| *** retreved has quit IRC | 00:04 | |
| *** olivier427 has joined #openstack-ansible | 00:06 | |
| *** skamithi has joined #openstack-ansible | 00:24 | |
| *** pegmanm_ has joined #openstack-ansible | 00:29 | |
| *** raddaoui has quit IRC | 00:30 | |
| *** pegmanm has quit IRC | 00:31 | |
| *** jthorne has joined #openstack-ansible | 00:36 | |
| *** jthorne has quit IRC | 00:38 | |
| *** eil397 has quit IRC | 00:41 | |
| *** weezS has quit IRC | 00:52 | |
| *** jamielennox|away is now known as jamielennox | 00:55 | |
| *** skamithi has quit IRC | 01:08 | |
| *** baker has joined #openstack-ansible | 01:12 | |
| *** skamithi has joined #openstack-ansible | 01:13 | |
| *** markvoelker has joined #openstack-ansible | 01:15 | |
| *** markvoelker has quit IRC | 01:20 | |
| *** kjelly has joined #openstack-ansible | 01:20 | |
| *** skamithi has left #openstack-ansible | 01:21 | |
| *** jamielennox is now known as jamielennox|away | 01:23 | |
| *** oneswig has joined #openstack-ansible | 01:27 | |
| *** pegmanm has joined #openstack-ansible | 01:30 | |
| *** oneswig has quit IRC | 01:32 | |
| *** pegmanm_ has quit IRC | 01:33 | |
| *** daneyon_ has quit IRC | 01:36 | |
| *** daneyon has joined #openstack-ansible | 01:36 | |
| *** kjelly has quit IRC | 01:39 | |
| *** marekd has quit IRC | 01:40 | |
| *** marekd has joined #openstack-ansible | 01:40 | |
| *** izaakk has quit IRC | 01:53 | |
| *** baker has quit IRC | 01:58 | |
| *** olivier427 has quit IRC | 02:01 | |
| *** baker has joined #openstack-ansible | 02:03 | |
| *** baker_ has joined #openstack-ansible | 02:04 | |
| *** phiche1 has joined #openstack-ansible | 02:06 | |
| *** baker has quit IRC | 02:07 | |
| *** woodard has quit IRC | 02:08 | |
| *** phiche has quit IRC | 02:10 | |
| *** olivier427 has joined #openstack-ansible | 02:13 | |
| *** markvoelker has joined #openstack-ansible | 02:16 | |
| *** markvoelker has quit IRC | 02:20 | |
| *** baker_ has quit IRC | 02:25 | |
| *** zaze89 has joined #openstack-ansible | 02:25 | |
| *** baker has joined #openstack-ansible | 02:26 | |
| *** skamithi has joined #openstack-ansible | 02:27 | |
| *** pegmanm_ has joined #openstack-ansible | 02:32 | |
| *** baker has quit IRC | 02:33 | |
| *** pegmanm has quit IRC | 02:34 | |
| *** jamielennox|away is now known as jamielennox | 02:44 | |
| *** skamithi has left #openstack-ansible | 02:45 | |
| *** baker has joined #openstack-ansible | 02:59 | |
| *** skamithi has joined #openstack-ansible | 03:01 | |
| *** skamithi has left #openstack-ansible | 03:01 | |
| *** baker has quit IRC | 03:05 | |
| *** baker has joined #openstack-ansible | 03:06 | |
| *** zaze89 has quit IRC | 03:07 | |
| *** baker_ has joined #openstack-ansible | 03:08 | |
| *** skamithi has joined #openstack-ansible | 03:09 | |
| *** zhangjn has joined #openstack-ansible | 03:09 | |
| *** baker has quit IRC | 03:11 | |
| *** skamithi has quit IRC | 03:11 | |
| *** skamithi has joined #openstack-ansible | 03:11 | |
| *** baker has joined #openstack-ansible | 03:16 | |
| *** oneswig has joined #openstack-ansible | 03:16 | |
| *** skamithi has quit IRC | 03:16 | |
| *** olivier427 has quit IRC | 03:17 | |
| *** baker_ has quit IRC | 03:19 | |
| *** skamithi has joined #openstack-ansible | 03:19 | |
| *** oneswig has quit IRC | 03:20 | |
| *** skamithi has quit IRC | 03:23 | |
| *** olivier427 has joined #openstack-ansible | 03:24 | |
| *** skamithi has joined #openstack-ansible | 03:31 | |
| *** pegmanm has joined #openstack-ansible | 03:32 | |
| *** skamithi has quit IRC | 03:32 | |
| *** skamithi has joined #openstack-ansible | 03:33 | |
| *** skamithi has quit IRC | 03:33 | |
| *** pegmanm_ has quit IRC | 03:34 | |
| *** skamithi has joined #openstack-ansible | 03:34 | |
| *** sdake has joined #openstack-ansible | 03:35 | |
| *** coolg has joined #openstack-ansible | 03:44 | |
| *** skamithi has left #openstack-ansible | 03:52 | |
| coolg | For liberty installation using openstack-ansible, whether the container address should be public ip? | 03:54 |
|---|---|---|
| *** CheKoLyN has joined #openstack-ansible | 03:55 | |
| coolg | I have used the same subnet Container management: 172.29.236.0/22 Tunnel: 172.29.240.0/22 Storage: 172.29.244.0/22 for my installation | 03:56 |
| coolg | which is given in installlation guide | 03:57 |
| *** baker has quit IRC | 03:59 | |
| *** skamithi has joined #openstack-ansible | 04:00 | |
| *** skamithi has quit IRC | 04:04 | |
| *** skamithi has joined #openstack-ansible | 04:05 | |
| *** cemmason has joined #openstack-ansible | 04:07 | |
| *** jamielennox is now known as jamielennox|away | 04:10 | |
| *** skamithi has left #openstack-ansible | 04:14 | |
| *** markvoelker has joined #openstack-ansible | 04:17 | |
| *** javeriak has joined #openstack-ansible | 04:18 | |
| *** markvoelker has quit IRC | 04:21 | |
| *** javeriak_ has joined #openstack-ansible | 04:23 | |
| *** grumpycatt has quit IRC | 04:23 | |
| *** javeriak has quit IRC | 04:24 | |
| *** grumpycatt has joined #openstack-ansible | 04:26 | |
| *** skamithi has joined #openstack-ansible | 04:32 | |
| *** pegmanm_ has joined #openstack-ansible | 04:32 | |
| *** pegmanm has quit IRC | 04:35 | |
| *** jamielennox|away is now known as jamielennox | 04:39 | |
| *** coolg has quit IRC | 04:44 | |
| *** zhangjn has quit IRC | 04:46 | |
| *** zhangjn has joined #openstack-ansible | 04:47 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Add support for the nova_api db https://review.openstack.org/274932 | 04:48 |
| *** oneswig has joined #openstack-ansible | 05:04 | |
| *** sdake has quit IRC | 05:05 | |
| *** oneswig has quit IRC | 05:09 | |
| *** jamielennox is now known as jamielennox|away | 05:13 | |
| *** pegmanm has joined #openstack-ansible | 05:34 | |
| *** pegmanm_ has quit IRC | 05:36 | |
| *** javeriak has joined #openstack-ansible | 05:44 | |
| *** javeriak_ has quit IRC | 05:47 | |
| *** javeriak has quit IRC | 05:52 | |
| *** skamithi has quit IRC | 05:53 | |
| *** CheKoLyN has quit IRC | 06:00 | |
| *** javeriak has joined #openstack-ansible | 06:01 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Add support for the nova_api db https://review.openstack.org/274932 | 06:05 |
| *** fawadkhaliq has joined #openstack-ansible | 06:11 | |
| *** javeriak has quit IRC | 06:11 | |
| *** fawadkhaliq has quit IRC | 06:11 | |
| *** markvoelker has joined #openstack-ansible | 06:18 | |
| *** javeriak has joined #openstack-ansible | 06:21 | |
| *** javeriak has quit IRC | 06:21 | |
| *** markvoelker has quit IRC | 06:22 | |
| *** pegmanm_ has joined #openstack-ansible | 06:34 | |
| *** cemmason has quit IRC | 06:36 | |
| *** pegmanm has quit IRC | 06:37 | |
| *** cemmason has joined #openstack-ansible | 06:48 | |
| *** javeriak has joined #openstack-ansible | 06:51 | |
| *** oneswig has joined #openstack-ansible | 06:52 | |
| *** olivier427 has quit IRC | 06:54 | |
| *** zhangjn has quit IRC | 06:54 | |
| *** phiche1 has quit IRC | 06:55 | |
| *** oneswig has quit IRC | 06:57 | |
| *** keedya has joined #openstack-ansible | 07:04 | |
| *** phiche has joined #openstack-ansible | 07:05 | |
| *** javeriak has quit IRC | 07:14 | |
| *** keedya has quit IRC | 07:26 | |
| *** javeriak has joined #openstack-ansible | 07:30 | |
| *** blackstar257 has joined #openstack-ansible | 07:36 | |
| *** pegmanm has joined #openstack-ansible | 07:37 | |
| *** javeriak_ has joined #openstack-ansible | 07:37 | |
| *** javeriak has quit IRC | 07:38 | |
| *** pegmanm_ has quit IRC | 07:38 | |
| *** sdake has joined #openstack-ansible | 08:09 | |
| *** agireud has quit IRC | 08:10 | |
| *** agireud has joined #openstack-ansible | 08:12 | |
| *** markvoelker has joined #openstack-ansible | 08:18 | |
| *** mikelk has joined #openstack-ansible | 08:22 | |
| *** markvoelker has quit IRC | 08:22 | |
| *** tlbr_ has quit IRC | 08:23 | |
| *** pegmanm_ has joined #openstack-ansible | 08:35 | |
| *** pegmanm has quit IRC | 08:38 | |
| *** oneswig has joined #openstack-ansible | 08:41 | |
| *** tlbr has joined #openstack-ansible | 08:42 | |
| *** oneswig has quit IRC | 08:45 | |
| *** blackstar257 has quit IRC | 08:49 | |
| *** marekd has quit IRC | 08:50 | |
| *** oneswig has joined #openstack-ansible | 08:51 | |
| *** marekd has joined #openstack-ansible | 08:52 | |
| *** marekd has quit IRC | 09:01 | |
| *** zhangjn has joined #openstack-ansible | 09:04 | |
| *** zhangjn has quit IRC | 09:04 | |
| *** blackstar257 has joined #openstack-ansible | 09:07 | |
| *** zhangjn has joined #openstack-ansible | 09:09 | |
| *** gparaskevas has joined #openstack-ansible | 09:11 | |
| *** mgoddard has joined #openstack-ansible | 09:30 | |
| *** blackstar257 has quit IRC | 09:35 | |
| *** grumpycatt has quit IRC | 09:41 | |
| odyssey4me | mattt hughsaunders git-harry andymccr FYI, interesting pattern: https://github.com/openstack/ansible-role-diskimage-builder | 09:44 |
| *** grumpycatt has joined #openstack-ansible | 09:47 | |
| *** zhangjn has quit IRC | 09:57 | |
| *** stian_ has joined #openstack-ansible | 10:09 | |
| *** zhangjn has joined #openstack-ansible | 10:11 | |
| *** grumpycatt has quit IRC | 10:14 | |
| *** andrei has quit IRC | 10:15 | |
| *** openstackgerrit has quit IRC | 10:17 | |
| *** openstackgerrit has joined #openstack-ansible | 10:17 | |
| *** markvoelker has joined #openstack-ansible | 10:19 | |
| *** markvoelker has quit IRC | 10:23 | |
| *** javeriak_ has quit IRC | 10:29 | |
| stian_ | Hi! What is the best way to rebuild a container on a different node? Example: I want to destroy galera on node 3 and rebuild it on node 4. | 10:30 |
| *** grumpycatt has joined #openstack-ansible | 10:31 | |
| *** evrardjp_ is now known as evrardjp | 10:34 | |
| *** javeriak has joined #openstack-ansible | 10:38 | |
| *** pegmanm has joined #openstack-ansible | 10:41 | |
| *** pegmanm_ has quit IRC | 10:43 | |
| *** stian_ has quit IRC | 10:52 | |
| *** electrofelix has joined #openstack-ansible | 10:53 | |
| *** zhangjn has quit IRC | 10:57 | |
| *** zhangjn has joined #openstack-ansible | 10:58 | |
| *** zhangjn has quit IRC | 10:59 | |
| *** zhangjn has joined #openstack-ansible | 10:59 | |
| *** zhangjn has quit IRC | 11:00 | |
| *** zhangjn has joined #openstack-ansible | 11:00 | |
| *** zhangjn has quit IRC | 11:01 | |
| *** cemmason has quit IRC | 11:01 | |
| *** stian_ has joined #openstack-ansible | 11:02 | |
| *** zhangjn has joined #openstack-ansible | 11:06 | |
| *** zhangjn has quit IRC | 11:06 | |
| odyssey4me | stian_ you simply add the group to your openstack_user_config.yml or to conf.d with the host in that group, as described in https://github.com/openstack/ansible-role-diskimage-builder | 11:06 |
| odyssey4me | oops - the right url is https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.example | 11:06 |
| odyssey4me | hughsaunders andymccr mattt looking for reviews for https://review.openstack.org/272723 please | 11:10 |
| *** cemmason has joined #openstack-ansible | 11:12 | |
| odyssey4me | stian_ it's probably best to setup the new container on the new host, then once that's up and working properly, remove the old container from the old host | 11:13 |
| *** zhangjn has joined #openstack-ansible | 11:13 | |
| *** zhangjn has quit IRC | 11:13 | |
| *** zhangjn has joined #openstack-ansible | 11:14 | |
| *** zhangjn has quit IRC | 11:14 | |
| stian_ | odyssey4me : What abount the inventory ? The old node is still defined there and is rebuilt when I run the playbook. | 11:15 |
| *** swati_ has joined #openstack-ansible | 11:15 | |
| *** zhangjn has joined #openstack-ansible | 11:15 | |
| odyssey4me | stian_ you can use the inventory-manage script to remove entries from the inventory | 11:15 |
| *** zhangjn has quit IRC | 11:15 | |
| *** zhangjn has joined #openstack-ansible | 11:16 | |
| odyssey4me | stian_ if you've used a host group that includes the container group among others, then you may have to set the affinity for that particular host down to 0 | 11:16 |
| stian_ | odyssey4me : ahh, that is exactly what I was looking for. Thank you very much! | 11:16 |
| *** zhangjn has quit IRC | 11:16 | |
| odyssey4me | stian_ for example, this is setting affinity: https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/openstack_user_config.yml.aio#L71-L77 | 11:17 |
| odyssey4me | in this case, though, you'd be setting the affinity to 0 instead of 3 | 11:17 |
| *** zhangjn has joined #openstack-ansible | 11:17 | |
| *** zhangjn has quit IRC | 11:17 | |
| odyssey4me | in that example all containers from the shared-infra_hosts group (which contains galera, rabbit, etc) would have the standard default affinity of 1 - but then rabbit and galera have affinities of 3 (ie 3 containers for each host) | 11:18 |
| stian_ | odyssey4me : I tried to set affinity to 1 galera container on node 1 and 2 galera containers on node 2. That worked fine. But when I set affinity back to 1 on node 2, it still built 2 an extra container on node2. | 11:19 |
| *** zhangjn has joined #openstack-ansible | 11:19 | |
| *** zhangjn has quit IRC | 11:19 | |
| odyssey4me | stian_ you never need to set an affinity of 1 - that's the default | 11:20 |
| odyssey4me | but yes, once you change the affinity you'll also need to remove any existing container from the inventory | 11:20 |
| *** zhangjn has joined #openstack-ansible | 11:20 | |
| *** zhangjn has quit IRC | 11:20 | |
| stian_ | Yes, the problem was that it still existed in the inventory. | 11:20 |
| stian_ | Thanks! | 11:21 |
| *** zhangjn has joined #openstack-ansible | 11:21 | |
| *** zhangjn has quit IRC | 11:21 | |
| odyssey4me | stian_ sure :) | 11:23 |
| *** zhangjn has joined #openstack-ansible | 11:24 | |
| *** zhangjn has quit IRC | 11:24 | |
| *** zhangjn has joined #openstack-ansible | 11:25 | |
| *** javeriak has quit IRC | 11:25 | |
| *** javeriak has joined #openstack-ansible | 11:27 | |
| swati_ | Hi Jesse, how are openstack components installed using ansible....like is nova installed using venv? | 11:27 |
| *** agireud has quit IRC | 11:27 | |
| *** agireud has joined #openstack-ansible | 11:29 | |
| odyssey4me | swati_ all openstack components are now, by default, installed into a venv - assuming the deployer has left the default in place | 11:34 |
| odyssey4me | swati_ I sent an email earlier this year suggesting that we import your current work into an openstack repo, then we can work on finalising it together | 11:35 |
| odyssey4me | I can then pair you up with one of the team who understands the framework well enough to help with that finalisation | 11:36 |
| *** oneswig has quit IRC | 11:39 | |
| *** pegmanm_ has joined #openstack-ansible | 11:40 | |
| *** pegmanm has quit IRC | 11:43 | |
| *** cemmason has quit IRC | 11:45 | |
| *** gus_ is now known as gus | 11:47 | |
| *** oneswig has joined #openstack-ansible | 11:54 | |
| *** skamithi has joined #openstack-ansible | 12:03 | |
| javeriak | hey folks, the galera-bootstrap tag doesnt seem to be available on kilo anymore? | 12:04 |
| odyssey4me | javeriak I'm not sure if it ever was - that's a carry over from Juno and I think the mechanism for that was changed for kilo. | 12:05 |
| odyssey4me | the docs need an update | 12:05 |
| javeriak | oh, what tag should we use then to rekick galera? | 12:05 |
| odyssey4me | javeriak try this? http://docs.openstack.org/developer/openstack-ansible/developer-docs/quickstart-aio.html#rebooting-an-aio | 12:05 |
| javeriak | odyssey4me but rerunning the infra playbook sometimes fails again even after doing that | 12:06 |
| odyssey4me | javeriak then you'll need to actually look into what's broken in the galera cluster | 12:06 |
| hughsaunders | odyssey4me: re linting patch, the gate-check-docs script has gone, so will there be a docs job to call the docs tox env? | 12:06 |
| odyssey4me | another option, if you're ok with it, is to try trashing the containers and rebuilding them | 12:07 |
| odyssey4me | the data stays, as that's bind-mounted to the host | 12:07 |
| javeriak | where does this come from 'wsrep_incomming_adress does not match accross cluster'? I only managed to sort this out by deleting the cotainers and rebuilding | 12:07 |
| odyssey4me | hughsaunders the docs job has been using the docs tox target for ages | 12:07 |
| odyssey4me | hughsaunders the gate-check-docs scripts hasn't been in use for a long, long time | 12:07 |
| odyssey4me | javeriak hmm, that's odd - sounds like a busted up inventory | 12:08 |
| odyssey4me | javeriak that's set here: https://github.com/openstack/openstack-ansible/blob/kilo/playbooks/roles/galera_server/templates/cluster.cnf.j2#L18 | 12:09 |
| *** swati_ has quit IRC | 12:09 | |
| odyssey4me | javeriak here's where that error comes from: https://github.com/openstack/openstack-ansible/blob/kilo/playbooks/galera-install.yml#L43-L50 | 12:09 |
| javeriak | odyssey4me yep see it, dont really see how that went out of sync, guess will have to dig deeper | 12:10 |
| *** oneswig has quit IRC | 12:16 | |
| Bofu2U | random Q ... if you deploy using HA controllers, and one of them goes down, does any part of the networking component go with it? | 12:18 |
| *** skamithi has quit IRC | 12:20 | |
| *** markvoelker has joined #openstack-ansible | 12:20 | |
| *** javeriak_ has joined #openstack-ansible | 12:20 | |
| *** javeriak has quit IRC | 12:24 | |
| *** markvoelker has quit IRC | 12:24 | |
| *** retreved has joined #openstack-ansible | 12:46 | |
| *** oneswig has joined #openstack-ansible | 12:50 | |
| javeriak_ | hey guys, im getting a 'ValueError: too many values to unpack ' in the keystone logs while failing to login to horizon, any ideas whats wrong? | 12:50 |
| *** oneswig has quit IRC | 12:54 | |
| cloudnull | morning | 12:55 |
| Bofu2U | mornin | 12:56 |
| *** logan- has quit IRC | 13:00 | |
| *** logan- has joined #openstack-ansible | 13:00 | |
| cloudnull | javeriak_: what release is that ? | 13:06 |
| javeriak_ | cloudnull im on 11.2.7 | 13:07 |
| cloudnull | do you have the py traceback handy? | 13:07 |
| javeriak_ | i saw something in the horizon logs too about wrong version | 13:07 |
| javeriak_ | yes let me find | 13:07 |
| javeriak_ | [pid 7749:tid 140630231602944] SSL Library Error: error:1408A10B:SSL routines:SSL3_GET_CLIENT_HELLO:wrong <-- thats from horizon logs | 13:08 |
| cloudnull | hum... | 13:12 |
| cloudnull | that i've not seen | 13:13 |
| javeriak_ | cloudnull: http://paste.ubuntu.com/14857007/ | 13:13 |
| mhayden | morning | 13:13 |
| javeriak_ | my horzion one got clipped, but this was it: SSL routines:SSL3_GET_CLIENT_HELLO:wrong version number | 13:14 |
| *** oneswig has joined #openstack-ansible | 13:14 | |
| cloudnull | lbragstad dolphm dstanek have you guys seen http://paste.ubuntu.com/14857007/ w/ keystone in Kilo ? | 13:15 |
| dstanek | cloudnull: looks like fernet is somehow broken | 13:17 |
| cloudnull | yea | 13:17 |
| dstanek | i don't recall that specific error though | 13:17 |
| javeriak_ | just checked my code base, and im on the kilo 11.2.8+ | 13:17 |
| *** oneswig has quit IRC | 13:18 | |
| *** grumpycatt has quit IRC | 13:18 | |
| cloudnull | javeriak_: would you mind doing a rebase from the branch kilo rebuilding your repo and rerunning the keystone playbook ? | 13:19 |
| javeriak_ | cloudnull sure; should i leave the keystone containers as is or try rebuilding first? | 13:20 |
| *** oneswig has joined #openstack-ansible | 13:20 | |
| *** markvoelker has joined #openstack-ansible | 13:21 | |
| dstanek | cloudnull: that's not a federated token is it? | 13:22 |
| cloudnull | javeriak_: i'd leave everything intact | 13:22 |
| cloudnull | dstanek: i dont think so, javeriak_ are you deploying federation ? | 13:22 |
| javeriak_ | k thanks cloudnull; nope no federation | 13:22 |
| cloudnull | javeriak_: you should be able to rebuild in place. | 13:22 |
| odyssey4me | javeriak_ that ssl version number issue is a non issue - it's most probably haproxy's checks causing that | 13:23 |
| javeriak_ | odyssey4me i seem to be getting version mismatch errors with keystone cli commands as well | 13:23 |
| javeriak_ | but will try rebasing and rerunning for now | 13:23 |
| odyssey4me | javeriak_ you've deployed with SSL then? | 13:23 |
| odyssey4me | SSL for both keystone and horizon? | 13:24 |
| javeriak_ | odyssey4me yes | 13:24 |
| *** grumpycatt has joined #openstack-ansible | 13:24 | |
| javeriak_ | WARNING: unsupported identity-api-version 3, falling back to 2.0 <-- this | 13:24 |
| odyssey4me | javeriak_ is the keystone ssl set on the keystone/apache or on haproxy? | 13:24 |
| odyssey4me | javeriak_ that's an openrc misconfiguration | 13:26 |
| *** markvoelker has quit IRC | 13:26 | |
| *** markvoelker_ has joined #openstack-ansible | 13:26 | |
| *** andrei__ has joined #openstack-ansible | 13:26 | |
| javeriak_ | odyssey4me not sure, it'll be the standard config that comes on kilo | 13:27 |
| odyssey4me | javeriak_ the standard config in kilo doesn't do SSL for keystone | 13:27 |
| odyssey4me | it can, but that's not enabled by default | 13:27 |
| dstanek | cloudnull: javeriak_: the best i can tell is that the token has a broken payload - maybe dolphm or lbragstad remember a bug? | 13:27 |
| javeriak_ | odyssey4me this doesnt set ssl for keystone then ssl_protocol: "ALL -SSLv2 -SSLv3" | 13:28 |
| cloudnull | javeriak_: are you using the keystone CLI or the openstack CLI ? | 13:28 |
| javeriak_ | cloudnull openstack cli | 13:28 |
| cloudnull | kk | 13:28 |
| javeriak_ | so cloudnull dstanek odyssey4me rebasing and running the keystone plays seems to have fixed it | 13:29 |
| odyssey4me | javeriak_ nope - there are quite a few more options | 13:29 |
| odyssey4me | javeriak_ that just sets the protocols allowed when SSL is used | 13:30 |
| javeriak_ | hmm dunno what went wrong in the first place | 13:30 |
| *** mikelk has quit IRC | 13:31 | |
| *** tlbr has quit IRC | 13:37 | |
| cloudnull | maybe something happened on the original wheel build that caused the failure? | 13:39 |
| *** keedya has joined #openstack-ansible | 13:40 | |
| *** woodard has joined #openstack-ansible | 13:41 | |
| *** cemmason has joined #openstack-ansible | 13:45 | |
| *** b3rnard0_away is now known as b3rnard0 | 13:46 | |
| *** keedya has quit IRC | 13:51 | |
| dolphm | cloudnull: do you have the token and fernet keys? | 14:00 |
| cloudnull | i do not. | 14:00 |
| cloudnull | javeriak_: was seeing that issue | 14:01 |
| dolphm | javeriak_: o/ | 14:01 |
| cloudnull | however it seems that on redeploy everything was right with the world. so idk | 14:01 |
| javeriak_ | hey dolphm; yep cloudnull redeploying fixed it; let me add that this was an operational system; i dont know what brought it to that state though | 14:02 |
| *** skamithi has joined #openstack-ansible | 14:10 | |
| *** tlbr has joined #openstack-ansible | 14:11 | |
| *** sdake has quit IRC | 14:11 | |
| openstackgerrit | git-harry proposed openstack/openstack-ansible: Upgrades: Cleanup RabbitMQ / vhost https://review.openstack.org/272652 | 14:13 |
| *** dmsimard has quit IRC | 14:15 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: DEFCORE: Updated tempest config and resources https://review.openstack.org/274248 | 14:19 |
| andrei__ | hi all, did anyone try to disable iscsi_helper? I see tgt is installing by default and running in cinder container. Is there any good way to disable it without changing playbooks? | 14:20 |
| *** mikelk has joined #openstack-ansible | 14:26 | |
| *** grumpycatt has quit IRC | 14:30 | |
| cloudnull | andrei__: i think thats some left over cruft that we need to add a conditional too | 14:34 |
| *** grumpycatt has joined #openstack-ansible | 14:35 | |
| *** javeriak_ has quit IRC | 14:35 | |
| cloudnull | andrei__: the cinder apt packages assume tgt will be installed from here https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_cinder/defaults/main.yml#L216 | 14:36 |
| cloudnull | we'd need to abstract that out when using cinder+lvm | 14:36 |
| *** KLevenstein has joined #openstack-ansible | 14:45 | |
| *** pegmanm has joined #openstack-ansible | 14:46 | |
| *** keedya has joined #openstack-ansible | 14:47 | |
| *** pegmanm_ has quit IRC | 14:49 | |
| *** jeh has joined #openstack-ansible | 14:49 | |
| andrei__ | thank you | 14:51 |
| *** Mudpuppy has joined #openstack-ansible | 14:53 | |
| openstackgerrit | Merged openstack/openstack-ansible: Update Defcore test list function https://review.openstack.org/270138 | 14:54 |
| *** Mudpuppy has quit IRC | 14:59 | |
| *** Mudpuppy has joined #openstack-ansible | 15:00 | |
| *** zhangjn has quit IRC | 15:00 | |
| *** dmsimard has joined #openstack-ansible | 15:02 | |
| *** bryan_att has joined #openstack-ansible | 15:03 | |
| *** olivier427 has joined #openstack-ansible | 15:08 | |
| *** michaelgugino has joined #openstack-ansible | 15:09 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 15:10 | |
| logan- | andrei__: i just went through that last week :) | 15:11 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Add support for the nova_api db https://review.openstack.org/274932 | 15:11 |
| logan- | re: disabling tgt in cinder | 15:11 |
| logan- | one moment I am finding the commit | 15:11 |
| logan- | the first thing you will need to do is override cinder_apt_packages in user_variables and remove tgt from that list: http://paste.gentoolinux.info/imoxudumac.axapta | 15:14 |
| logan- | then, if you run cinder you will see a failure on a task called "Ensure cinder tgt include" | 15:14 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Use Swift venv Python interpreter https://review.openstack.org/275256 | 15:14 |
| logan- | so a quick stopgap patch I put in place for now is: http://paste.gentoolinux.info/abakumotij.diff | 15:15 |
| logan- | then it will deploy without installing tgt anywhere. you will still have some junked iscsi_* settings in your cinder.conf as there are no conditionals there, but they don't appear to affect my environment at all (using rbd backends only) | 15:16 |
| *** cemmason has quit IRC | 15:17 | |
| andrei__ | setting have no 'disable' feature :) | 15:17 |
| andrei__ | thank you, I'll check | 15:17 |
| *** baker has joined #openstack-ansible | 15:19 | |
| *** dmsimard has quit IRC | 15:21 | |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Changed ceilometer configs to use keystonev3 https://review.openstack.org/274905 | 15:21 |
| cloudnull | logan- andrei__ idk you have time, but if you could parameterize the tgt install It'd be much appreciated =) | 15:23 |
| *** Bjoern has joined #openstack-ansible | 15:24 | |
| *** cemmason has joined #openstack-ansible | 15:25 | |
| logan- | I will try to circle back to that someday. will probably have to spend a little more time on that one because I am not very familiar with cinder/iscsi deployments | 15:31 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible: [WIP] Remove venv activation code https://review.openstack.org/275279 | 15:39 |
| andrei__ | me neither. Probably difficulty here is default values for iscsi_helper, so it's assumed that tgt is installed. On pure ceph installation for example tgt is not needed, nevertheless one can probably use different backends with different volume drivers. | 15:40 |
| *** olivier427 has quit IRC | 15:40 | |
| *** jthorne has joined #openstack-ansible | 15:44 | |
| *** raddaoui has joined #openstack-ansible | 15:45 | |
| *** Mudpuppy has quit IRC | 15:46 | |
| *** galstrom_zzz is now known as galstrom | 15:46 | |
| *** olivier427 has joined #openstack-ansible | 15:54 | |
| *** cemmason has quit IRC | 15:56 | |
| cloudnull | logan-: ++ happy to circle back on it whenever | 15:58 |
| *** dweaver has quit IRC | 15:59 | |
| *** dweaver has joined #openstack-ansible | 16:00 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Resolve bashate violation E005 https://review.openstack.org/275290 | 16:00 |
| odyssey4me | bug triage in this meeting room cloudnull, mattt, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, Sam-I-Am, odyssey4me, serverascode, rromans, erikmwilson, mancdaz, _shaps_, BjoernT, claco, echiu, dstanek, jwagner, ayoung, prometheanfire, evrardjp, arbrandes, mhayden, scarlisle, luckyinva, ntt, javeriak, automagically, spotz, vdo, jmccrory, alextricity25, jasondotstar, KLevenstein, admin0, michaelgugino | 16:01 |
| palendae | o/ | 16:01 |
| michaelgugino | here | 16:01 |
| * mhayden woots | 16:01 | |
| alextricity25 | o/ | 16:01 |
| odyssey4me | hello everyone, welcome :) | 16:02 |
| KLevenstein | . | 16:02 |
| odyssey4me | first up: https://bugs.launchpad.net/openstack-ansible/+bug/1535075 | 16:02 |
| openstack | Launchpad bug 1535075 in openstack-ansible trunk "extra kernel module checks for kernel version only done on the deploy node" [Undecided,New] - Assigned to Jimmy McCrory (jimmy-mccrory) | 16:02 |
| odyssey4me | jmccrory it looks like to found the root cause | 16:02 |
| mattt | \o | 16:02 |
| jmccrory | do we want some kind of forced re-caching during the host setup playbook to help avoid that? | 16:03 |
| odyssey4me | well, I saw that there was an option for the CLI to clear the fact cache | 16:04 |
| michaelgugino | I think rerunning setup with gather facts again | 16:04 |
| odyssey4me | but I don't have a CLI handy right now | 16:04 |
| michaelgugino | so, we could include that at the top of the playbook | 16:04 |
| odyssey4me | I don't think that we should just flush the fact cache willy nilly | 16:05 |
| odyssey4me | I think that we should make people aware that the fact cache can be flushed and that it should be part of the troubleshooting process. | 16:05 |
| cloudnull | we can do rm /etc/openstack_deploy/ansible_facts/* | 16:05 |
| cloudnull | itll then rebuild the facts | 16:05 |
| cloudnull | also the facts are cached for 24 hours | 16:06 |
| cloudnull | that can be tuned down in the ansible.cfg file | 16:06 |
| odyssey4me | yes, that can be done - but the more ansible way is --flush-cache on the CLI... or something to that effect | 16:06 |
| odyssey4me | can someone do an 'ansible-playbook --help' quickly? | 16:06 |
| michaelgugino | or call setup in the playbook | 16:06 |
| *** gparaskevas has quit IRC | 16:07 | |
| cloudnull | it doesnt seem that the flush-cache command works | 16:07 |
| *** weezS has joined #openstack-ansible | 16:08 | |
| michaelgugino | flush cache is indeed a command line option | 16:08 |
| prometheanfire | neato | 16:08 |
| cloudnull | http://cdn.pasteraw.com/i98ikveaeen46n1vceju5cnqzbbekcn | 16:08 |
| cloudnull | same thing when i run the command manually http://cdn.pasteraw.com/h72ig7b5sqf87xyh93v0zqzlc5cu7fk | 16:09 |
| cloudnull | w/out the wrapper | 16:09 |
| cloudnull | IMO, use "rm" technology and rerun play :) | 16:09 |
| odyssey4me | ok, I've added a note to that effect and marked the bug as invalid | 16:09 |
| odyssey4me | it's be nice if someone could add a note about this in http://docs.openstack.org/developer/openstack-ansible/install-guide/app-tips.html | 16:10 |
| michaelgugino | Maybe a known issues section | 16:10 |
| cloudnull | maybe mark that bug as a doc issue ? | 16:10 |
| odyssey4me | well, if someone has a gap to add a known issues section and add this, that'd be great | 16:12 |
| michaelgugino | there is a troubleshooting section | 16:13 |
| michaelgugino | http://docs.openstack.org/developer/openstack-ansible/install-guide/ops-troubleshooting.html | 16:13 |
| odyssey4me | sounds like a good section to add it | 16:13 |
| odyssey4me | any volunteers to put together a docs patch for it? | 16:13 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Allow apache file overrides https://review.openstack.org/267749 | 16:13 |
| jmccrory | i'll take that up | 16:14 |
| odyssey4me | great, thanks jimmy | 16:14 |
| odyssey4me | I'm going to mark it as triaged | 16:14 |
| michaelgugino | Perhaps would should instruct people to run apt-get upgrade while setting up target hosts? | 16:14 |
| odyssey4me | michaelgugino I think that perhaps there should be some sort of pre-requisites guide with some examples of things that should be done before executing a build. | 16:15 |
| *** dweaver has quit IRC | 16:15 | |
| michaelgugino | Well, there is: http://docs.openstack.org/developer/openstack-ansible/install-guide/targethosts-prepare.html | 16:16 |
| *** dweaver has joined #openstack-ansible | 16:16 | |
| odyssey4me | sounds like a good spot | 16:16 |
| odyssey4me | any volunteers to add another section there recommending an apt-get dist-upgrade ? | 16:17 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible-plugins: Allow apache file overrides https://review.openstack.org/275298 | 16:17 |
| *** skamithi has quit IRC | 16:17 | |
| *** skamithi has joined #openstack-ansible | 16:18 | |
| michaelgugino | I'll take that | 16:18 |
| odyssey4me | thanks michaelgugino | 16:18 |
| michaelgugino | yw | 16:19 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1538268 | 16:19 |
| openstack | Launchpad bug 1538268 in openstack-ansible "openstack-ansible-security newalias task fails if disable_ipv6 true" [Undecided,New] | 16:19 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Allow apache file overrides https://review.openstack.org/267749 | 16:19 |
| odyssey4me | mhayden ping? | 16:19 |
| odyssey4me | This looks valid to me, but I think mhayden will need to look into it unless someone else wants to give it a go. | 16:21 |
| *** sdake has joined #openstack-ansible | 16:21 | |
| odyssey4me | ok, moving on: https://bugs.launchpad.net/openstack-ansible/+bug/1539207 | 16:21 |
| openstack | Launchpad bug 1539207 in openstack-ansible "permission on swift log directories incorrect" [Undecided,New] | 16:21 |
| odyssey4me | wasn't there a patch in review for this? | 16:22 |
| cloudnull | i thought there was | 16:23 |
| cloudnull | this was one side https://review.openstack.org/#/c/273231/ | 16:23 |
| cloudnull | i want to say travis did the other patch | 16:24 |
| odyssey4me | ping automagically | 16:24 |
| odyssey4me | :/ not here | 16:24 |
| odyssey4me | https://review.openstack.org/273231 ? | 16:25 |
| odyssey4me | nope - ok, this looks like a valid bug then | 16:25 |
| odyssey4me | any volunteers to patch it up? the bug report appears to resolve it | 16:26 |
| mhayden | odyssey4me: sorry, on a conf call at the moment -- i'll look at the bug | 16:26 |
| cloudnull | the issue is that we need to adjust https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_swift/tasks/swift_pre_install.yml#L90-L91 | 16:26 |
| cloudnull | to be syslog | 16:26 |
| cloudnull | not swift the user name | 16:26 |
| odyssey4me | there's a patch in the report which says differently | 16:27 |
| openstackgerrit | Merged openstack/openstack-ansible: Use Swift venv Python interpreter https://review.openstack.org/273943 | 16:27 |
| odyssey4me | but yeah, I agree that would be a better course of action | 16:27 |
| mhayden | i'll take https://bugs.launchpad.net/openstack-ansible/+bug/1538268 | 16:27 |
| openstack | Launchpad bug 1538268 in openstack-ansible "openstack-ansible-security newalias task fails if disable_ipv6 true" [Undecided,New] - Assigned to Major Hayden (rackerhacker) | 16:27 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Use Swift venv Python interpreter https://review.openstack.org/275256 | 16:27 |
| odyssey4me | any volunteers to patch that up? | 16:27 |
| cloudnull | odyssey4me: i can bang that out rigth quick | 16:28 |
| odyssey4me | ok cloudnull assigned | 16:28 |
| odyssey4me | next: https://bugs.launchpad.net/openstack-ansible/+bug/1539210 | 16:29 |
| openstack | Launchpad bug 1539210 in openstack-ansible "empty log directories cause invalid logrotate files" [Undecided,New] | 16:29 |
| *** daneyon_ has joined #openstack-ansible | 16:29 | |
| odyssey4me | ah, https://review.openstack.org/#/c/273292/ is looking for another core reviewer | 16:30 |
| odyssey4me | mattt ^ | 16:30 |
| odyssey4me | hughsaunders andymccr d34dh0r53 stevelle ^ | 16:30 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Fix swift log perms https://review.openstack.org/275311 | 16:31 |
| *** spotz_zzz is now known as spotz | 16:31 | |
| *** dweaver has quit IRC | 16:31 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-rsyslog_client: Added conditional templating https://review.openstack.org/273292 | 16:32 |
| *** dweaver has joined #openstack-ansible | 16:32 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Added conditional templating https://review.openstack.org/273293 | 16:32 |
| spotz | hey guys sorry I'm late was driving to Austin, miss anything? | 16:32 |
| *** daneyon has quit IRC | 16:33 | |
| odyssey4me | ok, so that one is in progress | 16:33 |
| odyssey4me | welcome spotz :) | 16:33 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1539236 | 16:33 |
| openstack | Launchpad bug 1539236 in openstack-ansible "lxc_host cache issues" [Undecided,New] | 16:33 |
| *** skamithi has quit IRC | 16:34 | |
| cloudnull | the only way ive seen this issue in the wild is if the play was stopped while expanding the lxc cache | 16:34 |
| odyssey4me | this is a valid issue, and a bit of a hangover from Juno where all hosts had lxc installed | 16:35 |
| odyssey4me | and yes, what cloudnull said | 16:35 |
| *** Mudpuppy has joined #openstack-ansible | 16:35 | |
| alextricity25 | I thought there was already a sha check in place? | 16:35 |
| alextricity25 | or something like that | 16:35 |
| odyssey4me | mattt also discovered that the sha check in Ansible <2 doesn't work as you think it does | 16:35 |
| odyssey4me | we should perhaps be a little more robust in our checks | 16:36 |
| cloudnull | the download works but if the task is stopped when the tar is expanded it can cuase issues | 16:36 |
| odyssey4me | any volunteers to fix this up? | 16:36 |
| alextricity25 | I had an issue with the container image cache in Kilo | 16:36 |
| *** agireud has quit IRC | 16:36 | |
| *** skamithi has joined #openstack-ansible | 16:36 | |
| alextricity25 | the play wasn't stopped...the image was just corrupted in transit | 16:36 |
| alextricity25 | at least it seemed that way | 16:36 |
| cloudnull | odyssey4me: i think the real fix here is to begin using the download template and building without the rpc image download | 16:36 |
| odyssey4me | yeah, I actually have started doing some work on that again | 16:36 |
| odyssey4me | let me take this as it's complementary work | 16:37 |
| *** daneyon has joined #openstack-ansible | 16:37 | |
| *** agireud has joined #openstack-ansible | 16:37 | |
| cloudnull | i think if we hybridify the work you had before and this role https://github.com/os-cloud/lxc_image_cache we'd have a fairly robust solution | 16:37 |
| odyssey4me | cloudnull yeah, I've been looking at several options | 16:38 |
| cloudnull | we could also simply use the download template as is | 16:38 |
| cloudnull | but anyway ... i think the correct solution is to look at using the lxc community images instead | 16:39 |
| odyssey4me | I think next cycle we can build more of an image pipeline and look at working on prepping images per container profile or something | 16:39 |
| cloudnull | maybe by then lxd will more robust too | 16:39 |
| odyssey4me | but for now I just want to add a simple mechanism to use a downloaded image, or use an image from the lxc images | 16:39 |
| logan- | download template being the distro one cloudnull? are tehre any major differences beetween the rpc one and the regular trusty template that comes with ubuntu? | 16:39 |
| odyssey4me | both would then require some additional prep before deployment | 16:39 |
| *** daneyon_ has quit IRC | 16:40 | |
| odyssey4me | logan- FYI I've got a WIP here: https://review.openstack.org/272743 | 16:40 |
| palendae | cloudnull: I'm not sure that LXD replaces any of that logic. It might, I guess | 16:40 |
| cloudnull | logan-: not much difference. | 16:40 |
| cloudnull | the rpc image has additional packages within it to speed things up | 16:40 |
| odyssey4me | it's a long way from done, but essentially it grabs a raw image from the lxc daily builds, then does the updates we need to it before we can use it. | 16:40 |
| logan- | thanks odyssey4me, will look thru that | 16:40 |
| odyssey4me | it makes the whole image prep process transparent | 16:40 |
| cloudnull | ++ | 16:41 |
| odyssey4me | the objective is to make the image prep process completely transparent so that the rpc-prepped image is no longer required | 16:41 |
| odyssey4me | this should also speed things up, because the rpc image is old :p | 16:41 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1539246 | 16:42 |
| openstack | Launchpad bug 1539246 in openstack-ansible "only install one of nova_console_novnc or nova_console to haproxy" [Undecided,New] | 16:42 |
| *** izaakk has joined #openstack-ansible | 16:42 | |
| cloudnull | odyssey4me: the rpc image is up-to-date | 16:42 |
| cloudnull | just updated last week | 16:42 |
| cloudnull | odyssey4me: on that issue | 16:43 |
| odyssey4me | hmm, it would seem to me that we need to ensure that novnc and spice have their own groups | 16:43 |
| andymccr | transparency is the main reason i'd like to see that done. | 16:43 |
| cloudnull | if you have spice, and then change to novnc, you need to keep both endpoints | 16:44 |
| odyssey4me | or we need to consolidate https://github.com/openstack/openstack-ansible/blob/master/playbooks/vars/configs/haproxy_config.yml#L122-L139 somehow to do one or the other | 16:44 |
| odyssey4me | andymccr ++ | 16:44 |
| andymccr | well that also sounds like an haproxy issue and not a spice vs novnc issue | 16:44 |
| cloudnull | any vm built with a particular console type will remain using that console type forever | 16:44 |
| jthorne | i recently switched from spice to novnc in OSIC Cloud 1 | 16:44 |
| cloudnull | unless you modify the DB to change it | 16:44 |
| odyssey4me | cloudnull yuck, that's fun | 16:44 |
| andymccr | that sounds fine? | 16:45 |
| *** mgoddard_ has joined #openstack-ansible | 16:45 | |
| odyssey4me | ok cloudnull in that case we need to have different container profiles for each | 16:45 |
| andymccr | and cleaner tbh | 16:45 |
| andymccr | just remove the old no_vnc containers | 16:45 |
| andymccr | and create new ones | 16:45 |
| odyssey4me | unless it's possible to run both in the same container? | 16:45 |
| cloudnull | they can be in the same container | 16:45 |
| cloudnull | all that works | 16:45 |
| andymccr | do you really want old containers that "had" spice running, (and still have all the bits there) now running novnc? | 16:45 |
| andymccr | with the intention of no longer running spice console | 16:45 |
| odyssey4me | shouldn't we look at fronting novnc/spice with Apache/nginx? afaik both support it | 16:45 |
| cloudnull | they work on different ports and endpoints | 16:46 |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible: Update Defcore test list function https://review.openstack.org/275318 | 16:46 |
| Bjoern | put a proxy in front of a proxy, novnc and spice are already a proxy? | 16:46 |
| odyssey4me | using nginx/apache can provide SSL endpoints for both equally | 16:46 |
| Bjoern | spice has TLS builtin, novnc i'm not sure currently | 16:47 |
| odyssey4me | and when you set it up with a web server front-end, it doesn't use the python webserver (proxy) | 16:47 |
| *** dweaver has quit IRC | 16:47 | |
| odyssey4me | it's far more secure that way, IMO | 16:47 |
| cloudnull | andymccr: normally i say no. but if the deployment changes console types and has running vms itll orphan the old vms access to the console | 16:48 |
| *** dweaver has joined #openstack-ansible | 16:48 | |
| *** mgoddard has quit IRC | 16:48 | |
| Bjoern | if it works like keystone/wsgi then it's fine | 16:48 |
| andymccr | thats not a solvable problem within OSA though | 16:48 |
| cloudnull | true | 16:48 |
| odyssey4me | Bjoern I may stand under correction, but I've seen blog posts and stuff that give me that understanding. | 16:49 |
| cloudnull | i mean you can deploy one and then another and both will run | 16:49 |
| cloudnull | and will do so all within the same container. | 16:49 |
| *** skamithi has quit IRC | 16:49 | |
| odyssey4me | anyway, that's not the immediate issue here - it just might be a nice way to round out the config in a more robust way | 16:49 |
| odyssey4me | the ssl will scale better too | 16:49 |
| cloudnull | and when the deployer wants to clean it up they can nuke those containers and redeploy them for a clean env | 16:49 |
| odyssey4me | the bug in question specifies that haproxy isn't aware of which back-end is actually there | 16:50 |
| andymccr | or it is but doesnt clean it up? | 16:50 |
| cloudnull | ^ | 16:50 |
| andymccr | we don't have a mechanism to ensure certain files "aren't present" | 16:50 |
| andymccr | and im not sure we should add one. | 16:50 |
| cloudnull | ^ also | 16:50 |
| odyssey4me | ++ | 16:50 |
| openstackgerrit | Hugh Saunders proposed openstack/openstack-ansible: Update Defcore test list function https://review.openstack.org/275320 | 16:50 |
| andymccr | unless we want to do an "ensure disabled" or something | 16:50 |
| andymccr | but that sounds like more effort than is worth | 16:50 |
| cloudnull | i remember that from the chef days and I'd like not to repeate that | 16:51 |
| andymccr | its a bug IF you manually remove the haproxy conf for spice/novnc and it keeps re-adding it when the console isn't present i guess | 16:51 |
| cloudnull | i think our current haproxy role adds everything we support to the conf | 16:51 |
| cloudnull | even if its not present | 16:51 |
| andymccr | yeh true | 16:51 |
| andymccr | so you will have that problem for all services really | 16:51 |
| odyssey4me | https://github.com/openstack/openstack-ansible/blob/master/playbooks/vars/configs/haproxy_config.yml#L122-L139 will need to reference a different group for each console type so that it's more aware. Each console type would need to be in a container group that specifies the type. | 16:52 |
| andymccr | except that the ones with no containers won't have existing hosts i guess | 16:52 |
| evrardjp | hello, it just happens that I come by here | 16:52 |
| cloudnull | i like evrardjp approach better for a revamped haproxy role | 16:52 |
| cloudnull | speak of the devil :) | 16:52 |
| evrardjp | :) | 16:53 |
| evrardjp | sorry I'm not much available nowadays | 16:53 |
| cloudnull | this role https://github.com/evrardjp/ansible-haproxy | 16:53 |
| evrardjp | but I think it's the way to go | 16:53 |
| evrardjp | (don't check on the "build failed", it's just a problem in my CI... :p) | 16:54 |
| *** Nepoc has joined #openstack-ansible | 16:54 | |
| evrardjp | so what's the issue, except lack of modularity? | 16:54 |
| odyssey4me | ok, this is just a cosmetic error anyway | 16:54 |
| andymccr | if its not causing haproxy to fail, im inclined to not care as much about the bug. | 16:54 |
| andymccr | like wishlist item "haproxy plays could be better" | 16:54 |
| evrardjp | could you ping me the link? | 16:55 |
| odyssey4me | so I'd say that it's confirmed, but low priority - any objections/preferred suggestions? | 16:55 |
| cloudnull | ++ | 16:55 |
| logan- | maybe looking at calling haproxy as a conditional dependency to do service setup | 16:55 |
| odyssey4me | ok, I've added some comments | 16:56 |
| odyssey4me | I think we've spent enough time on it :) | 16:56 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1539803 | 16:57 |
| openstack | Launchpad bug 1539803 in openstack-ansible "enable support for availability zones for compute and services" [Undecided,New] | 16:57 |
| odyssey4me | this is already possible | 16:57 |
| cloudnull | admin0 is interested in that ^ | 16:57 |
| git-harry | scre | 16:58 |
| odyssey4me | http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-openstack.html | 16:58 |
| cloudnull | odyssey4me: is it manual or automated ? | 16:58 |
| evrardjp | az are already supported in openstack-ansible, maybe a doc is needed then? | 16:58 |
| odyssey4me | more specifically: http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-openstack.html#overriding-conf-files | 16:58 |
| *** mikelk has quit IRC | 16:58 | |
| odyssey4me | per host config overrides are already documented there | 16:59 |
| evrardjp | also this is documented: http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-cinder-az.html | 16:59 |
| openstackgerrit | Merged openstack/openstack-ansible: Fix rsync service restart in os_swift https://review.openstack.org/273149 | 16:59 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Fix rsync service restart in os_swift https://review.openstack.org/275324 | 17:00 |
| evrardjp | nova it's about 2 clicks in the web interface OR few api calls, already explained in the upstream openstack doc | 17:00 |
| odyssey4me | marking as invalid | 17:00 |
| odyssey4me | last one: https://bugs.launchpad.net/openstack-ansible/+bug/1540531 | 17:01 |
| openstack | Launchpad bug 1540531 in openstack-ansible "Upgrade issues with 'Create log aggregation links'" [Undecided,New] | 17:01 |
| *** oneswig has quit IRC | 17:01 | |
| *** skamithi has joined #openstack-ansible | 17:01 | |
| odyssey4me | palendae can I ask you to look into this as part of the upgrade testing/fixing? | 17:02 |
| palendae | odyssey4me: Sure thing | 17:02 |
| odyssey4me | it looks entirely plausible, but will likely be based on a situation where someone is coming from a pretty old build | 17:02 |
| *** raddaoui has quit IRC | 17:02 | |
| odyssey4me | thanks palendae | 17:02 |
| palendae | odyssey4me: Yeah, I can see a situation where our customers have environemtns that move from Icehouse/Juno to Kilo to Liberty | 17:03 |
| odyssey4me | palendae yeah, or who have pre-existing environments and convert to using OSA | 17:03 |
| odyssey4me | :) | 17:03 |
| *** dweaver has quit IRC | 17:03 | |
| *** scarlisle has joined #openstack-ansible | 17:03 | |
| *** mgoddard_ has quit IRC | 17:03 | |
| odyssey4me | thank you all for your time - we're a few mins over | 17:03 |
| *** mgoddard has joined #openstack-ansible | 17:04 | |
| *** dweaver has joined #openstack-ansible | 17:04 | |
| cloudnull | cheers | 17:04 |
| *** oneswig has joined #openstack-ansible | 17:06 | |
| Nepoc | Hello everyone, back with a question. How do I enable the firewall? The tab is absent from Horizon. | 17:06 |
| odyssey4me | Nepoc you're looking for 'Security Groups' | 17:08 |
| cloudnull | Nepoc: or are you looking for FWAAS ? | 17:08 |
| odyssey4me | Security Groups are basic firewalling which can be applied for internal and external rules | 17:08 |
| bgmccollum | anyone configured OSA to use LXC? | 17:08 |
| odyssey4me | bgmccollum yeah, some guy on the mailing list instrumented the lxc hypervisor | 17:09 |
| bgmccollum | odyssey4me interesting... *digs through ML* | 17:09 |
| logan- | that sounds interesting | 17:09 |
| Nepoc | If FWAAS is what provided the "Firewall" tab in horizon then yes that's what I'm looking for... One moment I'll try to get more information from the developer here who's looking for it. | 17:09 |
| odyssey4me | bgmccollum openstack-operators list - 'virt_type=lxc' | 17:10 |
| cloudnull | Nepoc: we dont have that option as something we deploy by default. however we could add it with little work . | 17:10 |
| *** oneswig has quit IRC | 17:10 | |
| odyssey4me | bgmccollum here're his notes: https://etherpad.openstack.org/p/osad-lxc-hyper-notes | 17:11 |
| cloudnull | at last check, in kilo, it wasnt the most function of neutron services | 17:11 |
| bgmccollum | odyssey4me thanks! | 17:11 |
| cloudnull | but maybe time to revise that | 17:11 |
| odyssey4me | bgmccollum caveat - please do an appendix entry or blog post showing how to do it right :) | 17:11 |
| logan- | odyssey4me, bgmccollum: thanks | 17:12 |
| bgmccollum | odyssey4me: git apply-patch ... | 17:12 |
| *** raddaoui has joined #openstack-ansible | 17:12 | |
| cloudnull | Nepoc: looking at whats required in master/liberty | 17:12 |
| odyssey4me | bgmccollum I don't think that modifying the source is necessary, as you'll see in my responses to him | 17:12 |
| odyssey4me | bgmccollum my responses are in green text | 17:13 |
| bgmccollum | got it | 17:13 |
| odyssey4me | logan- I challenge you to beat bgmccollum to the punch. :) | 17:13 |
| Nepoc | cloudnull: that would be great if we could enable that | 17:13 |
| odyssey4me | Nepoc It should be a pretty easy edit to enable it, and we'd be happy to help you put it together. | 17:14 |
| logan- | lol, I would love to but have to get kvm to production first :( that seems like a fun experiment though | 17:15 |
| *** sdake has quit IRC | 17:15 | |
| bgmccollum | odyssey4me logan- its also on my "i shouldn't be looking at this" list :/ | 17:16 |
| prometheanfire | why do we ignore errors in bootstraping ansible (during the galaxy install) https://github.com/openstack/openstack-ansible/blob/master/scripts/bootstrap-ansible.sh#L89 ? | 17:16 |
| logan- | haha | 17:16 |
| Nepoc | odyssey4me: Do you guys have an idea of how much time it might take? | 17:16 |
| prometheanfire | cloudnull: you had that in your megapatch :P | 17:17 |
| Nepoc | odyssey4me: Also one of the developers here asked if there is a way to enable it manually or are there things missing? | 17:18 |
| odyssey4me | Nepoc I expect that the basic neutron binaries should be there, but it may be missing drivers. | 17:18 |
| *** spotz is now known as spotz_zzz | 17:19 | |
| *** dweaver has quit IRC | 17:19 | |
| odyssey4me | Nepoc I'm not sure that any of us are familiar enough with FWAAS right now to provide a decent answer. | 17:19 |
| odyssey4me | any volunteers to see what it takes to get FWAAS going? :) | 17:19 |
| *** dweaver has joined #openstack-ansible | 17:20 | |
| Nepoc | I think cloudnull is taking a look? | 17:20 |
| *** skamithi has quit IRC | 17:20 | |
| palendae | What are the <service_name>_venv_tag variables for? Tagging to a specific release? https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_cinder/defaults/main.yml#L19-L20 | 17:23 |
| *** phiche has quit IRC | 17:24 | |
| palendae | Also noticing all the services are set to download from the hard-coded string "untagged" rather than the venv_tag value, is that intentional? https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_cinder/defaults/main.yml#L31 | 17:25 |
| odyssey4me | palendae the tag name is overridden in the group vars I think | 17:25 |
| odyssey4me | so the venv gets named according to the release tag number | 17:26 |
| palendae | Ok | 17:26 |
| odyssey4me | ie cinder_12.0.3 | 17:26 |
| palendae | But that download URL doesn't appear to | 17:26 |
| palendae | Gotcha | 17:26 |
| palendae | Hm, not seeing venv_tags in playbooks/defaults, inventory/group_vars, or playbooks/vars | 17:28 |
| odyssey4me | palendae playbooks/os-cinder-install.yml: cinder_venv_download_url: "{{ openstack_repo_url }}/venvs/{{ openstack_release }}/{{ ansible_distribution | lower }}/cinder-{{ openstack_release }}.tgz" | 17:28 |
| palendae | Ah, I see | 17:28 |
| odyssey4me | palendae I think it's all in the playbook | 17:28 |
| odyssey4me | sorry, misdirection | 17:28 |
| palendae | np | 17:28 |
| palendae | See it now, thanks | 17:28 |
| palendae | Doesn't actually look like the venv_tag value is used, then | 17:29 |
| palendae | Since the URLs use openstack_release | 17:29 |
| odyssey4me | palendae heh, sounds like we have a little bug there :) | 17:30 |
| palendae | Nope, not according to ack in playbooks/. It's set to openstack_release, but never actually used | 17:30 |
| jmccrory | anyone know why the default neutron device and VM interfaces are set to 1450 MTU here? https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_neutron/defaults/main.yml#L244-L252 | 17:30 |
| palendae | Oh, used with vevn_bin | 17:30 |
| palendae | Ok | 17:30 |
| odyssey4me | good catch - if you do find it's not being used then please feel free to clean it up | 17:30 |
| palendae | odyssey4me: Misread; used in exactly 1 spot per role :p | 17:31 |
| *** phiche has joined #openstack-ansible | 17:31 | |
| odyssey4me | jmccrory gre encapsulation? | 17:31 |
| odyssey4me | (or vxlan) | 17:31 |
| odyssey4me | no idea really, just throwing it out there | 17:31 |
| *** openstackgerrit has quit IRC | 17:32 | |
| *** openstackgerrit has joined #openstack-ansible | 17:32 | |
| *** dweaver has quit IRC | 17:35 | |
| *** dweaver has joined #openstack-ansible | 17:35 | |
| michaelgugino | gre needs mtu of 1450 | 17:36 |
| jmccrory | oh ok, so it's limiting there to account for encapsulation headers outside of VM? | 17:40 |
| bgmccollum | jmccrory yup | 17:41 |
| Bjoern | odeysee4me: 1540531 is an issue inside liberty and not just mitaka | 17:45 |
| palendae | Bjoern: I was going to treat that as a liberty problem | 17:46 |
| Bjoern | ok | 17:46 |
| *** javeriak has joined #openstack-ansible | 17:46 | |
| palendae | Bjoern: I'm not working on mitaka upgrades yet | 17:46 |
| *** sdake has joined #openstack-ansible | 17:48 | |
| *** javeriak has quit IRC | 17:49 | |
| *** javeriak has joined #openstack-ansible | 17:50 | |
| *** sdake has quit IRC | 17:50 | |
| *** javeriak_ has joined #openstack-ansible | 17:54 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Add notes about bashate E006 exception https://review.openstack.org/275349 | 17:55 |
| *** javeriak has quit IRC | 17:55 | |
| *** weezS has quit IRC | 17:56 | |
| odyssey4me | Bjoern yep, palendae is looking into it and it can be targeted once the triage is complete | 17:57 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible: Docs: encourage users to run apt-get dist-upgrade on target hosts https://review.openstack.org/275354 | 17:59 |
| lbragstad | is the ansible docs site down? | 18:00 |
| palendae | Not loading for me | 18:00 |
| palendae | Yeah, just got a cloudflare error | 18:01 |
| lbragstad | http://docs.ansible.com/list_of_files_modules.html | 18:01 |
| odyssey4me | lbragstad ask jmccrory - he knows them out of his head : | 18:01 |
| odyssey4me | :p | 18:01 |
| palendae | Hm, ansible.com loads and readthedocs.org loads | 18:01 |
| lbragstad | docs.ansible.com seems to load | 18:02 |
| palendae | Can't remember if they hosted on rtd, looks like not | 18:02 |
| palendae | lbragstad: docs.ansible.com loads for me out of cloudflare, but I get a banner saying the host is down | 18:02 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 18:03 | |
| lbragstad | hmm - okay. thanks for the sanity check! | 18:03 |
| palendae | I can browse it now | 18:03 |
| palendae | But still have the cloudflare banner | 18:04 |
| palendae | Aaaand spoke too soon | 18:04 |
| lbragstad | ah! | 18:04 |
| lbragstad | ba dum psh... | 18:04 |
| jmccrory | archive.org kinda works...slowly https://web.archive.org/web/20150405065537/http://docs.ansible.com/list_of_files_modules.html | 18:07 |
| jmccrory | odyssey4me heh not quite there yet | 18:07 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible: Docs: encourage users to run apt-get dist-upgrade on target hosts https://review.openstack.org/275354 | 18:08 |
| odyssey4me | michaelgugino thanks for ^ - I've left a comment with a question :) | 18:10 |
| michaelgugino | odyssey4me: replied | 18:11 |
| *** mariognarly has joined #openstack-ansible | 18:13 | |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible: Docs: encourage users to run apt-get dist-upgrade on target hosts https://review.openstack.org/275354 | 18:15 |
| *** shanec has joined #openstack-ansible | 18:15 | |
| michaelgugino | I see that reboot is already covered at the bottom of target-hosts prep. I just moved that step and added some text. | 18:15 |
| cloudnull | Nepoc: your welcome to pull the changes in and test them out to see it mets your needs https://github.com/os-cloud/os-ansible-deployment/tree/neutron-fwaas | 18:16 |
| cloudnull | ill be submitting it as a review once I can validate it works on a running system I have | 18:16 |
| cloudnull | but thats the WIP branch | 18:16 |
| Nepoc | cloudnull: Awesome that was fast | 18:16 |
| odyssey4me | Nepoc while cloudnull is probably a genius, OSA is also an easy framework to add things to once you know how :) | 18:17 |
| cloudnull | ^ that | 18:18 |
| cloudnull | because of the work odyssey4me has put into it | 18:18 |
| cloudnull | :) | 18:18 |
| shanec | hey all, when setting lxc_cache_resolvers in user_vars, what's the syntax? this is throwing a yaml error"xc_cache_resolvers: "nameserver 69.20.0.164","nameserver 69.20.0.196" | 18:18 |
| odyssey4me | lol, not me - all of us as a community :p | 18:18 |
| cloudnull | ^ that | 18:18 |
| odyssey4me | shanec - it looks like you've added a comma delimited string instead of a yaml list? | 18:19 |
| SamYaple | "nameserver 69.20.0.164","nameserver 69.20.0.196" should be "nameserver 69.20.0.164,nameserver 69.20.0.196" if i read your intent correcly shanec | 18:20 |
| odyssey4me | shanec it should be formatted like so: https://github.com/openstack/openstack-ansible-lxc_hosts/blob/master/defaults/main.yml#L84-L86 | 18:20 |
| Nepoc | Well I'll review it and let you know my success and hopefully glean some info of how I can easily make these changes as well. | 18:20 |
| shanec | awesome, thanks SamYaple and odyssey4me | 18:21 |
| odyssey4me | thanks for helping out SamYaple :) | 18:21 |
| Nepoc | Speaking of other changes, the SpiceVNC is ... unusable, it doesn't update the display in any browser, you have to type and hit refresh to get updates. | 18:21 |
| *** skamithi has joined #openstack-ansible | 18:22 | |
| Nepoc | Can I swap it with something else? | 18:22 |
| odyssey4me | Nepoc novnc is an option too, although I'm not sure whether anyone's put a config doc together for that yet | 18:22 |
| odyssey4me | it'd be rather lovely if someone did :) | 18:22 |
| odyssey4me | alright, I'm outta here - time to head home | 18:23 |
| odyssey4me | night all | 18:23 |
| cloudnull | nite | 18:25 |
| cloudnull | Nepoc: set "nova_console_type: novnc" | 18:26 |
| cloudnull | in your user_variables.yml | 18:26 |
| *** metral has quit IRC | 18:26 | |
| cloudnull | the option comes from https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/os_nova/defaults/main.yml#L156-L157 | 18:26 |
| Nepoc | Excellent! | 18:29 |
| cloudnull | once you set that, you'll want to run os-nova-install.yml | 18:30 |
| bgmccollum | odyssey4me logan- got LXC working...fiddled many bits, now need to figure out which was the correct combination | 18:33 |
| *** metral_zzz has joined #openstack-ansible | 18:35 | |
| *** metral_zzz is now known as metral | 18:35 | |
| Nepoc | cloudnull: looks like I don't have the driver No module named neutron_fwaas.services.firewall.fwaas_plugin | 18:41 |
| SamYaple | odyssey4me: i always help!... when i remember to log into this channel.... | 18:41 |
| cloudnull | Nepoc: likely need to make sure neutron-fwaas is installed within the venv | 18:42 |
| openstackgerrit | Merged openstack/openstack-ansible: Ensure that Heat role works with multiple Keystone back-ends https://review.openstack.org/272714 | 18:43 |
| *** raddaoui has quit IRC | 18:44 | |
| logan- | bgmccollum: nice! any massive changes needed or were you able to do most of it using variable overrides? | 18:48 |
| bgmccollum | mostly variables i think...i fiddles lots of things...back tracking now to get exact changes | 18:48 |
| logan- | cool | 18:48 |
| *** skamithi has quit IRC | 18:52 | |
| *** oneswig has joined #openstack-ansible | 18:54 | |
| Nepoc | cloudnull: yeah that was my problem... now to figure out how to get it in there automatically | 18:56 |
| *** oneswig has quit IRC | 18:59 | |
| *** electrofelix has quit IRC | 19:00 | |
| cloudnull | yea itll get in there if you have the venv rebuilt this can be done by simply setting repo_build_venv_force_rebuild=true | 19:00 |
| cloudnull | so the command: [ openstack-ansible repo-build.yml os-neutron-install.yml -e 'repo_build_venv_force_rebuild=true' ] should do what you wnat | 19:01 |
| cloudnull | *want | 19:01 |
| Nepoc | okay sounds good, just so I can learn something. How does it know to include that in the venv now? | 19:02 |
| cloudnull | its part of the way the repo-build play works with the plugin py_pkgs | 19:03 |
| *** raddaoui has joined #openstack-ansible | 19:03 | |
| cloudnull | this role https://github.com/openstack/openstack-ansible/tree/master/playbooks/roles/repo_build | 19:03 |
| Nepoc | Yeah I was just digging through trying to figure out how it was going to get pulled in :) | 19:04 |
| cloudnull | this plugin https://github.com/openstack/openstack-ansible-plugins/blob/master/lookups/py_pkgs.py | 19:04 |
| Nepoc | Looks like I have some reading to do :) | 19:05 |
| cloudnull | so what happens is the plugin iterates over a list of items which is the locations of various roles | 19:05 |
| cloudnull | https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L24 | 19:05 |
| cloudnull | which constructs a big json blob of every python type package it could find | 19:05 |
| *** richoid1 has joined #openstack-ansible | 19:05 | |
| cloudnull | which it then globs into role based requirements that are used for the individual venvs | 19:06 |
| *** pellaeon_ has joined #openstack-ansible | 19:06 | |
| Nepoc | Nice and complex but handy | 19:07 |
| cloudnull | it used to be simpler, however with the galaxy approach to roles we needed something a bit more complex. sadly. | 19:08 |
| Nepoc | Ah I see | 19:09 |
| *** itsuugo_ has joined #openstack-ansible | 19:10 | |
| *** evrardjp_ has joined #openstack-ansible | 19:11 | |
| *** lbragstad_ has joined #openstack-ansible | 19:11 | |
| *** pabelanger_ has joined #openstack-ansible | 19:11 | |
| *** mrodden_ has joined #openstack-ansible | 19:11 | |
| *** miguelgrinberg_ has joined #openstack-ansible | 19:11 | |
| *** charz has joined #openstack-ansible | 19:11 | |
| *** weezS has joined #openstack-ansible | 19:11 | |
| *** arbrandes has quit IRC | 19:12 | |
| *** evrardjp has quit IRC | 19:12 | |
| *** richoid has quit IRC | 19:12 | |
| *** pabelanger has quit IRC | 19:12 | |
| *** mattt has quit IRC | 19:12 | |
| *** itsuugo has quit IRC | 19:12 | |
| *** pellaeon has quit IRC | 19:12 | |
| *** charz_ has quit IRC | 19:12 | |
| *** lbragstad has quit IRC | 19:12 | |
| *** mrodden has quit IRC | 19:12 | |
| *** miguelgrinberg has quit IRC | 19:12 | |
| *** lbragstad_ is now known as lbragstad | 19:12 | |
| *** miguelgrinberg_ is now known as miguelgrinberg | 19:12 | |
| *** mrodden_ is now known as mrodden | 19:12 | |
| *** pabelanger_ is now known as pabelanger | 19:12 | |
| *** itsuugo_ is now known as itsuugo | 19:12 | |
| *** skamithi has joined #openstack-ansible | 19:13 | |
| Nepoc | I see how it's working (I think) the "neutron_fwaas: true" gets picked up by the script when it runs through the roles. | 19:14 |
| cloudnull | yes | 19:14 |
| bgmccollum | logan- updated etherpad, i think its as simple as that -- https://etherpad.openstack.org/p/osad-lxc-hyper-notes | 19:14 |
| logan- | tyvm! bookmarked for when I get some time :) | 19:14 |
| Nepoc | cloudnull: Another thanks for your help! | 19:15 |
| cloudnull | anytime | 19:15 |
| *** mattt1 has joined #openstack-ansible | 19:16 | |
| *** arbrandes has joined #openstack-ansible | 19:16 | |
| *** javeriak_ has quit IRC | 19:16 | |
| *** skamithi has left #openstack-ansible | 19:19 | |
| palendae | Bjoern: Left a question for you on https://bugs.launchpad.net/openstack-ansible/+bug/1540531 | 19:23 |
| openstack | Launchpad bug 1540531 in openstack-ansible "Upgrade issues with 'Create log aggregation links'" [Undecided,New] - Assigned to Nolan Brubaker (nolan-brubaker) | 19:23 |
| *** Bjoern is now known as Bjoern_zZzZzZzZ | 19:27 | |
| *** dmsimard has joined #openstack-ansible | 19:32 | |
| *** Bjoern_zZzZzZzZ is now known as Bjoern | 19:36 | |
| *** dmsimard has quit IRC | 19:37 | |
| *** jwagner is now known as jwanger_lunch | 19:39 | |
| *** dmsimard has joined #openstack-ansible | 19:41 | |
| *** oneswig has joined #openstack-ansible | 19:41 | |
| *** dmsimard has quit IRC | 19:41 | |
| Nepoc | cloudnull: I made that change to novnc but I get "console is currently unavailable" in the browser and I'm seeing this in the horizon logs: "Recoverable error: No available console found. | 19:44 |
| Nepoc | " | 19:44 |
| cloudnull | are you using haproxy ? | 19:44 |
| cloudnull | you may need to rerun that role to setup the vip . | 19:44 |
| Nepoc | ahh that's likely it | 19:44 |
| *** dmsimard has joined #openstack-ansible | 19:46 | |
| Nepoc | cloudnull: still the same issue, I see novnc running on the nova-console container and haproxy looks good | 19:54 |
| Nepoc | cloudnull: I can hit the novnc directly | 19:55 |
| cloudnull | can you build a new vm to test | 19:55 |
| Nepoc | indeed, I was just thinking of testing that too. | 19:55 |
| cloudnull | i bet its an issue with the old vm trying to hit a console type thats no longer available, which is an issue when switching console types | 19:56 |
| cloudnull | the libvirt.xml wont see the new changes. | 19:56 |
| cloudnull | for your old vms | 19:56 |
| Nepoc | So I would need to manually make those changes or just restart the vm? | 19:58 |
| *** KLevenstein has quit IRC | 20:01 | |
| cloudnull | Nepoc: i thnk the change has to be in the DB however i've not spent much time figuring out exactly what would be needed to migrate it | 20:02 |
| *** KLevenstein has joined #openstack-ansible | 20:02 | |
| Nepoc | Okay no problem, thank you for help | 20:03 |
| Nepoc | the help | 20:03 |
| bgmccollum | i wonder if migrating the VM would rewrite the libvirt.xml with the new console type... | 20:19 |
| palendae | cloudnull: With the upgrade to MariaDB 10, what was the reasoning for including xtrabackup2 by default? | 20:20 |
| cloudnull | we stopped using the system provided version and set it to a downloadable deb | 20:22 |
| palendae | Ok. Trying to provide notes for docs, all I can find is that it was dropped cause it was old | 20:23 |
| cloudnull | this was because the system version broke one night after the mariadb repos released a busted version to stable | 20:23 |
| cloudnull | that was right before the liberty release, like the evening of (if i remember right). | 20:23 |
| palendae | Ok. That was v1, which isn't compatible with the 10.x line anyway, right? | 20:24 |
| palendae | And iirc xtrabackup has something to do with replication? | 20:24 |
| cloudnull | it was v2 | 20:25 |
| cloudnull | but the maria folks released v2.3 which busted everything | 20:26 |
| cloudnull | so we stayed on 2.2 | 20:26 |
| palendae | Ok. So there's not much to it other than to keep MariaDB and xtrabackup in sync so updating one doesn't break the other? | 20:26 |
| cloudnull | yup | 20:27 |
| palendae | Cool, thanks | 20:27 |
| cloudnull | palendae: i'd find you the commit , but gerrit is sucking right now | 20:28 |
| SamYaple | cloudnull: you should gertty! | 20:28 |
| palendae | cloudnull: I found the update one and the spec, but they didn't really say much other than xtrabackup was old | 20:28 |
| palendae | I just did git log --grep locally | 20:28 |
| palendae | Documenting changed/added vars for docs | 20:28 |
| cloudnull | SamYaple: i gave gertty a go not long ago, it was ok. | 20:29 |
| palendae | https://review.openstack.org/#/c/227839/ <- gold standard | 20:29 |
| cloudnull | i found some of it frustrating | 20:29 |
| SamYaple | cloudnull: yea i use it about %70 of the time. its not a full replacement | 20:29 |
| palendae | SamYaple: Last I had looked, it was for syncing in-flight reviews. It doesn't have a search feature like the web interface, does it? | 20:30 |
| *** shanec has quit IRC | 20:32 | |
| SamYaple | palendae: no it doesnt have a search feature afaik you are right. honestly i just saw "gerrit problems" and said "gertty" :) not a huge fain of 2.0 web interface | 20:32 |
| palendae | :p | 20:32 |
| palendae | Yeah, for that, git log --grep anyway | 20:32 |
| *** shanec has joined #openstack-ansible | 20:33 | |
| Nepoc | If I have a single Galera container and when it gets rebooted I always need to run the galera-bootstrap to get it up and running. Is there a way to fix that? (not including adding more galera containers) | 20:33 |
| *** oneswig has quit IRC | 20:34 | |
| cloudnull | Nepoc: if you add >2 containers the cluster will stay online if you reboot a single node | 20:34 |
| palendae | Nepoc: afaik, if you only have one galera node, it'll have to be re-bootstrapped cause there's no other machines to get the cluster info from | 20:34 |
| palendae | The reference design does 3 for that reason, and to establish quorum | 20:34 |
| cloudnull | >=3 is the best way to maintain cluster integrity | 20:35 |
| cloudnull | what palendae said | 20:35 |
| SamYaple | palendae: no it shouldn't need to be bootstraped with one galera node | 20:35 |
| bgmccollum | gerrit 2.11 looks like web 1.0 | 20:35 |
| SamYaple | the config itself should have "gcomm://" with no address (essentially --wsrep-new-cluster) | 20:35 |
| palendae | I think that's what our "bootstrap" does | 20:36 |
| palendae | Basically start without trying to connect to an existing cluster cause there isn't one | 20:36 |
| *** oneswig has joined #openstack-ansible | 20:37 | |
| Nepoc | palendae: okay I'll try that out "wsrep_cluster_address = gcomm:// | 20:38 |
| Nepoc | " ? | 20:38 |
| *** cemmason has joined #openstack-ansible | 20:39 | |
| palendae | Nepoc: Sounds like | 20:39 |
| *** eil397 has joined #openstack-ansible | 20:44 | |
| izaakk | Hello all!, I'm trying to do an AIO behind a proxy, I did configure the /etc/openstack_deploy/user_variables.yml with the proxy env vars, but it seems to work only for the host, when the script starts the basic lxc host setup looks like the proxy settings are not applied since its not able to get pip, any suggestion? | 20:46 |
| *** sdake has joined #openstack-ansible | 20:47 | |
| palendae | izaakk: So the repo container doesn't get the env vars? | 20:48 |
| izaakk | palendae: I think so, how can I check that? | 20:49 |
| *** baker has quit IRC | 20:49 | |
| palendae | izaakk: lxc-ls to see if the repo container is there, lxc-attach -n <container_name> | 20:49 |
| palendae | Then you can look for the pip info | 20:50 |
| izaakk | palendae: uh, weird, when I do lxc-ls there is no output.. | 20:54 |
| palendae | izaakk: What did you run/ | 20:55 |
| palendae | izaakk: The containers should be made fairly early | 20:57 |
| izaakk | I first ran the scripts/bootstrap_ansible.sh & bootstrap_aio.sh, then went to the user_vars to set the proxy. I ran the run_playbooks.sh. After like one minute, when it's at the Basic lxc host setup fails at the Get Modern PIP TASK | 20:59 |
| palendae | Ah, ok | 21:00 |
| palendae | So it probably hasn't set up the containers yet | 21:00 |
| *** baker has joined #openstack-ansible | 21:02 | |
| *** mattt1 has quit IRC | 21:02 | |
| *** mattt1 has joined #openstack-ansible | 21:02 | |
| palendae | izaakk: What docs did you reference for the proxy settings? It's been a while since I've used htem | 21:02 |
| *** mattt1 is now known as mattt | 21:02 | |
| izaakk | palendae: I didn't find anything in the docs, but I found in the file /etc/openstack_deploy/user_variables.yml in the repo, the following: http://paste.openstack.org/show/485772/ | 21:07 |
| *** spotz_zzz is now known as spotz | 21:11 | |
| *** raddaoui has quit IRC | 21:12 | |
| palendae | izaakk: Ok. Mind pasting the error on paste.openstack.org? | 21:15 |
| *** CheKoLyN has joined #openstack-ansible | 21:15 | |
| izaakk | palendae: sure thing, http://paste.openstack.org/show/485773/ | 21:19 |
| palendae | izaakk: What's in your /root/.pip/pip.conf? | 21:21 |
| izaakk | palendae: then, retries with -vvvv: http://paste.openstack.org/show/485774/ | 21:21 |
| *** raddaoui has joined #openstack-ansible | 21:21 | |
| izaakk | palendae: where can I find the pip.conf? I don't have one in /root | 21:24 |
| palendae | Oh, hm. can you do a set and see if the proxy vars are really in the environment? | 21:25 |
| palendae | Though if they're in the user vars they might be passed in. Your verbose output doesn't appear to have any proxy info, but I don't remember what the output looks like | 21:25 |
| izaakk | palendae: you mean, set export? | 21:27 |
| palendae | Using `set` should dump all the environment variables | 21:27 |
| palendae | Might also help to see your user_variables | 21:28 |
| palendae | Also, what version are you deploying? | 21:28 |
| *** sdake has quit IRC | 21:29 | |
| izaakk | palendae: I'm deploying master | 21:31 |
| *** cemmason has quit IRC | 21:36 | |
| *** spotz is now known as spotz_zzz | 21:36 | |
| izaakk | palendae: for the user_variables, I just added the following lines : http://paste.openstack.org/show/485775/ | 21:37 |
| palendae | izaakk: Ok. Did you sanitize that proxy address, or is it literally proxy.server.com? | 21:38 |
| izaakk | palendae: I did sanitize it hehe | 21:39 |
| palendae | For sanity, can you telnet to that machine/port? | 21:39 |
| palendae | Or ping it | 21:39 |
| palendae | Re-reading your error, looks like the SSL cert coming back from bootstrap.pypa.io:443 for you is invalid. Wonder if the proxy's not passing that correctly | 21:40 |
| *** baker has quit IRC | 21:41 | |
| *** spotz_zzz is now known as spotz | 21:42 | |
| izaakk | palendae: yeah, that's the proxy server I use for other stuff and it's working.. huhh weird | 21:45 |
| palendae | izaakk: Is it working with get-pip.py outside of the OSA stuff? | 21:45 |
| odyssey4me | hmm, this is not a new issue | 21:46 |
| odyssey4me | it generally happens when the proxy is being naughty and intercepting the cert | 21:46 |
| palendae | It's not one I've encountered, I don't work with proxies much | 21:46 |
| palendae | Yeah, that's what I thought | 21:46 |
| palendae | izaakk: Is it a corporate or personal proxy? | 21:46 |
| izaakk | palendae: Corporate | 21:47 |
| palendae | Ok, figured :) | 21:47 |
| odyssey4me | we need to turn the lines like this into something configurable: https://github.com/openstack/openstack-ansible-pip_install/blob/master/tasks/main.yml#L46 | 21:47 |
| palendae | Ah, yeah - izaakk You could try changing that line to no =\ | 21:47 |
| odyssey4me | there're a few of those | 21:47 |
| odyssey4me | izaakk this would be a good first patch :) | 21:48 |
| palendae | Might get you further...and then blow up | 21:48 |
| palendae | But yeah, pip stuff's spread around a little | 21:48 |
| odyssey4me | no, I mean that there are a few of those in that task file | 21:48 |
| palendae | Oh | 21:48 |
| * palendae is running out of steam | 21:49 | |
| odyssey4me | izaakk something else we've learned is that you may need to set http_proxy, HTTP_PROXY, etc in order for it to work | 21:50 |
| *** oneswig has quit IRC | 21:50 | |
| odyssey4me | ie capitals and smalls... some linux bits respect the one set, and others the other set | 21:50 |
| palendae | The global_environment_variables bit only did the all caps | 21:50 |
| *** baker has joined #openstack-ansible | 21:51 | |
| palendae | odyssey4me: Before I forget, I tossed a note on https://etherpad.openstack.org/p/openstack-ansible-upgrade-framework about env.d possibly changing between releases. It occurred to me we hadn't accounted for it in Liberty yet, but I also figured it's gonna be something to at least check every version | 21:51 |
| izaakk | odyssey4me: alright, will try to change that and submit the patch. Will try to add both, upper and lower case. | 21:51 |
| izaakk | thanks palendae and odyssey4me !! | 21:51 |
| palendae | Not sure if it belongs on the same line or separate, since env.d changes could be to existing containers, not necessarily new ones | 21:52 |
| palendae | izaakk: Sure thing | 21:52 |
| odyssey4me | palendae yes, good call - the trouble with env.d is that it's up to the deployer how they want to lay things out... but we should have a way of informing them of new environment bits | 21:53 |
| palendae | odyssey4me: Right, that's what I was puzzling through over lunch | 21:53 |
| palendae | I know it needs to happen, I just don't know how to go about it yet | 21:53 |
| odyssey4me | palendae now with config_template I'm starting to wonder whether we shouldn't do away with env.d/conf.d and just use config_template to modify the stock setup | 21:53 |
| palendae | once copied, the project doesn't "own" env.d anymore | 21:53 |
| odyssey4me | palendae I would recommend, as a starting point, just release notes to inform of changes - from there, perhaps some sort of 'inspector' scripts could compare what a deployer has to the stock bits | 21:54 |
| *** metral has quit IRC | 21:54 | |
| odyssey4me | then they can actively make the changes they want, or spot changes they made on purpose | 21:54 |
| palendae | Yeah. I know git-harry did an 'update_yaml.py' script for rpc-openstack that merged/overwrote default values for the config | 21:55 |
| logan- | update-yaml.py is awesome | 21:55 |
| palendae | If not config_template, maybe something similar could be useful | 21:55 |
| odyssey4me | but yeah - perhaps we should switch to using the stock bits by default, and the conf.d/env.d entries can be overrides while we deprecate that functionality | 21:55 |
| palendae | Hm, yeah | 21:55 |
| palendae | So we have a stock thing that we own | 21:55 |
| palendae | user_env.d :p | 21:55 |
| odyssey4me | or yes - the update-yaml thing could have an 'inspector' mode which tells you how things will change | 21:56 |
| palendae | Ok, for liberty I think that's sailed | 21:56 |
| odyssey4me | yep, but these effects are things that need to feed into a more sustainable framework that's ongoing | 21:56 |
| palendae | Yeah, I've got some of that for the variable changes in defaults/main.yml for ansible roles, but I basically rely on anisgenome to fish that stuff out for me to do some comparisons on | 21:56 |
| palendae | Right | 21:56 |
| palendae | I' | 21:56 |
| *** sdake has joined #openstack-ansible | 21:57 | |
| palendae | I'm trying to keep a mind towards what we'll need in the future, even if I don't do it now | 21:57 |
| palendae | Like renaming/deprecating vars | 21:57 |
| odyssey4me | palendae I dunno if you've seen https://github.com/openstack/ansible-role-diskimage-builder - but I quite like how the defaults are things you override, and the vars are things you don't | 21:58 |
| palendae | I have a pipeline of scripts that'll compare variables now, but I'm not adding them to the project because I'm not sure they fit, and also they require doing git checkouts right now...and they won't get added back to kilo | 21:58 |
| palendae | I have not looked at those closely, no | 21:58 |
| *** jthorne has quit IRC | 21:59 | |
| odyssey4me | palendae keep them somewhere public - they can feed into the conversation when we get back to it | 21:59 |
| palendae | I'll toss em in a repo | 21:59 |
| odyssey4me | my ideal is that every patch considers docs/releasenotes/upgrades with it | 21:59 |
| palendae | Right | 21:59 |
| odyssey4me | but for that to happen, we need a framework to slot that consideration into it | 22:00 |
| palendae | The hope would be they'd not be relevant for too many releases | 22:00 |
| palendae | I'd hoped we could get upgrade stuff in with each thing in liberty, but obviously didn't happen | 22:00 |
| odyssey4me | yeah, such is life - but we're learning | 22:00 |
| odyssey4me | with each step we learn newer things | 22:01 |
| palendae | Yeah | 22:01 |
| palendae | This process, thus far, has been better than Juno -> Kilo | 22:01 |
| odyssey4me | we almost have some breathing room to actually step back and look around for inspiration... and I think that may just be possible next cycle... at least with one or two streams of work | 22:02 |
| *** shanec has quit IRC | 22:02 | |
| odyssey4me | I can see us doing quite a bit of refactoring to take advantage of Ansible 2, but also at the same time applying what we think is a good set of patterns consistently across roles. | 22:03 |
| stevelle | inspiration / consternation. same same. | 22:03 |
| palendae | The thing that's going to get tricky is the integration gate staying stable | 22:03 |
| *** metral_zzz has joined #openstack-ansible | 22:03 | |
| *** metral_zzz is now known as metral | 22:03 | |
| *** Mudpuppy_ has joined #openstack-ansible | 22:03 | |
| palendae | I've noticed when a separated role causes gate failure, it adds a step of tracking to get it fixed | 22:03 |
| palendae | Such is the nature though | 22:04 |
| palendae | And we're early days | 22:04 |
| odyssey4me | yeah, that's a new set of things to learn - but we'll get better at it | 22:04 |
| stevelle | should each role have an integration check run on osa? | 22:04 |
| odyssey4me | stevelle we need to figure out what we think is the right way to do it | 22:05 |
| stevelle | I like the idea of having it as an experimental check on each role repo at least | 22:05 |
| odyssey4me | in my mind it makes more sense to treat many of the roles like libraries - they must have their own tests, then when we tag them an integration test should run to validate that they work | 22:05 |
| *** woodard has quit IRC | 22:05 | |
| odyssey4me | many of the OpenStack roles rely on the other OS roles anyway, as do many of the non OS roles | 22:06 |
| palendae | Right | 22:06 |
| odyssey4me | so they do a sort-of integration test already - but just using a minimum set of roles to achieve the goal of functional testing | 22:06 |
| *** Mudpuppy has quit IRC | 22:06 | |
| odyssey4me | eg the Galera role implements a cluster using the lxc hosts and lxc container create roles | 22:07 |
| odyssey4me | it then runs a barrel of tests to validate that the cluster is working | 22:07 |
| odyssey4me | same with rabbit | 22:07 |
| stevelle | With gate split I suspect we might be able to do some interesting work to get tempest tests running for the os_role repos at least | 22:07 |
| odyssey4me | both of those have had more testing added since they were split out - with both we discovered shortfalls and ended up improving the setups in liberty & kilo as a result | 22:08 |
| *** Mudpuppy_ has quit IRC | 22:08 | |
| odyssey4me | yeah, cloudnull is trying to figure all that out | 22:08 |
| cloudnull | whats that ? | 22:08 |
| odyssey4me | it's likely to require some restructuring too | 22:08 |
| stevelle | It works for the infra, for sure, but you're not going to meaningfully test cinder by itself | 22:08 |
| stevelle | smoke test maybe | 22:09 |
| odyssey4me | speaking of which, we need some more feedback in https://review.openstack.org/272764 please | 22:09 |
| odyssey4me | palendae stevelle ^ | 22:09 |
| cloudnull | this one too https://review.openstack.org/#/c/271423/ =) | 22:09 |
| odyssey4me | stevelle well, cinder could conceivably tested with a minimal nova/glance build | 22:10 |
| odyssey4me | nova/glance/keystone | 22:10 |
| stevelle | odyssey4me: that's closer to my thinking yes. | 22:10 |
| odyssey4me | so that's the general idea | 22:10 |
| stevelle | one should be able to do tempest tests for all 3 of them in the one build | 22:10 |
| odyssey4me | which means that each role does an integration test of sorts - it's different to the production build-out though, and that's a good thing | 22:11 |
| stevelle | and should run that set of tempest tests for any change to any of the 3 roles | 22:11 |
| odyssey4me | no dynamic inventory, less containers, less complexity | 22:11 |
| *** raddaoui has quit IRC | 22:11 | |
| odyssey4me | quicker feedback | 22:11 |
| *** baker_ has joined #openstack-ansible | 22:13 | |
| *** baker has quit IRC | 22:13 | |
| odyssey4me | anyway, I came on for a cross project meeting which was postponed :) | 22:14 |
| cloudnull | stevelle: so far my plan of attack was to do the independent roles for all the things and add tests much like we have. being that I wasnt sure what to test I thought I'd have each role execute its part of the smoke tests we currently do to ensure functionality. | 22:14 |
| odyssey4me | so yeah, let me get back to !computering :) | 22:14 |
| cloudnull | chao | 22:14 |
| palendae | odyssey4me: Do that :) | 22:14 |
| *** KLevenstein has quit IRC | 22:14 | |
| odyssey4me | good chat all :) thanks for the heads up palendae :) | 22:15 |
| odyssey4me | ciao | 22:15 |
| *** rev_dr has quit IRC | 22:16 | |
| *** KLevenstein has joined #openstack-ansible | 22:18 | |
| *** retreved has quit IRC | 22:21 | |
| *** admin0 has joined #openstack-ansible | 22:22 | |
| cfarquhar | are there any playbooks in liberty that aren't idempotent? | 22:24 |
| *** sdake has quit IRC | 22:24 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: [WIP] Add support for upgrading from Kilo to Liberty https://review.openstack.org/272182 | 22:25 |
| *** sdake has joined #openstack-ansible | 22:26 | |
| *** phiche1 has joined #openstack-ansible | 22:33 | |
| admin0 | setup-hosts set's up the containers right ? but the IP is natted and not the br-mgmt , so setup-infra fails ?? | 22:33 |
| *** jwanger_lunch is now known as jwagner | 22:33 | |
| admin0 | my deploy server has access to just the mgmt range only | 22:33 |
| *** phiche has quit IRC | 22:34 | |
| *** oneswig has joined #openstack-ansible | 22:35 | |
| *** vCloudernBeer has joined #openstack-ansible | 22:38 | |
| vCloudernBeer | ping odyssey4me | 22:38 |
| admin0 | ping anyone :) | 22:39 |
| *** phiche has joined #openstack-ansible | 22:39 | |
| vCloudernBeer | admin0 I am here | 22:39 |
| *** oneswig has quit IRC | 22:39 | |
| *** phiche1 has quit IRC | 22:40 | |
| admin0 | vCloudernBeer: my ansible server has access to br-mgmt .... i run ansible playbooks .. setup-hosts work fine without issue .. i see containers running on all 3 infra servers , but they have the lxcbr0 ip only | 22:40 |
| admin0 | when i run setup-infra, it fails | 22:40 |
| admin0 | and this is ood | 22:40 |
| admin0 | odd | 22:40 |
| vCloudernBeer | admin0 I am new and do not know the answer. I am still exploring | 22:41 |
| *** jamielennox|away is now known as jamielennox | 22:41 | |
| admin0 | ok | 22:41 |
| vCloudernBeer | admin0 When I install the all-in-one, I see a whole bunch of new interfaces created on my Linux machine. | 22:41 |
| admin0 | vCloudernBeer: i am not using the all-in-one, but real different machines .. | 22:42 |
| admin0 | i have never even used the all-in-one actually | 22:42 |
| vCloudernBeer | admin0 Yeah, you mentioned that. I have limited hardware for my home lab | 22:42 |
| vCloudernBeer | so I can only do the AIO version to play around | 22:43 |
| admin0 | you can also do multiple virtualboxes | 22:43 |
| vCloudernBeer | admin0 I tried but not enough horse power (RAM on my machine). :( I am doing this on my own after work. No company sponsor | 22:44 |
| *** shanec has joined #openstack-ansible | 22:45 | |
| admin0 | this is also my home setup .. got a few boxes to play wtih .. so every evening after 11 when power is cheaper, turn ON ,, deploy .. test stuff .. power down :) | 22:46 |
| vCloudernBeer | admin0 that is good. :) | 22:47 |
| *** shanec has quit IRC | 22:52 | |
| *** pegmanm_ has joined #openstack-ansible | 22:53 | |
| admin0 | setup-hosts, it brings up the containers, but the containers have just 10.0 lxcbridge ip .. and in inventory json, the ssh host for the containers is null .. so my setup-infra fails | 22:53 |
| admin0 | but i fail to find out how to fix it | 22:53 |
| admin0 | since lxc-ls -f show the containers just have the nat range and no other ips in the br-mgmt range | 22:53 |
| *** pegmanm has quit IRC | 22:56 | |
| *** b3rnard0 is now known as b3rnard0_away | 22:58 | |
| *** CheKoLyN has quit IRC | 23:02 | |
| *** jamielennox is now known as jamielennox|away | 23:05 | |
| *** phiche has quit IRC | 23:05 | |
| *** scarlisle has quit IRC | 23:07 | |
| openstackgerrit | Bjoern Teipel proposed openstack/openstack-ansible: Replacing LBaaSv1 with v2 configuration https://review.openstack.org/275440 | 23:19 |
| *** shanec has joined #openstack-ansible | 23:19 | |
| *** admin0 has quit IRC | 23:20 | |
| *** vCloudernBeer has quit IRC | 23:21 | |
| *** michaelgugino has quit IRC | 23:23 | |
| *** KLevenstein has quit IRC | 23:44 | |
| *** galstrom is now known as galstrom_zzz | 23:44 | |
| *** baker_ has quit IRC | 23:46 | |
| *** baker has joined #openstack-ansible | 23:47 | |
| *** sdake has quit IRC | 23:48 | |
| *** baker has quit IRC | 23:54 | |
| *** jeh has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!