| sigmavirus24 | So apparently you can have a module param like this: https://review.openstack.org/#/c/196943/4/playbooks/roles/os_nova/tasks/nova_service_add.yml | 00:08 |
|---|---|---|
| sigmavirus24 | Fun stuff | 00:08 |
| *** alextricity_h has quit IRC | 00:09 | |
| *** javeriak has joined #openstack-ansible | 00:13 | |
| *** javeriak has quit IRC | 00:18 | |
| *** logan2 has quit IRC | 00:19 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 00:19 |
| *** javeriak has joined #openstack-ansible | 00:21 | |
| *** logan2 has joined #openstack-ansible | 00:22 | |
| *** javeriak has quit IRC | 00:24 | |
| *** sdake has quit IRC | 00:29 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 00:31 |
| *** annashen has joined #openstack-ansible | 00:34 | |
| *** annashen has quit IRC | 00:38 | |
| *** galstrom_zzz is now known as galstrom | 00:52 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 00:58 | |
| *** galstrom is now known as galstrom_zzz | 01:02 | |
| palendae | cloudnull: If I'm reading the list on line 34 and the --delete option correctly, those listed directories should be deleted on a sync, right? https://review.openstack.org/#/c/195399/2/playbooks/repo-clone-mirror.yml | 01:16 |
| palendae | Here's what I'm seeing https://gist.github.com/nrb/c5661fa55bb26119d967; I'd think I would have fewer items, not more, right? | 01:19 |
| palendae | Added that to my review | 01:23 |
| *** galstrom_zzz is now known as galstrom | 01:31 | |
| *** annashen has joined #openstack-ansible | 01:35 | |
| *** annashen has quit IRC | 01:39 | |
| *** sdake has joined #openstack-ansible | 01:57 | |
| *** sdake_ has joined #openstack-ansible | 01:59 | |
| *** sdake has quit IRC | 02:02 | |
| cloudnull | Palendae you'll likely see more entries as its a literal clone of what's upstream which has more bits in it than you need while the repo-build is literally only creating the things needed for a deployment. | 02:03 |
| cloudnull | So its likely to grow. | 02:04 |
| *** davidself has joined #openstack-ansible | 02:10 | |
| *** sdake_ has quit IRC | 02:10 | |
| *** sdake has joined #openstack-ansible | 02:10 | |
| *** annashen has joined #openstack-ansible | 02:36 | |
| *** annashen has quit IRC | 02:41 | |
| openstackgerrit | Miguel Grinberg proposed stackforge/os-ansible-deployment: [WIP] Keystone IdP configuration https://review.openstack.org/194259 | 02:49 |
| *** galstrom is now known as galstrom_zzz | 03:06 | |
| *** JRobinson__ is now known as JRobinson__afk | 03:33 | |
| *** annashen has joined #openstack-ansible | 03:37 | |
| *** annashen has quit IRC | 03:41 | |
| openstackgerrit | Miguel Grinberg proposed stackforge/os-ansible-deployment: [WIP] Keystone IdP configuration https://review.openstack.org/194259 | 04:11 |
| *** annashen has joined #openstack-ansible | 04:37 | |
| *** annashen has quit IRC | 04:42 | |
| *** fawadkhaliq has joined #openstack-ansible | 04:43 | |
| *** JRobinson__afk is now known as JRobinson__ | 04:45 | |
| *** annashen has joined #openstack-ansible | 05:39 | |
| *** annashen has quit IRC | 05:43 | |
| *** fawadkhaliq has quit IRC | 06:04 | |
| *** fawadkhaliq has joined #openstack-ansible | 06:25 | |
| *** fawadkhaliq has quit IRC | 06:31 | |
| *** annashen has joined #openstack-ansible | 06:39 | |
| *** annashen has quit IRC | 06:44 | |
| *** radek_ has joined #openstack-ansible | 06:48 | |
| *** radek_ has quit IRC | 06:52 | |
| *** radek_ has joined #openstack-ansible | 06:53 | |
| *** subscope has quit IRC | 07:24 | |
| *** subscope has joined #openstack-ansible | 07:37 | |
| *** annashen has joined #openstack-ansible | 07:40 | |
| *** vdo has joined #openstack-ansible | 07:40 | |
| *** annashen has quit IRC | 07:45 | |
| evrardjp | hello everyone | 07:47 |
| *** annashen has joined #openstack-ansible | 08:41 | |
| *** annashen has quit IRC | 08:46 | |
| openstackgerrit | Hugh Saunders proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 09:09 |
| evrardjp | still to talk about network models, is the scenario3b possible? (cf. networking guide) | 09:19 |
| openstackgerrit | Andy McCrae proposed stackforge/os-ansible-deployment: Allow Swift middleware to be set via a variable https://review.openstack.org/181560 | 09:30 |
| openstackgerrit | Andy McCrae proposed stackforge/os-ansible-deployment: Allow Swift middleware to be set via a variable https://review.openstack.org/181560 | 09:31 |
| openstackgerrit | Andy McCrae proposed stackforge/os-ansible-deployment: Allow Swift middleware to be set via a variable https://review.openstack.org/181560 | 09:34 |
| openstackgerrit | Hugh Saunders proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 09:36 |
| mattt | evrardjp: morning2u | 09:39 |
| mattt | what is the document that you speak of :) | 09:39 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Updated tempest isolation options https://review.openstack.org/195225 | 09:39 |
| evrardjp | hey mattt | 09:40 |
| evrardjp | this one http://docs.openstack.org/networking-guide/deploy_scenario3b.html | 09:40 |
| *** annashen has joined #openstack-ansible | 09:42 | |
| *** sdake has quit IRC | 09:46 | |
| *** annashen has quit IRC | 09:46 | |
| svg | hello and a goodday | 09:49 |
| *** eandersson_ has joined #openstack-ansible | 09:50 | |
| svg | I'm looking into defining my public endpoint with ssl - I assume the best is to define the public vip in osad config with a hostname instead of an ip | 09:50 |
| svg | and I also assume the internal vif should best be kept as a plain ip | 09:51 |
| svg | anyone other thoughts on this? | 09:51 |
| *** eandersson has quit IRC | 09:53 | |
| *** eandersson_ is now known as eandersson | 09:53 | |
| openstackgerrit | Darren Birkett proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 10:05 |
| evrardjp | svg: I don't know how it works with OSAD, but apparently there is a default variable keystone_service proto, which is by default used for keystone_service_publicuri_proto | 10:14 |
| svg | yes, that is about setting the protocol | 10:15 |
| svg | in latest master/kilo there is now also a general var to set that for all endpoints too | 10:16 |
| svg | question is about hostname vs ip | 10:16 |
| openstackgerrit | Andy McCrae proposed stackforge/os-ansible-deployment: Default to use host/port for glance_api_servers https://review.openstack.org/195496 | 10:17 |
| evrardjp | on your certificate you should always use names and altnames ;) | 10:17 |
| evrardjp | I still don't get the question :p | 10:18 |
| evrardjp | grepping external_lb_vip_address show things (urls) that could be replaced by a hostname... except for one thing: ec2_dmz_host... I'm not sure that could be a hostname instead of an ip | 10:24 |
| mattt | i'd have ot look at how this all pieces together, but typically you'd want your public services registering w/ a hostname since that's what your users will see in the catalog etc. ... internal not so much | 10:25 |
| evrardjp | so I guess we'll need an expert in OSAD to know the impact of using a hostname in external_lb_vip_address... | 10:25 |
| *** JRobinson__ has quit IRC | 10:27 | |
| svg | mattt: exactly + de external services are what will do ssl, and are the ports that will be opened up to external networks | 10:27 |
| *** annashen has joined #openstack-ansible | 10:43 | |
| *** annashen has quit IRC | 10:48 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Create Junit XML Report from tempest run https://review.openstack.org/197512 | 10:54 |
| *** eandersson_ has joined #openstack-ansible | 11:26 | |
| *** eandersson has quit IRC | 11:29 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Add support for specifying custom static routes https://review.openstack.org/197335 | 11:42 |
| *** annashen has joined #openstack-ansible | 11:44 | |
| *** annashen has quit IRC | 11:48 | |
| *** markvoelker has quit IRC | 12:14 | |
| *** markvoelker has joined #openstack-ansible | 12:14 | |
| *** raginbajin has quit IRC | 12:18 | |
| *** raginbajin has joined #openstack-ansible | 12:19 | |
| cloudnull | Morning | 12:28 |
| evrardjp | good morning cloudnull | 12:37 |
| cloudnull | How's it evrardjp | 12:37 |
| evrardjp | I'm following this documentation to setup IPv6 for openstack: http://docs.openstack.org/networking-guide/adv_config_ipv6.html ;) | 12:39 |
| Sam-I-Am | evrardjp: you could read the patch that might land there someday | 12:43 |
| Sam-I-Am | https://review.openstack.org/#/c/181049/4/doc/networking-guide/source/adv_config_ipv6.rst,cm | 12:43 |
| Sam-I-Am | and this | 12:44 |
| Sam-I-Am | https://review.openstack.org/#/c/178739/10/doc/networking-guide/source/adv_config_ipv6.rst,cm | 12:44 |
| *** annashen has joined #openstack-ansible | 12:45 | |
| *** annashen has quit IRC | 12:49 | |
| *** cbaesema has quit IRC | 13:22 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Configurable memcached connections limit & threads https://review.openstack.org/194499 | 13:35 |
| *** annashen has joined #openstack-ansible | 13:45 | |
| *** annashen has quit IRC | 13:50 | |
| *** jwagner_away is now known as jwagner | 13:50 | |
| *** TheIntern has joined #openstack-ansible | 13:54 | |
| palendae | cloudnull: Morning - so running clone mirror (as opposed to just building the mirror) is expected to have more entries? | 13:56 |
| *** yapeng has joined #openstack-ansible | 13:58 | |
| *** KLevenstein has joined #openstack-ansible | 13:58 | |
| *** alextricity has quit IRC | 14:01 | |
| *** subscope has quit IRC | 14:03 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:12 | |
| *** Mudpuppy has joined #openstack-ansible | 14:14 | |
| cloudnull | palendae: yes. | 14:17 |
| cloudnull | you dont need everything there. but there will be more things | 14:18 |
| cloudnull | in the bug I raise the question if we should have the "mirror-clone" process for kilo ? | 14:18 |
| palendae | I guess I'm still confused by the syntax. As I understood it, --exclude on rsync meant to NOT sync that directory from the source to the destination | 14:19 |
| palendae | Or is that to exclude them from the delete | 14:19 |
| cloudnull | it excludes the directory and or its contents. | 14:20 |
| palendae | From? | 14:20 |
| palendae | The sync or the delete? | 14:21 |
| palendae | I think my expectation's off - if it's excluding from the sync, then why would those directories be added when they're not there? | 14:21 |
| cloudnull | do they have anything within them ? | 14:21 |
| * palendae checks | 14:22 | |
| *** TheIntern has quit IRC | 14:23 | |
| palendae | Yes. container_images has rpc-trust-container.{,old,1}.tgz | 14:23 |
| cloudnull | that it should . | 14:24 |
| cloudnull | what about python_packages | 14:24 |
| palendae | index.html and pools | 14:25 |
| palendae | python_packages/pools/index.html | 14:25 |
| cloudnull | so they're empty otherwise? | 14:25 |
| cloudnull | you dont see all this http://rpc-repo.rackspace.com/python_packages/ | 14:26 |
| palendae | No | 14:26 |
| cloudnull | and in the mirror directory you shouldn't have the old mariadb bits http://rpc-repo.rackspace.com/mirror/ | 14:27 |
| palendae | mirror isn't present. openstackgit is but is empty | 14:27 |
| cloudnull | so these are the base directories that we create on a repo server install | 14:30 |
| cloudnull | https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/repo_server/tasks/repo_post_install.yml#L44-L57 | 14:30 |
| cloudnull | so it sounds like its working | 14:30 |
| cloudnull | you have the basic dirs created by the server setup and the exclude didn't populate anything that it shouldn't | 14:31 |
| palendae | cloudnull: Got it, thanks | 14:38 |
| *** matt____ has joined #openstack-ansible | 14:39 | |
| *** darrenc_ has joined #openstack-ansible | 14:39 | |
| mattt | cloudnull: see my feedback on https://review.openstack.org/#/c/193848/ ? | 14:40 |
| *** serverascode_ has joined #openstack-ansible | 14:41 | |
| *** rackertom_ has joined #openstack-ansible | 14:45 | |
| *** mrodden_ has joined #openstack-ansible | 14:45 | |
| *** eandersson_ has quit IRC | 14:46 | |
| *** errr has quit IRC | 14:46 | |
| *** mancdaz has quit IRC | 14:46 | |
| *** andymccr has quit IRC | 14:46 | |
| *** git-harry has quit IRC | 14:46 | |
| *** rackertom has quit IRC | 14:46 | |
| *** serverascode has quit IRC | 14:46 | |
| *** mrodden has quit IRC | 14:46 | |
| *** darrenc has quit IRC | 14:46 | |
| *** mattoliverau has quit IRC | 14:46 | |
| *** mrodden_ is now known as mrodden | 14:46 | |
| *** annashen has joined #openstack-ansible | 14:46 | |
| cloudnull | mattt: yup. i just got done checking in on the icehouse built wheels and resubmitted the review for recheck. | 14:50 |
| *** rackertom_ is now known as rackertom | 14:51 | |
| *** annashen has quit IRC | 14:51 | |
| *** git-harry has joined #openstack-ansible | 14:51 | |
| *** rackertom has quit IRC | 14:52 | |
| *** eandersson_ has joined #openstack-ansible | 14:52 | |
| *** TheIntern has joined #openstack-ansible | 14:53 | |
| *** errr has joined #openstack-ansible | 14:53 | |
| *** serverascode_ is now known as serverascode | 14:56 | |
| *** subscope has joined #openstack-ansible | 14:57 | |
| *** rackertom has joined #openstack-ansible | 15:00 | |
| *** andymccr has joined #openstack-ansible | 15:06 | |
| *** mancdaz has joined #openstack-ansible | 15:07 | |
| mattt | cloudnull: cool, thanks ! | 15:23 |
| *** weezS has joined #openstack-ansible | 15:30 | |
| *** weezS has quit IRC | 15:33 | |
| *** jaypipes has quit IRC | 15:37 | |
| openstackgerrit | Julian Montez proposed stackforge/os-ansible-deployment: Allow Swift middleware to be set via a variable https://review.openstack.org/181560 | 15:40 |
| sigmavirus24 | cloudnull: sweet | 15:46 |
| cloudnull | huh ? | 15:46 |
| *** annashen has joined #openstack-ansible | 15:47 | |
| jwagner | cloudnull hey weezS is having issues getting into the channel | 15:48 |
| *** gcivitella has joined #openstack-ansible | 15:48 | |
| cloudnull | # use the /join command | 15:48 |
| cloudnull | :) | 15:48 |
| jwagner | basically that dir wasnt a lxc container dir so it failed when trying to access the config file | 15:48 |
| palendae | /join #openstack-ansible | 15:48 |
| jwagner | he cant connect to freenode | 15:48 |
| jwagner | he is working on it | 15:49 |
| cloudnull | i cant fix that | 15:49 |
| jwagner | yeah me niehter | 15:49 |
| jwagner | what do u need from us | 15:49 |
| jwagner | is it just a bad assumption on the play side that all dir in /var/lib/lxc will be container dirs | 15:49 |
| cloudnull | wait, how can he be in rcbops if he cant connect to freenode ? | 15:49 |
| jwagner | dunno | 15:49 |
| palendae | cloudnull: I don't see him in there any more | 15:49 |
| cloudnull | ok anyway ... | 15:50 |
| palendae | jwagner: we got https://github.com/rcbops/rpc-openstack/issues/212, will likely make an LP issue out of it since osad makes that play | 15:50 |
| cloudnull | ^^ | 15:50 |
| palendae | (doing that now) | 15:50 |
| cloudnull | also jwagner no its not a nad assumption. | 15:50 |
| jwagner | yeah i told him to make the rpc-openstack issue then i realized that it was an osad issue, but you guys were already on it | 15:50 |
| cloudnull | we just need to more thoughful how the upgrade play is interating through those items | 15:50 |
| palendae | jwagner: Yeah, starting it there's fine | 15:51 |
| palendae | jwagner: Is this completely blocking your testing? | 15:51 |
| jwagner | ok so for now we went ahead and removed the lost+found dirs | 15:51 |
| jwagner | so we can move forward | 15:51 |
| palendae | i.e. does it need to get in for the Monday tag? | 15:51 |
| *** weezS has joined #openstack-ansible | 15:51 | |
| *** annashen has quit IRC | 15:52 | |
| jwagner | palendae not sure | 15:52 |
| palendae | Ok | 15:52 |
| jwagner | since it is something we can fix on our side to continue testing | 15:52 |
| jwagner | not a big deal to us | 15:52 |
| jwagner | but i can see it causeing ops headaches | 15:53 |
| palendae | Yeah | 15:53 |
| palendae | Ok | 15:53 |
| palendae | Trying to determine if that needs to be an rpc-openstack rc | 15:53 |
| *** jwagner is now known as jwagner_away | 15:53 | |
| *** jwagner_away is now known as jwagner | 15:53 | |
| jwagner | i dont think it will | 15:54 |
| jwagner | nvm yeah it probably will | 15:54 |
| palendae | Ok | 15:54 |
| jwagner | since the osad hash will need to be updated | 15:54 |
| palendae | Right | 15:55 |
| palendae | Ok | 15:55 |
| palendae | jwagner, cloudnull https://bugs.launchpad.net/openstack-ansible/+bug/1470571 | 15:55 |
| openstack | Launchpad bug 1470571 in openstack-ansible trunk "/tmp/fix_host_things.yml file not found issue." [Undecided,New] | 15:55 |
| jwagner | we need to repo_container lv size fix in as well | 15:55 |
| jwagner | we keep running out of space | 15:55 |
| jwagner | in our tests | 15:55 |
| jwagner | if we can get that green lit | 15:55 |
| palendae | jwagner: https://review.openstack.org/#/c/195399/ needs another +2 and a +W and it'll be in master | 15:56 |
| palendae | That should save some space | 15:56 |
| palendae | At least, that's the patch associated with the bug (https://bugs.launchpad.net/openstack-ansible/+bug/1468567) | 15:56 |
| openstack | Launchpad bug 1468567 in openstack-ansible "v11.0 - Repo Containers - No space left on device" [High,In progress] - Assigned to Kevin Carter (kevin-carter) | 15:56 |
| *** alextricity has joined #openstack-ansible | 15:59 | |
| *** yapeng has quit IRC | 16:05 | |
| *** jwagner is now known as jwagner_away | 16:11 | |
| *** eandersson_ has quit IRC | 16:22 | |
| *** sdake has joined #openstack-ansible | 16:24 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Handle proxies through environment variables https://review.openstack.org/192717 | 16:31 |
| *** jwagner_away is now known as jwagner | 16:32 | |
| *** vdo has quit IRC | 16:38 | |
| *** gcivitella has quit IRC | 16:48 | |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: Fixed the upgradde script for fix_host_thing.yml https://review.openstack.org/197657 | 16:48 |
| *** annashen has joined #openstack-ansible | 16:48 | |
| cloudnull | jwagner: ^^ patch for the upgrade | 16:51 |
| jwagner | ty sir | 16:51 |
| *** annashen has quit IRC | 16:53 | |
| sigmavirus24 | hughsaunders: have you pushed what you have? | 16:55 |
| sigmavirus24 | or are you still tinkering? | 16:56 |
| hughsaunders | sigmavirus24: yeah still tinkering | 16:56 |
| sigmavirus24 | well I'm going to go ride a bicycle for a short bit. ping me if you push something up in the meantime | 16:58 |
| sigmavirus24 | hughsaunders: on a bright not | 16:58 |
| sigmavirus24 | *note | 16:58 |
| sigmavirus24 | I *just* reproduced the gate failure | 16:59 |
| hughsaunders | sigmavirus24: great :) | 16:59 |
| sigmavirus24 | all the -vvvvs | 16:59 |
| *** Mudpuppy_ has joined #openstack-ansible | 17:00 | |
| *** Mudpuppy_ has quit IRC | 17:01 | |
| *** Mudpuppy_ has joined #openstack-ansible | 17:01 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 17:04 | |
| *** Mudpuppy has quit IRC | 17:04 | |
| jwagner | cloudnull with the removal of those dir and some cleaning up behind weezS the upgrade seems to be going | 17:08 |
| jwagner | well | 17:08 |
| *** annashen has joined #openstack-ansible | 17:10 | |
| *** sdake_ has joined #openstack-ansible | 17:18 | |
| *** sdake has quit IRC | 17:21 | |
| *** jaypipes has joined #openstack-ansible | 17:22 | |
| *** Mudpuppy_ has quit IRC | 17:33 | |
| *** sdake_ is now known as sdake | 17:39 | |
| openstackgerrit | Hugh Saunders proposed stackforge/os-ansible-deployment: [WIP] Add v3 calls for federation to keystone module https://review.openstack.org/197677 | 17:44 |
| *** sdake has quit IRC | 17:44 | |
| hughsaunders | sigmavirus24_awa: Have pushed, but not much I'm afraid. Will carry on tomorrow. | 17:45 |
| *** sdake has joined #openstack-ansible | 17:46 | |
| cloudnull | hughsaunders: https://review.openstack.org/#/c/197677/ needs to be rebased | 17:47 |
| *** sdake has quit IRC | 17:48 | |
| *** sdake has joined #openstack-ansible | 17:49 | |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: [WIP] Add v3 calls for federation to keystone module https://review.openstack.org/197677 | 17:50 |
| *** tbyrom138 has joined #openstack-ansible | 17:50 | |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: Remove unused swift vars https://review.openstack.org/197094 | 17:54 |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: Updated tempest isolation options https://review.openstack.org/197342 | 17:56 |
| cloudnull | cores please do the needful for icehouse https://review.openstack.org/#/c/193848/ | 17:57 |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: Fixed the upgrade script edge case https://review.openstack.org/197657 | 17:59 |
| *** tbyrom138_ has joined #openstack-ansible | 18:01 | |
| *** tbyrom138 has quit IRC | 18:04 | |
| *** tbyrom138_ is now known as tbyrom138 | 18:04 | |
| *** yapeng has joined #openstack-ansible | 18:06 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 18:06 | |
| *** yapeng has quit IRC | 18:11 | |
| *** galstrom_zzz is now known as galstrom | 18:27 | |
| *** sdake has quit IRC | 18:38 | |
| *** sdake has joined #openstack-ansible | 18:38 | |
| *** sdake_ has joined #openstack-ansible | 18:41 | |
| *** sdake has quit IRC | 18:44 | |
| *** javeriak has joined #openstack-ansible | 18:50 | |
| sigmavirus24 | hughsaunders: I know why that's failing | 18:54 |
| *** Mudpuppy has joined #openstack-ansible | 18:54 | |
| *** tbyrom138 has quit IRC | 18:58 | |
| sigmavirus24 | svg: ping | 18:59 |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment: Update port definition for PLUMgrid in Juno https://review.openstack.org/197705 | 19:00 |
| svg | sigmavirus24: pong | 19:00 |
| sigmavirus24 | Mind if I DM you? | 19:00 |
| cloudnull | you have to follow first | 19:02 |
| cloudnull | :D | 19:02 |
| palendae | cloudnull: Nope! | 19:03 |
| palendae | There's a 'allow randos to DM me' option | 19:03 |
| cloudnull | ah. | 19:03 |
| palendae | Fairly new | 19:03 |
| * cloudnull goes to make sure thats off | 19:03 | |
| sigmavirus24 | lol | 19:03 |
| palendae | Yeah, I had it turned off then interacted with a hotel via their twitter account | 19:04 |
| *** tbyrom138 has joined #openstack-ansible | 19:05 | |
| cloudnull | yup its good to go | 19:06 |
| *** yapeng has joined #openstack-ansible | 19:06 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 19:08 |
| *** yapeng has quit IRC | 19:11 | |
| svg | sigmavirus24: of course, no need to ask | 19:12 |
| svg | sigmavirus24: palendae tbh, asking permission to PM is fairly new to me, is that some dark part of netiquette I don't know about? Pretty rare from my experience. | 19:19 |
| palendae | svg: Not that I know of | 19:20 |
| palendae | I was speaking of DMs on twitter myself - they require that you're followed by the person in order to send, afaik | 19:21 |
| svg | :) | 19:21 |
| *** sdake has joined #openstack-ansible | 19:21 | |
| *** sdake has quit IRC | 19:23 | |
| *** sdake has joined #openstack-ansible | 19:23 | |
| sigmavirus24 | svg: it's netiquette that I follow mainly because: A) Freenode and Mozilla's IRC networks have become overwhelmed with spambots that DM you spam links. B) It seems polite when I don't know the person very well | 19:25 |
| *** sdake_ has quit IRC | 19:25 | |
| svg | ok, if you dm'ed me, for me both cases are not an issue :) | 19:26 |
| *** annashen has quit IRC | 19:29 | |
| *** annashen has joined #openstack-ansible | 19:30 | |
| *** annashen has quit IRC | 19:34 | |
| *** galstrom is now known as galstrom_zzz | 19:49 | |
| *** galstrom_zzz is now known as galstrom | 19:58 | |
| alextricity | Are there any tools for writing the .rst documents under the doc/ directory? | 19:58 |
| alextricity | And validating syntax | 19:58 |
| Sam-I-Am | alextricity: its pretty much straight up rst | 20:01 |
| palendae | alextricity: Only thing for validating the syntax right now is running the make | 20:01 |
| palendae | make html | 20:01 |
| Sam-I-Am | if you 'make html' and it breaks, you did bad | 20:01 |
| palendae | Yeah, pretty much that | 20:01 |
| Sam-I-Am | you can probably glean a lot from the existing docs | 20:01 |
| palendae | No other tools that I'm aware of | 20:01 |
| Sam-I-Am | not that my rst is perfect | 20:02 |
| alextricity | cool. Thanks guys | 20:02 |
| Sam-I-Am | palendae: eventually we may lint it | 20:02 |
| palendae | Sam-I-Am: Sure | 20:02 |
| palendae | Though we don't even have the build job running on commits yet | 20:02 |
| palendae | I'm not aware of rst linters that don't just build, but I could be mistaken | 20:03 |
| Sam-I-Am | palendae: we have niceness and syntax linters | 20:03 |
| Sam-I-Am | er, upstream docs does | 20:04 |
| palendae | Ok | 20:04 |
| palendae | So they exist, it's just my ignorance | 20:04 |
| Sam-I-Am | i suspect we'll just copy that shit over | 20:04 |
| palendae | Yeah | 20:04 |
| palendae | Copy it over and see if anything doesn't work | 20:04 |
| alextricity | odyssey4me: cloudnull: Would I just add an .rst in the install-guide/ directory for the ceilometer bits? | 20:07 |
| alextricity | Or are there docs somewhere else that I have to consider? | 20:08 |
| Sam-I-Am | alextricity: what are you trying to document? | 20:10 |
| Sam-I-Am | alextricity: how to deploy ceilometer with osad? | 20:10 |
| alextricity | Sam-I-Am: configuring ceilometer with osad, and installing it | 20:10 |
| alextricity | would be something like config-ceilometer.rst, and install-ceilometer.rst, right? | 20:11 |
| Sam-I-Am | uhh... let me look | 20:11 |
| *** daneyon has joined #openstack-ansible | 20:11 | |
| Sam-I-Am | yeah i'd say a configure-ceilometer.rst | 20:12 |
| Sam-I-Am | and thats probably it | 20:12 |
| Sam-I-Am | unless there playbooks for it are not included in one of the existing meta-playbooks | 20:12 |
| Sam-I-Am | configure-ceilometer would go with the rest of the configuration sections | 20:12 |
| Sam-I-Am | it would also need to be marked as optional (see configure.rst) | 20:13 |
| alextricity | It is its own playbook, called os-install-ceilometer.yml | 20:13 |
| alextricity | Oh yeah..good point | 20:13 |
| Sam-I-Am | alextricity: so its not included in the usual openstack playbooks? | 20:13 |
| alextricity | It would not be included in the 'setup-openstack.yml' playbook | 20:14 |
| alextricity | it's an optional playbook that the deployer can run if they choose to deploy ceilometer | 20:14 |
| Sam-I-Am | iirc is should be 'os-ceilometer-install.yml' for consistency | 20:14 |
| alextricity | The PR has it that way...i just couldn't remember ;) | 20:15 |
| alextricity | lol | 20:15 |
| Sam-I-Am | you might want to add an optional step for the ceilometer playbook to install-openstack-run.rst if thats where it needs to run (not earlier) | 20:16 |
| alextricity | Thanks a good idea. you know what..I might just do what swift does | 20:21 |
| alextricity | the swift playbooks don't do anything if the hosts are not defined in openstack_user_config. Maybe I should just do that | 20:21 |
| palendae | alextricity: That's a safe bet | 20:22 |
| Sam-I-Am | alextricity: that would be a better way | 20:23 |
| *** tbyrom138 has quit IRC | 20:25 | |
| *** galstrom is now known as galstrom_zzz | 20:25 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 20:33 |
| *** openstackgerrit has quit IRC | 20:37 | |
| *** openstackgerrit has joined #openstack-ansible | 20:37 | |
| *** tbyrom138 has joined #openstack-ansible | 20:39 | |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Configurable memcached connections limit & threads https://review.openstack.org/197729 | 20:40 |
| *** annashen has joined #openstack-ansible | 20:46 | |
| *** KLevenstein has quit IRC | 20:47 | |
| *** daneyon has quit IRC | 20:47 | |
| *** jwagner is now known as jwagner_away | 20:54 | |
| palendae | cloudnull: Do you have any suggestions for handling overrideable variables in Juno? iirc, most of them are in group_vars, and those aren't overriden with rpc_user_config, right? | 20:57 |
| *** matt____ is now known as mattoliverau | 20:58 | |
| cloudnull | What do you need to override ? | 21:00 |
| palendae | stevelle was asking me | 21:00 |
| palendae | So dunno :p | 21:00 |
| stevelle | cloudnull: it's for the memcaced connections limit | 21:00 |
| palendae | I thought, in general, group_vars were harder to override, hence moving most variables out into defaults/main.yml of their role | 21:00 |
| cloudnull | change the role to support defaults and it should be ok. | 21:01 |
| stevelle | I was doing that but it was getting really ugly, repeating the default 4 or 5 times | 21:01 |
| palendae | Because you were dragging it across services? | 21:02 |
| stevelle | no because the connection limit is used, counting, 4 times across the plays and templates | 21:03 |
| palendae | With different names? | 21:04 |
| palendae | So if you define memcache_connection_limit in memcache/defaults/main.yml, those 4 other places will need to refer to it, sure | 21:04 |
| stevelle | so introduce a defaults/main.yml then | 21:05 |
| palendae | Yeah, that's what I'm hearing | 21:05 |
| palendae | Stuff in there can then be overriden | 21:05 |
| palendae | It should probably follow whatever convention for that role is present in kilo+ | 21:05 |
| stevelle | ok, now that I understood. will go with that | 21:05 |
| palendae | Redoing the whole role *might* be too much for that backport (I'm assuming it's a backport), but we'll see | 21:06 |
| *** javeriak has quit IRC | 21:06 | |
| *** sdake_ has joined #openstack-ansible | 21:07 | |
| *** sdake has quit IRC | 21:09 | |
| *** javeriak has joined #openstack-ansible | 21:14 | |
| cloudnull | palendae: stevelle: thats what I'd do, and what I had done with the juno backport for tempest . | 21:17 |
| jmccrory | palendae: you can override in user_variables, the global_overrides block in rpc_user_config isn't really true. | 21:18 |
| stevelle | I initially misunderstood the use of defaults as repeating |default('stuff') but just adding the defaults/main.yml is fine with me | 21:18 |
| palendae | Oooh, yeah, no | 21:18 |
| palendae | That would be horrible | 21:18 |
| palendae | jmccrory: Yeah, it'd have to be top-level, not inside global_overrides | 21:19 |
| jmccrory | i think global_overrides eventually feed into inventory, but role, playbook, and extra vars all override inventory...so not a great name | 21:21 |
| palendae | Right | 21:21 |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Configurable memcached connections limit & threads https://review.openstack.org/197740 | 21:24 |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Configurable memcached connections limit & threads https://review.openstack.org/197740 | 21:25 |
| *** javeriak has quit IRC | 21:27 | |
| *** javeriak has joined #openstack-ansible | 21:28 | |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Clear expired Django sessions regularly from DB https://review.openstack.org/197744 | 21:38 |
| svg | cloudnull: stevelle palendae catching up on ^^ but if you think group_vars is hard, that's because the groups parent/child structure in osad is non-existent | 21:41 |
| svg | I noticed kilo goes even further into not using group vars | 21:42 |
| palendae | svg: It's way better in Kilo | 21:42 |
| svg | palendae: how is that? | 21:42 |
| palendae | Well, "better" in that we've moved things out into defaults/main.yml in the roles | 21:42 |
| palendae | Not necessarily that we've changed any of the grouping | 21:42 |
| svg | k, but that is not better as in making use of the ansible inventory features :) | 21:42 |
| palendae | Can you give an example of where parent/child would make sense? | 21:43 |
| svg | overruling vars in group vars strongly depends on the group structure, you basically need a tree | 21:43 |
| palendae | Yeah, I'm not sure what that tree would look like | 21:43 |
| stevelle | I'm not sure that a parent/child structure would make that much sense for OpenStack. | 21:43 |
| stevelle | The shared infra, particularly, works against the model that I would image makes that valuable. | 21:44 |
| palendae | Initial thought would be the 'infra' stuff like cache, MQ, DB, at root, then service, then individual APIs | 21:44 |
| palendae | But I kind of agree with stevelle. Totally possible I've just been too close to how osad does it for too long, though | 21:44 |
| svg | my knowledge of how osad works is still too little to fully grasp how and why things are done now tbh | 21:44 |
| palendae | Fair | 21:44 |
| palendae | I could see it in some situations | 21:44 |
| stevelle | palendae: mongodb! | 21:45 |
| palendae | Where you're controlling OS-level stuff, then branching for DBs/web servers | 21:45 |
| svg | palendae: i come from a point where i extensively use group dependencies to overrules things | 21:46 |
| svg | think a tomcat group with different applivcatsion sub groups wit each a sb group per dev/test/prod envronment | 21:47 |
| palendae | Ah, ok | 21:47 |
| palendae | So you override as you go down | 21:47 |
| svg | so coming here and seeing that not used at all, is a bit counterintuitive | 21:47 |
| svg | yes | 21:47 |
| *** javeriak has quit IRC | 21:48 | |
| palendae | So you'd have a tomcat role and group? | 21:48 |
| svg | yes | 21:48 |
| svg | and that tomcat role manages hundreds of similar apps, but which have a bunch os separate things depending on the app and on the environment | 21:49 |
| stevelle | OpenStack relies on shared infra and one-offs as opposed to dedicated infra per service which makes that hard to apply | 21:50 |
| svg | perhaps an example: there was recently two patches that 1 allow to set http vs https for endpoints per application, and 2 that sets a default for all endpoints | 21:50 |
| svg | tis is now handled by all those vars being prefixed | 21:50 |
| svg | could have been a default in all.yml, and overriding it on openstack app level | 21:51 |
| svg | which would make the code simpler imho, no need to var|default() etc | 21:51 |
| svg | not saying current way is bad though, just a very new approach to me | 21:52 |
| svg | there was a time in ansible where this group dependency and vars overruling was flawed though, and lots of people turned away from it | 21:53 |
| svg | fixed since 1.7 | 21:53 |
| palendae | Yeah, osad started prior to 1.6 iirc, so that may be why | 21:54 |
| svg | I thinkn cloudnull told me something that was kind of a reason, yes | 21:54 |
| palendae | I'm out for the evening, night all | 21:54 |
| svg | but using inventory can be quie powerfull | 21:54 |
| svg | bye | 21:54 |
| *** javeriak has joined #openstack-ansible | 21:55 | |
| *** TheIntern has quit IRC | 22:01 | |
| *** Mudpuppy has quit IRC | 22:03 | |
| *** errr has quit IRC | 22:12 | |
| openstackgerrit | Ian Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3 https://review.openstack.org/196943 | 22:15 |
| *** sdake_ is now known as sdake | 22:17 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Updated icehouse to the latest SHAs - 06.20.2015 https://review.openstack.org/193848 | 22:18 |
| *** annashen has quit IRC | 22:20 | |
| *** annashen has joined #openstack-ansible | 22:21 | |
| *** javeriak has quit IRC | 22:21 | |
| *** annashen has quit IRC | 22:25 | |
| *** sdake is now known as sdake_ | 22:29 | |
| *** darrenc_ is now known as darrenc | 22:37 | |
| *** sdake_ has quit IRC | 22:54 | |
| *** radek_ has quit IRC | 23:05 | |
| *** javeriak has joined #openstack-ansible | 23:16 | |
| *** tbyrom138 has quit IRC | 23:24 | |
| *** weezS has quit IRC | 23:28 | |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: Updated tempest isolation options https://review.openstack.org/197342 | 23:28 |
| sigmavirus24 | burned about an hour on https://bugs.launchpad.net/keystone/+bug/1470670 this afternoon | 23:36 |
| openstack | Launchpad bug 1470670 in Keystone "When creating a user with v3, if the domain is excluded a 401 Unauthenticated is returned" [Undecided,New] | 23:36 |
| *** annashen has joined #openstack-ansible | 23:51 | |
| *** JRobinson__ has joined #openstack-ansible | 23:55 | |
| *** annashen has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!