| *** markvoelker has quit IRC | 00:01 | |
| *** chandanc_ has joined #openstack-ansible | 00:53 | |
| *** chandanc has joined #openstack-ansible | 00:54 | |
| *** thorst has joined #openstack-ansible | 01:01 | |
| *** fishcried has joined #openstack-ansible | 01:04 | |
| *** thorst has quit IRC | 01:04 | |
| *** thorst has joined #openstack-ansible | 01:05 | |
| *** thorst has quit IRC | 01:13 | |
| *** cfarquhar has quit IRC | 01:24 | |
| *** thorst has joined #openstack-ansible | 01:26 | |
| *** chandanc has quit IRC | 01:26 | |
| *** chandanc_ has quit IRC | 01:26 | |
| *** thorst has quit IRC | 01:26 | |
| *** thorst has joined #openstack-ansible | 01:27 | |
| *** cfarquhar has joined #openstack-ansible | 01:29 | |
| *** cfarquhar has quit IRC | 01:29 | |
| *** cfarquhar has joined #openstack-ansible | 01:29 | |
| *** thorst has quit IRC | 01:31 | |
| *** fishcried1 has joined #openstack-ansible | 01:45 | |
| *** jamesdenton has joined #openstack-ansible | 01:46 | |
| *** fishcried has quit IRC | 01:47 | |
| *** fishcried1 is now known as fishcried | 01:47 | |
| *** markvoelker has joined #openstack-ansible | 01:58 | |
| *** markvoelker has quit IRC | 02:02 | |
| *** jamesdenton has quit IRC | 02:15 | |
| *** sacharya has joined #openstack-ansible | 02:27 | |
| *** thorst has joined #openstack-ansible | 02:30 | |
| *** thorst has quit IRC | 02:30 | |
| *** thorst has joined #openstack-ansible | 02:30 | |
| *** thorst has quit IRC | 02:36 | |
| *** chandanc has joined #openstack-ansible | 02:36 | |
| *** chandanc_ has joined #openstack-ansible | 02:36 | |
| *** chandanc__ has joined #openstack-ansible | 03:19 | |
| *** chandanc_ has quit IRC | 03:22 | |
| *** chandanc has quit IRC | 03:22 | |
| *** chandanc has joined #openstack-ansible | 03:22 | |
| *** Guest10310 has quit IRC | 03:28 | |
| *** thorst has joined #openstack-ansible | 03:34 | |
| *** thorst has quit IRC | 03:41 | |
| *** fishcried1 has joined #openstack-ansible | 03:50 | |
| *** fishcried has quit IRC | 03:51 | |
| *** fishcried1 has quit IRC | 03:52 | |
| *** markvoelker has joined #openstack-ansible | 03:59 | |
| *** markvoelker has quit IRC | 04:03 | |
| *** sacharya has quit IRC | 04:10 | |
| *** bbmbx__ has joined #openstack-ansible | 04:24 | |
| *** bbmbx has quit IRC | 04:27 | |
| *** thorst has joined #openstack-ansible | 04:39 | |
| *** thorst has quit IRC | 04:46 | |
| *** fishcried has joined #openstack-ansible | 05:25 | |
| *** saneax_AFK is now known as saneax | 05:32 | |
| *** saneax is now known as saneax_AFK | 05:32 | |
| *** thorst has joined #openstack-ansible | 05:43 | |
| *** McMurlock has joined #openstack-ansible | 05:48 | |
| *** saneax_AFK is now known as saneax | 05:48 | |
| *** thorst has quit IRC | 05:51 | |
| *** M00nr41n has quit IRC | 05:56 | |
| *** markvoelker has joined #openstack-ansible | 05:59 | |
| *** fishcried has quit IRC | 06:02 | |
| *** markvoelker has quit IRC | 06:04 | |
| *** thorst has joined #openstack-ansible | 06:50 | |
| *** thorst has quit IRC | 06:56 | |
| *** george_goh has quit IRC | 06:57 | |
| *** george_goh has joined #openstack-ansible | 06:58 | |
| *** javeriak has joined #openstack-ansible | 07:00 | |
| *** agireud has quit IRC | 07:04 | |
| *** agireud has joined #openstack-ansible | 07:07 | |
| *** javeriak has quit IRC | 07:07 | |
| *** elopez has joined #openstack-ansible | 07:15 | |
| *** elopez is now known as Guest90337 | 07:16 | |
| *** zerda2 has joined #openstack-ansible | 07:26 | |
| *** bsv has joined #openstack-ansible | 07:29 | |
| *** mikelk has joined #openstack-ansible | 07:38 | |
| *** mikelk has quit IRC | 07:41 | |
| *** Oku_OS-away is now known as Oku_OS | 07:45 | |
| *** mikelk has joined #openstack-ansible | 07:47 | |
| *** chandanc__ has quit IRC | 07:48 | |
| *** chandanc has quit IRC | 07:49 | |
| *** thorst has joined #openstack-ansible | 07:53 | |
| *** M00nr41n has joined #openstack-ansible | 07:55 | |
| *** permalac has joined #openstack-ansible | 08:00 | |
| *** markvoelker has joined #openstack-ansible | 08:00 | |
| *** thorst has quit IRC | 08:01 | |
| *** markvoelker has quit IRC | 08:05 | |
| *** electrofelix has joined #openstack-ansible | 08:06 | |
| *** javeriak has joined #openstack-ansible | 08:08 | |
| *** chandanc__ has joined #openstack-ansible | 08:12 | |
| *** chandanc has joined #openstack-ansible | 08:15 | |
| *** javeriak has quit IRC | 08:21 | |
| *** zerda2 has quit IRC | 08:27 | |
| *** thng has joined #openstack-ansible | 08:27 | |
| thng | Hi all | 08:27 |
|---|---|---|
| thng | Anyone can tell me where can I set globally proxy settings ? I did it in users_variables.yml but it does not work for pip_install role while trying to get "Modern PIP" | 08:28 |
| *** javeriak has joined #openstack-ansible | 08:29 | |
| *** javeriak has quit IRC | 08:30 | |
| *** javeriak has joined #openstack-ansible | 08:30 | |
| pjm6 | Good morning | 08:43 |
| pjm6 | thng, are you using haproxy?= | 08:43 |
| thng | nope | 08:44 |
| thng | I try to use aio | 08:44 |
| thng | and my server is behind a proxy | 08:45 |
| pjm6 | you use this guide | 08:46 |
| pjm6 | http://docs.openstack.org/developer/openstack-ansible/install-guide/app-no-internet-connectivity.html#practice-b-proxy-access-to-internet-resources ? | 08:46 |
| *** tiagogomes has joined #openstack-ansible | 08:47 | |
| thng | pjm6: Yes, I set all proxy settings already in /etc/openstack_deploy/user_variables.yml. But it still does not have any effect | 08:48 |
| pjm6 | thng, did you try enter in the container and try to access the internet? | 08:49 |
| pjm6 | that configuration was in the beginning ? Before you run setup-hosts | 08:49 |
| pjm6 | (forget about setup-hosts, you're using AIO) | 08:50 |
| thng | I think I did an error in user_variables.yml. I set up variables HTTP_PROXY etc. but I didnot see to uncomment global_environment_variables: | 08:54 |
| thng | I will do a test | 08:54 |
| *** admin0 has joined #openstack-ansible | 08:57 | |
| pjm6 | that could be a reason for it not being working | 08:58 |
| pjm6 | automagically, are you therE?= | 08:58 |
| *** thorst has joined #openstack-ansible | 09:00 | |
| thng | I still doesnot work... I got now http_proxy etc. in /etc/environment | 09:02 |
| thng | TASK: [pip_install | Get Modern PIP] ****************************************** <172.29.236.100> ESTABLISH CONNECTION FOR USER: root <172.29.236.100> REMOTE_MODULE get_url validate_certs=yes force=yes dest=/opt/get-pip.py url=https://bootstrap.pypa.io/get-pip.py <172.29.236.100> EXEC ssh -C -v -o ControlMaster=auto -o ControlPersist=60s -o ControlPath="/root/.ansible/cp/ansible-ssh-%h-%p-%r" -o StrictHostKeyChecking=no -o Kb | 09:02 |
| thng | 3rd time already | 09:03 |
| *** javeriak has quit IRC | 09:03 | |
| *** javeriak has joined #openstack-ansible | 09:03 | |
| pjm6 | weird >S | 09:04 |
| thng | Result from run 1 is: {u'msg': u'Failed to validate the SSL certificate for bootstrap.pypa.io:443. Make sure your managed systems have a valid CA certificate installed. If the website serving the url uses SNI you need python >= 2.7.9 on your managed machine. You can use validate_certs=False if you do not need to confirm the server\\s identity but this is unsafe and not recommended Paths checked for this platform: /etc/ssl/cer | 09:05 |
| thng | I got this error while enabling -vvv in ansible : | 09:05 |
| *** thorst has quit IRC | 09:06 | |
| pjm6 | well it seems that you've problems with certificates | 09:07 |
| pjm6 | if you do lxc-ls do you see any container? | 09:07 |
| thng | No, lxc-ls isnot already installed yet at this state. I found the problem (i hope) | 09:12 |
| thng | I must disable and validate_certs in all roles | 09:12 |
| thng | while using proxy | 09:12 |
| thng | pip_install works now. | 09:13 |
| pjm6 | yes, for some reason the CA isn't being readed correctly | 09:16 |
| thng | hoho I have to modify some roles manually :D | 09:19 |
| pjm6 | if you just disable HTTPS | 09:20 |
| pjm6 | don't solve your probleM? | 09:20 |
| thng | https of which ? | 09:21 |
| pjm6 | apt repositories | 09:21 |
| pjm6 | i'm trying to find that config file | 09:21 |
| thng | it is not easier because this option is set by default to yes in lxc_hosts, galera_server and pip_install at the first sign in /etc/ansible/roles | 09:23 |
| *** pcaruana has joined #openstack-ansible | 09:27 | |
| pjm6 | yeah, it seems that are not a easy way to do it :s | 09:28 |
| *** M00nr41n has quit IRC | 09:28 | |
| thng | :) anyway, thanks for your help :D | 09:31 |
| thng | it is dling lxc_cache.... donnt know if there is any prob else :D | 09:32 |
| *** pjm6_ has joined #openstack-ansible | 09:34 | |
| *** pjm6 has quit IRC | 09:38 | |
| *** gparaskevas has joined #openstack-ansible | 09:38 | |
| *** zerda2 has joined #openstack-ansible | 09:42 | |
| ioni | hey guys | 09:49 |
| ioni | if i want to install additional packages on compute hosts, how should I do that? | 09:49 |
| ioni | i need to add a new role in /etc/ansible right | 09:51 |
| *** Bofu2U has quit IRC | 09:51 | |
| *** thorst has joined #openstack-ansible | 10:04 | |
| *** pjm6_ is now known as pjm6 | 10:10 | |
| *** thorst has quit IRC | 10:11 | |
| *** javeriak has quit IRC | 10:13 | |
| evrardjp | good morning everyone | 10:15 |
| *** javeriak has joined #openstack-ansible | 10:26 | |
| *** karimb has joined #openstack-ansible | 10:29 | |
| *** zerda2 has quit IRC | 10:34 | |
| *** mikelk has quit IRC | 10:41 | |
| pjm6 | good morning evrardjp | 10:46 |
| pjm6 | evrardjp, it was you that have done some work in horizon? | 10:46 |
| evrardjp | depends | 10:55 |
| evrardjp | I know the role a little | 10:55 |
| evrardjp | :D | 10:55 |
| evrardjp | pjm6: why? | 10:58 |
| pjm6 | I was talking about the project xd | 10:58 |
| pjm6 | because i'm trying to do testing in a dashboard that I created and the lack of docs (or my methods of search are bad) are not the best | 10:58 |
| *** chandanc__ has quit IRC | 10:59 | |
| *** chandanc has quit IRC | 10:59 | |
| *** admin0 has quit IRC | 11:01 | |
| *** mikelk has joined #openstack-ansible | 11:04 | |
| *** thorst has joined #openstack-ansible | 11:08 | |
| *** thorst has quit IRC | 11:15 | |
| evrardjp | stevelle: when you're there: do you remember the step where the timeout appeared for before your "recheck timeout ovh" on this commit? https://review.openstack.org/#/c/328274/ | 11:15 |
| evrardjp | It currently have failed for tempest reasons, I'll recheck. | 11:15 |
| evrardjp | Guys bad news, I think the gate is broken | 11:23 |
| evrardjp | or more, completely BORKED | 11:23 |
| evrardjp | (for the integrated builds) | 11:24 |
| pjm6 | because i'm trying to do testing in a dashboard that I created and the lack of docs (or my methods of search are bad) are not the best | 11:24 |
| pjm6 | ups sorry | 11:24 |
| pjm6 | btw: evrardjp i could do it =) | 11:24 |
| pjm6 | i'm using openstack_dashboard.test.helpers | 11:25 |
| pjm6 | the TestCase ( to simulate a normal user) and test.BaseAdminViewTests (admin users) | 11:25 |
| evrardjp | it's been a while and it probably has changed now pjm6 :D | 11:25 |
| pjm6 | evrardjp, true :) but anything might help, anyway using that I could figure out =) | 11:25 |
| pjm6 | there is a docs about how doing tests? | 11:25 |
| evrardjp | for django panel/dashboard? | 11:26 |
| pjm6 | yes | 11:26 |
| pjm6 | the docs I saw say don't have much examples, and i'm doing by seeing examples from tests of openstack_dashboard | 11:26 |
| evrardjp | I don;t know where the docs are for that | 11:26 |
| evrardjp | did you check latest dev horizon docs? | 11:26 |
| pjm6 | http://docs.openstack.org/developer/horizon/topics/testing.html | 11:27 |
| pjm6 | http://docs.openstack.org/developer/horizon/ref/test.html | 11:27 |
| pjm6 | this ones | 11:27 |
| pjm6 | maybe lack me knowledge of django :) | 11:28 |
| pjm6 | for instance, I didn't know about self.client.get and self.client.post | 11:28 |
| pjm6 | (maybe its used in django by default in tests) | 11:28 |
| evrardjp | I'm definitely not the right person for that | 11:30 |
| evrardjp | all I was using was standard unittest/mox testing with basic python scripts | 11:31 |
| evrardjp | that were used to populate my dashboard | 11:31 |
| *** Guest__ has joined #openstack-ansible | 11:32 | |
| evrardjp | I'd be happy to have feedback about the latest practices 'though | 11:33 |
| *** thorst has joined #openstack-ansible | 11:33 | |
| *** psilvad has joined #openstack-ansible | 11:38 | |
| pjm6 | evrardjp, i'm searching about how to do and what pratices xD | 11:39 |
| pjm6 | but when I know something more, i could share yes :) | 11:39 |
| evrardjp | maybe you should contact horizon guys on irc? | 11:40 |
| pjm6 | yes, already ask there, but no one replied yet | 11:41 |
| pjm6 | btw: the mox is the mock? | 11:41 |
| evrardjp | yup | 11:42 |
| *** retreved has joined #openstack-ansible | 11:43 | |
| pjm6 | that is used when we want to "abstract" an event? | 11:43 |
| pjm6 | for instance, i have a API that i communicate with an external server | 11:43 |
| pjm6 | the mock could be used to simulate that? | 11:44 |
| evrardjp | there is no magic behind it | 11:45 |
| evrardjp | so you'd have to write a mock API | 11:46 |
| *** elo has quit IRC | 11:46 | |
| evrardjp | IIRC, swagger and apiary were good for writing APIs, because it allowed to generate these | 11:47 |
| evrardjp | but it's been REALLY a while that I did this | 11:47 |
| evrardjp | and I didn't do that for long, so I am a noob on these | 11:47 |
| pjm6 | but the general idea is to reproduce/simulate events right? | 11:48 |
| evrardjp | so basically here if you didn't start with that, you'd have to create fake local results for your api, to explain what your code should expect, and then you assert the results | 11:48 |
| evrardjp | I don't know if I'm clear | 11:48 |
| evrardjp | I'm definitely not the right person :p | 11:49 |
| evrardjp | I think google will help you better than me :p | 11:50 |
| evrardjp | I am a noob on that level :p | 11:50 |
| pjm6 | yes yes evrardjp , been great help =) it was what i'm thinking | 11:51 |
| pjm6 | i didn't have sure | 11:51 |
| *** admin0 has joined #openstack-ansible | 11:51 | |
| evrardjp | pjm6: it's better to talk here | 11:56 |
| evrardjp | so you can set the affinity for your containers | 11:56 |
| evrardjp | http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-initial.html#affinity | 11:56 |
| evrardjp | this way you'll run multiple containers on a host | 11:56 |
| *** weshay has joined #openstack-ansible | 11:56 | |
| evrardjp | run 3 containers on your node, and this way you'll test the redundancy | 11:56 |
| evrardjp | ofc it's better to have 3 physical nodes instead of 3 containers on the same node | 11:57 |
| evrardjp | but it gives you an idea of how it works | 11:57 |
| *** evrardjp has left #openstack-ansible | 11:57 | |
| *** evrardjp has joined #openstack-ansible | 11:57 | |
| pjm6 | yeah, and can do that in current config? | 12:01 |
| *** markvoelker has joined #openstack-ansible | 12:01 | |
| pjm6 | it will cause no problems with sync? | 12:01 |
| mhayden | happy monday | 12:02 |
| pjm6 | hi mhayden | 12:02 |
| pjm6 | i have to run, come back later. thanks evrardjp :) | 12:03 |
| evrardjp | np, yw pjm6! | 12:04 |
| evrardjp | happy mondayversay mhayden! | 12:04 |
| evrardjp | I got my yubikeys! woot | 12:05 |
| *** markvoelker has quit IRC | 12:06 | |
| *** elo has joined #openstack-ansible | 12:08 | |
| odyssey4me | thng to turn off cert validation (because you have a naughty proxy), check http://docs.openstack.org/developer/openstack-ansible/install-guide/app-no-internet-connectivity.html#considerations-when-proxying-tls-traffic | 12:09 |
| odyssey4me | ioni there are many ways to install additional packages on your compute hosts - you can make your own playbook to do it and target the appropriate group | 12:10 |
| *** thorst has quit IRC | 12:11 | |
| odyssey4me | o/ all | 12:11 |
| odyssey4me | evrardjp yeah, I saw that something's gone broken on the weekend due to some upstream changes, I think | 12:11 |
| odyssey4me | have you found out anything interesting in your log reviews? | 12:12 |
| *** thorst has joined #openstack-ansible | 12:15 | |
| *** markvoelker has joined #openstack-ansible | 12:25 | |
| *** woodard has joined #openstack-ansible | 12:26 | |
| *** woodard has quit IRC | 12:26 | |
| thng | odyssey4me: thanks. I also got it. Now, I have just gotten this error : | 12:26 |
| thng | + ansible neutron_agent -m shell -a 'DEBIAN_FRONTEND=noninteractive apt-get install iptables-persistent' -t /openstack/log/ansible_cmd_logs//add_neutron_agent_checksum_rule ++ exit_fail 262 0 | 12:26 |
| *** woodard has joined #openstack-ansible | 12:27 | |
| thng | I havent figured out what it is | 12:27 |
| thng | just came back from lunch :D | 12:27 |
| *** elo has quit IRC | 12:30 | |
| *** Guest90337 has quit IRC | 12:31 | |
| thng | Ah it cannot download iptables-persistent from ubuntu repo because network is unreacheable.... weird. | 12:31 |
| *** Qiming_ has joined #openstack-ansible | 12:31 | |
| odyssey4me | thng it would seem that you re-ran run-playbooks a few times, so the containers have lost connectivity | 12:31 |
| odyssey4me | thng I'd suggest using the lxc-container-destroy (or some name like that) to destroy the containers, then you can re-run run-playbooks again | 12:32 |
| *** klamath has joined #openstack-ansible | 12:34 | |
| *** klamath has quit IRC | 12:34 | |
| *** klamath has joined #openstack-ansible | 12:35 | |
| thng | I re-ran several times run-playbooks but only when It couldnot download pip_install and at that moment, no container is installed.... I thought that it did not affect anything.... After, when the validation cert is disabled, I didnt rerun anymore | 12:35 |
| thng | I will destroy container as you told and rerun run-playbook as you adviced | 12:35 |
| *** Qiming_ has quit IRC | 12:35 | |
| odyssey4me | thng if the containers have been created, run-playbooks restarts the lxc bridge, which breaks container connectivity | 12:35 |
| odyssey4me | so you can either restart all the containers, or just destroy and recreate them | 12:36 |
| *** javeriak has quit IRC | 12:36 | |
| *** woodard_ has joined #openstack-ansible | 12:38 | |
| *** woodard has quit IRC | 12:38 | |
| thng | odyssey4me : do I need to stop all services, uninstall core service and pip configuration file as in the rebuilding AIO section ? | 12:41 |
| odyssey4me | thng no, that shouldn't be necessary - your install didn't get very far | 12:41 |
| thng | oki. I did it and reruning playbook | 12:44 |
| *** jamesdenton has joined #openstack-ansible | 12:45 | |
| odyssey4me | thng which branch are you running? | 12:45 |
| thng | mitaka | 12:50 |
| mrhillsman | g'morning | 12:54 |
| thng | hi | 12:54 |
| mhayden | odyssey4me: i'd like to find a proper home for this thing -> https://review.openstack.org/#/c/328469/ | 12:58 |
| mhayden | automagically and i were doing some head scratching last week on where it belongs | 12:58 |
| *** karimb has quit IRC | 12:59 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins: Provide an empty default for image properties https://review.openstack.org/328990 | 13:00 |
| odyssey4me | hughsaunders does https://review.openstack.org/328990 make sense? | 13:00 |
| hughsaunders | odyssey4me: looks reasonable | 13:01 |
| hughsaunders | odyssey4me: so the default default is None... Would be better if the default key was required whenever required=false. | 13:03 |
| odyssey4me | yeah, that certainly is a good review guideline | 13:03 |
| evrardjp | odyssey4me: sorry I am back only now | 13:05 |
| evrardjp | yes it's I guess one of the two recent glance changes | 13:05 |
| evrardjp | I marked it in a commit | 13:05 |
| odyssey4me | evrardjp the issue should be solved by https://review.openstack.org/328990 | 13:06 |
| evrardjp | I don't see that issue coming on my recently updated aio | 13:06 |
| evrardjp | ok good | 13:07 |
| evrardjp | it makes sense | 13:07 |
| evrardjp | thanks for checking that for me | 13:07 |
| evrardjp | :D | 13:07 |
| *** messy has joined #openstack-ansible | 13:07 | |
| thng | odyssey4me: I got the same error after destroying container | 13:07 |
| odyssey4me | thng attach into a container and check whether it has connectivity | 13:08 |
| *** Zucan has joined #openstack-ansible | 13:08 | |
| odyssey4me | thng check that it can speak to DNS and resolve DNS | 13:08 |
| odyssey4me | you're more than likely in a situation where you have duplicated IP addresses, or some other network connectivity problem | 13:09 |
| thng | oki I will try it | 13:09 |
| odyssey4me | hughsaunders evrardjp lol, it looks like jmccrory had already pushed a patch up to solve the same problem: https://review.openstack.org/328671 | 13:12 |
| odyssey4me | which do you think is the better approach? | 13:12 |
| odyssey4me | I think I might prefer jmccrory's mechanism better. | 13:13 |
| thng | odysset4me : I think becasue there is no proxy settings in container. I have just attached to utility_container and done a apt-get install | 13:14 |
| thng | it didnot work | 13:14 |
| odyssey4me | thng did you configure the proxy settings in user_variables? https://github.com/openstack/openstack-ansible/blob/master/etc/openstack_deploy/user_variables.yml#L137-L146 | 13:15 |
| odyssey4me | thng can you pastebin your user_variables for inspection? | 13:15 |
| thng | yes I did it | 13:16 |
| thng | ## PROXY proxy_env_url: http://80.78.6.10:8080 no_proxy_env: "localhost,127.0.0.1,{% for host in groups['all_containers'] %}{{ hostvars[host]['container_address'] }}{% if not loop.last %},{% endif %}{% endfor %}" global_environment_variables: HTTP_PROXY: "{{ proxy_env_url }}" HTTPS_PROXY: "{{ proxy_env_url }}" NO_PROXY: "{{ no_proxy_env }}" http_proxy: "{{ proxy_env_url }}" https_proxy: "{{ proxy_env_url }}" no_pr | 13:16 |
| odyssey4me | jmccrory automagically cloudnull d34dh0r53 stevelle mattt hughsaunders andymccr mhayden can we get another review on https://review.openstack.org/328671 please? | 13:16 |
| evrardjp | odyssey4me: I think this is the best approach | 13:18 |
| evrardjp | but to be honest, it's almost the same | 13:18 |
| * mhayden looks | 13:18 | |
| evrardjp | odyssey4me: do you abandon your change? | 13:19 |
| odyssey4me | thng can you please pastebin it instead of just pasting it here | 13:19 |
| odyssey4me | evrardjp yes, I think jmccrory's is better | 13:19 |
| evrardjp | when we are at speaking about plugins | 13:20 |
| evrardjp | what do you think of using ARA callbacks? | 13:20 |
| evrardjp | https://github.com/openstack/ara | 13:20 |
| mhayden | odyssey4me: hah, checked commas a hundred times in that diff | 13:21 |
| mhayden | :P | 13:21 |
| *** oneswig has joined #openstack-ansible | 13:21 | |
| *** tiagogomes has quit IRC | 13:21 | |
| mhayden | something in the openstack ci gate image is preventing cron jobs from being provisioned | 13:22 |
| odyssey4me | evrardjp that looks like something good to submit tooling for in openstack-ansible-ops | 13:22 |
| evrardjp | odyssey4me: because that's what will be the successor of human log callback | 13:22 |
| evrardjp | it is | 13:22 |
| evrardjp | but it could also be interesting for our gate stats | 13:23 |
| odyssey4me | evrardjp we can't really use it for the gate | 13:23 |
| odyssey4me | it requires a running database to view, so it has no value in the gate | 13:23 |
| evrardjp | nothing prevents using an external DB :D | 13:23 |
| evrardjp | but ok | 13:23 |
| odyssey4me | it looks like dmsimard has done a good job there :) | 13:23 |
| evrardjp | also it could be useful to just replace what we have | 13:24 |
| evrardjp | you know him/her? | 13:24 |
| evrardjp | apparently its him | 13:24 |
| odyssey4me | I know dmsimard about as well as anyone else who lurks here. :) While he doesn't really work on OSA specifically, he does lurk and sometimes highlights issues coming up in Ansible or related to Ansible things. :) | 13:25 |
| thng | odyssey4me: http://pastebin.com/Vvpeqd2n | 13:25 |
| evrardjp | odyssey4me: ok | 13:26 |
| *** Guest__ is now known as Bofu2U | 13:26 | |
| odyssey4me | thng can you attach into a container and check that /etc/environment has the appropriate content? | 13:27 |
| thng | Yes I did, I will show you : http://pastebin.com/qg4sZ9ud | 13:30 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins: Optionally include image properties https://review.openstack.org/328671 | 13:30 |
| odyssey4me | thng ok, so with that in place from inside a container can you curl a web address? | 13:31 |
| thng | yes | 13:33 |
| odyssey4me | thng ok, and can you apt-get update? | 13:33 |
| thng | nope | 13:34 |
| thng | like apt-get install | 13:34 |
| odyssey4me | thng ok, have you configured the hosts to work through the proxy? | 13:34 |
| thng | I will try to add apt.conf for proxy to see | 13:34 |
| thng | yes | 13:34 |
| *** Ashana has joined #openstack-ansible | 13:35 | |
| odyssey4me | thng ok, for newton we've implemented an automatic configuration of the containers in the same way as the host | 13:35 |
| thng | it works now if I add apt.conf | 13:35 |
| dmsimard | oh hai | 13:36 |
| odyssey4me | but for mitaka you'll have to give it the extra apt config | 13:36 |
| thng | where can I do it ? | 13:36 |
| dmsimard | odyssey4me, evrardjp: let me know if you happen to have any questions about ara :) | 13:36 |
| evrardjp | :D | 13:36 |
| evrardjp | tbh, I wanted to do something similar recently for a personal project of mine | 13:36 |
| odyssey4me | thng you can have the correct files on your deployment host, then add them into this dict: https://github.com/openstack/openstack-ansible-lxc_hosts/blob/stable/mitaka/defaults/main.yml#L82-L87 | 13:37 |
| automagically | Morning all | 13:37 |
| *** permalac has quit IRC | 13:37 | |
| evrardjp | dmsimard: without the web interface | 13:37 |
| odyssey4me | thng that way all containers will contain the configs you want | 13:37 |
| dmsimard | evrardjp: oh, then do please contribute if you'd like :) | 13:37 |
| evrardjp | dmsimard: I'll be happy to talk with you about the callbacks, and what we can extend | 13:37 |
| evrardjp | cool | 13:37 |
| dmsimard | We worked on the ARA gate jobs last week, ARA *just* joined the OpenStack umbrella | 13:38 |
| dmsimard | Here's our own integration tests output http://logs.openstack.org/65/327965/1/gate/gate-ara-integration-ubuntu-trusty/cefee0b/logs/build/ | 13:38 |
| evrardjp | it maybe out of scope, so I'll discuss that later with you, but anyway I see good reasons to consume at least parts of it | 13:38 |
| odyssey4me | dmsimard boo :( http://logs.openstack.org/65/327965/1/gate/gate-ara-integration-ubuntu-trusty/cefee0b/logs/build/playbook/index.html | 13:38 |
| dmsimard | huh, what ? That worked before o_O | 13:39 |
| dmsimard | It looks like some stuff got deleted | 13:39 |
| evrardjp | :D | 13:39 |
| evrardjp | it looks like it! | 13:39 |
| evrardjp | else it's a nice html page | 13:39 |
| thng | odyssey4me : ok thanks. I will try to do it :) | 13:39 |
| *** galstrom_zzz is now known as galstrom | 13:39 | |
| evrardjp | :D | 13:39 |
| odyssey4me | dmsimard ah, everything may have been targzipped - it happens after a few days | 13:39 |
| dmsimard | odyssey4me: interesting, so html stuff becomes html.gz ? | 13:40 |
| dmsimard | didn't know that. | 13:40 |
| odyssey4me | dmsimard all logs on the openstack-ci log servers get gzipped after a few days | 13:40 |
| *** tiagogomes has joined #openstack-ansible | 13:41 | |
| dmsimard | good to know, I thought it was expected of jobs to gzip their own stuff | 13:41 |
| odyssey4me | here's an example: http://logs.openstack.org/73/324173/7/gate/gate-openstack-ansible-dsvm-commit/c8d1dd9/ | 13:41 |
| odyssey4me | you can see there that anything in the logs dir is gzipped, but any subdirectories are not | 13:42 |
| odyssey4me | I do think that may be a bug though -they actually should be recursively compressing all files. | 13:42 |
| *** schwicht has joined #openstack-ansible | 13:49 | |
| *** michaelgugino has joined #openstack-ansible | 13:49 | |
| *** flaviodsr has joined #openstack-ansible | 13:50 | |
| *** fedruantine has quit IRC | 13:50 | |
| dmsimard | odyssey4me, evrardjp: here, this one isn't zipped http://logs.openstack.org/08/329008/1/check/gate-ara-integration-ubuntu-trusty/8fde562/logs/build/ | 13:54 |
| evrardjp | good | 13:55 |
| evrardjp | :D | 13:55 |
| odyssey4me | dmsimard yeah, that's neat - far better than that horrible profiler callback | 13:55 |
| evrardjp | odyssey4me: that's why I talked to you about ARA ! | 13:55 |
| dmsimard | there's definitely a lot of UI/UX work needed to improve usability | 13:56 |
| evrardjp | :D | 13:56 |
| dmsimard | I'm exploring options but I'm definitely not a frontend guy | 13:56 |
| evrardjp | dmsimard: for my usage, it was drilling down to the variables | 13:56 |
| evrardjp | which is what this is doing | 13:56 |
| *** tiagogomes has quit IRC | 13:56 | |
| odyssey4me | my only wish is a more intelligent log output that automatically sees newlines and pretty-prints them properly | 13:56 |
| dmsimard | evrardjp: yeah but it requires too many clicks and stuff :) | 13:56 |
| evrardjp | odyssey4me: when you have the db and structure, I guess it's not that hard :D | 13:57 |
| dmsimard | odyssey4me: I think it was you or cloudnull but someone pulled my human_log plugin which does a better job at pretty printing stuff | 13:57 |
| dmsimard | it's not perfect but it's better | 13:57 |
| dmsimard | for example: https://ci.centos.org/view/rdo/view/weirdo/job/weirdo-mitaka-periodic-packstack-scenario001/89/consoleFull | 13:58 |
| odyssey4me | dmsimard yeah, I recall seeing that - as I recall we're in a bit of a stuck situation because we don't understand the licensing issues relating to callbacks and such | 13:58 |
| dmsimard | the problem is that it doesn't suppress the verbose output, it prints the verbose output AND the pretty printed output so it's super confusing | 13:58 |
| odyssey4me | see the comments in https://review.openstack.org/#/c/321331/ | 13:58 |
| odyssey4me | dmsimard yeah, what I was hoping for is something that only outputs stderror if there is content in it and the task has failed | 13:59 |
| *** bsv has quit IRC | 13:59 | |
| dmsimard | We had some discussions around licensing in ARA recently with Ansible Inc, and Red Hat legal | 13:59 |
| odyssey4me | dmsimard I was also thinking that stdout should also only be pretty printed if the task fails | 13:59 |
| odyssey4me | dmsimard so the general idea is that you don't do verbose logging, and the human readable log will output the right things if/when you need them... not for everything | 14:00 |
| dmsimard | This is definitely a recurring topic, I'll try and get them to put some clear guidelines out | 14:00 |
| odyssey4me | dmsimard thanks, that would be most helpful - we need a clear set of guidelines about what Ansible bits can be included in OpenStack projects | 14:00 |
| odyssey4me | whatever is put together needs to be clear and cover each of the plugin types - libraries, callbacks, etc | 14:01 |
| odyssey4me | we need to know definitively whether the license is compatible to include in an OpenStack project | 14:01 |
| *** TxGirlGeek has joined #openstack-ansible | 14:03 | |
| *** adrian_otto has joined #openstack-ansible | 14:05 | |
| odyssey4me | alextricity25 happy to see you putting your review capacity to good use :) | 14:05 |
| alextricity25 | hehe | 14:08 |
| *** tiagogomes has joined #openstack-ansible | 14:08 | |
| dmsimard | odyssey4me: fwiw you're including a GPL project in your requirements.txt of openstack-ansible :) | 14:09 |
| dmsimard | nitpicking on a callback sounds odd to me in that context | 14:10 |
| *** gregfaust has joined #openstack-ansible | 14:10 | |
| odyssey4me | dmsimard as I understand it there's a difference between adding a package requirements versus subclassing an object | 14:11 |
| dmsimard | I see. | 14:11 |
| odyssey4me | as I understand it using a package does not result in license contamination, but subclassing does | 14:11 |
| odyssey4me | when we last looked into this, the feedback we got was that Ansible libraries are fine because they stand alone... but Ansible callbacks have license contamination due to the subclassing | 14:12 |
| *** Mudpuppy has joined #openstack-ansible | 14:13 | |
| dmsimard | I can't quite share the info I have since it's not the same context and IANAL and all that, I poked the beehive to try and see if we can come up with some clear guidelines | 14:13 |
| dmsimard | The gist of the issue is that the definition of "code linking" is mostly up to interpretation in GPLv3 | 14:13 |
| dmsimard | The Zuulv3 callbacks are GPL'd but then again it's not a "big tent" project | 14:14 |
| odyssey4me | dmsimard yeah, I hate licensing... I personally think that patenting (the root cause for the need of licensing) in software is stupid. I wish that licensing would die in a fire. | 14:14 |
| dmsimard | ARA callbacks are Apache v2 (hence why the topic came up internally) and they'll stay that way for the time being | 14:15 |
| odyssey4me | ok, so that's where we need clarity | 14:15 |
| dmsimard | however, that particular callback originates from a GPL open sourced plugin anyway (author GPL'd it and I improved on top of what he did) | 14:16 |
| dmsimard | so I doubt we'd be able to change the licensing on that particular one | 14:16 |
| evrardjp | sorry I was busy, I didn't see the rest of the conversation | 14:16 |
| evrardjp | yes indeed, you can write filters etc without caring about the license, because you're not importing ansible code | 14:17 |
| evrardjp | which isn't the case for callbacks | 14:17 |
| evrardjp | at least that's what I understood | 14:17 |
| odyssey4me | I know that we can easily consume a callback regardless of the licensing. So if your callback was hosted somewhere we could just pull it in from there. The things we need clarification on is whether we really can have an Apache2 license for a callback (or whatever plugin/library) and whether we can hold something with a different license in our repo. As far as I know the second question is already answered by | 14:17 |
| odyssey4me | https://wiki.openstack.org/wiki/LegalIssuesFAQ#Licensing_of_library_dependencies | 14:17 |
| openstackgerrit | Ian Cordasco proposed openstack/openstack-ansible: Use correct keystone auth parameters https://review.openstack.org/328430 | 14:18 |
| openstackgerrit | Ian Cordasco proposed openstack/openstack-ansible: Use correct keystone auth parameters https://review.openstack.org/328432 | 14:18 |
| evrardjp | the problem, ppl use GPL, and they don't know how to use this | 14:18 |
| evrardjp | LGPL should have been used | 14:19 |
| evrardjp | or another license :p | 14:19 |
| evrardjp | anyway | 14:19 |
| evrardjp | I'm not a lawyer | 14:19 |
| evrardjp | :p | 14:19 |
| *** TxGirlGeek has quit IRC | 14:20 | |
| *** TxGirlGeek has joined #openstack-ansible | 14:20 | |
| *** permalac has joined #openstack-ansible | 14:21 | |
| *** TheIntern has joined #openstack-ansible | 14:22 | |
| odyssey4me | dmsimard thanks for poking the hornets nest :) | 14:22 |
| dmsimard | hornets ? bah. bees. honey bees. hard workers :) | 14:22 |
| openstackgerrit | Merged openstack/openstack-ansible-os_aodh: Convert default from MongoDB to MySQL storage https://review.openstack.org/328510 | 14:22 |
| *** TxGirlGeek has quit IRC | 14:25 | |
| *** TxGirlGeek has joined #openstack-ansible | 14:27 | |
| *** spotz_zzz is now known as spotz | 14:27 | |
| *** TxGirlGeek has quit IRC | 14:29 | |
| *** TxGirlGeek has joined #openstack-ansible | 14:30 | |
| *** fedruantine has joined #openstack-ansible | 14:31 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:32 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts: Implement base container for overlayfs-backed containers https://review.openstack.org/329029 | 14:38 |
| odyssey4me | alextricity25 stevelle I expect that the os-aodh-install playbook may require some db tasks to go with https://review.openstack.org/328510 ? | 14:39 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts: Implement base container for overlayfs-backed containers https://review.openstack.org/329029 | 14:41 |
| *** galstrom is now known as galstrom_zzz | 14:42 | |
| *** jthorne has joined #openstack-ansible | 14:45 | |
| *** sacharya has joined #openstack-ansible | 14:47 | |
| evrardjp | odyssey4me: overlayfs <3 | 14:51 |
| odyssey4me | evrardjp yeah, I did some testing with it over the weekend and it's pretty awesome | 14:54 |
| evrardjp | yup | 14:54 |
| odyssey4me | using overlayfs for the AIO should cut the build time by 10-15 mins | 14:54 |
| evrardjp | I got issues on the past with locks 'though | 14:54 |
| *** gparaskevas has quit IRC | 14:54 | |
| evrardjp | but I think it's kinda perfect for our use case | 14:55 |
| odyssey4me | we can try it out for the AIO for a while and see how it goes, but if we find it works well then I think we should consider making it the only way we deploy containers | 14:55 |
| odyssey4me | it'll totally change how we deploy packages, do upgrades, etc | 14:55 |
| *** phalmos has joined #openstack-ansible | 14:56 | |
| alextricity25 | odyssey4me: You're right. Originally we planned on the deployer already having the DB and DB users setup in MongoDB, but now that we are using MySQL it would be good to make that part of the playbooks | 14:57 |
| odyssey4me | alextricity25 can you whip up a patch for that quickly, seeing as I expect that with that patch merged it might cause a gate blocker | 14:58 |
| *** permalac has quit IRC | 14:58 | |
| *** Ashana has quit IRC | 14:58 | |
| alextricity25 | odyssey4me: sure thing! | 14:58 |
| *** KLevenstein has joined #openstack-ansible | 14:58 | |
| *** galstrom_zzz is now known as galstrom | 14:58 | |
| odyssey4me | evrardjp I think that perhaps we could also use lvm snapshots to improve any deployments using lvm-backed containers | 15:00 |
| odyssey4me | but I'll test that out another time - I just want to push some patches in which enable overlayfs for now | 15:00 |
| evrardjp | we are using flat files to store that | 15:01 |
| evrardjp | it's not that great | 15:01 |
| evrardjp | for using lvm snapshot backend feature IIRC | 15:01 |
| evrardjp | but yes, the most important point is to create this container, at a stopped state, to use it for differentials | 15:02 |
| *** kstev has joined #openstack-ansible | 15:02 | |
| evrardjp | it increase the storage needed to start (one more container) but it also decreases afterwards (just the diff) | 15:02 |
| evrardjp | but bringing this needs careful maintaing later: how do we updates the basic container? | 15:03 |
| *** michaelgugino has quit IRC | 15:03 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_container_create: Implement overlayfs-backed containers https://review.openstack.org/329045 | 15:04 |
| evrardjp | my point would be, shouldn't we use the template of the other containers, instead of having a new one? | 15:04 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Destroy repo containers before upgrade https://review.openstack.org/328485 | 15:05 |
| evrardjp | not "template", cache or something | 15:05 |
| odyssey4me | evrardjp I did some tests yesterday. The base container can be changed, and the changes are reflected in the implemented containers, but I actually think that it would be better for us to move towards considering the containers more disposable. | 15:05 |
| odyssey4me | So basically for major upgrades we should be deleting all containers and building them off a fresh base container. | 15:06 |
| *** zerda2 has joined #openstack-ansible | 15:06 | |
| *** michaelgugino has joined #openstack-ansible | 15:07 | |
| evrardjp | odyssey4me: In all cases, they should be diposed of on a regular basis | 15:07 |
| evrardjp | else the overhead becomes too big | 15:07 |
| odyssey4me | yeah, with overlayfs I actually think it's quite practical for us to destroy all containers even in patch upgrades | 15:07 |
| evrardjp | I'm just thinking of a way to quickly benefit from os changes (ssh security update for example) without a need of rebuilding | 15:08 |
| *** thng has quit IRC | 15:08 | |
| odyssey4me | as long as the container IP's don't change (as this impacts external LB configuration) I think that moving towards full container disposability is a great way to get rid of some of the issue we're seeing in upgrades | 15:09 |
| evrardjp | I fully agree | 15:09 |
| evrardjp | but I still don't get why we don't do that on the container that is using as base for the other containers | 15:10 |
| evrardjp | the cache container if I recall the name correctly | 15:11 |
| odyssey4me | evrardjp well, that's what I'm thinking - if we're disposing of containers regularly then there's no reason why we can't implement a fresh base container perhaps even for each tag, we then destroy and rebuild the containers using that base in serial (to maintain uptime) | 15:11 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Fix keystone DB Access variable https://review.openstack.org/327660 | 15:12 |
| *** permalac has joined #openstack-ansible | 15:13 | |
| evrardjp | I think we agree, but we are not speaking about the same things | 15:13 |
| odyssey4me | evrardjp if we do that then we're always in a place where all containers are always up to date, without even trying | 15:14 |
| evrardjp | I agree | 15:15 |
| evrardjp | It's good to do it that way | 15:15 |
| evrardjp | let me read the lxc_hosts again to be sure my idea isn't stupid | 15:15 |
| odyssey4me | evrardjp I did try something else though - I implemented a trusty base container on xenial, then created an overlayfs-backed container based on a snapshot of the base... then I chrooted into the base container and upgraded it to Xenial | 15:15 |
| *** weezS has joined #openstack-ansible | 15:16 | |
| odyssey4me | after a reboot of the overlayfs-backed container it actually showed that it was now a xenial container | 15:16 |
| odyssey4me | I was surprised that it worked so well. | 15:16 |
| evrardjp | you mean you're surpised the ubuntu upgrade works ? | 15:16 |
| evrardjp | or you're suprised overlayfs works? | 15:16 |
| evrardjp | because I'm super confident in the latter :p | 15:16 |
| odyssey4me | evrardjp FYI the container 'cache' (ie the rootfs that all containers use as the base) is refreshed every single time the lxc_hosts role is executed | 15:17 |
| evrardjp | also because it's been a long time I'm using it too | 15:17 |
| evrardjp | odyssey4me: YUP | 15:17 |
| evrardjp | that's why! | 15:17 |
| evrardjp | I mean | 15:17 |
| evrardjp | that's why I would rather have it there | 15:17 |
| odyssey4me | evrardjp nah, I was just surprised that even though I'd installed some packages in the container from trusty everything still worked | 15:18 |
| odyssey4me | obviously my packages needed an upgrade | 15:18 |
| evrardjp | every single time you run, you have the latest cache, which makes it easy for security updates | 15:18 |
| evrardjp | that was my whole point | 15:18 |
| evrardjp | odyssey4me: it's FS :D | 15:18 |
| evrardjp | it has bitten me a few times 'though | 15:19 |
| odyssey4me | evrardjp yeah, so we can't really do it in the cache - it'd have to be done by updating the base container on which all the containers are based | 15:19 |
| evrardjp | that's what I don't get | 15:19 |
| evrardjp | oh you mean the clean cache container ? | 15:20 |
| evrardjp | but it doesn't matter | 15:20 |
| odyssey4me | evrardjp the base container is supposed to be immutable, so I think a better process is to build fresh base containers regularly and the roll rebuilds of all the containers based on the new fresh base | 15:20 |
| alextricity25 | odyssey4me: The aodh-dbsync script runs the initial database bits | 15:21 |
| odyssey4me | I did some some stuff which we'd have to change if we want to work this way - we'd have to try to limit the unnecessary changes we make in the container itself. For example, we shouldn't be running 'apt-get update' in the container - we should do it in the base instead. | 15:21 |
| alextricity25 | odyssey4me: I don't think there is anything else to do apart from that. Did you have something else in mind? | 15:22 |
| evrardjp | I guess we have a problem of vocabulary here | 15:22 |
| evrardjp | base is the container you are creating again in your last commit, while cache is the one you consider clean downloaded and stopped | 15:22 |
| evrardjp | right? | 15:23 |
| odyssey4me | alextricity25 hmm, I think you're missing the point - essentially we need the tasks as are present in https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-cinder-install.yml#L146-L173 to also be in https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-aodh-install.yml | 15:23 |
| odyssey4me | alextricity25 the issue is that there is no aodh db or db user in mysql right now | 15:23 |
| alextricity25 | Ah. That's right | 15:24 |
| alextricity25 | Thanks odyssey4me | 15:24 |
| evrardjp | alextricity25: also please change the doc that goes with it if it's an IRR (not the case here) | 15:24 |
| evrardjp | but this way you don't forget :p | 15:24 |
| *** jmckind has joined #openstack-ansible | 15:24 | |
| odyssey4me | evrardjp the container cache is just a zipped tarball of the rootfs, the base container is the one created as a basis for the snapshot | 15:24 |
| evrardjp | odyssey4me: you agree with my terminology ? | 15:24 |
| evrardjp | ok | 15:24 |
| evrardjp | That's the issue | 15:24 |
| evrardjp | let's rollback to explanation | 15:24 |
| evrardjp | WHAT IF | 15:25 |
| evrardjp | we download a container (the container cache), this container having overlayfs as backing store. Then we do stuff, but don't zip the tarball | 15:25 |
| evrardjp | we keep the container as it is, and you've got your "base" template | 15:26 |
| *** weshay is now known as weshay_lunch_brb | 15:26 | |
| *** Drago has joined #openstack-ansible | 15:27 | |
| evrardjp | if you want to transfer to multiple nodes, you can still compress-send/decompress-receive on nodes | 15:27 |
| mhayden | darrenc: you around? | 15:27 |
| *** Drago has quit IRC | 15:27 | |
| *** Drago has joined #openstack-ansible | 15:27 | |
| odyssey4me | yeah, I looked into that - the trouble with implementing a folder which is used as the backing for overlayfs for the containers is that lxc has no control mechanisms for it - so you may completely break that folder... the way I'm doing it now there are protections in place implemented by LXC | 15:27 |
| evrardjp | oh | 15:28 |
| evrardjp | that's where I am lacking the docs | 15:28 |
| odyssey4me | but yeah, I did think of that - but it's a complete break of the process we have right now, and I'd rather we didn't break the process until we're certain that we want to move over to this method as *the* method for all production environments | 15:28 |
| evrardjp | it's not really a change, we just don't notify the Destroy base container :D | 15:29 |
| odyssey4me | something that we could do as a neat extension of this idea is to actually prep different base containers for each service | 15:29 |
| *** johnmilton has joined #openstack-ansible | 15:30 | |
| evrardjp | you mean an upper overlay or a different base ? | 15:30 |
| evrardjp | I mean, using a different lowerdir, or have a higher lowerdir | 15:30 |
| odyssey4me | evrardjp yeah, it can be that simple but I didn't want to use LXC_NAME as the base container - I wanted something that's a bit more explicitly named | 15:31 |
| evrardjp | OMG terminology will be difficult | 15:31 |
| evrardjp | odyssey4me: understood then | 15:31 |
| odyssey4me | evrardjp well, it seems that LXC can't use more than one layer at this stage | 15:31 |
| evrardjp | what? | 15:31 |
| evrardjp | that sucks | 15:31 |
| odyssey4me | not that I could see - but I didn't try very hard | 15:31 |
| evrardjp | fplay | 15:32 |
| evrardjp | let me get info on this | 15:32 |
| odyssey4me | but it'd be quite nice to have the repo server hold a set of base images which could be used for all container builds | 15:32 |
| evrardjp | we are reinventing docker registry ! | 15:33 |
| evrardjp | :p | 15:33 |
| *** al_loew has joined #openstack-ansible | 15:33 | |
| odyssey4me | so we could layer base + nova software = nova-base | 15:34 |
| evrardjp | except docker used AuFS which was supposed to be doomed | 15:34 |
| evrardjp | I hate aufs | 15:34 |
| evrardjp | :p | 15:34 |
| odyssey4me | then the deploy is to deploy a nova-base, then simply to configure things on it (no software install takes place) | 15:34 |
| evrardjp | odyssey4me: yeah, I'll get info on this, because I never runned with multiple bases for lxc | 15:34 |
| evrardjp | I had that for other things though | 15:34 |
| *** johnmilton has quit IRC | 15:35 | |
| *** TxGirlGeek has quit IRC | 15:35 | |
| *** javeriak has joined #openstack-ansible | 15:35 | |
| *** agireud has quit IRC | 15:36 | |
| odyssey4me | well, deploy nova-base - then create containers based on that as a snapshot, then configure the software in the containers (but no need to install any software) | 15:36 |
| odyssey4me | rumour has it that iscsi may work from inside a container in xenial | 15:36 |
| evrardjp | This is like a completely different process | 15:36 |
| *** mikelk has quit IRC | 15:37 | |
| odyssey4me | yeah, but we wouldn't have to change all that much | 15:37 |
| odyssey4me | the tasks in the roles can stay as-is... we'd just have to change the playbooks | 15:37 |
| odyssey4me | the tasks in the role to install the software would just need to be properly idempotent so that the software installs just end up being noops because the software's already there | 15:38 |
| evrardjp | anyway, I don't think it's a good idea to allow the change of the backingstore without a proper recent kernel | 15:38 |
| evrardjp | maybe we should have a check for that | 15:38 |
| *** agireud has joined #openstack-ansible | 15:38 | |
| evrardjp | or at least a note | 15:38 |
| odyssey4me | yeah, I was thinking about implementing that - but I made sure that it was noted in the release notes. | 15:38 |
| *** fedruantine has quit IRC | 15:38 | |
| evrardjp | because 3. is definitely not as good as recent ones on overlayfs | 15:39 |
| evrardjp | ok | 15:39 |
| evrardjp | let me review then | 15:39 |
| evrardjp | :D | 15:39 |
| odyssey4me | I want to use this for the AIO, even on trusty, so I didn't want to put a check in place that hard failed. | 15:39 |
| evrardjp | good we aren't moving that much files around | 15:39 |
| evrardjp | (just create/copy and delete) | 15:40 |
| *** admin0 has quit IRC | 15:41 | |
| evrardjp | quick question, where do you plan to define the FS? | 15:42 |
| evrardjp | because here, we just create the thing in a dir, so I guess the mount is lacking | 15:43 |
| *** javeriak_ has joined #openstack-ansible | 15:43 | |
| *** tiagogomes has quit IRC | 15:44 | |
| *** javeriak has quit IRC | 15:44 | |
| *** weshay_lunch_brb is now known as weshay | 15:45 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_container_create: Implement overlayfs-backed containers https://review.openstack.org/329045 | 15:47 |
| *** johnmilton has joined #openstack-ansible | 15:47 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Implement overlayfs as the backing store for the AIO https://review.openstack.org/329066 | 15:47 |
| michaelgugino | odyssey4me: if you could look at https://review.openstack.org/#/c/324045/ I would appreciate it. Tempest check is failing in the CI environment complaining about availability zone, but I cannot replicate the failure locally | 15:47 |
| *** tiagogomes has joined #openstack-ansible | 15:48 | |
| evrardjp | odyssey4me: it's 3.18 btw IIRC | 15:48 |
| evrardjp | haha: https://en.wikipedia.org/wiki/OverlayFS | 15:48 |
| odyssey4me | evrardjp ok, please note that in the review so that I change the comments - I was just going off the info provided in the footnotes of https://www.hastexo.com/blogs/florian/2016/02/21/containers-just-because-everyone-else/ | 15:49 |
| odyssey4me | evrardjp what do you mean the mount is lacking? the container create implements the overlayfs mount all on its own | 15:49 |
| evrardjp | I like this title | 15:49 |
| automagically | mhayden Can you give this a quick review at some point today? https://review.openstack.org/#/c/327331/ | 15:49 |
| odyssey4me | evrardjp you'll see the container has a config line like: lxc.rootfs = overlayfs:/var/lib/lxc/cache/rootfs:/var/lib/lxc/container1/delta0 | 15:50 |
| automagically | I appreciate all your other reviews | 15:50 |
| evrardjp | odyssey4me: haha that's what I did home exactly for the same purpose at home. Love this post already! | 15:50 |
| evrardjp | or close | 15:50 |
| evrardjp | odyssey4me: didn't see the line in the commit | 15:50 |
| evrardjp | will re-review | 15:51 |
| evrardjp | thanks for the explanation | 15:51 |
| *** TxGirlGeek has joined #openstack-ansible | 15:52 | |
| odyssey4me | evrardjp no, it's not in the commit - it's just how lxc configures it when you do the lxc-clone based on a snapshot and implement overlayfs as the backing store | 15:52 |
| odyssey4me | it's really, really simple - lxc does all the heavy lifting | 15:52 |
| evrardjp | oh yes, true it does that | 15:54 |
| evrardjp | my bad | 15:54 |
| evrardjp | like I said, I'm mostly using that on host directly... I had that setup long ago and then I forgot :p | 15:54 |
| evrardjp | it just works so transparently... | 15:55 |
| *** sacharya_ has joined #openstack-ansible | 15:55 | |
| odyssey4me | well, if we can get that series of patches merged I think it'll be quite interesting | 15:55 |
| evrardjp | good to have a technology refresh once in a while :D | 15:55 |
| *** TheIntern has quit IRC | 15:56 | |
| *** TheIntern has joined #openstack-ansible | 15:56 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Docs: Specify supported distributions https://review.openstack.org/329072 | 15:57 |
| openstackgerrit | Merged openstack/openstack-ansible: Use default metadata_backlog value from Neutron https://review.openstack.org/328503 | 15:57 |
| *** sacharya has quit IRC | 15:58 | |
| evrardjp | mhayden: do you need this to merge at some point? https://review.openstack.org/#/c/310067/2 | 15:59 |
| mhayden | automagically: lookin' | 16:05 |
| mhayden | automagically: i'll trade you for a gander at https://review.openstack.org/329072 | 16:05 |
| *** tiagogomes has quit IRC | 16:06 | |
| mhayden | evrardjp: i do, but i think that rolls into my bigger problem of "i need to figure out logging" | 16:06 |
| evrardjp | ara ? :D | 16:06 |
| mhayden | possibly | 16:06 |
| evrardjp | :p | 16:06 |
| mhayden | i'm eager to get something that is easy to import into a siem... perhaps CSV for that | 16:07 |
| mhayden | well a SIEM or a vulnerability management platform | 16:07 |
| evrardjp | makes sense | 16:07 |
| evrardjp | but you'll need to link the inventory | 16:07 |
| evrardjp | for that to makes sense | 16:07 |
| evrardjp | it is a whole new world I guess | 16:07 |
| evrardjp | dmsimard: is csv export possible with ARA? | 16:07 |
| evrardjp | else mhayden building docs for security role gives a serious load of warnings | 16:09 |
| evrardjp | documents not included in any toctree | 16:09 |
| mhayden | yeah, that's on my list of to-do's... sphinx complains about that | 16:09 |
| mhayden | i have a special way of including everything | 16:09 |
| evrardjp | ok you're already ware of it | 16:09 |
| mhayden | which throws off sphinx | 16:09 |
| dmsimard | evrardjp: I've never really tried poking around csv stuff, technically the ARA cli is built on top of cliff (same python framework as python-openstackclient) so if you do "-f csv" to any command it *should* work | 16:10 |
| dmsimard | evrardjp: otherwise, you could probably export the database (sqlite or mysql) as csv and do things with it | 16:10 |
| dmsimard | evrardjp: what's your use case around wanting csv ? | 16:11 |
| evrardjp | not mine, for mhayden | 16:11 |
| evrardjp | he is looking about improving his ansible playbooks run reporting | 16:11 |
| *** electrofelix has quit IRC | 16:11 | |
| *** TxGirlGeek has quit IRC | 16:12 | |
| evrardjp | not pushing anything, just trying to make ppl work together if possible | 16:12 |
| mhayden | dmsimard: long story short, i'm looking to generate some logs from the security role that could tell a deployer, or a deployer's security team, about which stig configurations were applied to an openstack environment | 16:13 |
| mhayden | to hand over to an auditor and/or corp security team | 16:13 |
| evrardjp | and the tasks with the stig configurations are marked with tags in ansible, right? | 16:13 |
| evrardjp | (for ease of understanding for everyone) | 16:13 |
| odyssey4me | hughsaunders you may find https://review.openstack.org/#/q/status:open+topic:overlayfs-implementation interesting :) | 16:16 |
| evrardjp | I think he has left for today | 16:17 |
| dmsimard | mhayden, evrardjp: hm, we don't give tags any special treatment in ara right now, it's probably a good idea to leverage tags somehow. | 16:17 |
| flaviodsr | odyssey4me I have checked and the sahara role is already using the multi-distro framework btw | 16:19 |
| mhayden | dmsimard: yeah, i'm thinking about registering a variable that i could use to make a log at the end | 16:19 |
| *** woodard_ has quit IRC | 16:20 | |
| evrardjp | dmsimard: I think tags should appear if possible, somehow :D | 16:20 |
| *** woodard has joined #openstack-ansible | 16:20 | |
| dmsimard | evrardjp: yes, it's a good idea I'm taking notes | 16:20 |
| evrardjp | I'll read ARA's code, I hope soon, but it's not gonna be until at least 2 week-ends :/ | 16:21 |
| *** woodard has quit IRC | 16:21 | |
| *** tiagogomes has joined #openstack-ansible | 16:21 | |
| *** woodard has joined #openstack-ansible | 16:21 | |
| mhayden | automagically: reviewed the vagrant file stuff | 16:21 |
| mhayden | i guess i should check out this vagrant thing | 16:21 |
| mhayden | then again, virtualbox makes me cry tears | 16:22 |
| dmsimard | evrardjp: I like to think it's a fairly simple application | 16:22 |
| evrardjp | how does playbook nesting work? if the playbook was an include of another one, would it be the invokee that will log the tasks/data? | 16:22 |
| evrardjp | just curious | 16:23 |
| *** michaelgugino has quit IRC | 16:24 | |
| dmsimard | Right now we're working on making role and includes more obvious (and filter-able in general) and file-based task filtering. I used to have a playbook for a project that included other playbooks but I got rid of that. | 16:25 |
| dmsimard | As far as ARA is concerned, it'd probably record that as being just one playbook | 16:25 |
| dmsimard | But it's an interesting thought, I'll try and test it | 16:25 |
| evrardjp | that's what I would have thought | 16:25 |
| dmsimard | It'd be one playbook but multiple plays | 16:25 |
| dmsimard | Just like they'd be in the same file | 16:26 |
| dmsimard | and you'd see the include tasks. | 16:26 |
| evrardjp | yes, that's what I thought too | 16:26 |
| evrardjp | it is what ansible is doing and reporting anyway | 16:26 |
| dmsimard | evrardjp: http://logs.openstack.org/08/329008/1/check/gate-ara-integration-ubuntu-trusty/8fde562/logs/build/playbook/63127a1a-acf7-48ec-93a4-6186eb706113/index.html | 16:26 |
| dmsimard | This is an example of playbook with different plays | 16:26 |
| dmsimard | If you drill down on the details of each play, you see the tasks related to it | 16:26 |
| dmsimard | You can see the beginning of the work we started around file-based filtering here http://logs.openstack.org/08/329008/1/check/gate-ara-integration-ubuntu-trusty/8fde562/logs/build/playbook/63127a1a-acf7-48ec-93a4-6186eb706113/file/index.html | 16:27 |
| dmsimard | We're storing the content of the files now too, so eventually we'll have a UI for that, maybe even diffing files across runs (i.e, to try and tell what changed between two runs to see what failed) | 16:28 |
| evrardjp | yes, I understand, that's what I expected | 16:28 |
| *** TxGirlGeek has joined #openstack-ansible | 16:28 | |
| *** fedruantine has joined #openstack-ansible | 16:28 | |
| evrardjp | dmsimard: ahah that's the feature I wanted to implement for my home project | 16:28 |
| evrardjp | good :D | 16:28 |
| *** alikins has joined #openstack-ansible | 16:29 | |
| *** johnmilton has quit IRC | 16:30 | |
| *** metral_zzz is now known as metral | 16:32 | |
| *** tiagogomes has quit IRC | 16:36 | |
| *** mummer has joined #openstack-ansible | 16:37 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Address low-hanging TODO items https://review.openstack.org/329098 | 16:38 |
| *** pcaruana has quit IRC | 16:38 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Address low-hanging TODO items https://review.openstack.org/329098 | 16:39 |
| mhayden | do your items hang low, can you swing them to and fro | 16:41 |
| *** al_loew has quit IRC | 16:43 | |
| mhayden | automagically: your vagrant stuff looks good -- hopefully the middle ground makes sense | 16:46 |
| *** michaelgugino has joined #openstack-ansible | 16:46 | |
| *** M00nr41n has joined #openstack-ansible | 16:47 | |
| *** zerda2 has quit IRC | 16:49 | |
| *** tiagogomes has joined #openstack-ansible | 16:50 | |
| *** bbmbx__ has quit IRC | 16:51 | |
| *** bbmbx__ has joined #openstack-ansible | 16:51 | |
| *** Drago has quit IRC | 16:52 | |
| *** woodard_ has joined #openstack-ansible | 17:00 | |
| *** woodard has quit IRC | 17:03 | |
| *** kstev has quit IRC | 17:05 | |
| *** sacharya has joined #openstack-ansible | 17:13 | |
| *** sacharya_ has quit IRC | 17:13 | |
| *** Zucan has quit IRC | 17:15 | |
| *** sacharya_ has joined #openstack-ansible | 17:16 | |
| *** Drago has joined #openstack-ansible | 17:17 | |
| *** sacharya has quit IRC | 17:19 | |
| *** Drago has quit IRC | 17:19 | |
| *** Drago has joined #openstack-ansible | 17:19 | |
| *** scarlisle has joined #openstack-ansible | 17:20 | |
| *** TheIntern has quit IRC | 17:32 | |
| *** gregfaust has quit IRC | 17:32 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_sahara: Use master git branches https://review.openstack.org/329123 | 17:34 |
| michaelgugino | I think I figured out what's going on with my Nova patch for 16.04 | 17:35 |
| michaelgugino | Looks like the handler is only firing once for the openstack host, and not the localhost (where the compute service lives) | 17:36 |
| jmccrory | palendae: have you looked into adding tests around osa-plugins? | 17:36 |
| michaelgugino | which makes about zero sense... must be a bug in ansible, that's the only thing I can guess | 17:38 |
| michaelgugino | although, I'm not seeing the problem locally. | 17:39 |
| *** pcaruana has joined #openstack-ansible | 17:39 | |
| michaelgugino | hmm, looks like it's trying to setup UCA repo and failing... | 17:41 |
| michaelgugino | well, that's interesting. It was automatically rebasing my patch in gerrit :/ | 17:44 |
| openstackgerrit | Michael Gugino proposed openstack/openstack-ansible-os_nova: Implement Xenial Support https://review.openstack.org/324045 | 17:46 |
| *** TheIntern has joined #openstack-ansible | 17:47 | |
| palendae | jmccrory, I haven't looked at it yet | 17:49 |
| openstackgerrit | Merged openstack/openstack-ansible-memcached_server: Cleanup/standardize usage of tags https://review.openstack.org/328309 | 17:49 |
| palendae | jmccrory, I'd think the first step would be adding in relevant tox info and then setting up imports | 17:49 |
| *** fedruantine has quit IRC | 17:51 | |
| *** weezS has quit IRC | 17:52 | |
| evrardjp | odyssey4me: I replied to your concern for https://review.openstack.org/#/c/328274/3 | 17:54 |
| evrardjp | jmccrory: I think osa-plugins could have a proper rewrite | 17:54 |
| *** Drago has quit IRC | 17:55 | |
| evrardjp | proper testing, but also proper CRUD operations for anything we are doing (for example modules) | 17:55 |
| evrardjp | proper for anything we are doing, and proper CRUD idempotency for modules* | 17:55 |
| evrardjp | darn it I can't write! | 17:56 |
| *** raddaoui has joined #openstack-ansible | 17:56 | |
| *** weezS has joined #openstack-ansible | 17:57 | |
| *** albertcard has joined #openstack-ansible | 17:57 | |
| *** kstev has joined #openstack-ansible | 17:58 | |
| *** jthorne has quit IRC | 17:59 | |
| *** TxGirlGeek has quit IRC | 17:59 | |
| jmccrory | evrardjp: hoping to eventually be able to just remove some of the libraries for ansible 2 modules as well, glance -> os_image for example | 17:59 |
| *** jthorne has joined #openstack-ansible | 17:59 | |
| mhayden | so there are three hotels here in SAT that will give special rates (around $100/night) for the mid-cycle | 18:00 |
| mhayden | and one provides a free shuttle to/from Rackspace's HQ | 18:00 |
| palendae | yeah, conributing to and using the upstream ones would be great | 18:00 |
| palendae | mhayden, lemme guess, Aloft? | 18:00 |
| mhayden | si | 18:00 |
| mhayden | there's also the Valencia and one other | 18:00 |
| *** saneax is now known as saneax_AFK | 18:01 | |
| mhayden | i'm wondering if we could cobble together some funds from the locals to do a cookout | 18:01 |
| mhayden | i don't mind offering up my place for that | 18:01 |
| evrardjp | true | 18:03 |
| *** jthorne has quit IRC | 18:04 | |
| evrardjp | jmccrory: and with 2.1 a few modules were improved, I should read them to see the extend of the improvements | 18:04 |
| *** cloader89 has joined #openstack-ansible | 18:06 | |
| *** jthorne has joined #openstack-ansible | 18:06 | |
| *** Zucan has joined #openstack-ansible | 18:07 | |
| *** TxGirlGeek has joined #openstack-ansible | 18:17 | |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Configuring AODH DB now that it uses MySQL. https://review.openstack.org/329136 | 18:17 |
| *** Drago has joined #openstack-ansible | 18:19 | |
| *** sdake_ has joined #openstack-ansible | 18:23 | |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible-os_aodh: AODH dbsync script needs pymongo https://review.openstack.org/329137 | 18:23 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Configuring AODH DB now that it uses MySQL. https://review.openstack.org/329136 | 18:24 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Fix documentation warnings from sphinx https://review.openstack.org/329139 | 18:29 |
| *** TxGirlGeek has quit IRC | 18:31 | |
| *** TxGirlGeek has joined #openstack-ansible | 18:32 | |
| *** berendt has joined #openstack-ansible | 18:33 | |
| stevelle | thanks alextricity25 for those | 18:33 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Fix documentation warnings from sphinx https://review.openstack.org/329139 | 18:36 |
| *** TxGirlGeek has quit IRC | 18:42 | |
| *** fedruantine has joined #openstack-ansible | 18:43 | |
| *** TxGirlGeek has joined #openstack-ansible | 18:43 | |
| *** Zucan has quit IRC | 18:49 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Fix documentation warnings from sphinx https://review.openstack.org/329139 | 18:50 |
| stevelle | alextricity25: I'm not seeing that failure. | 18:51 |
| alextricity25 | stevelle: Really...hmmm..are you using virtualenvs? | 18:51 |
| stevelle | alextricity25: yes | 18:51 |
| stevelle | https://gist.github.com/stevelle/43e5873807f9432b397bdfe5132f3a4b | 18:52 |
| alextricity25 | This is what I'm seeing: | 18:52 |
| alextricity25 | http://paste.openstack.org/show/515798/ | 18:52 |
| alextricity25 | Oh wow that formatting | 18:53 |
| stevelle | alextricity25: that's a full OSA, and that isn't using mysql b/c of group vars, right? | 18:54 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Add key fields to audit rules https://review.openstack.org/328288 | 18:54 |
| alextricity25 | Hmm...let me check something | 18:55 |
| alextricity25 | stevelle: you're right | 18:57 |
| stevelle | jmccrory gets the credit | 18:57 |
| stevelle | see his comment on /329136/1 | 18:58 |
| alextricity25 | stevelle: nice. hey...are we getting rid of the aodh_db_port variable all together | 18:59 |
| alextricity25 | ? | 18:59 |
| alextricity25 | https://review.openstack.org/#/c/328510/7/defaults/main.yml | 19:00 |
| stevelle | alextricity25: I am inclined toward it, tough call whether to retain and fix the default or not | 19:00 |
| alextricity25 | that works for me :) | 19:00 |
| mhayden | jmccrory: thanks for the quick look -- this review should be identical https://review.openstack.org/328288 | 19:00 |
| alextricity25 | urgggg...but that should probably be configurable, no? What if the deployer wants AODH to use a different db? | 19:01 |
| alextricity25 | on a different port | 19:01 |
| jmccrory | mhayden: cool, got it | 19:02 |
| mhayden | thanks! | 19:03 |
| mhayden | evrardjp: i went through the security role docs and fixed the warnings -> https://review.openstack.org/329139 | 19:03 |
| evrardjp | oh thank you ! | 19:03 |
| evrardjp | I'm sorry but it's Belgium playing today @Euro, I can't review | 19:05 |
| evrardjp | later :D | 19:05 |
| stevelle | alextricity25: aodh_db_address: "mongodb:27017" | 19:05 |
| stevelle | ez user_var | 19:05 |
| alextricity25 | try! | 19:05 |
| alextricity25 | true* | 19:05 |
| stevelle | alextricity25: work remaining is to write the upgrade step to run the migration, I'll submit a release note today | 19:08 |
| stevelle | that after we get the playbook / group_vars fixed | 19:08 |
| alextricity25 | stevelle: Great! Thanks! I'll update my commit to make the group_var changes. | 19:08 |
| alextricity25 | I'm doing that now | 19:09 |
| *** javeriak has joined #openstack-ansible | 19:12 | |
| *** javeriak_ has quit IRC | 19:15 | |
| alextricity25 | stevelle: Should aodh_db_address here be "aodh_galera_address"? | 19:20 |
| alextricity25 | https://review.openstack.org/#/c/328510/7/defaults/main.yml | 19:20 |
| alextricity25 | should I set it like so in defaults/main.yml: aodh_db_address: {{ aodh_galera_address }} | 19:21 |
| stevelle | errr, um, that won't work as typed :) | 19:25 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Docs: Specify supported distributions https://review.openstack.org/329072 | 19:25 |
| alextricity25 | aodh_db_address: "{{ aodh_galera_address }}"? | 19:26 |
| stevelle | alextricity25: remove them from group_vars, they shouldn't be needed | 19:28 |
| stevelle | the galera vars should declared locally in the playbook, like other roles | 19:28 |
| *** bsv has joined #openstack-ansible | 19:28 | |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible-security: Docs: Specify supported distributions https://review.openstack.org/329072 | 19:30 |
| alextricity25 | stevelle: Okay I see. At this point it will connect to galera the exact same way all other services do | 19:32 |
| alextricity25 | stevelle: Maybe we should get rid of "aodh_db_type" all together | 19:33 |
| stevelle | alextricity25: from group vars, yes | 19:33 |
| openstackgerrit | Christopher Hultin proposed openstack/openstack-ansible-os_magnum: Specifying Keystone Authentication Version https://review.openstack.org/329162 | 19:34 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: [WIP] Update lists of skipped security role tasks https://review.openstack.org/327732 | 19:37 |
| chris_hultin | mhayden: Yep, just another small little 'gotcha' we're running into with Magnum. | 19:38 |
| alextricity25 | stevelle: Were you testing against a local mysql DB or galera? | 19:44 |
| *** woodard_ has quit IRC | 19:44 | |
| stevelle | local | 19:44 |
| palendae | mhayden, automagically Have we seen other patches need this many rechecked? https://review.openstack.org/#/c/325380/ | 19:44 |
| *** javeriak has quit IRC | 19:44 | |
| alextricity25 | Ahhhh okay | 19:44 |
| stevelle | it was containered though | 19:44 |
| *** woodard has joined #openstack-ansible | 19:45 | |
| alextricity25 | Should we have to connect to galera by default you think? | 19:45 |
| mhayden | palendae: i have seen a few take that many | 19:45 |
| stevelle | alextricity25: that's why the role has mysql+pymysql default now | 19:45 |
| mhayden | however, something is slow in gate-land because even odyssey4me's overlayfs patch is taking a while | 19:45 |
| *** woodard has quit IRC | 19:45 | |
| stevelle | only reason any of that was in group vars before was because mongo | 19:46 |
| *** woodard has joined #openstack-ansible | 19:46 | |
| palendae | mhayden, Recently? Just wanna make sure that change didn't introduce additional time, even if that's unlikely | 19:46 |
| alextricity25 | stevelle: I'm getting this strange error when I run aodh-dbsync | 19:46 |
| mhayden | palendae: i'd be surprised if it did, but anyhting is possible | 19:46 |
| alextricity25 | ERROR aodh NoMatches: No 'aodh.storage' driver found, looking for '' | 19:46 |
| alextricity25 | THat's when I try to connect to galera | 19:46 |
| palendae | mhayden, Ok. Just to be safe I'm gonna run some local tests for speed of head of master vs that | 19:46 |
| palendae | Inventory should generally be a tiny fraction of runtime, but seeing as it hasn't verified even once, I'm skeptical | 19:47 |
| stevelle | alextricity25: strange | 19:49 |
| *** fedruantine has quit IRC | 19:53 | |
| palendae | mhayden, In terms of test execution time that patch only added 1 second locally, so I don't think that's it | 20:00 |
| *** woodard_ has joined #openstack-ansible | 20:05 | |
| *** sdake_ is now known as sdake | 20:06 | |
| *** woodard has quit IRC | 20:09 | |
| *** McMurlock has quit IRC | 20:09 | |
| *** woodard_ has quit IRC | 20:14 | |
| *** woodard has joined #openstack-ansible | 20:14 | |
| *** woodard has quit IRC | 20:15 | |
| *** woodard has joined #openstack-ansible | 20:15 | |
| mhayden | just sent an email to the list about the mid-cycle dates | 20:21 |
| mhayden | there is an etherpad on there | 20:21 |
| mhayden | waiting on the hotel discount contract now, but i need to know dates | 20:21 |
| michaelgugino | mhayden: what is the email list? | 20:21 |
| mhayden | openstack-dev | 20:22 |
| openstackgerrit | Miguel Alex Cantu (alextricity25) proposed openstack/openstack-ansible: Configuring AODH DB now that it uses MySQL. https://review.openstack.org/329136 | 20:22 |
| michaelgugino | what is the mailing list for this project? | 20:22 |
| mhayden | michaelgugino: http://lists.openstack.org/pipermail/openstack-dev/2016-June/097218.html | 20:22 |
| mhayden | michaelgugino: you can email openstack-dev with the [openstack-ansible] tag in the subject line | 20:22 |
| michaelgugino | gotcha | 20:22 |
| michaelgugino | thanks | 20:22 |
| mhayden | de nada! | 20:22 |
| *** jwitko has joined #openstack-ansible | 20:23 | |
| *** TxGirlGeek has quit IRC | 20:24 | |
| *** M00nr41n has quit IRC | 20:25 | |
| *** TxGirlGeek has joined #openstack-ansible | 20:30 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-security: Ensure aide-common package is installed https://review.openstack.org/329177 | 20:33 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: [WIP] Update lists of skipped security role tasks https://review.openstack.org/327732 | 20:33 |
| *** TxGirlGeek has quit IRC | 20:44 | |
| openstackgerrit | Cameron Loader proposed openstack/openstack-ansible-os_keystone: Updated the keystone WSGI application locations in Mitaka https://review.openstack.org/329182 | 20:44 |
| *** TxGirlGeek has joined #openstack-ansible | 20:52 | |
| *** cfarquhar has quit IRC | 20:53 | |
| *** psilvad has quit IRC | 20:54 | |
| *** bbmbx__ has quit IRC | 20:55 | |
| *** thorst has quit IRC | 20:56 | |
| *** TxGirlGeek has quit IRC | 20:57 | |
| *** Mudpuppy has quit IRC | 20:59 | |
| *** TxGirlGeek has joined #openstack-ansible | 21:00 | |
| *** thorst has joined #openstack-ansible | 21:03 | |
| *** TheIntern has quit IRC | 21:04 | |
| *** adrian_otto has quit IRC | 21:07 | |
| *** thorst has quit IRC | 21:07 | |
| openstackgerrit | Cameron Loader proposed openstack/openstack-ansible-os_keystone: Update the keystone WSGI application locations https://review.openstack.org/329182 | 21:15 |
| *** jamesdenton has quit IRC | 21:24 | |
| *** woodard_ has joined #openstack-ansible | 21:26 | |
| *** woodard__ has joined #openstack-ansible | 21:28 | |
| *** woodard has quit IRC | 21:29 | |
| *** woodard_ has quit IRC | 21:31 | |
| *** schwicht has quit IRC | 21:32 | |
| *** TxGirlGeek has quit IRC | 21:32 | |
| *** woodard__ has quit IRC | 21:32 | |
| *** thorst has joined #openstack-ansible | 21:35 | |
| *** thorst_ has joined #openstack-ansible | 21:39 | |
| *** thorst has quit IRC | 21:39 | |
| *** cfarquhar has joined #openstack-ansible | 21:40 | |
| *** cfarquhar has quit IRC | 21:40 | |
| *** cfarquhar has joined #openstack-ansible | 21:40 | |
| *** michaelgugino has quit IRC | 21:41 | |
| *** adrian_otto has joined #openstack-ansible | 21:42 | |
| *** thorst_ has quit IRC | 21:43 | |
| *** galstrom is now known as galstrom_zzz | 21:44 | |
| *** admin0 has joined #openstack-ansible | 21:45 | |
| mrda | Morning | 21:46 |
| *** schwicht has joined #openstack-ansible | 21:47 | |
| bsv | mrda: evenings :) | 21:49 |
| spotz | hey mrda | 21:49 |
| *** weezS has quit IRC | 21:49 | |
| *** bsv has quit IRC | 21:53 | |
| mrda | o/ | 21:54 |
| *** messy has quit IRC | 21:55 | |
| *** jamesdenton has joined #openstack-ansible | 21:57 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: [POC] Remove most I/O inventory tests https://review.openstack.org/329201 | 21:57 |
| *** jmckind has quit IRC | 21:57 | |
| palendae | stevelle, sigmavirus24, automagically, izaakk, raddaoui ^ curious about thoughts on that approach. The patch seems kind of messy with mocks being peppered everywhere, but that's partially a result of how we're testing everything from the top level righ tnow | 21:58 |
| *** cloader89 has quit IRC | 21:59 | |
| *** schwicht has quit IRC | 22:09 | |
| jmccrory | chris_hultin: there's quite a bit of config_overrides in the magnum playbook. is it planned to go back and add those directly to the magnum.conf template at some point? | 22:14 |
| *** KLevenstein has quit IRC | 22:16 | |
| *** galstrom_zzz is now known as galstrom | 22:17 | |
| palendae | alextricity25, Are your patches meant to fix this kind of test failure? http://logs.openstack.org/98/329098/2/check/gate-openstack-ansible-dsvm-commit/683c7f4/console.html#_2016-06-13_21_00_26_210 | 22:18 |
| *** galstrom is now known as galstrom_zzz | 22:19 | |
| *** spotz is now known as spotz_zzz | 22:19 | |
| *** saneax_AFK is now known as saneax | 22:21 | |
| *** TxGirlGeek has joined #openstack-ansible | 22:22 | |
| *** schwicht has joined #openstack-ansible | 22:28 | |
| *** retreved has quit IRC | 22:29 | |
| *** schwicht has quit IRC | 22:32 | |
| *** TxGirlGeek has quit IRC | 22:37 | |
| *** saneax is now known as saneax_AFK | 22:40 | |
| *** jamesdenton has quit IRC | 22:46 | |
| *** admin0 has quit IRC | 22:46 | |
| *** jamesdenton has joined #openstack-ansible | 22:50 | |
| *** kstev has quit IRC | 22:52 | |
| *** jamesdenton has quit IRC | 22:54 | |
| *** schwicht has joined #openstack-ansible | 22:54 | |
| *** woodard has joined #openstack-ansible | 22:59 | |
| *** adrian_otto has quit IRC | 22:59 | |
| *** adrian_otto has joined #openstack-ansible | 22:59 | |
| *** schwicht has quit IRC | 23:01 | |
| *** saneax_AFK is now known as saneax | 23:06 | |
| stevelle | palendae: yes | 23:07 |
| *** schwicht has joined #openstack-ansible | 23:08 | |
| *** jthorne has quit IRC | 23:23 | |
| *** phalmos has quit IRC | 23:27 | |
| *** alikins has quit IRC | 23:34 | |
| *** Nepoc has quit IRC | 23:40 | |
| *** mummer has quit IRC | 23:54 | |
| *** weshay has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!