depricated ansible_ssh_host variable https://review.openstack.org/390646 | 19:36 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_cloudkitty: Remove depricated ansible_ssh_host variable https://review.openstack.org/390647 | 19:36 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_designate: Remove depricated ansible_ssh_host variable https://review.openstack.org/390648 | 19:36 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_ironic: Remove depricated ansible_ssh_host variable https://review.openstack.org/390649 | 19:37 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_magnum: Remove depricated ansible_ssh_host variable https://review.openstack.org/390650 | 19:37 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_rally: Remove depricated ansible_ssh_host variable https://review.openstack.org/390651 | 19:37 |
| *** johnmilton has quit IRC | 19:38 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_zaqar: Remove depricated ansible_ssh_host variable https://review.openstack.org/390652 | 19:38 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Remove depricated ansible_ssh_host variable https://review.openstack.org/390653 | 19:38 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-galera_server: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390641 | 19:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_aodh: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390644 | 19:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_barbican: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390645 | 19:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_ceilometer: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390646 | 19:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_cloudkitty: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390647 | 19:41 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_designate: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390648 | 19:42 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-haproxy_server: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390642 | 19:42 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_ironic: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390649 | 19:42 |
| *** tschacara has joined #openstack-ansible | 19:43 | |
| *** Drago has joined #openstack-ansible | 19:43 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_rally: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390651 | 19:43 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-ops: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390643 | 19:43 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_zaqar: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390652 | 19:43 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390653 | 19:43 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_magnum: Remove deprecated ansible_ssh_host variable https://review.openstack.org/390650 | 19:44 |
| cloudnull | sorry for the spam. | 19:44 |
| *** Drago has quit IRC | 19:45 | |
| *** Drago has joined #openstack-ansible | 19:45 | |
| cloudnull | palendae: https://review.openstack.org/#/q/topic:bug/1636606 resolves the rest of the roles however I didn't bulk job the integrated repo as it touches a bit more of the internals. | 19:45 |
| cloudnull | -cc stevelle | 19:45 |
| palendae | Cool | 19:46 |
| stevelle | will grind through those during some meeting | 19:46 |
| cloudnull | esberglu: interesting that it's writing 127.0.0.1 in the file. | 19:46 |
| cloudnull | what is the value prior to running that script? | 19:47 |
| Matias | cloudnull: the nova-api-metadata-container has /var/lock/nova owned by nova:nova, however I don't know what is creating the dir | 19:47 |
| Matias | (this even after a reboot) | 19:47 |
| esberglu | cloudnull: Our two internal nameservers. | 19:47 |
| *** tschacara has quit IRC | 19:48 | |
| cloudnull | Matias: maybe nova is creating that space automatically ? | 19:48 |
| Matias | yes, probably through the rootwrap? I'm not sure of what's happening though | 19:49 |
| Matias | there's a /var/lock/nova/nova-iptables file inside | 19:49 |
| Matias | need to go away now, i'll be back in a couple hours | 19:50 |
| *** weezS has joined #openstack-ansible | 19:50 | |
| *** tschacara has joined #openstack-ansible | 19:51 | |
| *** tschacara has quit IRC | 19:52 | |
| cloudnull | see ya in a while Matias | 19:52 |
| cloudnull | esberglu: are you running the unbound role? | 19:52 |
| cloudnull | how is this aio being setup / | 19:52 |
| cloudnull | **? | 19:53 |
| *** cathrichardson has quit IRC | 19:55 | |
| Drago | cloudnull: The BOOTSTRAP_OPTS looks to be working. OSA is working through the services | 19:55 |
| *** johnmilton has joined #openstack-ansible | 19:56 | |
| *** aludwar has quit IRC | 19:56 | |
| *** aludwar has joined #openstack-ansible | 19:56 | |
| *** javeriak has quit IRC | 19:57 | |
| esberglu | cloudnull: I followed the quickstart guide. So I run bootstrap-ansible.sh, then bootstrap-aio.sh. Then I add my specific variables to user_variables.yml. Then run run-playbooks.sh | 19:57 |
| esberglu | The only variables I added were as follows | 19:57 |
| esberglu | http://paste.openstack.org/show/586982/ | 19:58 |
| *** hughmFLEXin has quit IRC | 19:59 | |
| *** hughmFLEXin has joined #openstack-ansible | 19:59 | |
| *** cathrichardson has joined #openstack-ansible | 20:00 | |
| *** BjoernT has quit IRC | 20:03 | |
| *** javeriak has joined #openstack-ansible | 20:05 | |
| *** javeriak has quit IRC | 20:05 | |
| *** javeriak has joined #openstack-ansible | 20:05 | |
| *** alij has joined #openstack-ansible | 20:06 | |
| *** dxiri has quit IRC | 20:06 | |
| *** dxiri has joined #openstack-ansible | 20:07 | |
| cloudnull | best I can tell, the only role that would mess with that on a physical host would be the resolveconf role which is being pulled in from unbound. | 20:09 |
| *** alij has quit IRC | 20:10 | |
| esberglu | cloudnull: That was my thinking as well. I tried adding | 20:10 |
| cloudnull | which would not run unless there were hosts within the unbound_all group | 20:10 |
| esberglu | resolvconf_reslover_ips: xxxx | 20:10 |
| Drago | cloudnull: A question - how do you override variables in inventory/group_vars? | 20:11 |
| cloudnull | Drago: whatever the var is that you want to set redefine it within the /etc/openstack_deploy/user_variables.yml file | 20:11 |
| esberglu | to my config which is what the resolvconf role uses but no luck. I will take a look and see if this is coming from unbound | 20:12 |
| cloudnull | esberglu: do you by chance have verbose task output we could look through ? | 20:12 |
| Drago | I put openstack_service_publicuri_proto: http into user_variables before run-playbooks ran, but it doesn't seem to have had an effect | 20:12 |
| Drago | Well, I mean it sort of worked | 20:13 |
| Drago | Let me check something | 20:13 |
| cloudnull | if this is an AIO and it was done by the playbooks there should be a log under /openstack/log/ansible-logging/ansible.log | 20:13 |
| cloudnull | esberglu: ^ | 20:13 |
| Drago | cloudnull: Oh, haproxy_ssl is what I forgot | 20:15 |
| *** javeriak has quit IRC | 20:16 | |
| esberglu | cloudnull: I changed my resolvconf back and reran some of the playbooks. Looks like the output from this issue is no longer there. | 20:17 |
| cloudnull | interesting. | 20:17 |
| cathrichardson | /msg NickServ identify <password> | 20:17 |
| *** javeriak has joined #openstack-ansible | 20:17 | |
| esberglu | I am planning on recreating this either tonight or tomorrow and will be sure to keep the logs this time | 20:17 |
| cloudnull | cathrichardson: access denied. please try again | 20:18 |
| cloudnull | :) | 20:18 |
| cloudnull | esberglu: let us know what you find out. | 20:18 |
| *** berendt has joined #openstack-ansible | 20:18 | |
| *** Hosam has joined #openstack-ansible | 20:18 | |
| cathrichardson | cloudnull: Just ignore me/that! | 20:19 |
| * cloudnull nothing to see, moving along | 20:20 | |
| esberglu | cloudnull: Is there a config option for verbose output through the run-playbooks.sh script? Or what's the way to go about doing that? | 20:20 |
| cloudnull | esberglu: it this is all just an AIO i'd run the gate-check-commit.sh script. | 20:20 |
| *** cathrichardson has quit IRC | 20:20 | |
| cloudnull | it's the one stop shop for running an aio | 20:20 |
| *** cathrichardson has joined #openstack-ansible | 20:21 | |
| cloudnull | if you need to do something custom I'd avoid the run-playbooks script and just use the playbooks directly. | 20:21 |
| cloudnull | I know the docs call out using run-playbooks.sh but gate-check-commit.sh will do all that for you. | 20:22 |
| esberglu | cloudnull: Okay. Thanks for the help and advice! | 20:23 |
| cloudnull | anytime | 20:24 |
| Drago | cloudnull: Another question. What is the difference between, say `openstack-ansible os-heat-install.yml` and `openstack-ansible setup-everything.yml --limit heat_all`? | 20:25 |
| cloudnull | in the end, nothing; for that specific usecase. | 20:26 |
| Drago | cloudnull: So the scope would be more limited if it were setup-openstack.yml? | 20:27 |
| cloudnull | however if you were to run openstack-ansible setup-everything.yml --limit nova_compute you'd run the neutron and nova playbooks as neutron and nova share hosts | 20:27 |
| cloudnull | yes the limit will scope the the group | 20:27 |
| cloudnull | you can achieve some of the same stuff with tags too. which would be *NAME, *NAME-config, *NAME-install | 20:28 |
| Drago | cloudnull: Okay | 20:28 |
| Drago | cloudnull: So if I have a config-only change, I can use --limit heat-config? | 20:29 |
| cloudnull | so say you wanted to reconfigure heat, you could run openstack-ansible os-heat-install.yml --tags heat-config | 20:29 |
| Drago | Oh, —tags, got it | 20:29 |
| cloudnull | last one, but say you wanted to scope everything to neutron and you didn't know which host group it was in you could use tags; like so: openstack-ansible setup-everything.yml --tags neutron | 20:30 |
| cloudnull | that would rip through all of the playbooks and do only thing things that are tagged neutron | 20:31 |
| cloudnull | **only the things... | 20:31 |
| Drago | cloudnull: Ooh, that's cool! | 20:32 |
| Drago | cloudnull: Is there a list of nifty things in OSA for people that are not familiar with it, or possibly even with Ansible? | 20:32 |
| *** javeriak has quit IRC | 20:35 | |
| cloudnull | idk if there's anything specific within the docs that denote that. | 20:37 |
| *** admin0 has joined #openstack-ansible | 20:42 | |
| dxiri | seeing this on "repo_build : Create OpenStack-Ansible requirement wheels" is this normal? | 20:46 |
| dxiri | Could not find a tag or branch '4404ae6a7e2b8d96db102ada8aaf07cbc92e256b', assuming commit. | 20:46 |
| *** Hosam has quit IRC | 20:57 | |
| *** Hosam_ has joined #openstack-ansible | 20:57 | |
| *** v1k0d3n has joined #openstack-ansible | 20:59 | |
| *** jheroux has quit IRC | 21:03 | |
| dxiri | any ideas on how can I fix this? | 21:03 |
| dxiri | http://paste.ubuntu.com/23380841/ | 21:03 |
| dxiri | for some reason pyyaml doesn't want to instal | 21:03 |
| dxiri | install | 21:03 |
| dxiri | on the repo container | 21:03 |
| Drago | dxiri: Are you trying to install a custom version of something? | 21:03 |
| dxiri | nope | 21:03 |
| dxiri | I just saw ansible fail at that point | 21:04 |
| dxiri | and since the stdout its garbled inside the ansible command I tried it myself | 21:04 |
| *** askb has joined #openstack-ansible | 21:05 | |
| dxiri | I have tried removing /tmp/openstack_builder/ folder and re-running the command that ansible runs, but it always fails on pyyaml | 21:05 |
| dxiri | the rest are installed ok | 21:05 |
| dxiri | pip wheel --timeout 120 --wheel-dir /tmp/openstack-wheel-output --find-links /var/www/repo/links --find-links /tmp/openstack-wheel-output --constraint /var/www/repo/os-releases/14.0.0/requirements_constraints.txt --build /tmp/openstack-builder --log /var/log/repo/repo_builder.log --requirement /var/www/repo/os-releases/14.0.0/requirements.txt | 21:05 |
| dxiri | that is the full command ansible is sending over | 21:05 |
| *** thorst_ has joined #openstack-ansible | 21:09 | |
| cloudnull | dxiri: the only time i've seen something like that was when the system pip was using python-pip from an apt package. | 21:09 |
| cloudnull | which was version 1.5.6 | 21:10 |
| cloudnull | however if that's executing in a container it should be pip from upstream | 21:10 |
| dxiri | cloudnull: so do you suggest I reinstall pip? | 21:12 |
| dxiri | on this container? | 21:12 |
| cloudnull | i wouldn't think that's needed. | 21:12 |
| cloudnull | can you pip install pyyaml in a venv manually ? | 21:12 |
| cloudnull | does that produce the same error ? | 21:12 |
| cloudnull | according to https://pypi.python.org/pypi/PyYAML/3.12 | 21:13 |
| cloudnull | that version has been there since august so i'd not assume that was the issue either. | 21:13 |
| *** thorst_ has quit IRC | 21:13 | |
| dxiri | root@infra1-repo-container-e85f8cd9:/tmp/openstack-builder/pyyaml/test# bin/pip install pyyaml | 21:14 |
| dxiri | Collecting pyyaml | 21:14 |
| dxiri | Using cached PyYAML-3.12.tar.gz | 21:14 |
| dxiri | Building wheels for collected packages: pyyaml | 21:14 |
| dxiri | Running setup.py bdist_wheel for pyyaml ... done | 21:14 |
| dxiri | Stored in directory: /root/.cache/pip/wheels/2c/f7/79/13f3a12cd723892437c0cfbde1230ab4d82947ff7b3839a4fc | 21:14 |
| dxiri | Successfully built pyyaml | 21:14 |
| dxiri | Installing collected packages: pyyaml | 21:14 |
| dxiri | Successfully installed pyyaml-3.12 | 21:14 |
| dxiri | inside virtualenv works ok | 21:14 |
| dxiri | I just did virtualenv test, then cd to test dir, then bin/pip install pyyaml | 21:14 |
| cloudnull | same test w/ a new venv | 21:15 |
| cloudnull | pip wheel --timeout 120 --wheel-dir /tmp/openstack-wheel-output --find-links /var/www/repo/links --find-links /tmp/openstack-wheel-output pyyaml | 21:15 |
| cloudnull | from within the reop container | 21:15 |
| *** esberglu has left #openstack-ansible | 21:16 | |
| dxiri | root@infra1-repo-container-e85f8cd9:/tmp/openstack-builder/pyyaml/test2# ./bin/pip wheel --timeout 120 --wheel-dir /tmp/openstack-wheel-output --find-links /var/www/repo/links --find-links /tmp/openstack-wheel-output pyyaml | 21:19 |
| dxiri | Collecting pyyaml | 21:19 |
| dxiri | File was already downloaded /tmp/openstack-wheel-output/PyYAML-3.12-cp27-cp27mu-linux_x86_64.whl | 21:19 |
| dxiri | Skipping pyyaml, due to already being wheel. | 21:19 |
| *** h5t4 has quit IRC | 21:21 | |
| cloudnull | can you do the same w/ an install | 21:21 |
| cloudnull | ./bin/pip install --timeout 120 --wheel-dir /tmp/openstack-wheel-output --find-links /var/www/repo/links --find-links /tmp/openstack-wheel-output pyyaml | 21:21 |
| openstackgerrit | Michael Carden proposed openstack/openstack-ansible-plugins: Updated for deprecated "ansible_ssh_user" https://review.openstack.org/390636 | 21:22 |
| *** thorst_ has joined #openstack-ansible | 21:22 | |
| dxiri | interesting got "no such option: --wheel-dir" | 21:24 |
| dxiri | probably install does not have a --wheel-dir argument? | 21:25 |
| *** Hosam_ has quit IRC | 21:28 | |
| dxiri | will run the exact same command ansible is trying to run but on the virtualenv, one min | 21:28 |
| *** schwicht has quit IRC | 21:31 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add command to remove IPs from inventory https://review.openstack.org/390375 | 21:35 |
| dxiri | ./bin/pip wheel --timeout 120 --wheel-dir /tmp/openstack-wheel-output --find-links /var/www/repo/links --find-links /tmp/openstack-wheel-output --constraint /var/www/repo/os-releases/14.0.0/requirements_constraints.txt --build /tmp/openstack-builder --log /var/log/repo/repo_builder.log --requirement /var/www/repo/os-releases/14.0.0/requirements.txt | 21:35 |
| dxiri | ran that...completed without issues | 21:36 |
| *** alij has joined #openstack-ansible | 21:36 | |
| cloudnull | rerun repo-build and see if its still complaining? | 21:37 |
| *** v1k0d3n has quit IRC | 21:37 | |
| dxiri | re-run repo build? do you mean the ansible task? | 21:38 |
| dxiri | how? | 21:38 |
| cloudnull | openstack-ansible repo-build.myl | 21:39 |
| cloudnull | openstack-ansible repo-build.yml | 21:39 |
| *** alij has quit IRC | 21:40 | |
| *** retreved_ has quit IRC | 21:41 | |
| *** v1k0d3n has joined #openstack-ansible | 21:42 | |
| *** schwicht has joined #openstack-ansible | 21:47 | |
| *** v1k0d3n has quit IRC | 21:47 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Remove ansible_ssh_host variable from inventory https://review.openstack.org/390672 | 21:49 |
| palendae | Ah poop needed the bug reference | 21:49 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Remove ansible_ssh_host variable from inventory https://review.openstack.org/390672 | 21:54 |
| *** agireud has quit IRC | 21:54 | |
| *** cathrichardson has quit IRC | 21:54 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins: Resolve distro package index issue https://review.openstack.org/390635 | 21:54 |
| palendae | Oh sweet, Gerrit fixes the branch now | 21:55 |
| *** berendt has quit IRC | 21:56 | |
| *** alij has joined #openstack-ansible | 21:57 | |
| *** schwicht has quit IRC | 21:59 | |
| *** alij has quit IRC | 22:02 | |
| *** chris_hultin is now known as chris_hultin|AWA | 22:02 | |
| *** schwicht has joined #openstack-ansible | 22:07 | |
| *** d3n14l has quit IRC | 22:25 | |
| *** alij has joined #openstack-ansible | 22:29 | |
| *** schwicht has quit IRC | 22:33 | |
| *** weezS has quit IRC | 22:33 | |
| *** alij has quit IRC | 22:34 | |
| *** javeriak has joined #openstack-ansible | 22:39 | |
| *** agireud has joined #openstack-ansible | 22:44 | |
| *** agireud has quit IRC | 22:48 | |
| *** agireud has joined #openstack-ansible | 22:50 | |
| *** agireud has quit IRC | 22:53 | |
| *** javeriak has quit IRC | 22:54 | |
| *** javeriak has joined #openstack-ansible | 22:54 | |
| *** agireud has joined #openstack-ansible | 22:56 | |
| *** agireud has quit IRC | 22:57 | |
| *** alij has joined #openstack-ansible | 22:57 | |
| *** agireud has joined #openstack-ansible | 22:58 | |
| *** javeriak has quit IRC | 22:58 | |
| *** schwicht has joined #openstack-ansible | 23:00 | |
| *** Mudpuppy has joined #openstack-ansible | 23:01 | |
| *** alij has quit IRC | 23:02 | |
| *** klamath has quit IRC | 23:02 | |
| *** schwicht has quit IRC | 23:11 | |
| *** Mudpuppy has quit IRC | 23:12 | |
| *** Mudpuppy has joined #openstack-ansible | 23:13 | |
| *** Drago has quit IRC | 23:15 | |
| c-mart | may be a dumb question, does anyone know how to an OSA-deployed OpenStack cloud to generate an admin openrc file? | 23:16 |
| c-mart | the openrc files in the utility containers are configured for internal URLs. I need externally accessible URLs | 23:16 |
| *** schwicht has joined #openstack-ansible | 23:16 | |
| *** Mudpuppy has quit IRC | 23:17 | |
| *** agrebennikov has quit IRC | 23:17 | |
| stevelle | c-mart: log into horizon as admin and download one? | 23:19 |
| *** hughmFLEXin has quit IRC | 23:23 | |
| *** hughmFLEXin has joined #openstack-ansible | 23:24 | |
| *** agireud has quit IRC | 23:24 | |
| c-mart | stevelle, I see a place to get an openrc under the "Compute" project, and that gets me closer but I only see OS_AUTH_URL, not OS_ADMIN_URL. also, when I try to use that openrc file to do something like `openstack user list` on the command line, it tries to redirect to an internal URL (on the container management network), which doesn't work from outside the cluster | 23:28 |
| c-mart | so, if there's another place to get an openrc file for administrative purposes, I'd be keen to know about it | 23:28 |
| *** hughmFLEXin has quit IRC | 23:28 | |
| stevelle | thinking about it, I'm pretty sure that admin urls are not being set up to work from external | 23:29 |
| *** agireud has joined #openstack-ansible | 23:30 | |
| stevelle | I'm not sure I'd want them to work except on mgmt network anyway | 23:30 |
| c-mart | well, my "external" URL is still only internally reachable | 23:31 |
| dxiri | do I need my external IP to be tied to a specific interface for haproxy to work? | 23:31 |
| dxiri | getting this on the haproxy.log | 23:31 |
| dxiri | Oct 25 16:10:47 sjociara11 haproxy-systemd-wrapper[10069]: [ALERT] 298/161047 (10070) : Starting frontend repo_git-front-1: cannot bind socket [172.25.190.253:9418] | 23:31 |
| stevelle | c-mart: I'm strange perhaps but I'd still not expose it externally. if someone needs admin cmds they can probably issue them from the deploy host or util container | 23:32 |
| stevelle | but I think the endpoints are just not set up for it, which might be what you're seeing happen on the redirect | 23:33 |
| palendae | iirc the catalog is set up to only have the admin URLs on interal IPs | 23:34 |
| *** agireud has quit IRC | 23:34 | |
| c-mart | my application that interacts with OpenStack (and needs to issue admin API calls) is on another subnet, in another rack. any idea if that behavior is easy to reconfigure? | 23:35 |
| *** agireud has joined #openstack-ansible | 23:35 | |
| stevelle | c-mart: I'd be guessing to answer, but might be as simple as $ openstack endpoint create ... | 23:36 |
| palendae | https://github.com/openstack/openstack-ansible-os_keystone/blob/4cb21da177af4c889832a7556256601053820089/tasks/keystone_service_setup.yml#L39 - set that variable | 23:36 |
| palendae | Or yeah, create a new endpoint | 23:36 |
| palendae | keystone_service_adminurl is the one that's set by OSA | 23:37 |
| c-mart | thank you palendae and stevelle | 23:38 |
| stevelle | dxiri: no straight forward hard coding or anything I could point to. | 23:38 |
| *** thorst_ has quit IRC | 23:39 | |
| *** thorst has joined #openstack-ansible | 23:40 | |
| *** v1k0d3n has joined #openstack-ansible | 23:44 | |
| *** thorst has quit IRC | 23:45 | |
| *** v1k0d3n has quit IRC | 23:49 | |
| c-mart | this is not strictly an OSA-related question - but is there a sane way to browse the OpenStack docs? Like, limit a search to docs within a specific release, or view an entire table of contents in the navigation pane? :) | 23:52 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!