| *** hongbin has quit IRC | 00:02 | |
| *** fragatin_ has quit IRC | 00:05 | |
| *** vijendar has quit IRC | 00:05 | |
| *** vijendar has joined #openstack-containers | 00:06 | |
| *** askb has joined #openstack-containers | 00:06 | |
| *** fragatina has joined #openstack-containers | 00:08 | |
| *** vijendar has quit IRC | 00:21 | |
| *** vijendar has joined #openstack-containers | 00:22 | |
| *** gluytium has joined #openstack-containers | 00:23 | |
| *** adrian_otto has joined #openstack-containers | 00:36 | |
| *** adrian_otto has quit IRC | 00:39 | |
| *** adrian_otto has joined #openstack-containers | 00:41 | |
| *** adrian_otto has quit IRC | 00:42 | |
| *** duonghq has joined #openstack-containers | 00:43 | |
| *** adrian_otto has joined #openstack-containers | 00:43 | |
| *** syed_ has quit IRC | 00:45 | |
| *** adrian_otto has quit IRC | 00:46 | |
| *** adrian_otto has joined #openstack-containers | 00:46 | |
| *** tkarke has quit IRC | 01:00 | |
| *** tkarke has joined #openstack-containers | 01:01 | |
| *** tkarke has quit IRC | 01:02 | |
| *** chetna has quit IRC | 01:11 | |
| *** adrian_otto has quit IRC | 01:16 | |
| *** trananhkma has quit IRC | 01:22 | |
| *** adrian_otto has joined #openstack-containers | 01:27 | |
| *** vijendar has quit IRC | 01:30 | |
| *** vijendar has joined #openstack-containers | 01:30 | |
| *** adrian_otto has quit IRC | 01:33 | |
| *** trananhkma has joined #openstack-containers | 01:39 | |
| *** vijendar has quit IRC | 01:46 | |
| *** vijendar has joined #openstack-containers | 01:46 | |
| *** vijendar has quit IRC | 02:02 | |
| *** muralia_ has joined #openstack-containers | 02:12 | |
| *** muralia has quit IRC | 02:14 | |
| *** shu-mutou has joined #openstack-containers | 02:20 | |
| *** hongbin has joined #openstack-containers | 02:24 | |
| *** muralia has joined #openstack-containers | 02:27 | |
| *** muralia_ has quit IRC | 02:27 | |
| *** fragatina has quit IRC | 02:50 | |
| *** fragatina has joined #openstack-containers | 02:55 | |
| *** fragatina has quit IRC | 02:56 | |
| *** adrian_otto has joined #openstack-containers | 03:07 | |
| *** adrian_otto has quit IRC | 03:26 | |
| *** vijendar has joined #openstack-containers | 03:26 | |
| *** vijendar_ has joined #openstack-containers | 03:33 | |
| *** vijendar has quit IRC | 03:33 | |
| *** adisky has joined #openstack-containers | 03:40 | |
| *** tonanhngo has quit IRC | 03:53 | |
| *** vijendar_ has quit IRC | 04:00 | |
| *** dave-mccowan has quit IRC | 04:01 | |
| *** tonanhngo has joined #openstack-containers | 04:02 | |
| *** vijendar has joined #openstack-containers | 04:04 | |
| *** tonanhngo has quit IRC | 04:04 | |
| *** vijendar has quit IRC | 04:08 | |
| *** narayrak has joined #openstack-containers | 04:17 | |
| *** tonanhngo has joined #openstack-containers | 04:24 | |
| *** tonanhngo has quit IRC | 04:25 | |
| *** hongbin has quit IRC | 04:32 | |
| *** tonanhngo has joined #openstack-containers | 04:44 | |
| *** tonanhngo has quit IRC | 04:46 | |
| *** tonanhngo has joined #openstack-containers | 05:05 | |
| *** tonanhngo has quit IRC | 05:06 | |
| *** tonanhngo has joined #openstack-containers | 05:20 | |
| *** tonanhngo has quit IRC | 05:20 | |
| *** ramishra has quit IRC | 05:30 | |
| *** ramishra has joined #openstack-containers | 05:31 | |
| *** chetna has joined #openstack-containers | 05:33 | |
| *** jperry has joined #openstack-containers | 05:42 | |
| *** fragatina has joined #openstack-containers | 05:43 | |
| *** takashi has joined #openstack-containers | 05:46 | |
| *** fragatina has quit IRC | 05:47 | |
| *** rcernin has joined #openstack-containers | 05:54 | |
| *** vimal has joined #openstack-containers | 05:58 | |
| *** chetna has quit IRC | 05:59 | |
| *** takashi has quit IRC | 06:00 | |
| *** GB21 has joined #openstack-containers | 06:01 | |
| *** vmud213 has joined #openstack-containers | 06:35 | |
| *** dimtruck is now known as zz_dimtruck | 06:36 | |
| *** chetna has joined #openstack-containers | 06:44 | |
| *** chetna has quit IRC | 06:45 | |
| *** belmoreira has joined #openstack-containers | 06:50 | |
| *** pcaruana has joined #openstack-containers | 06:53 | |
| *** chetna has joined #openstack-containers | 06:53 | |
| *** chetna has quit IRC | 06:54 | |
| *** snarwade has quit IRC | 06:59 | |
| *** david-lyle has quit IRC | 07:03 | |
| *** snarwade has joined #openstack-containers | 07:11 | |
| openstackgerrit | yatin proposed openstack/python-magnumclient: Remove invalid check for 'manifest' path https://review.openstack.org/393207 | 07:21 |
|---|---|---|
| openstackgerrit | yatin proposed openstack/python-magnumclient: Increase UT coverage https://review.openstack.org/394798 | 07:21 |
| *** chetna has joined #openstack-containers | 07:54 | |
| *** Jeffrey4l has quit IRC | 07:54 | |
| *** chetna has quit IRC | 07:55 | |
| *** pcastets has joined #openstack-containers | 07:56 | |
| *** takashi has joined #openstack-containers | 07:56 | |
| *** wangbo has quit IRC | 08:04 | |
| *** sahilsinha has quit IRC | 08:07 | |
| openstackgerrit | yatin proposed openstack/python-magnumclient: Increase UT coverage https://review.openstack.org/394798 | 08:09 |
| *** wangbo has joined #openstack-containers | 08:18 | |
| *** jperry has quit IRC | 08:21 | |
| *** GheRivero has joined #openstack-containers | 08:27 | |
| *** tovin07_ has joined #openstack-containers | 08:31 | |
| *** zenoway has joined #openstack-containers | 08:36 | |
| *** sahilsinha has joined #openstack-containers | 08:37 | |
| openstackgerrit | Spyros Trigazis proposed openstack/magnum: Make cinder volume optional https://review.openstack.org/391830 | 08:40 |
| *** chetna has joined #openstack-containers | 08:40 | |
| *** chetna has quit IRC | 08:41 | |
| *** narayrak has quit IRC | 08:43 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/magnum-ui: Imported Translations from Zanata https://review.openstack.org/394822 | 08:43 |
| *** narayrak has joined #openstack-containers | 08:44 | |
| *** shu-mutou is now known as shu-mutou-AWAY | 09:02 | |
| *** khamtamtun has joined #openstack-containers | 09:07 | |
| *** khamtamtun has quit IRC | 09:10 | |
| openstackgerrit | Merged openstack/magnum-ui: Imported Translations from Zanata https://review.openstack.org/394822 | 09:11 |
| GheRivero | morning | 09:21 |
| *** strigazi_AFK is now known as strigazi | 09:25 | |
| strigazi | GheRivero, hi | 09:28 |
| *** narayrak has quit IRC | 09:38 | |
| *** narayrak has joined #openstack-containers | 09:38 | |
| *** pcaruana has quit IRC | 09:51 | |
| *** Jeffrey4l has joined #openstack-containers | 09:55 | |
| *** narayrak has quit IRC | 10:00 | |
| *** narayrak has joined #openstack-containers | 10:00 | |
| *** Jeffrey4l has quit IRC | 10:00 | |
| *** narayrak has quit IRC | 10:01 | |
| *** narayrak has joined #openstack-containers | 10:01 | |
| *** tovin07_ has quit IRC | 10:03 | |
| openstackgerrit | Merged openstack/magnum: Remove unused configure-flannel.sh https://review.openstack.org/394495 | 10:03 |
| *** Jeffrey4l has joined #openstack-containers | 10:08 | |
| *** Jeffrey4l has quit IRC | 10:14 | |
| *** askb has quit IRC | 10:16 | |
| *** duonghq has quit IRC | 10:26 | |
| openstackgerrit | fengbeihong proposed openstack/magnum: Support dcos installation on centos cluster https://review.openstack.org/386435 | 10:29 |
| openstackgerrit | fengbeihong proposed openstack/magnum: [WIP] Add test for supporting dcos in magnum https://review.openstack.org/394270 | 10:30 |
| *** jperry has joined #openstack-containers | 10:32 | |
| *** maestropandy has joined #openstack-containers | 10:38 | |
| *** maestropandy has left #openstack-containers | 10:41 | |
| *** chhavi has joined #openstack-containers | 10:58 | |
| *** chetna has joined #openstack-containers | 11:26 | |
| *** chetna has quit IRC | 11:27 | |
| *** dfflanders has quit IRC | 11:33 | |
| *** khamtamtun has joined #openstack-containers | 12:00 | |
| *** khamtamtun has quit IRC | 12:05 | |
| *** khamtamtun has joined #openstack-containers | 12:05 | |
| *** khamtamtun has quit IRC | 12:07 | |
| *** khamtamtun has joined #openstack-containers | 12:17 | |
| *** david-lyle has joined #openstack-containers | 12:37 | |
| *** GB21 has quit IRC | 12:38 | |
| *** wangbo has quit IRC | 12:42 | |
| *** Drago has joined #openstack-containers | 12:46 | |
| *** zz_dimtruck is now known as dimtruck | 12:56 | |
| *** snarwade has quit IRC | 13:00 | |
| *** chhavi has quit IRC | 13:01 | |
| *** trananhkma has quit IRC | 13:02 | |
| *** khamtamtun has quit IRC | 13:02 | |
| *** takashi has quit IRC | 13:09 | |
| *** jperry has quit IRC | 13:09 | |
| *** chhavi has joined #openstack-containers | 13:10 | |
| *** DanyC has joined #openstack-containers | 13:13 | |
| *** vmud213 has quit IRC | 13:24 | |
| *** david-lyle has quit IRC | 13:29 | |
| *** jwcroppe has quit IRC | 13:35 | |
| *** jwcroppe has joined #openstack-containers | 13:36 | |
| *** jwcroppe has quit IRC | 13:40 | |
| *** khamtamtun has joined #openstack-containers | 13:42 | |
| *** khamtamtun has quit IRC | 13:43 | |
| *** DanyC has left #openstack-containers | 13:43 | |
| *** openstackgerrit has quit IRC | 13:48 | |
| *** openstackgerrit has joined #openstack-containers | 13:49 | |
| *** jwcroppe has joined #openstack-containers | 13:57 | |
| *** fragatina has joined #openstack-containers | 13:57 | |
| *** fragatina has quit IRC | 14:02 | |
| *** _mbelanger has joined #openstack-containers | 14:09 | |
| *** chetna has joined #openstack-containers | 14:13 | |
| *** chetna has quit IRC | 14:14 | |
| *** dave-mccowan has joined #openstack-containers | 14:14 | |
| *** jperry has joined #openstack-containers | 14:23 | |
| *** hieulq_ has joined #openstack-containers | 14:31 | |
| *** hieulq_ has quit IRC | 14:32 | |
| *** hieulq_ has joined #openstack-containers | 14:34 | |
| *** absubram has joined #openstack-containers | 14:39 | |
| *** absubram_ has joined #openstack-containers | 14:41 | |
| *** absubram has quit IRC | 14:44 | |
| *** absubram_ is now known as absubram | 14:44 | |
| *** mjura has joined #openstack-containers | 14:44 | |
| *** mjura has quit IRC | 14:45 | |
| *** mjura has joined #openstack-containers | 14:45 | |
| *** hongbin has joined #openstack-containers | 14:46 | |
| *** Drago has quit IRC | 14:49 | |
| *** FengShengqin has joined #openstack-containers | 14:54 | |
| *** _mbelanger has quit IRC | 14:55 | |
| *** mjura has quit IRC | 14:57 | |
| *** catintheroof has joined #openstack-containers | 14:57 | |
| *** vimal has quit IRC | 15:01 | |
| *** adrian_otto has joined #openstack-containers | 15:01 | |
| openstackgerrit | Spyros Trigazis proposed openstack/magnum: [WIP] Add cluster_attributes table https://review.openstack.org/395012 | 15:03 |
| *** vijendar has joined #openstack-containers | 15:06 | |
| *** hieulq_ has quit IRC | 15:12 | |
| *** hieulq_ has joined #openstack-containers | 15:13 | |
| *** mastermind has joined #openstack-containers | 15:22 | |
| *** absubram has quit IRC | 15:22 | |
| *** zenoway has quit IRC | 15:25 | |
| *** zenoway has joined #openstack-containers | 15:28 | |
| *** david-lyle has joined #openstack-containers | 15:29 | |
| *** vijendar has quit IRC | 15:31 | |
| *** vijendar has joined #openstack-containers | 15:32 | |
| *** adrian_otto has quit IRC | 15:32 | |
| *** mastermind has quit IRC | 15:32 | |
| *** tonanhngo has joined #openstack-containers | 15:33 | |
| *** david-lyle has quit IRC | 15:34 | |
| *** narayrak has quit IRC | 15:35 | |
| *** tonanhngo has quit IRC | 15:37 | |
| *** _mbelanger has joined #openstack-containers | 15:38 | |
| *** Drago has joined #openstack-containers | 15:38 | |
| *** jmckind has joined #openstack-containers | 15:39 | |
| *** vijendar has quit IRC | 15:42 | |
| *** Drago has quit IRC | 15:42 | |
| *** vijendar has joined #openstack-containers | 15:43 | |
| *** chris_hultin|AWA is now known as chris_hultin | 15:43 | |
| *** chris_hultin is now known as chris_hultin|AWA | 15:44 | |
| *** rpothier has joined #openstack-containers | 15:45 | |
| *** chris_hultin|AWA is now known as chris_hultin | 15:46 | |
| *** adrian_otto has joined #openstack-containers | 15:46 | |
| adrian_otto | Our team meeting will begin in just under 15 minutes in #openstack-meeting-alt (note time shift for those who use DST in their local timezones) | 15:47 |
| adrian_otto | We will begin at 1600 UTC | 15:47 |
| *** adrian_otto has quit IRC | 15:47 | |
| *** Drago has joined #openstack-containers | 15:48 | |
| *** syed_ has joined #openstack-containers | 15:49 | |
| strigazi | adrian_otto https://review.openstack.org/#/c/391770/ | 15:49 |
| strigazi | adrian_otto requires a PTL +1 | 15:50 |
| *** muralia has quit IRC | 15:52 | |
| *** randallburt has joined #openstack-containers | 15:55 | |
| *** adrian_otto has joined #openstack-containers | 15:57 | |
| *** belmoreira has quit IRC | 15:58 | |
| *** chetna has joined #openstack-containers | 15:58 | |
| *** chetna has quit IRC | 15:59 | |
| *** randallburt has quit IRC | 16:00 | |
| *** randallburt has joined #openstack-containers | 16:07 | |
| *** anush has quit IRC | 16:13 | |
| *** zenoway has quit IRC | 16:17 | |
| *** ramishra has quit IRC | 16:17 | |
| *** zenoway has joined #openstack-containers | 16:18 | |
| *** ramishra has joined #openstack-containers | 16:19 | |
| *** rcernin has quit IRC | 16:20 | |
| *** zenoway has quit IRC | 16:22 | |
| *** chhavi has quit IRC | 16:22 | |
| *** hieulq_ has quit IRC | 16:37 | |
| *** jasond has joined #openstack-containers | 16:43 | |
| *** muralia has joined #openstack-containers | 16:47 | |
| *** FengShengqin has quit IRC | 16:49 | |
| *** tonanhngo has joined #openstack-containers | 16:52 | |
| *** tonanhngo has quit IRC | 16:53 | |
| Drago | hongbin: Let's continue the discussion here :) | 16:53 |
| hongbin | Drago: i will go to lunch in about 1 hour | 16:54 |
| Drago | hongbin: I will too | 16:54 |
| hongbin | Drago: ok, we can discuss it after an hour i guess | 16:54 |
| *** tonanhngo has joined #openstack-containers | 16:54 | |
| Drago | hongbin: We can discuss until one of us has to leave, and then continue after lunch | 16:55 |
| hongbin | Drago: oh, i mean i will go to lunch now, and back in an hour | 16:55 |
| hongbin | Drago: sorry, i said that wrong in before | 16:55 |
| *** rcernin has joined #openstack-containers | 16:56 | |
| Drago | hongbin: Oh okay. I will be at lunch in an hour, so I will be back in 2+ hours | 16:56 |
| hongbin | Drago: ok | 16:56 |
| *** GheRivero has left #openstack-containers | 16:59 | |
| strigazi | adrian_otto, https://review.openstack.org/#/c/391770/ requires a PTL +1 | 17:02 |
| adrian_otto | ok strigazi I will have a look. Thanks for the heads-up. | 17:02 |
| *** Drago1 has joined #openstack-containers | 17:02 | |
| *** Drago has quit IRC | 17:06 | |
| *** Drago1 has quit IRC | 17:06 | |
| jvgrant__ | strigazi: ping | 17:07 |
| *** muralia1 has joined #openstack-containers | 17:08 | |
| strigazi | jvgrant__ 64 bytes from strigazi : icmp_seq=1 ttl=124 time=0.441 ms | 17:08 |
| *** muralia has quit IRC | 17:09 | |
| jvgrant__ | :) | 17:09 |
| jvgrant__ | strigazi: you worked on the install guide right? | 17:09 |
| strigazi | yes | 17:09 |
| jvgrant__ | strigazi: i was helping someone yesterday that came to the channel and he ran into 2 issues. wonderign if you saw these in your testing or had work arounds | 17:10 |
| jvgrant__ | 1) driver template yaml files missing when using apt-get install | 17:10 |
| jvgrant__ | 2) using "controller" in the address for keystone registration cause the nodes to be unable to find the endpoints since they couldn't resolve "controller" | 17:11 |
| strigazi | 2) is easy, the controller is a convention of the install guide across OS projects | 17:11 |
| strigazi | means the hostname or ip of the controller node | 17:12 |
| jvgrant__ | right, but they resolve it on each node using a modification in the etc files, but the new master/minion nodes don't have that when created | 17:12 |
| strigazi | the IG assumes a controller and a compute node | 17:12 |
| *** Drago has joined #openstack-containers | 17:13 | |
| jvgrant__ | so cluster-creates would time out | 17:13 |
| jvgrant__ | when they changed the keystone entries to actual IP in keystone instead of controller everything worked | 17:13 |
| strigazi | yes, controller must be replaced with the ip or hostname of the controller node, that is the convention | 17:14 |
| strigazi | eg | 17:14 |
| strigazi | http://docs.openstack.org/newton/install-guide-rdo/glance-install.html#install-and-configure-components | 17:14 |
| Drago | strigazi: I missed the first part of this conversation. Are you saying that a real cloud should never have "controller" in its endpoints? | 17:15 |
| strigazi | controller is like an alias | 17:15 |
| jvgrant__ | right, which is what all the install guide instructions tell you to do | 17:16 |
| jvgrant__ | but it seems like Magnum doesn't work if we do that, as the nodes we create need those alias's as well but they don't have them | 17:16 |
| strigazi | you must not put controller anywhere | 17:17 |
| Drago | strigazi: Is that a yes or no? | 17:17 |
| strigazi | yes never | 17:18 |
| jvgrant__ | so they should replace the controller with the actual IP addresss | 17:18 |
| Drago | strigazi: Then the docs should be updated so that's obvious. The thing that further confuses it is that the docs also say to put a "controller" entry into /etc/hosts that will resolve "controller" to a real IP | 17:18 |
| strigazi | or hostname | 17:18 |
| Drago | So inside your cloud, it works | 17:18 |
| strigazi | Drago, in all the IGs cotroller is mentioned | 17:19 |
| strigazi | I'm trying to find where it is explained | 17:19 |
| jvgrant__ | i see the mentions of it being used and the instructions on /etc/hosts. | 17:20 |
| jvgrant__ | in our guide is shows controller not being replaced with ip as the return of the example keystone entry shows "controller" not the IP | 17:20 |
| Drago | Gotta run, I'll be back in about an hour and a half | 17:20 |
| strigazi | in all guides it's the same | 17:21 |
| *** vijendar_ has joined #openstack-containers | 17:22 | |
| *** vijendar_ has quit IRC | 17:22 | |
| *** Drago has quit IRC | 17:22 | |
| *** vijendar_ has joined #openstack-containers | 17:22 | |
| jvgrant__ | yeah, ours matches them. And everything works for that entry until... you create a cluster and those nodes try to send to http://controller:9511 which they can't resolve | 17:23 |
| *** vijendar has quit IRC | 17:23 | |
| jvgrant__ | those vms | 17:23 |
| strigazi | in all guides the endpoints are registered using controller. In a real deployment controller is like a variable where you have set the appriate ip or hostname | 17:25 |
| strigazi | I'll add a note to avoid confusion | 17:26 |
| jvgrant__ | except that since they have instructions to modify the /etc/hosts file in the install guide the using "controller" seems like it works | 17:26 |
| jvgrant__ | so i note will probably hep | 17:26 |
| jvgrant__ | help | 17:26 |
| strigazi | The thing is that the guide are officially named "Installation Tutorials", it's expected to modify the commands according to your evironment | 17:28 |
| strigazi | About the 1, was it in ubuntu? | 17:28 |
| strigazi | or debian? | 17:29 |
| jvgrant__ | i understand and it makes sense, but since it caused a user confusing we should probably try to avoid it | 17:29 |
| *** rcernin has quit IRC | 17:29 | |
| jvgrant__ | ubuntu | 17:29 |
| jvgrant__ | https://bugs.launchpad.net/magnum/+bug/1638393 | 17:29 |
| openstack | Launchpad bug 1638393 in Magnum "Launch an instance in Installation Guide for Container Infrastructure Management Service" [Undecided,New] | 17:29 |
| jvgrant__ | that is the bug for that one | 17:29 |
| jvgrant__ | the cluster.yaml file was missing | 17:30 |
| strigazi | interesting, I wrote the guide before the final release of packages so I had built the packages myself, this is a bug in the packages. | 17:31 |
| strigazi | I'll check with the packager | 17:31 |
| jvgrant__ | yeah, i wasn't sure how to check to see if the file was there in the package | 17:32 |
| *** chetna has joined #openstack-containers | 17:33 | |
| *** vijenda__ has joined #openstack-containers | 17:36 | |
| *** vijenda__ has quit IRC | 17:36 | |
| *** vijenda__ has joined #openstack-containers | 17:36 | |
| *** vijendar_ has quit IRC | 17:37 | |
| *** muralia1 has quit IRC | 17:38 | |
| openstackgerrit | Spyros Trigazis proposed openstack/magnum: Convert floating_ip_enabled from boolean to enum https://review.openstack.org/395095 | 17:40 |
| *** Drago has joined #openstack-containers | 17:40 | |
| *** chhavi has joined #openstack-containers | 17:42 | |
| openstackgerrit | Spyros Trigazis proposed openstack/magnum: Convert floating_ip_enabled from boolean to enum https://review.openstack.org/395095 | 17:42 |
| *** jmckind has quit IRC | 17:43 | |
| *** vijenda__ has quit IRC | 17:45 | |
| *** tonanhngo_ has joined #openstack-containers | 17:46 | |
| *** tonanhngo_ has quit IRC | 17:46 | |
| *** tonanhngo_ has joined #openstack-containers | 17:47 | |
| *** tonanhngo has quit IRC | 17:48 | |
| *** vijendar_ has joined #openstack-containers | 17:49 | |
| *** chhavi has quit IRC | 17:51 | |
| *** jmckind has joined #openstack-containers | 17:56 | |
| *** vijendar_ has quit IRC | 17:58 | |
| *** vijendar has joined #openstack-containers | 17:59 | |
| *** jperry has quit IRC | 18:00 | |
| *** pcastets has quit IRC | 18:01 | |
| *** vijendar has quit IRC | 18:06 | |
| *** vijendar has joined #openstack-containers | 18:06 | |
| *** fragatina has joined #openstack-containers | 18:08 | |
| *** vijendar has quit IRC | 18:09 | |
| *** jperry has joined #openstack-containers | 18:10 | |
| *** vijendar_ has joined #openstack-containers | 18:10 | |
| *** jvgrant__ has quit IRC | 18:13 | |
| *** jvgrant__ has joined #openstack-containers | 18:13 | |
| *** adrian_otto has quit IRC | 18:15 | |
| *** vijendar_ has quit IRC | 18:19 | |
| *** vijendar_ has joined #openstack-containers | 18:22 | |
| *** vijendar_ has quit IRC | 18:37 | |
| *** vijendar_ has joined #openstack-containers | 18:41 | |
| *** vijendar_ has quit IRC | 18:42 | |
| *** vijendar has joined #openstack-containers | 18:43 | |
| *** tkarke has joined #openstack-containers | 18:43 | |
| *** fragatina has quit IRC | 18:46 | |
| *** fragatina has joined #openstack-containers | 18:46 | |
| *** david-lyle_ has joined #openstack-containers | 18:47 | |
| *** vijendar has quit IRC | 18:48 | |
| *** vijendar has joined #openstack-containers | 18:48 | |
| Drago | Hi hongbin | 18:52 |
| *** vijendar has quit IRC | 18:52 | |
| hongbin | Drago: hi | 18:52 |
| hongbin | Want to start? | 18:52 |
| *** vijendar has joined #openstack-containers | 18:53 | |
| *** adisky has quit IRC | 18:53 | |
| *** david-lyle_ has quit IRC | 18:53 | |
| *** adrian_otto has joined #openstack-containers | 18:55 | |
| Drago | hongbin: Okay, now I'm ready | 18:55 |
| hongbin | Drago: you want to start first? | 18:55 |
| *** adrian_otto has quit IRC | 18:55 | |
| Drago | hongbin: I had a question for you, about the current iteration of the spec | 18:56 |
| hongbin | go ahead | 18:56 |
| Drago | hongbin: It is about exposing the state synchronizer class for drivers to implement. From the way it is written, it seems like drivers would not be forced to write a poller. Instead, they would be able to provide a class that could use a push mechanism | 18:58 |
| hongbin | my understanding is taht it is a *force* | 18:59 |
| Drago | hongbin: My question is, is there something in the current spec would prevent drivers from doing this? | 19:00 |
| hongbin | because magnum is calling the interface to sync state, then every drivers need to implement that interface. in that sense, it is a force | 19:00 |
| *** fragatin_ has joined #openstack-containers | 19:00 | |
| *** rpothier has quit IRC | 19:00 | |
| hongbin | Drago: however, my major concern is why the state synchronizer is there in the first iteration | 19:01 |
| hongbin | Drago: i am trying to get the rational of why this class is in the interface | 19:02 |
| Drago | hongbin: Can we talk about that in a minute? I would like to make sure I understand HeatPoller correctly | 19:02 |
| hongbin | Drago: sure | 19:02 |
| *** fragatina has quit IRC | 19:03 | |
| *** adrian_otto has joined #openstack-containers | 19:04 | |
| Drago | hongbin: (Imagine that we are not discussing the spec at all). If we wanted to implement a pull-based cluster state synchronization, could our changes be limited to HeatPoller? | 19:04 |
| hongbin | not sure exactly, need to check the code | 19:04 |
| hongbin | maybe not | 19:05 |
| hongbin | Drago: your point is? | 19:05 |
| Drago | hongbin: I think you are right, there is at least a bit more code that would need to change, but most of it is in HeatPoller | 19:06 |
| Drago | So assume we could get it all into HeatPoller, and magnum could simply ask HeatPoller to sync the cluster state | 19:07 |
| Drago | If we exposed HeatPoller to drivers, they could actually implement a non-poller | 19:07 |
| hongbin | you want to combine a poller and a pusher to the same interface? | 19:08 |
| Drago | hongbin: If we could, it would decouple polling from Magnum | 19:09 |
| hongbin | Drago: ok, then the state synchronizer interface need to evolve | 19:09 |
| hongbin | Drago: right now, it looks like a poller | 19:09 |
| Drago | hongbin: I think a little bit of code outside of HeatPoller needs to be worked with so that the *concept* of StateSynchronizer would work | 19:10 |
| Drago | hongbin: But do you see where the idea is going? | 19:10 |
| hongbin | ?? | 19:10 |
| Drago | hongbin: What are you confused about? | 19:11 |
| hongbin | "do you see where the idea is going?" | 19:11 |
| hongbin | what idea? | 19:12 |
| Drago | hongbin: I see. I mean the idea of StateSynchronizer being a generic interface that can be a poller or a pusher, depending on the implementation of the driver | 19:12 |
| hongbin | ok | 19:12 |
| hongbin | get that | 19:12 |
| hongbin | yes, if that can be done, it would be great | 19:12 |
| openstackgerrit | Vijendar Komalla proposed openstack/magnum: Specification for Magnum Admin API https://review.openstack.org/391537 | 19:12 |
| hongbin | maybe call it statesynchronizer is confusing | 19:13 |
| hongbin | need a better name | 19:13 |
| hongbin | and it should implements a generic logic to wait for a cluster to become active | 19:13 |
| Drago | hongbin: If it is a generic interface, it is at least better than calling it "HeatPoller" :) | 19:13 |
| Drago | hongbin: Can you help us figure out the minimal changes needed to achieve this? | 19:14 |
| hongbin | Drago: i think we need a better name, even more, i doubt if it worthes to create a dedicated class for it, why not just a method (calling it "wait_for_active" or something) | 19:14 |
| Drago | hongbin: I think the idea is that the logic for "wait for active" is currently in the HeatPoller class, basically | 19:15 |
| hongbin | yes, the logic is here, but moving it to interface is a different thing | 19:15 |
| Drago | hongbin: I am not sure, perhaps going one level up has the method we want? We could instead call the driver here? https://github.com/openstack/magnum/blob/7fc0935a3ee2a7d6d8fb4e2c404b16214a110c1e/magnum/conductor/handlers/cluster_conductor.py#L171 | 19:16 |
| randallburt | ok, so talked with Adrian and this afternoon, I'm going to add a stubbed out example of how to implement the proposed interface without Heat using pseudocode | 19:16 |
| Drago | hongbin: _poll_and_check is what uses HeatPoller | 19:17 |
| randallburt | and I removed _poll_and_check | 19:17 |
| *** vijendar_ has joined #openstack-containers | 19:17 | |
| hongbin | Drago: yes, how about promote _poll_and_check to the interface instead (renaming it to "wait_for_active" to indicate it can be pushed) | 19:18 |
| randallburt | hongbin: I'm not moving the poll_and_check to the interface. | 19:18 |
| hongbin | Drago: then, hide the entire HeatPoller class | 19:18 |
| randallburt | hongbin: not the logic of it anyway | 19:19 |
| hongbin | How it works is that | 19:19 |
| randallburt | hongbin: the logic of it will move to the Heat-specific implementation of the synchronizer | 19:19 |
| hongbin | In cluster_create, it does the logic to create a cluster, then call "wait_for_active" | 19:19 |
| Drago | hongbin: What would the driver use if HeatPoller is hidden? I'm not sure I understand | 19:19 |
| *** vijendar has quit IRC | 19:19 | |
| hongbin | Drago: i mean hiding it from the interface | 19:19 |
| randallburt | hongbin: and wait_for_active will call the synchronizer then check the status of the cluster | 19:20 |
| hongbin | Drago: that means if you are using Heat driver, you call the class | 19:20 |
| Drago | hongbin: But a driver could use it in its logic if it wanted? Okay I see. | 19:20 |
| hongbin | Drago: for other drivers, implement the "wait_for_active" method | 19:20 |
| *** vijendar_ has quit IRC | 19:20 | |
| hongbin | Yes, each driver are free to implement the method | 19:21 |
| randallburt | hongbin: so you want the exact same thing I proposed just as a single method on the driver rather than returning a class? How is that "better"? | 19:21 |
| hongbin | use a state synchronier or others (the point is to hide it from the interface) | 19:21 |
| randallburt | hongbin: there's absolutely zero increase in information hiding. | 19:21 |
| hongbin | it make alternative implementation easier and looks simplier and easier to understand | 19:22 |
| randallburt | hongbin: how does it make it "easier" | 19:22 |
| randallburt | hongbin: unless you're saying implementing a method is substantially "easier" than implementing a class with a single method? | 19:22 |
| Drago | It seems like using a method would make the interface smaller (fewer touchpoints) | 19:22 |
| hongbin | this is the middle ground i can think of | 19:22 |
| hongbin | it is up to you to accept it or not | 19:22 |
| hongbin | another alternative is to implement a method to sync state | 19:23 |
| hongbin | like "refresh_cluster" | 19:23 |
| randallburt | hongbin: sure, its fine. its just more disruptive (more code moving around/changing than has to). But if you agree its "functionally" equivalent, I'm at a loss as to why there is so much fuss and a -2 | 19:23 |
| randallburt | hongbin: that's *exactly* what the synchronizer DOES> | 19:24 |
| hongbin | please understand my -2 is procedural | 19:24 |
| hongbin | it blocks the patch until it gets to a point to consider merging | 19:25 |
| *** vijendar_ has joined #openstack-containers | 19:25 | |
| randallburt | hongbin: its disruptive and obstructionist in this case. but w/e. So if I remove the synchronizer in favor of a method on the driver class that server the *exact same purpose* but increases the implementation burden and effects tons more tests, you'll remove your -2? | 19:25 |
| randallburt | s/server/serves | 19:26 |
| hongbin | i remove my -2 until the spec has been thoughfully discussed | 19:26 |
| randallburt | hongbin: its been thoughtfully discussed for days. | 19:26 |
| hongbin | maybe up to the next meeting (if the whole team agree on something) | 19:26 |
| randallburt | hongbin: you're kidding | 19:26 |
| hongbin | i am serious | 19:27 |
| randallburt | fine. can you at least tell me if you agree to my previous statement? | 19:27 |
| randallburt | hongbin: ^^ | 19:27 |
| hongbin | depending on how to revise the spec | 19:28 |
| randallburt | hongbin: meaning? | 19:28 |
| hongbin | if you upload a new patch, i will do another review and tell you if i personally agree on | 19:28 |
| randallburt | hongbin: fine. can't wait. | 19:29 |
| Drago | randallburt: Look at this code. I think _poll_and_check is the right place to define the interface. Otherwise, StateSynchronizer would be stuffed into a loopingcall. That work is going to have to be done anyway to enable a non-polling solution https://github.com/openstack/magnum/blob/master/magnum/conductor/handlers/cluster_conductor.py#L175-L178 | 19:29 |
| randallburt | Drago: no, it isn't | 19:29 |
| randallburt | Drago: but I'm done arguing. | 19:29 |
| *** vijendar_ has quit IRC | 19:33 | |
| *** vijendar_ has joined #openstack-containers | 19:36 | |
| Drago | randallburt: I want to note that this is _poll_and_check in Handler, not poll_and_check in HeatPoller | 19:36 |
| Drago | hongbin: cluster_conductor.Handler._poll_and_check is blocking, right? | 19:38 |
| hongbin | Drago: AFAIK, no | 19:39 |
| hongbin | Drago: if you want it to be block, add lc.wait(..) after | 19:39 |
| Drago | hongbin: I see | 19:39 |
| *** vijenda__ has joined #openstack-containers | 19:49 | |
| *** vijendar_ has quit IRC | 19:53 | |
| *** fragatin_ has quit IRC | 19:58 | |
| *** fragatina has joined #openstack-containers | 19:58 | |
| *** fragatina has quit IRC | 20:02 | |
| *** fragatina has joined #openstack-containers | 20:14 | |
| *** vijenda__ has quit IRC | 20:25 | |
| *** vijendar has joined #openstack-containers | 20:26 | |
| *** fragatina has quit IRC | 20:26 | |
| *** fragatina has joined #openstack-containers | 20:26 | |
| tkarke | Hi, I have a question about the Creation of cluster without floating ip. As the bug exists in Newton Stable too, will the fix be backported? and if yes, will it be type enum or boolean for stable? | 20:32 |
| jvgrant__ | tkarke, i haven't seen the fix submitted to Newton Stable as well so for now, I"m guess no. | 20:40 |
| jvgrant__ | tkarke, the fix just went up for review though. i'm not sure what the intentions are | 20:40 |
| hongbin | Drago: for the design summit session about Creds/Certs Management (https://etherpad.openstack.org/p/ocata-magnum-tls), do we reach any conclusion on the approach? (sorry i was not there) | 20:54 |
| Drago | hongbin: In my mind, we really didn't besides committing to the nuclear option. The question of whether we should submit code upstream (to the COEs themselves), but I don't remember there being any consensus | 20:56 |
| hongbin | Drago: ack | 20:56 |
| hongbin | btw, option #3 is an interesting approach | 20:56 |
| hongbin | it might work for k8s, but hard for other COEs IMO | 20:57 |
| Drago | hongbin: Thanks. I came up with it but didn't think it would be seriously considered | 20:57 |
| hongbin | ok | 20:58 |
| Drago | hongbin: That approach will probably require submitting code to COEs as well, so option 2 is easier, I think | 20:59 |
| *** rcernin has joined #openstack-containers | 20:59 | |
| hongbin | Drago: not sure how option #2 is going to work, but happy to review a spec if there is any | 21:01 |
| Drago | hongbin: The COEs would natively support a revocation list. Magnum would have nothing to do with it | 21:02 |
| hongbin | ok | 21:03 |
| Drago | hongbin: Well, besides magnum asking the COE to revoke a user | 21:04 |
| hongbin | get that | 21:04 |
| *** EricGonczer_ has joined #openstack-containers | 21:05 | |
| *** askb has joined #openstack-containers | 21:11 | |
| *** vijendar has quit IRC | 21:17 | |
| *** vijendar has joined #openstack-containers | 21:18 | |
| *** vijendar has quit IRC | 21:25 | |
| *** vijendar has joined #openstack-containers | 21:27 | |
| *** _mbelanger has quit IRC | 21:40 | |
| openstackgerrit | Vijendar Komalla proposed openstack/magnum: Restart swarm infra containers if deleted https://review.openstack.org/395209 | 21:55 |
| *** vijendar has quit IRC | 22:00 | |
| *** chris_hultin is now known as chris_hultin|AWA | 22:08 | |
| *** tonanhngo_ has quit IRC | 22:09 | |
| *** tonanhngo has joined #openstack-containers | 22:14 | |
| *** tonanhngo has quit IRC | 22:18 | |
| *** catintheroof has quit IRC | 22:22 | |
| *** catintheroof has joined #openstack-containers | 22:22 | |
| *** adrian_otto has quit IRC | 22:24 | |
| *** catintheroof has quit IRC | 22:27 | |
| *** adrian_otto has joined #openstack-containers | 22:30 | |
| *** tonanhngo has joined #openstack-containers | 22:42 | |
| *** tonanhngo has quit IRC | 22:42 | |
| *** adrian_otto has quit IRC | 22:46 | |
| *** tonanhngo has joined #openstack-containers | 22:46 | |
| *** adrian_otto has joined #openstack-containers | 22:47 | |
| *** adrian_otto has quit IRC | 22:49 | |
| *** david-lyle_ has joined #openstack-containers | 22:50 | |
| randallburt | gotta run take care of some family stuff. will update spec first thing tomorrow | 22:54 |
| *** david-lyle_ has quit IRC | 22:54 | |
| *** randallburt has quit IRC | 22:56 | |
| *** jperry has quit IRC | 22:57 | |
| *** chris_hultin|AWA is now known as chris_hultin | 23:01 | |
| *** jasond has quit IRC | 23:15 | |
| *** jmckind has quit IRC | 23:17 | |
| *** rcernin has quit IRC | 23:18 | |
| *** absubram has joined #openstack-containers | 23:21 | |
| *** chris_hultin is now known as chris_hultin|AWA | 23:25 | |
| *** jwcroppe has quit IRC | 23:27 | |
| *** jwcroppe has joined #openstack-containers | 23:27 | |
| *** catintheroof has joined #openstack-containers | 23:29 | |
| *** catintheroof has quit IRC | 23:31 | |
| *** jwcroppe has quit IRC | 23:31 | |
| *** catintheroof has joined #openstack-containers | 23:31 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!