| *** achanda has quit IRC | 00:03 | |
| *** achanda_ has quit IRC | 00:04 | |
| *** banix has quit IRC | 00:10 | |
| klindgren | hello - trying to get magnum started up for the first time... I mist be totally daft on this - but it seems like logging is totally fubar? | 00:15 |
|---|---|---|
| klindgren | use_stderr = false | 00:15 |
| klindgren | state_path = /var/lib/magnum | 00:15 |
| klindgren | logfile=magnum.log | 00:15 |
| klindgren | logdir=/var/log/magnum | 00:15 |
| klindgren | verbose = true | 00:15 |
| klindgren | debug = true | 00:15 |
| klindgren | results in nothing in /var/log/magnum being created | 00:15 |
| klindgren | tried log_dir, tried without log_file tryied passing --log-file on the CLI | 00:16 |
| klindgren | I dont see any initialization message or anything | 00:16 |
| *** kebray has quit IRC | 00:16 | |
| *** yuanying has quit IRC | 00:19 | |
| *** yuanying_ is now known as yuanying | 00:19 | |
| klindgren | apparently has to be specifically started with --log_dir on the CLI | 00:21 |
| *** yuanying_ has joined #openstack-containers | 00:24 | |
| openstackgerrit | Surojit Pathak proposed openstack/python-magnumclient: Adding 'magnum service-list' https://review.openstack.org/226106 | 00:25 |
| *** suro-patz has quit IRC | 00:26 | |
| klindgren | actually - spoke too soon that doesn't even seem to work | 00:26 |
| *** yuanying has quit IRC | 00:29 | |
| *** yuanying_ is now known as yuanying | 00:29 | |
| *** yuanying_ has joined #openstack-containers | 00:30 | |
| *** Tango has quit IRC | 00:40 | |
| *** dims_ has quit IRC | 00:42 | |
| *** vilobhmm_111 has quit IRC | 00:47 | |
| *** vilobhmm_111 has joined #openstack-containers | 00:48 | |
| *** SourabhP has joined #openstack-containers | 00:49 | |
| *** hongbin has joined #openstack-containers | 00:51 | |
| *** harshs has quit IRC | 00:53 | |
| *** harshs has joined #openstack-containers | 00:53 | |
| *** suro-patz has joined #openstack-containers | 00:57 | |
| *** sdake_ has joined #openstack-containers | 01:04 | |
| *** sdake has quit IRC | 01:06 | |
| *** sdake_ has quit IRC | 01:07 | |
| *** sdake has joined #openstack-containers | 01:07 | |
| *** ganeshna has joined #openstack-containers | 01:07 | |
| *** ganeshna has quit IRC | 01:07 | |
| *** SourabhP has quit IRC | 01:07 | |
| *** ganeshna has joined #openstack-containers | 01:08 | |
| *** vilobhmm_111 has quit IRC | 01:10 | |
| *** ganeshna has quit IRC | 01:11 | |
| hongbin | yuanying: yuanying_ ping | 01:12 |
| *** vilobhmm_11 has quit IRC | 01:12 | |
| yuanying_ | hongbin: hi | 01:12 |
| *** ganeshna has joined #openstack-containers | 01:12 | |
| *** suro-patz1 has joined #openstack-containers | 01:12 | |
| hongbin | yuanying_: I want to talk about the review | 01:12 |
| hongbin | https://review.openstack.org/#/c/223025/1/magnum/conductor/template_definition.py | 01:12 |
| hongbin | what I don't like is the extra processing in template definition doesn't read very well | 01:13 |
| *** ganeshna has quit IRC | 01:13 | |
| hongbin | If you can make it read better, I am fine with the approach | 01:13 |
| *** suro-patz has quit IRC | 01:13 | |
| hongbin | yuanying_: ^^ | 01:14 |
| *** ganeshna has joined #openstack-containers | 01:14 | |
| yuanying_ | hongbin: you mean, code is difficult to read? | 01:14 |
| hongbin | maybe not, but the style is not consistent | 01:14 |
| hongbin | with the rest of the code | 01:15 |
| yuanying_ | ok | 01:15 |
| hongbin | yes, if the style can be more consistent, then it is fine | 01:16 |
| yuanying_ | I’ll try to update OutputMapping | 01:16 |
| hongbin | k. thx | 01:16 |
| yuanying_ | to introduce extra config | 01:16 |
| *** gangil has quit IRC | 01:18 | |
| *** ganeshna has quit IRC | 01:19 | |
| *** dims has joined #openstack-containers | 01:26 | |
| *** sthillma has quit IRC | 01:31 | |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Add functional test cases(containers actions) for swarm https://review.openstack.org/226125 | 01:37 |
| openstackgerrit | Surojit Pathak proposed openstack/magnum: Backend support for magnum service healthcheck https://review.openstack.org/220702 | 01:52 |
| *** erkules_ has joined #openstack-containers | 01:58 | |
| *** eghobo has quit IRC | 01:59 | |
| *** erkules has quit IRC | 02:01 | |
| openstackgerrit | Surojit Pathak proposed openstack/magnum: Documentation update for 'magnum service-list' https://review.openstack.org/226133 | 02:13 |
| *** muralia has quit IRC | 02:15 | |
| openstackgerrit | Motohiro/Yuanying Otsuka proposed openstack/magnum: [WIP] Add TLS support in heat kubernetes template https://review.openstack.org/202873 | 02:18 |
| *** yuanying_ has quit IRC | 02:19 | |
| *** suro-patz1 has quit IRC | 02:22 | |
| *** zz_dimtruck is now known as dimtruck | 02:27 | |
| *** achanda has joined #openstack-containers | 02:33 | |
| *** dims has quit IRC | 02:33 | |
| *** hongbin has quit IRC | 02:34 | |
| *** unicell has quit IRC | 02:36 | |
| *** sdake has quit IRC | 02:39 | |
| *** david-ly_ has joined #openstack-containers | 02:39 | |
| *** unicell has joined #openstack-containers | 02:40 | |
| *** david-lyle has quit IRC | 02:40 | |
| *** harshs has quit IRC | 02:40 | |
| *** unicell has quit IRC | 02:44 | |
| *** _hanhart has quit IRC | 02:47 | |
| *** tobe has joined #openstack-containers | 02:48 | |
| *** tobe has quit IRC | 02:48 | |
| *** yuanying_ has joined #openstack-containers | 02:49 | |
| *** muralia has joined #openstack-containers | 02:49 | |
| *** yuanying_ has quit IRC | 02:55 | |
| *** yuanying_ has joined #openstack-containers | 02:56 | |
| *** yuanying_ has quit IRC | 03:00 | |
| *** deepthi has joined #openstack-containers | 03:04 | |
| *** ganeshna has joined #openstack-containers | 03:06 | |
| *** kebray has joined #openstack-containers | 03:10 | |
| *** kebray has quit IRC | 03:10 | |
| *** kebray has joined #openstack-containers | 03:11 | |
| *** david-ly_ is now known as david-lyle | 03:17 | |
| *** ganeshna has quit IRC | 03:21 | |
| *** ganeshna has joined #openstack-containers | 03:23 | |
| *** Marga__ has joined #openstack-containers | 03:31 | |
| *** madhuri has joined #openstack-containers | 03:34 | |
| *** dims has joined #openstack-containers | 03:34 | |
| *** Marga_ has quit IRC | 03:34 | |
| *** Marga__ has quit IRC | 03:35 | |
| madhuri | yuanying: Hi | 03:37 |
| *** ganeshna_ has joined #openstack-containers | 03:43 | |
| *** ganeshna has quit IRC | 03:43 | |
| *** Tango has joined #openstack-containers | 03:46 | |
| *** eghobo has joined #openstack-containers | 03:49 | |
| *** ganeshna_ has quit IRC | 03:49 | |
| *** suro-patz has joined #openstack-containers | 03:49 | |
| madhuri | Tango: Hi | 04:08 |
| Tango | Hi Madhuri | 04:08 |
| *** harshs has joined #openstack-containers | 04:08 | |
| madhuri | I just wanted to talk about the k8s v1 client | 04:08 |
| madhuri | I saw the new code | 04:09 |
| *** dims has quit IRC | 04:09 | |
| madhuri | But other implementation is not updated | 04:09 |
| *** dimtruck is now known as zz_dimtruck | 04:09 | |
| madhuri | Do you know about it? | 04:09 |
| Tango | Other implementation? | 04:09 |
| madhuri | Are you or hongbin working on it? | 04:09 |
| madhuri | Yes the new code is not used in k8s_conductor | 04:10 |
| madhuri | Still old code is used | 04:10 |
| Tango | We have a patch that's has been under review for awhile that has this code: 207299 | 04:10 |
| Tango | It's a big patch because we have to switch everything at once | 04:11 |
| madhuri | Ok thanks. I will check it | 04:12 |
| madhuri | Tango: I will review it :) | 04:13 |
| Tango | That would be very helpful, thanks madhuri | 04:14 |
| *** vilobhmm_11 has joined #openstack-containers | 04:28 | |
| *** deepthi has quit IRC | 04:29 | |
| *** vilobhmm_111 has joined #openstack-containers | 04:30 | |
| *** deepthi has joined #openstack-containers | 04:30 | |
| *** vilobhmm_111 has quit IRC | 04:32 | |
| *** vilobhmm11 has joined #openstack-containers | 04:32 | |
| *** vilobhmm_11 has quit IRC | 04:33 | |
| openstackgerrit | Merged openstack/magnum: Change ignore-errors to ignore_errors https://review.openstack.org/225747 | 04:35 |
| *** harshs has quit IRC | 04:36 | |
| openstackgerrit | Merged openstack/magnum: Validates baymodel network_driver requests https://review.openstack.org/222337 | 04:36 |
| *** wenchma has joined #openstack-containers | 04:39 | |
| *** Marga_ has joined #openstack-containers | 04:47 | |
| *** Marga_ has quit IRC | 04:49 | |
| *** Marga_ has joined #openstack-containers | 04:49 | |
| *** suro-patz has quit IRC | 04:55 | |
| *** diga has joined #openstack-containers | 05:02 | |
| openstackgerrit | Merged openstack/magnum: Migrate to Kubernetes Release 1 https://review.openstack.org/207299 | 05:02 |
| *** Marga_ has quit IRC | 05:05 | |
| Tango | madhuri: Thanks Madhuri! | 05:17 |
| madhuri | Most welcome Tango :) | 05:17 |
| Tango | It's a big patch so it probably scared off everyone :) | 05:17 |
| madhuri | Yes I guess so | 05:17 |
| madhuri | But it had to go all together | 05:18 |
| *** vimal has joined #openstack-containers | 05:20 | |
| *** ganeshna has joined #openstack-containers | 05:21 | |
| *** eghobo has quit IRC | 05:22 | |
| *** ganeshna has quit IRC | 05:26 | |
| *** chandankumar has joined #openstack-containers | 05:34 | |
| *** kebray has quit IRC | 05:37 | |
| *** tbh has joined #openstack-containers | 05:52 | |
| *** yuanying_ has joined #openstack-containers | 05:56 | |
| madhuri | yuanying_: Hi | 06:04 |
| yuanying_ | madhuri: hi | 06:04 |
| madhuri | I was looking for you so long :) | 06:04 |
| yuanying_ | sorry, I had a cold last week | 06:05 |
| yuanying_ | I remaind a lot of tasks.. | 06:05 |
| *** preilly has quit IRC | 06:07 | |
| *** sthillma has joined #openstack-containers | 06:07 | |
| *** sthillma_ has joined #openstack-containers | 06:08 | |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Functional: Split python client functional testing case https://review.openstack.org/223907 | 06:09 |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Add functional test cases(containers actions) for swarm https://review.openstack.org/226125 | 06:09 |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Add functional test cases for swarm https://review.openstack.org/221633 | 06:09 |
| *** sthillma has quit IRC | 06:12 | |
| *** sthillma_ is now known as sthillma | 06:12 | |
| *** preilly has joined #openstack-containers | 06:12 | |
| *** preilly is now known as Guest59200 | 06:13 | |
| *** sthillma has quit IRC | 06:23 | |
| *** j___ has joined #openstack-containers | 06:24 | |
| *** BertrandN has joined #openstack-containers | 06:43 | |
| *** shuquan has joined #openstack-containers | 06:48 | |
| *** BertrandN has quit IRC | 06:52 | |
| *** shuquan has quit IRC | 06:54 | |
| *** Tango has quit IRC | 06:54 | |
| *** shuquan has joined #openstack-containers | 06:58 | |
| *** david-lyle has quit IRC | 06:58 | |
| *** shuquan has quit IRC | 07:03 | |
| *** subscope has quit IRC | 07:03 | |
| *** BertrandN has joined #openstack-containers | 07:08 | |
| *** vilobhmm11 has quit IRC | 07:09 | |
| *** david-lyle has joined #openstack-containers | 07:14 | |
| *** achanda has quit IRC | 07:17 | |
| *** vilobhmm11 has joined #openstack-containers | 07:21 | |
| *** vilobhmm11 has quit IRC | 07:28 | |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Fix container status when showing a paused containers https://review.openstack.org/226184 | 07:30 |
| *** yuanying_ has quit IRC | 07:32 | |
| *** erw has quit IRC | 07:41 | |
| *** ewindisch has quit IRC | 07:41 | |
| *** ramishra has quit IRC | 07:41 | |
| *** zhenguo has quit IRC | 07:41 | |
| *** briancurtin has quit IRC | 07:41 | |
| *** briancurtin has joined #openstack-containers | 07:43 | |
| *** zhenguo has joined #openstack-containers | 07:45 | |
| *** ramishra has joined #openstack-containers | 07:49 | |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Fix container action debug log not accurate https://review.openstack.org/226194 | 07:56 |
| *** chandankumar has quit IRC | 07:59 | |
| eliqiao | hello, any one know why function failed due to "Cannot fulfill request with a Unset network_driver, expecting a flannel network_driver. (HTTP 400)" | 08:04 |
| *** halfss has joined #openstack-containers | 08:11 | |
| *** i316653 has joined #openstack-containers | 08:11 | |
| *** mathspanda has joined #openstack-containers | 08:19 | |
| *** Shijia has joined #openstack-containers | 08:19 | |
| eliqiao | can anyone gives a hand? | 08:21 |
| *** Marga_ has joined #openstack-containers | 08:27 | |
| *** Marga_ has quit IRC | 08:28 | |
| *** Marga_ has joined #openstack-containers | 08:29 | |
| *** Shijia has quit IRC | 08:36 | |
| *** Shijia has joined #openstack-containers | 08:36 | |
| *** Shijia has quit IRC | 08:37 | |
| *** Shijia has joined #openstack-containers | 08:37 | |
| eliqiao | sigh ..... | 08:38 |
| *** Shijia_ has joined #openstack-containers | 08:42 | |
| *** Shijia has quit IRC | 08:42 | |
| *** Shijia_ has quit IRC | 08:43 | |
| *** Shijia has joined #openstack-containers | 08:43 | |
| *** mathspanda has quit IRC | 08:46 | |
| *** mathspanda has joined #openstack-containers | 08:47 | |
| *** Shijia_ has joined #openstack-containers | 08:56 | |
| *** Shijia has quit IRC | 08:56 | |
| *** Marga_ has quit IRC | 09:05 | |
| *** alex_klimov has joined #openstack-containers | 09:06 | |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Functional: Split python client functional testing case https://review.openstack.org/223907 | 09:08 |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Add functional test cases(containers actions) for swarm https://review.openstack.org/226125 | 09:08 |
| openstackgerrit | Eli Qiao proposed openstack/magnum: Add functional test cases for swarm https://review.openstack.org/221633 | 09:08 |
| *** alex_klimov has quit IRC | 09:09 | |
| *** alex_klimov has joined #openstack-containers | 09:10 | |
| *** mathspan_ has joined #openstack-containers | 09:14 | |
| *** mathspanda has quit IRC | 09:18 | |
| *** shuquan has joined #openstack-containers | 09:21 | |
| *** mathspan_ has quit IRC | 09:22 | |
| *** wanghua has joined #openstack-containers | 09:24 | |
| *** diga has quit IRC | 09:25 | |
| openstackgerrit | Hua Wang proposed openstack/magnum: Add registry to template https://review.openstack.org/223526 | 09:30 |
| madhuri | yuanying: Hi, you there? | 09:34 |
| openstackgerrit | Ma Wen Cheng proposed openstack/magnum: Sync the description with Kubernetes Release 1 version https://review.openstack.org/226222 | 09:34 |
| openstackgerrit | Hua Wang proposed openstack/magnum: Code refactor for keystoneclient https://review.openstack.org/218699 | 09:37 |
| *** shuquan has quit IRC | 09:45 | |
| *** shuquan has joined #openstack-containers | 09:47 | |
| *** dims has joined #openstack-containers | 09:47 | |
| openstackgerrit | Merged openstack/python-magnumclient: Change ignore-errors to ignore_errors https://review.openstack.org/225818 | 09:48 |
| *** Shijia_ has quit IRC | 09:54 | |
| *** erw has joined #openstack-containers | 09:54 | |
| *** chandankumar has joined #openstack-containers | 09:58 | |
| *** ewindisch has joined #openstack-containers | 09:58 | |
| *** halfss has quit IRC | 10:03 | |
| *** shuquan has quit IRC | 10:04 | |
| *** shuquan has joined #openstack-containers | 10:11 | |
| *** shuquan has quit IRC | 10:22 | |
| i316653 | Is anyone here to help with an installation? I've just a few questions. I think most people here are US western time so GMT 17:00 is possibly a better time to get envolved! | 10:22 |
| i316653 | *involved | 10:22 |
| wanghua | i316653: hi | 10:24 |
| i316653 | @wanghua Hi, Could you help me with some questions about Magnum? | 10:25 |
| wanghua | i316653:ok, if I know | 10:25 |
| i316653 | I have a source install of openstack (not via devstack) and I'm trying to get Magnum working. I can create a baymodel no problem using a fedora-21-atomic-3 image, but when I create a bay it errors out with no valid hosts found? I haven't a clue what needs setup from there to allow it to find hosts on the netowork | 10:27 |
| *** ganeshna has joined #openstack-containers | 10:27 | |
| i316653 | Do I need to create a nova instance and set it up to work with Magnum? Does it manage this itself using heat? | 10:28 |
| wanghua | You need not to create nova instance manually. | 10:29 |
| wanghua | It is created by magnum automatically. | 10:29 |
| wanghua | How many cores you env has | 10:30 |
| wanghua | I think you can look into nova log to see why no valid host found | 10:30 |
| wanghua | nova-scheduler.log will tell you why | 10:31 |
| i316653 | My env has 80 cores | 10:32 |
| i316653 | I'll try running again and studying the logs. It looks likely because I setup a nova instance on that environment, it may of taken all resources? | 10:35 |
| openstackgerrit | Merged openstack/magnum: Sync the description with Kubernetes Release 1 version https://review.openstack.org/226222 | 10:36 |
| wanghua | how many host do you have | 10:38 |
| wanghua | If one, you can use nova hypervisor-show `hostname` to query the resources info | 10:38 |
| i316653 | what value am I looking for? | 10:39 |
| i316653 | I have got http://paste.openstack.org/show/lB3bHX5fne0JSi9fVclf/ | 10:40 |
| wanghua | core, ram, disk | 10:40 |
| wanghua | hypervisor_type ironic? | 10:41 |
| i316653 | Yeah | 10:42 |
| wanghua | As I know, magnum has some bugs on ironic | 10:42 |
| i316653 | What other approaches would you recommend? | 10:43 |
| wanghua | Why you need ironic | 10:43 |
| wanghua | Or you just want to play with magnum | 10:43 |
| wanghua | https://bugs.launchpad.net/magnum?field.searchtext=ironic&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field.omit_dupes=on&field.has_patch=&field.has_no_package= | 10:44 |
| i316653 | It's part of our larger openstack build | 10:45 |
| wanghua | The ironic bugs have not been fixed. | 10:45 |
| i316653 | Oh, I see that now, what other ways is there to get magnum working? Is it possible without using ironic or would I need to do a devstack install of openstack? | 10:47 |
| wanghua | http://docs.openstack.org/developer/magnum/dev/dev-quickstart.html | 10:48 |
| wanghua | You can install magnum according to this tutorial | 10:48 |
| i316653 | Thanks a million! @wanghua I will have a look into devstack to continue CaaS with openstack and revisit where I was in a while! | 11:03 |
| *** ganeshna has quit IRC | 11:07 | |
| *** erkules_ is now known as erkules | 11:30 | |
| *** erkules has joined #openstack-containers | 11:30 | |
| *** deepthi has quit IRC | 11:32 | |
| *** julim has joined #openstack-containers | 11:37 | |
| *** zz_dimtruck is now known as dimtruck | 11:59 | |
| openstackgerrit | Merged openstack/magnum: Functional tests with Tempest - BayModel CRUD https://review.openstack.org/224170 | 12:00 |
| *** dimtruck is now known as zz_dimtruck | 12:02 | |
| *** zz_dimtruck is now known as dimtruck | 12:34 | |
| *** achanda has joined #openstack-containers | 12:41 | |
| *** achanda has quit IRC | 12:42 | |
| *** xek has joined #openstack-containers | 12:47 | |
| *** alejandrito has joined #openstack-containers | 12:47 | |
| *** Tango has joined #openstack-containers | 12:52 | |
| *** rlrossit has joined #openstack-containers | 12:58 | |
| *** yuanying_ has joined #openstack-containers | 13:01 | |
| madhuri | yuanying: Hi | 13:01 |
| *** tbh has quit IRC | 13:02 | |
| madhuri | yuanying_: Hi | 13:03 |
| *** chandankumar has quit IRC | 13:05 | |
| *** dims has quit IRC | 13:14 | |
| *** dims has joined #openstack-containers | 13:15 | |
| yuanying_ | hi yuanying | 13:21 |
| yuanying_ | misstake | 13:21 |
| yuanying_ | hi madhuri | 13:21 |
| madhuri | yuanying_: Hi, I got into an issue | 13:25 |
| madhuri | Where kube-apiserver was not running | 13:25 |
| madhuri | But I think I get the issue | 13:25 |
| madhuri | I had to use fedora-atomic-5 | 13:26 |
| madhuri | Thanks :) | 13:26 |
| yuanying_ | maybe, kube v1 issue | 13:26 |
| yuanying_ | I still use fedora-atomic-3 | 13:26 |
| madhuri | Yes it is | 13:27 |
| madhuri | I will leave in 10 mins. | 13:28 |
| madhuri | Have a good night yuanying_ | 13:28 |
| yuanying_ | good night madhuri | 13:28 |
| madhuri | yuanying_: I think there is still issue with it. | 13:32 |
| *** hongbin has joined #openstack-containers | 13:32 | |
| madhuri | I see kube-apiserver running but all commands fail | 13:32 |
| madhuri | Some issue with kubernetes vq | 13:33 |
| yuanying_ | at fedora-atomic-3? | 13:33 |
| madhuri | Sorry v1 | 13:33 |
| madhuri | fedora-atomic-5 | 13:33 |
| madhuri | [minion@fi-6lqbvanhh4-0-4fh4cja6js3k-kube-master-ddgqkm5fq6tz ~]$ kubectl get pods error: couldn't read version from server: Get http://localhost:8080/api: dial tcp 127.0.0.1:8080: connection refused | 13:33 |
| yuanying_ | hmm | 13:34 |
| madhuri | hongbin: might have idea about it | 13:34 |
| yuanying_ | who doesn implement kube v1 heat templates? | 13:34 |
| madhuri | Tango and hongbin | 13:35 |
| madhuri | hongbin will you be around for an hour or so? | 13:35 |
| hongbin | yes | 13:35 |
| madhuri | yuanying_: I have to rush for cab. | 13:35 |
| madhuri | Catch you later | 13:36 |
| yuanying_ | ok | 13:36 |
| madhuri | Bye hongbin yuanying_ | 13:36 |
| *** madhuri has quit IRC | 13:36 | |
| hongbin | yuanying_: hey, what is your question? | 13:36 |
| yuanying_ | actually, I didn’t check around kube api v1 | 13:37 |
| hongbin | ok | 13:37 |
| yuanying_ | madhuri said, my templates doesn’t work at fedora-atomic-5 | 13:37 |
| yuanying_ | I’m not sure what change is exist between v1 to v1beta3 | 13:38 |
| hongbin | yuanying_: check this review I046931ad491e8b7ee45943852901eac5c3df913e | 13:39 |
| hongbin | Tongo modified the template to make it work with the *-5 image | 13:39 |
| hongbin | You possibly need to apply his change. | 13:40 |
| yuanying_ | thanks | 13:40 |
| yuanying_ | oh, this templates is merged today | 13:40 |
| Tango | Hi everyone, the templates that are merged already in patch 207299 should work | 13:40 |
| Tango | If you get error, please let me know | 13:41 |
| Tango | One thing I do run into occasionally is that etcd fails to come up, and that breaks k8s | 13:42 |
| yuanying_ | hmm | 13:42 |
| hongbin | I guess it has something to do with the etcd discovery endpoint | 13:42 |
| hongbin | The etcd cluster need to access public etcd discovery service to bootstrap. If the internet access is unstable, that might fail | 13:43 |
| hongbin | Need to setup a local etcd discovery service to mitigate that | 13:44 |
| *** fawadkhaliq has joined #openstack-containers | 13:44 | |
| Tango | I do know my environment has a flaky switch so the external connection is flaky, so the etcd problem I see is probably just due to my set up. | 13:45 |
| Tango | usually I just restart the etcd service and it would be fine | 13:46 |
| hongbin | k | 13:46 |
| *** rpothier has joined #openstack-containers | 13:48 | |
| openstackgerrit | Motohiro/Yuanying Otsuka proposed openstack/magnum: [WIP] Add TLS support in heat kubernetes template https://review.openstack.org/202873 | 13:53 |
| openstackgerrit | Hongbin Lu proposed openstack/magnum: Docs update for new fedora atomic 1.0.4 https://review.openstack.org/221504 | 13:54 |
| *** fawadkhaliq has quit IRC | 13:59 | |
| *** fawadkhaliq has joined #openstack-containers | 14:01 | |
| *** fawadk has joined #openstack-containers | 14:02 | |
| *** dane_leblanc has joined #openstack-containers | 14:02 | |
| *** fawadkhaliq has quit IRC | 14:05 | |
| openstackgerrit | Murali Allada proposed openstack/magnum: WIP: Adding support for public baymodels https://review.openstack.org/226335 | 14:07 |
| *** irenab_ has joined #openstack-containers | 14:08 | |
| *** fawadkhaliq has joined #openstack-containers | 14:09 | |
| *** irenab has quit IRC | 14:11 | |
| *** irenab_ is now known as irenab | 14:11 | |
| *** fawadk has quit IRC | 14:11 | |
| *** vimal has quit IRC | 14:17 | |
| *** wshao has joined #openstack-containers | 14:28 | |
| *** harshs has joined #openstack-containers | 14:29 | |
| *** banix has joined #openstack-containers | 14:31 | |
| *** Marga_ has joined #openstack-containers | 14:37 | |
| *** fawadkhaliq has quit IRC | 14:39 | |
| Tango | hi hongbin | 14:44 |
| Tango | I am looking for your earlier patch where you added the option for discovery_url for etcd | 14:45 |
| Tango | wonder if you might have the link | 14:45 |
| * hongbin finding the link | 14:45 | |
| hongbin | Tango: Here you go I9dd3a47f6d50ebadf74c4ee65701183f18c9d629 | 14:46 |
| Tango | I want to look at an example how the api code handles the new option. I am trying to pass the password and got the client code, but somehow the api code is not forwarding to the RPC request | 14:47 |
| *** harshs has quit IRC | 14:47 | |
| hongbin | Here is the github commit link https://github.com/openstack/magnum/commit/bcdd70cf1e1fd1c1c52587202bf5a9a38703a44d | 14:47 |
| hongbin | Tango: Another example is the adding of network_driver https://github.com/openstack/magnum/commit/c0ad7c376e18c500136fa39f8097b340cfe40fbf | 14:49 |
| Tango | ok, thanks hongbin | 14:50 |
| hongbin | welcome | 14:50 |
| klindgren | hello - what release of the magnum client should be used with a kilo release of magnum. Seems like there are only a few tags? I was trying to use 0.0.1 - because everything else has liberty requirements. | 14:53 |
| klindgren | However I couldn't setup a baymodel - because the client didn't support the required field --coe | 14:54 |
| *** banix has quit IRC | 15:06 | |
| *** yuanying_ has quit IRC | 15:07 | |
| *** banix has joined #openstack-containers | 15:12 | |
| *** harshs has joined #openstack-containers | 15:12 | |
| *** annegentle has joined #openstack-containers | 15:14 | |
| *** Drago has joined #openstack-containers | 15:15 | |
| *** ganeshna has joined #openstack-containers | 15:16 | |
| *** dims has quit IRC | 15:20 | |
| *** Drago has quit IRC | 15:22 | |
| *** Drago has joined #openstack-containers | 15:23 | |
| *** j___ has quit IRC | 15:25 | |
| *** puranamr has joined #openstack-containers | 15:27 | |
| *** annegent_ has joined #openstack-containers | 15:27 | |
| *** puranamr has quit IRC | 15:27 | |
| *** dims has joined #openstack-containers | 15:28 | |
| *** harshs has quit IRC | 15:28 | |
| *** banix has quit IRC | 15:30 | |
| *** annegentle has quit IRC | 15:30 | |
| *** banix has joined #openstack-containers | 15:35 | |
| *** annegent_ has quit IRC | 15:36 | |
| *** annegentle has joined #openstack-containers | 15:36 | |
| *** i316653 has quit IRC | 15:39 | |
| *** Drago has quit IRC | 15:43 | |
| *** Drago1 has joined #openstack-containers | 15:43 | |
| hongbin | klindgren: I saw there is a stable/kilo branch. Have you tried it? | 15:44 |
| klindgren | can't pip install a stable/kilo branch of the client ;-) or atleast as far as I know you can't | 15:46 |
| openstackgerrit | Dane LeBlanc proposed openstack/python-magnumclient: Magnum show cmds display dictionaries with unicode u chars https://review.openstack.org/226076 | 15:51 |
| *** daneyon has joined #openstack-containers | 15:54 | |
| *** daneyon_ has joined #openstack-containers | 16:02 | |
| *** alex_klimov has quit IRC | 16:03 | |
| *** kebray has joined #openstack-containers | 16:03 | |
| *** kebray has quit IRC | 16:03 | |
| *** whydidyoustealmy is now known as shakamunyi | 16:03 | |
| *** kebray has joined #openstack-containers | 16:03 | |
| *** daneyon has quit IRC | 16:05 | |
| *** annegent_ has joined #openstack-containers | 16:06 | |
| *** eghobo has joined #openstack-containers | 16:06 | |
| openstackgerrit | Hongbin Lu proposed openstack/magnum: Implement bay monitoring and notifications https://review.openstack.org/215335 | 16:06 |
| *** annegentle has quit IRC | 16:07 | |
| *** eghobo_ has joined #openstack-containers | 16:09 | |
| *** devkulkarni has joined #openstack-containers | 16:11 | |
| *** eghobo has quit IRC | 16:13 | |
| *** Marga_ has quit IRC | 16:13 | |
| devkulkarni | hey folks.. I am looking for information about parameters that are passed in to magnum cli commands. Specifically, the magnum container-create command optionally takes in a json representation of the container. Where can I find information about what should be passed in the json body? | 16:14 |
| openstackgerrit | Merged openstack/magnum-ui: Add magnum api client https://review.openstack.org/222617 | 16:17 |
| *** puranamr has joined #openstack-containers | 16:17 | |
| *** chandankumar has joined #openstack-containers | 16:19 | |
| *** madhuri has joined #openstack-containers | 16:24 | |
| madhuri | hongbin: Hi | 16:24 |
| *** sthillma has joined #openstack-containers | 16:25 | |
| *** sthillma_ has joined #openstack-containers | 16:26 | |
| *** BertrandN has quit IRC | 16:27 | |
| *** sthillma has quit IRC | 16:29 | |
| *** sthillma_ is now known as sthillma | 16:29 | |
| *** gangil has joined #openstack-containers | 16:32 | |
| *** Marga_ has joined #openstack-containers | 16:33 | |
| madhuri | hongbin: Are you around? | 16:39 |
| *** annegent_ has quit IRC | 16:40 | |
| *** annegent_ has joined #openstack-containers | 16:42 | |
| madhuri | Tango: You around? | 16:45 |
| devkulkarni | madhuri: hi | 16:45 |
| devkulkarni | madhuri: I am looking for information about parameters that are passed in to magnum cli commands. Specifically, the magnum container-create command optionally takes in a json representation of the container. Where can I find information about what should be passed in the json body? | 16:46 |
| madhuri | Hi devkulkarni | 16:46 |
| *** harshs has joined #openstack-containers | 16:46 | |
| madhuri | Please wait | 16:46 |
| devkulkarni | sure | 16:46 |
| *** david-lyle has quit IRC | 16:46 | |
| madhuri | devkulkarni: Did you check quickstart? | 16:47 |
| madhuri | devkulkarni: https://github.com/openstack/magnum/blob/master/doc/source/dev/dev-quickstart.rst | 16:47 |
| *** achanda has joined #openstack-containers | 16:47 | |
| madhuri | This gives a simple example of creating container | 16:48 |
| *** annegent_ has quit IRC | 16:48 | |
| devkulkarni | madhuri: thanks. I was following this: http://docs.openstack.org/cli-reference/content/magnumclient_commands.html | 16:48 |
| devkulkarni | madhuri: I see an example of container-create in quick-start. cool. I assume that the json body is essentially contains as all the flags that are specified in the example | 16:50 |
| devkulkarni | madhuri: regarding the 'image' flag in container-create .. where is that image located? is it glance? | 16:51 |
| madhuri | madhuri: These can be any docker container | 16:54 |
| *** fawadkhaliq has joined #openstack-containers | 16:56 | |
| madhuri | devkulkarni: This can be any docker container image | 16:56 |
| madhuri | devkulkarni: gkance image is just to deploy a bay | 16:56 |
| *** Marga_ has quit IRC | 16:59 | |
| *** vilobhmm11 has joined #openstack-containers | 17:00 | |
| *** Marga_ has joined #openstack-containers | 17:00 | |
| *** adrian_otto has joined #openstack-containers | 17:01 | |
| *** Marga_ has quit IRC | 17:02 | |
| *** Marga_ has joined #openstack-containers | 17:03 | |
| *** gangil has left #openstack-containers | 17:05 | |
| *** Marga_ has quit IRC | 17:05 | |
| Tango | madhuri: Hi Madhuri | 17:05 |
| *** Marga_ has joined #openstack-containers | 17:06 | |
| *** banix has quit IRC | 17:07 | |
| *** kebray has quit IRC | 17:09 | |
| *** ganeshna has quit IRC | 17:10 | |
| *** puranamr has quit IRC | 17:10 | |
| *** ganeshna has joined #openstack-containers | 17:10 | |
| *** suro-patz has joined #openstack-containers | 17:11 | |
| *** dimtruck is now known as zz_dimtruck | 17:12 | |
| *** devkulkarni has quit IRC | 17:14 | |
| Tango | madhuri: going to the office now, be back in 30 mins | 17:14 |
| *** unicell has joined #openstack-containers | 17:14 | |
| *** achanda_ has joined #openstack-containers | 17:17 | |
| *** achanda has quit IRC | 17:17 | |
| *** Tango has quit IRC | 17:19 | |
| *** Marga_ has quit IRC | 17:20 | |
| *** Marga_ has joined #openstack-containers | 17:20 | |
| *** Marga_ has quit IRC | 17:21 | |
| *** diga has joined #openstack-containers | 17:21 | |
| *** kebray has joined #openstack-containers | 17:22 | |
| *** devkulkarni has joined #openstack-containers | 17:26 | |
| *** kebray has quit IRC | 17:26 | |
| hongbin | madhuri: just came back from lunch | 17:28 |
| madhuri | Hi hongbin | 17:30 |
| hongbin | madhuri: hey | 17:30 |
| madhuri | I created a k8s bay with fedora-atomic-5 | 17:30 |
| madhuri | And the k8s API doesn't seems to work | 17:30 |
| hongbin | ..... | 17:31 |
| hongbin | any error? | 17:31 |
| madhuri | Yes I sent it earlier | 17:31 |
| *** dims has quit IRC | 17:31 | |
| madhuri | I don't have it now | 17:31 |
| *** puranamr has joined #openstack-containers | 17:31 | |
| madhuri | Its late here, so won't be able to share again | 17:31 |
| madhuri | What is time there? | 17:32 |
| *** dims has joined #openstack-containers | 17:32 | |
| hongbin | 1:32 pm | 17:32 |
| hongbin | you are in indian? | 17:32 |
| madhuri | Yes | 17:32 |
| hongbin | I see | 17:32 |
| madhuri | So can we discuss tomorrow morning | 17:32 |
| hongbin | sure | 17:32 |
| madhuri | What time zone are you in? | 17:32 |
| hongbin | Toronto | 17:32 |
| *** Marga_ has joined #openstack-containers | 17:33 | |
| *** puranamr has quit IRC | 17:33 | |
| *** Marga_ has quit IRC | 17:34 | |
| *** Marga_ has joined #openstack-containers | 17:34 | |
| devkulkarni | madhuri: thanks | 17:34 |
| madhuri | devkulkarni: you're welcome | 17:34 |
| madhuri | hongbin: I will leave a mail to you or if possible I will catch you tomorrow morning at 8:30 IST | 17:35 |
| hongbin | madhuri: K. See you later | 17:36 |
| devkulkarni | madhuri: still one more question actually. I understand that for bay creation the image is read from glance. but for the container-create command can the container image be in glance? | 17:36 |
| hongbin | devkulkarni: No, container image is downloaded from docker hub | 17:36 |
| devkulkarni | hongbin: I see | 17:37 |
| madhuri | No devkulkarni, docker hub is the registry | 17:37 |
| madhuri | hongbin: I will catch you tomorrow | 17:37 |
| devkulkarni | hongbin: is there any plan for supporting glance/swift in the container-create? | 17:37 |
| madhuri | devkulkarni: hongbin bye | 17:37 |
| devkulkarni | later madhuri | 17:37 |
| *** puranamr has joined #openstack-containers | 17:37 | |
| hongbin | madhuri: good night | 17:37 |
| *** madhuri has quit IRC | 17:38 | |
| *** snrism has joined #openstack-containers | 17:39 | |
| hongbin | devkulkarni: Check out this blueprint, it suggest to create a private docker registry and use swift as backend https://blueprints.launchpad.net/magnum/+spec/registryv2-in-master | 17:40 |
| devkulkarni | hongbin: thanks for the link | 17:41 |
| devkulkarni | looking | 17:41 |
| *** sthillma has quit IRC | 17:44 | |
| *** zz_dimtruck is now known as dimtruck | 17:44 | |
| openstackgerrit | Merged openstack/python-magnumclient: Magnum show cmds display dictionaries with unicode u chars https://review.openstack.org/226076 | 17:44 |
| *** Tango has joined #openstack-containers | 17:49 | |
| devkulkarni | hongbin: seems interesting. have few questions. how would one push custom docker images to the docker registry on each node of a bay? or, that is not necessary as the custom image can be stored in swift and Docker Distribution magic will make it appear in the docker registry on each node? | 17:49 |
| hongbin | devkulkarni: I am not working on that bp, need someone else to clarify it | 17:51 |
| devkulkarni | hongbin: ok | 17:51 |
| hongbin | devkulkarni: maybe you can send an email for the question | 17:51 |
| devkulkarni | hongbin: sure. I will also add myself to receive notifications on that bp | 17:52 |
| hongbin | cool | 17:52 |
| devkulkarni | and will add my question there | 17:52 |
| hongbin | sure | 17:52 |
| devkulkarni | thanks hongbin | 17:52 |
| *** chandankumar has quit IRC | 17:52 | |
| *** chandankumar has joined #openstack-containers | 17:55 | |
| *** rlrossit1 has joined #openstack-containers | 17:57 | |
| *** fawadkhaliq has quit IRC | 17:58 | |
| *** rlrossit has quit IRC | 17:58 | |
| *** ganeshna has quit IRC | 18:06 | |
| *** sthillma has joined #openstack-containers | 18:07 | |
| *** snrism has quit IRC | 18:08 | |
| *** chandankumar has quit IRC | 18:08 | |
| *** puranamr has quit IRC | 18:09 | |
| *** snrism has joined #openstack-containers | 18:11 | |
| *** achanda has joined #openstack-containers | 18:23 | |
| *** achanda_ has quit IRC | 18:24 | |
| *** snrism_ has joined #openstack-containers | 18:25 | |
| *** diga has quit IRC | 18:27 | |
| *** snrism has quit IRC | 18:28 | |
| *** jwcroppe has joined #openstack-containers | 18:29 | |
| *** vilobhmm11 has quit IRC | 18:32 | |
| *** vilobhmm11 has joined #openstack-containers | 18:32 | |
| openstackgerrit | Surojit Pathak proposed openstack/magnum: Backend support for magnum service healthcheck https://review.openstack.org/220702 | 18:35 |
| vilobhmm11 | hongbin : ping | 18:36 |
| hongbin | vilobhmm11: yo | 18:36 |
| vilobhmm11 | looks like the k8s_api changed a lot after your v1 patch merged in | 18:36 |
| vilobhmm11 | just rebasing my change according to that | 18:36 |
| hongbin | ack | 18:37 |
| vilobhmm11 | I did saw your comment on the objects from bay patch | 18:37 |
| *** snrism has joined #openstack-containers | 18:38 | |
| hongbin | Sorry, I should ask you to set a dependency on v1 patch | 18:38 |
| hongbin | hope rebasing won't incur too much overhead | 18:39 |
| openstackgerrit | Surojit Pathak proposed openstack/magnum: Adding API support for magnum service https://review.openstack.org/226078 | 18:39 |
| *** snrism_ has quit IRC | 18:41 | |
| *** ganeshna has joined #openstack-containers | 18:41 | |
| openstackgerrit | Hongbin Lu proposed openstack/magnum: Fix funtional gate: specify missing network_driver https://review.openstack.org/226491 | 18:42 |
| *** achanda has quit IRC | 18:42 | |
| *** ganeshna has quit IRC | 18:43 | |
| *** ganeshna has joined #openstack-containers | 18:44 | |
| *** ganeshna has quit IRC | 18:47 | |
| *** ganeshna has joined #openstack-containers | 18:47 | |
| *** achanda has joined #openstack-containers | 18:48 | |
| openstackgerrit | Surojit Pathak proposed openstack/magnum: Documentation update for 'magnum service-list' https://review.openstack.org/226133 | 18:48 |
| klindgren | I have a question regarding the heat templates for a kube cluster creation. It looks like there is a strong bias towards having end users create routers/private networks. We don't do that in our environment, we are bascially doing flat networks with neutron. | 18:49 |
| klindgren | I am looking at changing the heat templates - to support not doing private networks, it looks like in the comments for the templates this could be supported - but.. I want to make sure I understand what the req is for private networks and how they are used | 18:51 |
| *** ganeshna has quit IRC | 18:51 | |
| openstackgerrit | Daneyon Hansen proposed openstack/magnum: WIP: Updates Swarm Heat Templates for Container Network Model https://review.openstack.org/224367 | 18:53 |
| suro-patz | daneyon_: adrian_otto: you want to answer the question above from klindgren:? klindgren: is looking for a rollout of magnum | 18:55 |
| * adrian_otto looking at scrollback | 18:56 | |
| daneyon_ | suro-patz thx for the heads-up | 18:56 |
| klindgren | I just don't want to go in all crazy pants changing heat templates :-) - without understanding why they are doing what they are doing | 18:56 |
| daneyon_ | klindgren I suggest sending a message to the ML before making changes to the heat templates. | 18:57 |
| suro-patz | daneyon_: np | 18:57 |
| adrian_otto | klindgren: great question. The issue is not so much the creation of isolated networks per bay, but how to make the bays accessible…. It's really more of an IPAM concern. | 18:58 |
| klindgren | I assume opestanck-dev with the [magnum] tag | 18:58 |
| klindgren | openstack-dev* | 18:59 |
| adrian_otto | depending on how you set up Nova, you might not be able to access your nova instances unless they have neutron assigned addresses | 18:59 |
| daneyon_ | klindgren magnum will leverage whatever neutron networking you have implemented in your OS cloud. All you need to do is provide the id or name to external_network_id. external_network_id is the id/name of your neutron network that will assign floating ip's to the kube master/minions. | 18:59 |
| adrian_otto | keep in mind that external-lb assumes that a neutron LB will be used to front-end apps that run within the k8s cluster on the bay. | 19:00 |
| *** jcrubio has joined #openstack-containers | 19:00 | |
| adrian_otto | klindgren: will you be running apps that you expect will be load balanced using k8s "services"? | 19:01 |
| suro-patz | daneyon_: I think the question is do we absolutely need neutron to deploy magnum. As there are a good number of nova-network deployment. Can magnum be rolled over nova-network? | 19:01 |
| adrian_otto | suro-patz: strictly speaking, no we don't support nova-network today | 19:01 |
| klindgren | daneyon_, see the thing is our networking setup is pretty unique. Well neutron unique (not unique in the datacenter space). So we already have a router created that ties the floating-ips network with the networks that vm's get. | 19:01 |
| adrian_otto | but there are some simple use cases that would still work without neutron | 19:02 |
| *** banix has joined #openstack-containers | 19:02 | |
| klindgren | But we also route floating_ips to vm's vs's creating a L3 router to do it | 19:02 |
| *** snrism has quit IRC | 19:02 | |
| adrian_otto | klindgren: ok, so it sounds like further work would need to be done to integrate the Kubernetes external-lb feature with what you are doing | 19:03 |
| suro-patz | adrian_otto: I think that might help some people to do POC deployment of magnum with nova-network | 19:03 |
| adrian_otto | unless you did all that as a neutron plugin, which it does not sound like what you described | 19:03 |
| klindgren | adrian_otto, working on the POC to try to answer that. I assume yes - we are working on getting a PCO using LBaaS working - but we identified a number of gaps with our LB vendor and their LbaaS integration | 19:03 |
| adrian_otto | suro-patz: yes, that would be helpful. I don't think we are there yet though. | 19:04 |
| daneyon_ | klindgren the neutron routers created by the heat templates are not used for floating-ip's. magnum expects an existing neutron floating-ip network (external_network_id) to be available. | 19:04 |
| daneyon_ | How this existing floating-ip network is implemented does not matter to Magnum. Magnum will simply make the Neutron API call's to ask for a floating-ip from this network you provide and assign the ip to the kube master and minions. | 19:05 |
| daneyon_ | klindgren I would take some more time diving into the heat-templates, do some test deployments, etc.. before considering any changes. | 19:06 |
| * adrian_otto stepping out for lunch, then meetings adfter | 19:07 | |
| *** kebray has joined #openstack-containers | 19:07 | |
| daneyon_ | as for supporting nova-network, I would +1 if it's implementation is aligned with the goals of the magnum project. | 19:07 |
| *** adrian_otto has quit IRC | 19:08 | |
| klindgren | I guess I am totally confused why the router is created then? The issue is the default templates are pretty biased to services that you have to have running in the cloud (LBaaS, Cinder) so no matter what, to even poc I need to make changes to the template since we currently provide neither | 19:08 |
| *** kebray has quit IRC | 19:08 | |
| *** kebray has joined #openstack-containers | 19:08 | |
| daneyon_ | suro-patz we do need neutron b/c we make neutron-specific calls through our heat templates. We could support nova-net, it just comes down to implementing the codes and the implementation needs to align with the goals of the magnum project. | 19:08 |
| suro-patz | daneyon_: Agree, but that code/template does not exist today. As a community, we have not invested in that | 19:09 |
| *** jcrubio has quit IRC | 19:09 | |
| suro-patz | Klindgren: as adrian_otto: suggested for POC, those changes/additional templates are welcome | 19:10 |
| daneyon_ | klindgren cinder is needed to attach a vol to the nodes for mounting the docker container ns share. | 19:11 |
| daneyon_ | klindgren a router is created, kube master/nodes are attached, etc.. | 19:12 |
| suro-patz | daneyon_: but cinder volume attachment is again not a necessity, as I understand | 19:13 |
| daneyon_ | klindgren for POC purposes, why not use devstack? | 19:13 |
| klindgren | is that volume moved around, when clusters are updated/changed or is it that you just wanted an extra disk associated with the instances? | 19:13 |
| daneyon_ | lbass for ha/scaling kube services such as the api, etcd, etc.. | 19:14 |
| daneyon_ | the templates are focused on providing prod services and not a demo env. That is why we mount a cinder vol for the docker container ns. | 19:15 |
| daneyon_ | the vol allows a user to take advantage of a cinder backend for storage mgt instead of having the containers running on ephemeral storage. | 19:16 |
| daneyon_ | for poc/demo purposes, I suggest using devstack. It will provide a full-featured magnum experience. | 19:17 |
| daneyon_ | I'm -1 for supporting demo/poc focused heat templates that strip out functionality. IMO we already have too many templates to support. | 19:18 |
| eghobo_ | daneyon_: +1 for # templates | 19:19 |
| daneyon_ | We are in the process of refactoring the templates to simplify things and reduce the number of templates. Here is the bp: https://blueprints.launchpad.net/magnum/+spec/generate-heat-templates Feel free to connect with tcammann1 tcammann_ on ideas related to improving the templates. | 19:20 |
| *** rlrossit1 has quit IRC | 19:23 | |
| daneyon_ | For example, Ii added network-driver and labels support to the kube templates. I'm in the process of adding network-driver and labels support to swarm. The current swarm templates are quite different than the kube one's and it's taking me far too much time. If we execute on https://blueprints.launchpad.net/magnum/+spec/generate-heat-templates adding a feature across all bay types should be very simple. | 19:23 |
| *** eghobo_ has quit IRC | 19:24 | |
| *** rlrossit has joined #openstack-containers | 19:25 | |
| *** puranamr has joined #openstack-containers | 19:27 | |
| *** puranamr has quit IRC | 19:28 | |
| *** wshao has quit IRC | 19:30 | |
| openstackgerrit | Vilobh Meshram proposed openstack/magnum: Objects from Bay : Retrieve k8s API client https://review.openstack.org/220344 | 19:31 |
| *** snrism has joined #openstack-containers | 19:32 | |
| *** adrian_otto has joined #openstack-containers | 19:33 | |
| *** rlrossit has quit IRC | 19:36 | |
| *** snrism has quit IRC | 19:37 | |
| *** achanda has quit IRC | 19:37 | |
| *** rlrossit has joined #openstack-containers | 19:38 | |
| *** bradjones has joined #openstack-containers | 19:46 | |
| *** bradjones has quit IRC | 19:46 | |
| *** bradjones has joined #openstack-containers | 19:46 | |
| *** kebray has quit IRC | 19:57 | |
| *** kebray has joined #openstack-containers | 19:58 | |
| *** banix has quit IRC | 20:01 | |
| *** annegentle has joined #openstack-containers | 20:02 | |
| klindgren | sorry - back daneyon_ demo/poc in devstack is useless to me and my company. Part of the POC is can I actually get magnum to work in my environment, and if I can with what feature set limitations (if any). We are trying to roll out a company wide answer for containers, be that magnum (which we are rooting for) - nova-docker - docker via heat templates. | 20:04 |
| klindgren | All I am trying to do right now is figure out the requirement for kube cluster to be in its own private network with its own route attached to a provider network. Seems like this based upon the assumption that this is how networking is done in the cloud, and this just creates those primitives to give the kube network a way to talk to the internet/ rest of the network. I assume however, these primitives get used later on w | 20:05 |
| klindgren | hen trying to do floating ips/lbaas for services (jsut what I am attempting to glean from looking at the heat templates). | 20:05 |
| *** dane_leblanc has quit IRC | 20:05 | |
| klindgren | In our cloud, every vm is plugged into the provider network, so they have network access through a real network gateway (vs's a linux box running hte l3 agent and iptables in a network name space) similarly floating ip support is handled via routing the IP in the network to the fixed_ip of the vm' vs's added a dnat rule on a router. | 20:06 |
| klindgren | So my main concern is how inside magnum and kube is the private network expected to be used? | 20:06 |
| klindgren | Are the minions and the kube master tied to the same internal network with flanneld running on top of that and then micro-services exposed to end users via a floating ip? What are the possible down sides of having multiple bay's attempting to run on the same provider network - without their own private network? | 20:07 |
| klindgren | I guess what I really need is a design doc/spec that talks about what a bay is suppose to provide of kube - so I can answer some of these questions my self, and another on how containers will use the network provided for them by neutron | 20:08 |
| daneyon_ | klindgren I have not deployed magnum on a neutron setup that uses provider networks, but it should work. You need to provide the name or uuid of the provider network to magnum when instantiating your baymodel using the external_network_id baymodel config flag. | 20:08 |
| daneyon_ | klindgren it's been a while since i've used provider networks, but it's starting to come back to me now. I used to worked with provider networks quite a bit b/c of the l3-agent ha/scale issue. | 20:11 |
| *** sthillma has quit IRC | 20:11 | |
| *** snrism has joined #openstack-containers | 20:12 | |
| *** eghobo has joined #openstack-containers | 20:13 | |
| *** achanda has joined #openstack-containers | 20:15 | |
| klindgren | daneyon_, I do that however - normal users on our cloud can't create a private network or a router. So in the templates: https://github.com/openstack/magnum/blob/master/magnum/templates/heat-kubernetes/kubecluster-coreos.yaml#L120-L149 | 20:16 |
| daneyon_ | klindgren give me 5 min to look through the provider network doc's again. | 20:16 |
| *** sthillma has joined #openstack-containers | 20:18 | |
| *** manjeets has joined #openstack-containers | 20:18 | |
| manjeets | hongbin: ping | 20:18 |
| klindgren | seems like thats exactly what thats trying to do? But the comment makes it seem like this is only done for ease of use? | 20:19 |
| klindgren | daneyon_, our network config is bascially: http://docs.openstack.org/havana/install-guide/install/apt/content/section_use-cases-multi-flat.html | 20:19 |
| hongbin | manjeets: what's up | 20:20 |
| klindgren | with some additional tweaks/restrictions thrown in | 20:20 |
| manjeets | you removed dependency from patch ? | 20:21 |
| manjeets | should i remove WIP from it as well since its implemented ? | 20:21 |
| manjeets | hongbin: | 20:21 |
| hongbin | manjeets: you are refering a review? which one? | 20:22 |
| manjeets | https://review.openstack.org/#/c/221504/ | 20:22 |
| daneyon_ | klindgren thaks for the info. the magnum router gets attached to your provider network. You need to supply magnum the name or id of the provider network using the external_network_id of your desired provider network when creating your baymodel. | 20:23 |
| hongbin | daneyon_: I think klindgren problem is that users don't have permission to create private net and router | 20:24 |
| daneyon_ | The router will use IP's from the associated subnet of the provider network to assign to the neutron lbaas services and assign a floating-ip to the kube master/nodes. | 20:25 |
| klindgren | I think we are talking past each other. In our clouds users dont create networks and they dont create routers. So the router create in the heat template fails. | 20:26 |
| manjeets | hongbin: https://review.openstack.org/#/c/221504/ | 20:27 |
| klindgren | so it sounds like, in my env I can change the assumption to that a floating_ip is all thats ultimately need for the lbaas services and the the kube masters/nodes and as long as that is able to provided it should work | 20:27 |
| *** wshao has joined #openstack-containers | 20:28 | |
| klindgren | the details of how a floating_ip is allocated is currently hard-coded into the template under the assumption that a router + external network id is needed for a flaoting IP to get allocated to a fixed_ip of an instance | 20:28 |
| hongbin | klindgren: so if you can change the heat template, what you would like to change? | 20:29 |
| daneyon_ | klindgren see if the existing templates will work in your env. I understand magnum does not deploy the cluster in the same manner as you deploy vm's in your current env. I don't think magnum will ever replicate all the diff ways to deploy an OS cloud. Try playing around with the templates to meet the needs of your environment. | 20:31 |
| daneyon_ | Before submitting a review to get any custom changes merged, I suggest filing a bp b/c imo it will take more than heat template changes to support different network topologies, etc.. | 20:32 |
| *** wshao has quit IRC | 20:32 | |
| hongbin | Well, if klindgren env is a typical setup, then I think it is fine to make the adjustment on the template | 20:33 |
| hongbin | The template should represent a setup that works in most of the cases | 20:34 |
| daneyon_ | ^ simply changing the heat templates can work, but more is needed by the conductor and api. | 20:35 |
| daneyon_ | hongbin agreed | 20:35 |
| daneyon_ | our template does work in most cases. | 20:35 |
| klindgren | hongbin, thats a good question - I am not sure how to answer that at this point. I get what what is done - has been done. At this point I am just trying to figure out if this is done a specific way because thats the typical way to do it in neutron - or if their is additional assumptions made about tenant isolation | 20:35 |
| daneyon_ | I'm +1 for adding provider networking support (if changes are even needed), but if we simply change the templates to work with provider networks, then they will not work in non-provider network setups. | 20:36 |
| klindgren | I know some other larger enterprise deployments have similar kinda of networking to what we do. I htink if this can work on nva-networking then it would work in my env as well. | 20:36 |
| klindgren | daneyon_, agreed | 20:37 |
| daneyon_ | imo we would need to expose an attr in the baymodel api such as --neutron-type that would allow a user to keep the default, select provider-networking, etc.. | 20:37 |
| klindgren | I think it comes down to are you doing mutli-tenant networking or not | 20:38 |
| daneyon_ | I think trying to replicate a magnum cluster deployment to look just like the OS undercloud is a mistake. Too many variations. The cluster (i.e. bay type) should just work, calling the api's of the OS services. Granted, certain OS undercloud services must be available (i.e. Heat, Neutron, etc..). | 20:39 |
| daneyon_ | adrian_otto feel free to chime in. | 20:39 |
| klindgren | where multi-tenant allows everyone to create their own networking stack and non-multi-tenant is use the networking stack thats provided | 20:40 |
| daneyon_ | klindgren agreed. I worked with several operators in the past that implemented provider networks. I created several OS HA guides that leveraged provider networks for HA and scale. | 20:40 |
| klindgren | I would say at this point the majority of my end-users use cases (99%) are give me a compute resource thats has an ip in this network (which is typically more of a set of cidr's that have special meaning in the company (dev/test/prod/pci/pki) | 20:41 |
| daneyon_ | I think magnum should be able to run on an OS cloud that uses provider networks. However, I don't believe the magnum bay deployment should be = to how existing operators deploy. It's a slippery slope of supporting a million different combinations. | 20:42 |
| klindgren | daneyon_, I agree with you. | 20:43 |
| daneyon_ | We should be able to specify the neutron external-network-id within the baymodel and have magnum plug bay's (i.e. the router's gw interface, master/minion floating-ip's) into that network. | 20:44 |
| daneyon_ | klindgren, so does each of your tenant's have their own provider network? | 20:45 |
| klindgren | thats something that we are restling with as well - as an enterprise do we really want every tenant to spinup a k8 cluster (and have to support that many possible clusters) or do we want to use magnum to spinup a few large enterprise clusters and ideally have minons in each tenant that can be used to house those projects containers | 20:45 |
| klindgren | daneyon_, nope their are shared across every tenant | 20:45 |
| openstackgerrit | Manjeet Singh Bhatia proposed openstack/magnum: Docs update for new fedora atomic 1.0.4 https://review.openstack.org/221504 | 20:47 |
| klindgren | at current count we have like 4.5k tenants - if half of those want to do containers - we are duplicating a ton of infrastructure for possibly 3-4 containers | 20:47 |
| adrian_otto | ok, so I'm partial to keeping it simple. Supporting every network topology under the sun is intractable. | 20:47 |
| adrian_otto | in terms of whether an enterprise should allow its developers to have their own bays, or share them... | 20:47 |
| adrian_otto | that's a decision that requires asking a lot of questions | 20:48 |
| adrian_otto | in general, the overhead of running a bay is relatively low, but if you're going to be running thousands of them, then it will gradually add up. | 20:48 |
| klindgren | adrian_otto, agreed - we (the openstack team at our company) are still trying to figure out how the people at our company are using containers | 20:48 |
| klindgren | and what do they want to see - offset by the possible support/infra overhead on our end | 20:49 |
| adrian_otto | klindgren: are you primarily concerned with offering business agility, standardizing support, or maximizing operational efficiency? | 20:49 |
| klindgren | I am not opposed to the bay per tenant - but I know the "death by a thousand cuts" | 20:50 |
| daneyon_ | klindgren I have to get ready for a meeting. Try the existing templates to see if you can simply plug magnum into your existing provider network. It would be helpful to know if that works, as it should. If you need to make changes to the templates to meet your needs, send a message to the ML and let's see how we can proceed in a manner that is consistent with the goals of the project. | 20:50 |
| adrian_otto | My suggestion would be to allow a bay per team at least, and see what sorts of bays they create, and ask them why they make those choices | 20:50 |
| klindgren | Can I pick all three :-) | 20:51 |
| adrian_otto | yes. | 20:51 |
| klindgren | I would say the biggest two are business agility and standardization of support | 20:51 |
| klindgren | our end users would love to try to stop woring about the infra pieces and start caring more about deploying code | 20:51 |
| adrian_otto | if you want all three, then you probably want to have a nova that can create containers with the libvirt/lxc driver | 20:52 |
| adrian_otto | and use those nodes to compose bays | 20:52 |
| adrian_otto | that way you will get more dense packing of bays per compute node | 20:52 |
| adrian_otto | and relatively high performance. | 20:52 |
| adrian_otto | give your devs the ability to create their own bays, and customize their own bay node images. | 20:53 |
| adrian_otto | then they can use the native tooling of their choice to interact with those bays. | 20:53 |
| *** annegentle has quit IRC | 20:54 | |
| adrian_otto | for standardization of support, you can use a business policy (may not need technical enforcement) to suggest that a particular bay type (or list of types) will be supported by your team, and others will be use-at-your-own-risk…. or whatever makes sense in your corporate culture. | 20:56 |
| manjeets | adrian: i agree there should some set of supported images and they should also be documentation if user wants own custom image | 20:57 |
| manjeets | at own risk | 20:58 |
| adrian_otto | if you find that for you it's important to enforce a more limited selection of choice, we should propose those as features for Magnum. Something like provider-supplied baymodels (prohibiting the creation of new baymodels unless you have a particular RBAC role in the identity system, such that normal users do not have the role.) | 20:58 |
| adrian_otto | sometimes you have really great developers that have a good reason to do something non-standard… it might be nice to have a way to allow that, as manjeets suggested. | 20:59 |
| *** suro-patz has quit IRC | 21:00 | |
| manjeets | i feel this support should be there to provide customers or users with more flexibility and they should every reason to use our product | 21:00 |
| manjeets | they should have* every reason | 21:01 |
| *** suro-patz has joined #openstack-containers | 21:02 | |
| *** rebase33 has joined #openstack-containers | 21:05 | |
| *** manjeets has quit IRC | 21:09 | |
| *** adrian_otto has quit IRC | 21:09 | |
| *** rpothier has quit IRC | 21:20 | |
| *** rlrossit has left #openstack-containers | 21:20 | |
| *** Marga_ has quit IRC | 21:21 | |
| *** kebray has quit IRC | 21:21 | |
| *** achanda has quit IRC | 21:21 | |
| *** Marga__ has joined #openstack-containers | 21:22 | |
| *** annegentle has joined #openstack-containers | 21:24 | |
| *** eghobo has quit IRC | 21:25 | |
| *** alejandrito has quit IRC | 21:32 | |
| *** adrian_otto has joined #openstack-containers | 21:32 | |
| *** Drago1 has quit IRC | 21:32 | |
| *** alejandrito has joined #openstack-containers | 21:33 | |
| *** eghobo has joined #openstack-containers | 21:40 | |
| openstackgerrit | Ton Ngo proposed openstack/magnum: Configure CoreOS for Kubernetes load balancer feature https://review.openstack.org/192425 | 21:43 |
| openstackgerrit | Ton Ngo proposed openstack/magnum: Configure Ironic for Kubernetes load balancer feature https://review.openstack.org/192426 | 21:43 |
| openstackgerrit | Ton Ngo proposed openstack/magnum: Configure Fedora Atomic for Kubernetes load balancer feature https://review.openstack.org/191878 | 21:43 |
| adrian_otto | Our team meeting will begin at 22:00 UTC (15 min from now) in #openstack-meeting-alt | 21:45 |
| vilobhmm11 | hongbin : ping | 21:47 |
| hongbin | vilobhmm11: yo | 21:47 |
| vilobhmm11 | here https://review.openstack.org/#/c/213368/8/magnum/api/controllers/v1/replicationcontroller.py #80 | 21:47 |
| *** snrism has quit IRC | 21:49 | |
| hongbin | vilobhmm11: I saw it linked to my comment. You have question? | 21:51 |
| vilobhmm11 | you wanted the bay_uuid also to be passed right…which makes sense.. | 21:52 |
| vilobhmm11 | http://paste.openstack.org/show/473631/ | 21:52 |
| hongbin | yes, since you add bay_uuid to the endpoint | 21:52 |
| vilobhmm11 | rcs/{uuid}/rcs/{bay_uuid} | 21:53 |
| vilobhmm11 | or any particular format | 21:53 |
| vilobhmm11 | sorry my n/w is diconnecting today | 21:53 |
| hongbin | I don't know the format | 21:53 |
| vilobhmm11 | sorry for later reply | 21:53 |
| hongbin | NP | 21:53 |
| hongbin | I think the link returned have to be correct | 21:54 |
| vilobhmm11 | ok…np..will figure it out…just wanted to be sure…that if you wanted some specific format | 21:54 |
| vilobhmm11 | sure | 21:54 |
| *** annegentle has quit IRC | 21:54 | |
| hongbin | If the link is correct, then I am happy | 21:54 |
| vilobhmm11 | cool thanks! | 21:54 |
| vilobhmm11 | hongbin : ^^ | 21:54 |
| *** annegentle has joined #openstack-containers | 21:54 | |
| *** snrism_ has joined #openstack-containers | 21:57 | |
| *** wshao has joined #openstack-containers | 21:58 | |
| *** suro-patz has quit IRC | 22:01 | |
| *** Drago has joined #openstack-containers | 22:01 | |
| *** suro-patz has joined #openstack-containers | 22:03 | |
| *** wshao has quit IRC | 22:04 | |
| *** hongbin_ has joined #openstack-containers | 22:06 | |
| *** devkulkarni has quit IRC | 22:06 | |
| *** dimtruck is now known as zz_dimtruck | 22:06 | |
| *** hongbin has quit IRC | 22:07 | |
| *** devkulkarni has joined #openstack-containers | 22:07 | |
| *** jerrygb has joined #openstack-containers | 22:08 | |
| *** wshao has joined #openstack-containers | 22:11 | |
| *** zhenguo has quit IRC | 22:12 | |
| *** devkulkarni has quit IRC | 22:12 | |
| *** gus has quit IRC | 22:12 | |
| *** zhenguo has joined #openstack-containers | 22:13 | |
| *** gus has joined #openstack-containers | 22:14 | |
| *** sthillma has quit IRC | 22:19 | |
| *** achanda has joined #openstack-containers | 22:22 | |
| *** jwcroppe has quit IRC | 22:25 | |
| *** jwcroppe has joined #openstack-containers | 22:26 | |
| *** briancurtin has quit IRC | 22:26 | |
| *** puranamr has joined #openstack-containers | 22:26 | |
| *** banix has joined #openstack-containers | 22:27 | |
| *** achanda has quit IRC | 22:27 | |
| *** briancurtin has joined #openstack-containers | 22:29 | |
| *** puranamr has quit IRC | 22:29 | |
| *** jwcroppe has quit IRC | 22:30 | |
| *** eghobo has quit IRC | 22:35 | |
| *** vilobhmm11 has quit IRC | 22:36 | |
| *** vilobhmm11 has joined #openstack-containers | 22:36 | |
| *** devkulkarni has joined #openstack-containers | 22:48 | |
| *** sthillma has joined #openstack-containers | 22:48 | |
| *** zz_dimtruck is now known as dimtruck | 22:50 | |
| *** alejandrito has quit IRC | 22:55 | |
| *** wshao has quit IRC | 22:56 | |
| *** annegentle has quit IRC | 22:56 | |
| *** sthillma has quit IRC | 22:58 | |
| *** wanghua has quit IRC | 23:00 | |
| *** hongbin_ has quit IRC | 23:01 | |
| *** adrian_otto has quit IRC | 23:02 | |
| openstackgerrit | Merged openstack/magnum: Fix funtional gate: specify missing network_driver https://review.openstack.org/226491 | 23:05 |
| *** sthillma has joined #openstack-containers | 23:08 | |
| *** snrism_ has quit IRC | 23:08 | |
| *** eghobo has joined #openstack-containers | 23:10 | |
| *** achanda has joined #openstack-containers | 23:12 | |
| *** Drago has quit IRC | 23:16 | |
| *** vilobhmm11 has quit IRC | 23:17 | |
| *** vilobhmm11 has joined #openstack-containers | 23:18 | |
| *** dims has quit IRC | 23:18 | |
| *** dimtruck is now known as zz_dimtruck | 23:18 | |
| *** vilobhmm111 has joined #openstack-containers | 23:19 | |
| *** vilobhmm11 has quit IRC | 23:22 | |
| *** zz_dimtruck is now known as dimtruck | 23:29 | |
| *** rebase33 has quit IRC | 23:48 | |
| *** dims has joined #openstack-containers | 23:54 | |
| *** dims has quit IRC | 23:55 | |
| *** Guest18166 has joined #openstack-containers | 23:55 | |
| *** eghobo has quit IRC | 23:56 | |
| *** ganeshna has joined #openstack-containers | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!