| *** britthouser has quit IRC | 00:01 | |
| *** fangfenghua_ is now known as fangfenghua | 00:02 | |
| *** britthouser has joined #openstack-ansible | 00:17 | |
| *** alejandrito has quit IRC | 00:19 | |
| *** appprod0 has joined #openstack-ansible | 00:28 | |
| *** Mudpuppy has joined #openstack-ansible | 00:31 | |
| *** Mudpuppy has quit IRC | 00:32 | |
| *** britthouser has quit IRC | 00:50 | |
| *** radek__ has joined #openstack-ansible | 01:02 | |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment: Modularizing Neutron playbooks https://review.openstack.org/184665 | 01:09 |
|---|---|---|
| *** radek__ has quit IRC | 01:31 | |
| *** serverascode has quit IRC | 01:48 | |
| *** sacharya has quit IRC | 01:48 | |
| *** serverascode has joined #openstack-ansible | 01:49 | |
| *** fangfenghua has quit IRC | 01:53 | |
| *** javeriak has quit IRC | 02:14 | |
| *** javeriak has joined #openstack-ansible | 02:43 | |
| *** jwagner is now known as jwagner_away | 02:45 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 04:20 | |
| *** sacharya has joined #openstack-ansible | 04:40 | |
| *** fangfenghua has joined #openstack-ansible | 04:47 | |
| *** sacharya has quit IRC | 05:30 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 05:32 | |
| *** appprod0 has quit IRC | 06:41 | |
| *** jcannava has quit IRC | 06:43 | |
| *** mordred has quit IRC | 06:43 | |
| *** galstrom_zzz has quit IRC | 06:43 | |
| *** galstrom_zzz has joined #openstack-ansible | 06:43 | |
| *** mordred has joined #openstack-ansible | 06:43 | |
| *** jcannava has joined #openstack-ansible | 06:43 | |
| *** radek__ has joined #openstack-ansible | 06:59 | |
| *** radek__ has quit IRC | 07:34 | |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment-specs: Add PLUMgrid Neutron Plugin to neutron playbooks https://review.openstack.org/184726 | 07:37 |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment-specs: Add PLUMgrid Neutron Plugin to neutron playbooks https://review.openstack.org/184726 | 07:41 |
| svg | mattt: thanks for the reviews; about the commit message, the problem is I didn't add some detailed description? ( I didn't because it looked so obvious); if I'm missing something else, I'd be glad to hear it. | 08:03 |
| *** javeriak has quit IRC | 08:08 | |
| mattt | svg: i think the generally accepted structure is a brief description on first line (limited to 50 chars), then a detailed description that is wrapped at 72 chars | 08:17 |
| mattt | it seems a bit pedantic when you're dealing w/ such trivial changes but it helps when someone's sorting through chanelog to see the intent without looking at actual code | 08:18 |
| mattt | i +2d anyway, more as an FYI for future reviews | 08:18 |
| svg | np[ & ack; I left out the desccription as it seemed so obvious, and such a small patch, but I understand; I'm happy to learn this more professional way of working, been spoiled with github before :) | 08:19 |
| mattt | svg: did you figure out that networking issue ? | 08:22 |
| *** daneyon has joined #openstack-ansible | 08:23 | |
| svg | not yet, but we stumped into other weird things - suspecting a problem with interfaces and/or bonding config or hw issue | 08:23 |
| svg | I started doing a kilo setup yesterday on similar hardware, fresh setup, and am having issues with connectivity at deploy time | 08:24 |
| svg | ansible suddely cannot connect to containers, while it worked fine in the previosu task, etc | 08:24 |
| mattt | svg: https://github.com/stackforge/os-ansible-deployment/commit/59c1a7f1c1e6f53b151a175482d471a3c086ecc9 | 08:30 |
| mattt | it may have something to do w/ this | 08:30 |
| *** daneyon has quit IRC | 08:31 | |
| svg | don't think so, I had it with juno too | 08:31 |
| svg | if ssh fails at the first try, not sure why it should work at the second of third time | 08:32 |
| svg | btw, do you know how mtu's are handled within containers? especially with regards to the vxlan tunneling | 08:33 |
| mattt | svg: i know there's a network_device_mtu = 1450 set in neutron.conf, but don't know much more than that unfortunately | 08:38 |
| svg | sounds about right, is one of the paths I'm looking at | 08:38 |
| svg | mattt: did you actually finished a setup with ceph support based on my patch? | 09:01 |
| mattt | svg: i think i got as far as nova | 09:10 |
| mattt | so glance and cinder were working on rbd correctly | 09:10 |
| svg | ok | 09:10 |
| svg | I'll see when I can lok at your remarks I still have to address | 09:10 |
| mattt | cool, i will do further testing when that review comes out of WIP | 09:11 |
| mattt | it feels like it's 95% there | 09:11 |
| *** sdake has joined #openstack-ansible | 09:40 | |
| svg | in Kilo, I want to override some (roles) defaults | 10:17 |
| svg | more specific from the nova role | 10:17 |
| svg | I'm not sure where to add an overrides or hostvars stanza in openstack_user_config.yml | 10:18 |
| *** vincent_vdk has quit IRC | 10:21 | |
| *** vincent_vdk has joined #openstack-ansible | 10:24 | |
| *** sdake has quit IRC | 10:26 | |
| svg | would that ba a container_vars section on compute_hosts, or infra_hosts, or both? (it's about vars that set the max ratio's for mem and cpu and number of instances etc.) | 10:31 |
| mattt | i typically add those to user_variables.yml | 10:40 |
| mattt | which var specifically? | 10:40 |
| svg | # nova_cpu_allocation_ratio: 2.0 | 10:51 |
| svg | # nova_max_instances_per_host: 50 | 10:51 |
| svg | # nova_ram_allocation_ratio: 1.0 | 10:51 |
| mattt | yeah i don't see why you can't bung those into user_variables.yml | 10:51 |
| mattt | then they'll apply to anything nova-related | 10:51 |
| svg | hm, in user vars, that would have the advantage to not need to repet them per host | 10:52 |
| svg | guess I could do the same for cinder_backends | 10:55 |
| mattt | i think cinder_backends makes more sense in openstack_user_config.yml | 11:01 |
| mattt | since it typically doesn't apply to everything | 11:01 |
| svg | So does nova? | 11:58 |
| svg | either way most variables exist pretty much everywhere :) | 11:59 |
| svg | luckily kilo/master is already a lot more simple than juno | 11:59 |
| svg | but I could rant on the way the ansible inventory is setup :p | 12:00 |
| mattt | hehehe | 12:00 |
| mattt | yeah it is a lot cleaner | 12:00 |
| mattt | you should see icehouse, that was horrible | 12:00 |
| *** shausy has joined #openstack-ansible | 12:02 | |
| *** shausy has left #openstack-ansible | 12:03 | |
| *** Mudpuppy has joined #openstack-ansible | 14:18 | |
| *** jwagner_away is now known as jwagner | 14:23 | |
| *** radek__ has joined #openstack-ansible | 14:32 | |
| *** fangfenghua has quit IRC | 15:13 | |
| *** radek__ has quit IRC | 15:13 | |
| *** britthouser has joined #openstack-ansible | 15:37 | |
| *** britthou_ has joined #openstack-ansible | 15:39 | |
| *** britthouser has quit IRC | 15:41 | |
| *** sdake has joined #openstack-ansible | 15:54 | |
| *** sdake has quit IRC | 16:00 | |
| *** javeriak has joined #openstack-ansible | 16:06 | |
| svg | mattt now I think of it, using rbd as cinder backend does not needs nova needs an rbd backend, but nova, or better libvirt, does need to get configured with the ceph client to access volumes | 16:08 |
| *** Mudpuppy has quit IRC | 16:30 | |
| *** jlvillal has quit IRC | 16:37 | |
| *** jimi|ansible has joined #openstack-ansible | 16:46 | |
| *** britthou_ has quit IRC | 16:46 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 16:47 | |
| jimi|ansible | does anyone have an example of a "bindep" file I can test? | 16:47 |
| *** britthouser has joined #openstack-ansible | 16:52 | |
| svg | WHat's a bindep? (hi jimi) | 17:04 |
| jimi|ansible | svg: it's a file format they use for package deps across platforms | 17:05 |
| svg | jimi|ansible: you know if that is available somewehere on an openstack cluster? I can look for it | 17:11 |
| *** daneyon has joined #openstack-ansible | 17:11 | |
| *** daneyon_ has joined #openstack-ansible | 17:15 | |
| *** daneyon has quit IRC | 17:18 | |
| *** daneyon_ has quit IRC | 17:34 | |
| *** britthouser has quit IRC | 17:36 | |
| openstackgerrit | Serge van Ginderachter proposed stackforge/os-ansible-deployment: Ceph/RBD support https://review.openstack.org/181957 | 17:42 |
| *** daneyon has joined #openstack-ansible | 17:43 | |
| *** britthouser has joined #openstack-ansible | 17:44 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 17:49 | |
| *** javeriak has quit IRC | 17:50 | |
| *** javeriak has joined #openstack-ansible | 18:07 | |
| javeriak | hey guys, so I comited https://review.openstack.org/#/c/184665/ for addressing the BP: https://blueprints.launchpad.net/openstack-ansible/+spec/modularize-neutron-plays | 18:15 |
| javeriak | my question about the neutron-agent service still stands so it isn't include it in my changes, any more info would be appreciated, I can then push in changes | 18:17 |
| *** britthouser has quit IRC | 18:30 | |
| *** Mudpuppy has joined #openstack-ansible | 18:33 | |
| *** appprod0 has joined #openstack-ansible | 18:40 | |
| *** jlvillal has joined #openstack-ansible | 18:51 | |
| *** sdake has joined #openstack-ansible | 18:57 | |
| *** mattt has quit IRC | 19:28 | |
| *** sdake has quit IRC | 19:31 | |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment: Modularizing Neutron playbooks https://review.openstack.org/184665 | 19:36 |
| *** daneyon has quit IRC | 19:43 | |
| *** Mudpuppy_ has joined #openstack-ansible | 19:59 | |
| *** Mudpuppy has quit IRC | 20:03 | |
| *** Mudpuppy_ has quit IRC | 20:04 | |
| *** javeriak has quit IRC | 20:08 | |
| *** appprod0 has quit IRC | 20:08 | |
| *** appprod0 has joined #openstack-ansible | 20:12 | |
| *** daneyon has joined #openstack-ansible | 20:30 | |
| *** mattt has joined #openstack-ansible | 20:35 | |
| *** mattt has quit IRC | 20:35 | |
| *** mattt has joined #openstack-ansible | 20:35 | |
| mattt | javeriak: haven't a clue to be honest, let me spin up a juno AIO to check | 20:36 |
| *** daneyon_ has joined #openstack-ansible | 20:42 | |
| *** daneyon_ has quit IRC | 20:43 | |
| *** daneyon_ has joined #openstack-ansible | 20:44 | |
| *** daneyon has quit IRC | 20:45 | |
| *** javeriak has joined #openstack-ansible | 20:59 | |
| *** appprod0 has quit IRC | 21:04 | |
| openstackgerrit | Javeria Khan proposed stackforge/os-ansible-deployment: Modularizing Neutron playbooks https://review.openstack.org/184665 | 21:06 |
| *** daneyon_ has quit IRC | 21:16 | |
| *** britthouser has joined #openstack-ansible | 21:20 | |
| *** JRobinson__ has joined #openstack-ansible | 21:29 | |
| *** britthouser has quit IRC | 21:31 | |
| mattt | javeriak: i don't see a neutron-agent service anywhere | 21:35 |
| javeriak | matt: take a look at 'service_names' in 'rpc_deployment/inventory/group_vars/neutron_all.yml' in juno branch, its listed as a service, there is also an emptry service_var for it under: rpc_deployment/vars/openstack_service_vars/neutron_agent.yml | 21:37 |
| mattt | javeriak: oh no i mean i don't see any such service running on a deployed AIO | 21:38 |
| javeriak | matt: yes exactly, as far i know, there is no such openstack service, however its been declared in both juno and master, and therefore the handler tries to restart it on neutron nodes | 21:39 |
| mattt | how does that not fail the run? | 21:39 |
| javeriak | matt: I'm not sure, it just always passes as 'ok' during restart tasks | 21:40 |
| javeriak | Also I don't mean to modify the original playbook structure in my commit, so I've put it back in, in my latest patch set. But if its an empty service, it should probably be removed | 21:43 |
| mattt | javeriak: i dunno i don't see anything in my run referencing neutron-agent | 21:57 |
| *** appprod0 has joined #openstack-ansible | 21:59 | |
| *** jwagner is now known as jwagner_away | 21:59 | |
| mattt | javeriak: which is odd because it should in theory! | 21:59 |
| javeriak | matt: oh okay, I'll just leave it in by default... btw i need to push another commit that will be dependent on a current one, do i need to wait until the first one is merged? | 22:00 |
| mattt | javeriak: i don't see why it needs to be there, it's probably safe to remove, but let's circle back when more people are online after the summit to confirm | 22:01 |
| javeriak | matt: oh so that's why there was no meeting this morning! | 22:01 |
| mattt | javeriak: yeah most of the team are busy w/ the summit! | 22:01 |
| mattt | javeriak: have a look at http://docs.openstack.org/infra/manual/developers.html, "Adding a Dependency" | 22:02 |
| javeriak | matt: cool, will I also have to wait for reviews on my new bp and commit till next week i suppose | 22:02 |
| mattt | javeriak: iirc you git review -d <parent>, then git checkout -b branch .... if you do it in that order when you do a git review to send up change it will confirm that you're making a dependent change | 22:03 |
| mattt | javeriak: you can submit the next one, it's no issue, but each time the parent changes you'll have to rebase the other commit | 22:03 |
| javeriak | matt: yes i can add the dependency, was just wondering if there was some kind of community policy against it | 22:03 |
| mattt | javeriak: nope i've not seen anyone mind at all | 22:04 |
| *** britthouser has joined #openstack-ansible | 22:05 | |
| javeriak | matt: okay cool, glad to see there are people still around holding the fort like me :P | 22:05 |
| mattt | javeriak: :P ok heading out for the night, we'll get some eyes on those reviews soon! | 22:08 |
| javeriak | matt: thanks! | 22:08 |
| palendae | svg, I misspoke yesterday. odyssey4me and cloudnull just set me straight. repo_hosts is optional, but if undefined you need to have a repo_url that's accessible | 22:26 |
| mattt | "openstack_repo_url" | 22:26 |
| mattt | was just looking at that | 22:26 |
| palendae | ^ | 22:26 |
| palendae | Yeah, I misunderstood what cloudnull said yesterday | 22:27 |
| mattt | i assumed this thing was mandatory too | 22:27 |
| *** appprod0 has quit IRC | 22:35 | |
| *** sdake has joined #openstack-ansible | 22:42 | |
| *** britthouser has quit IRC | 22:47 | |
| *** JRobinson__ is now known as JRobinson__afk | 22:48 | |
| *** appprod0 has joined #openstack-ansible | 22:54 | |
| *** britthouser has joined #openstack-ansible | 22:55 | |
| cloudnull | mattt: you dont need it, i would recommend it, it makes the whole redeployment faster, however it can be pointed to an upstream repo or elsewhere. | 22:57 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Update all services in master to current head https://review.openstack.org/182476 | 23:07 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Re-enable building of heat_repo_plugins https://review.openstack.org/181946 | 23:07 |
| *** jimi|ansible has quit IRC | 23:09 | |
| *** JRobinson__afk has quit IRC | 23:10 | |
| *** daneyon has joined #openstack-ansible | 23:21 | |
| *** daneyon has quit IRC | 23:35 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 23:42 | |
| *** britthouser has quit IRC | 23:44 | |
| *** britthouser has joined #openstack-ansible | 23:44 | |
| *** javeriak has quit IRC | 23:57 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 23:58 | |
| *** britthouser has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!