| *** Qiming has joined #senlin | 00:10 | |
| *** Qiming has quit IRC | 00:18 | |
| *** xuhaiwei has joined #senlin | 00:19 | |
| *** Qiming has joined #senlin | 01:01 | |
| *** Yanyan has joined #senlin | 01:36 | |
| *** mathspanda has joined #senlin | 01:47 | |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: [WIP]Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 01:49 |
|---|---|---|
| *** elynn has joined #senlin | 01:55 | |
| xuhaiwei | hi Qiming, Yanyan, I am doing exception handling of lbaas module, and found something strange, https://github.com/stackforge/senlin/blob/master/senlin/drivers/openstack/lbaas.py#L148 | 01:59 |
| xuhaiwei | should line148 be like line117? | 01:59 |
| Yanyan | let me have a look | 02:00 |
| Yanyan | you mean we should get and check the return value from _wait_for_lb_ready? | 02:01 |
| xuhaiwei | yes | 02:02 |
| xuhaiwei | dont need to? | 02:02 |
| xuhaiwei | and I also dont know when deleting the listener, why should we also check whether lb is ready? | 02:03 |
| Yanyan | hmm, look like we should do that | 02:04 |
| Yanyan | that is because all lbaas related operation will cause loadbalancer resource be updated | 02:04 |
| Yanyan | sorry, not all, just those creation/deletion/update operations | 02:04 |
| xuhaiwei | ok | 02:05 |
| Yanyan | xuhaiwei, I think you're right. We should also check lb ready in lb_delete methods | 02:05 |
| xuhaiwei | not only lb_delete, there are many places where _wait_for_lb_ready is called | 02:06 |
| Yanyan | if res is not True, we should fail this lb_delete operation | 02:07 |
| Yanyan | you mean in lb_create or member_add? | 02:07 |
| Yanyan | and member_remove | 02:08 |
| xuhaiwei | yes | 02:08 |
| xuhaiwei | all the places | 02:08 |
| Yanyan | since we need to check it in those cases | 02:08 |
| Yanyan | so I think we may need to add this logic in lb_delete() as well | 02:08 |
| Yanyan | and add some checks in lb_policy | 02:09 |
| xuhaiwei | lb_policy? | 02:10 |
| Yanyan | yes, since lb_policy will invoke lb_delete in some cases, we may need to check whether lb_delete succeeded | 02:12 |
| xuhaiwei | ok | 02:13 |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: [WIP]Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 02:13 |
| *** elynn has quit IRC | 02:35 | |
| *** elynn has joined #senlin | 02:35 | |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: [WIP]Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 02:44 |
| openstackgerrit | Yanyan Hu proposed stackforge/python-senlinclient: Avoid deleting metadata incorrectly when doing profile_update https://review.openstack.org/215412 | 02:55 |
| openstackgerrit | Merged stackforge/python-senlinclient: Avoid deleting metadata incorrectly when doing profile_update https://review.openstack.org/215412 | 03:06 |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: Correct description about marker option in getting_started doc https://review.openstack.org/215418 | 03:25 |
| openstackgerrit | xu-haiwei proposed stackforge/senlin: Check wait_for_lb_ready method's return value https://review.openstack.org/215430 | 04:18 |
| xuhaiwei | Yanyan | 04:25 |
| Yanyan | hi, xuhaiwei, just came back from lunch | 04:40 |
| *** xuhaiwei_ has joined #senlin | 04:41 | |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: [WIP]Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 04:42 |
| *** xuhaiwei has quit IRC | 04:42 | |
| *** Shijia has joined #senlin | 04:43 | |
| xuhaiwei_ | Yanyan, about this patch https://review.openstack.org/#/c/215418/ | 04:44 |
| xuhaiwei_ | I think the original doc is not bad | 04:45 |
| Yanyan | hmm, but some doc in senlin also give the description like revised version | 04:46 |
| xuhaiwei_ | you mean about 'marker'? | 04:47 |
| Yanyan | http://git.openstack.org/cgit/stackforge/senlin/tree/doc/source/developer/node.rst#n129 | 04:47 |
| Yanyan | yes, like this | 04:47 |
| Yanyan | http://git.openstack.org/cgit/openstack/glance/tree/glance/registry/client/v1/client.py#n100 | 04:48 |
| Yanyan | and some other services' decription about this option | 04:48 |
| Yanyan | and based on the code, I think the description in development doc should be correct | 04:49 |
| xuhaiwei_ | yes, I think the senlin's doc means the same meaning | 04:52 |
| Yanyan | yes, I guess so | 04:53 |
| Yanyan | but really not sure whether the expression is ok... | 04:54 |
| xuhaiwei_ | the problem is the command function | 04:54 |
| Yanyan | command function? | 04:55 |
| xuhaiwei_ | you mean this line 'from which you want tosee the returned list starts'? | 04:55 |
| Yanyan | yes | 04:56 |
| xuhaiwei_ | yes, the '--marker' option doesn't run well | 04:56 |
| xuhaiwei_ | that is a little difficult English sentence I think :) | 04:56 |
| Yanyan | em, so maybe we wait for other guys' suggestion :) | 04:57 |
| xuhaiwei_ | the original sequence should be 'you want to see the returned list starts from'?? | 04:57 |
| xuhaiwei_ | anyway, there is a bug in the source code, '--marker' option doesn't work now | 04:58 |
| xuhaiwei_ | this maybe SDK's bug | 04:58 |
| Yanyan | oh, no, xuhaiwei_, it's just an error in doc decription I think | 04:59 |
| Yanyan | the --marker option actually work | 04:59 |
| xuhaiwei_ | really?? | 05:00 |
| Yanyan | just the way it works is different from the description in getting_startted document | 05:00 |
| xuhaiwei_ | but it makes no difference even if I used --marker option | 05:00 |
| Yanyan | hmm, it works in my env | 05:01 |
| Yanyan | I just referred to the test result from chris http://paste.openstack.org/show/412476/ | 05:01 |
| xuhaiwei_ | got it | 05:03 |
| xuhaiwei_ | it only works when UUID is set, not works for short-id | 05:03 |
| Yanyan | yep | 05:04 |
| xuhaiwei_ | that's a bug? | 05:04 |
| Yanyan | hmm, I think it is a place we can improve our design | 05:04 |
| Yanyan | we query the Full id based on short ID first and then do query based on it | 05:05 |
| xuhaiwei_ | why short id doesn't work here? | 05:08 |
| Yanyan | I think the marker param can only accept full-ID | 05:09 |
| Yanyan | so if user directly gives a short-ID as the marker, it actually doesn't take effect | 05:09 |
| Yanyan | since short-ID is an conception that handled by senlin's code | 05:10 |
| xuhaiwei_ | yes | 05:10 |
| xuhaiwei_ | sdk, only handles full-id maybe | 05:10 |
| Yanyan | oh, you mean this. I think this is not SDK's problem. It just pass the request to senlin-api | 05:11 |
| Yanyan | without any translation | 05:11 |
| Yanyan | e.g. if you give a short-id here, what senlin API accept is the short-id | 05:12 |
| Yanyan | this short-id should be handled by senlin's engine service logic | 05:12 |
| Yanyan | actually this is handled by senlin's sqlalchemy api's implementation | 05:13 |
| xuhaiwei_ | got it | 05:13 |
| Yanyan | something like this :) | 05:14 |
| Yanyan | http://git.openstack.org/cgit/stackforge/senlin/tree/senlin/db/sqlalchemy/api.py#n131 | 05:14 |
| openstackgerrit | OpenStack Proposal Bot proposed stackforge/senlin: Updated from global requirements https://review.openstack.org/214993 | 05:14 |
| openstackgerrit | Merged stackforge/senlin: Set cluster status to warning after deletion failed https://review.openstack.org/215006 | 05:21 |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: [WIP]Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 05:27 |
| *** Shijia has quit IRC | 05:37 | |
| *** Shijia has joined #senlin | 05:37 | |
| *** Yanyan has quit IRC | 05:42 | |
| *** Yanyan has joined #senlin | 05:42 | |
| *** zhenguo has joined #senlin | 05:42 | |
| *** Shijia_ has joined #senlin | 05:51 | |
| *** Shijia has quit IRC | 05:55 | |
| *** Shijia_ has quit IRC | 06:11 | |
| *** Shijia has joined #senlin | 06:12 | |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: Add fake nova_v2 driver for functional test https://review.openstack.org/214090 | 06:13 |
| *** Shijia has quit IRC | 06:16 | |
| *** Shijia has joined #senlin | 06:17 | |
| Qiming | hi | 06:18 |
| Qiming | https://review.openstack.org/215466 | 06:18 |
| Qiming | just heads up | 06:18 |
| Qiming | Zhenguo has kindly offered to start a dashboard project for Senlin | 06:18 |
| zhenguo | hello senlin | 06:19 |
| zhenguo | yeah, I would like to add a new project for UI component of Senlin | 06:19 |
| Yanyan | hi, zhenguo, welcome :) | 06:20 |
| zhenguo | :) | 06:20 |
| Yanyan | so this UI project will be part of horizon? | 06:22 |
| *** Shijia has quit IRC | 06:23 | |
| zhenguo | it's a horizon plugin | 06:23 |
| Yanyan | ok, cool ;) | 06:23 |
| *** Shijia has joined #senlin | 06:23 | |
| *** kun_huang has joined #senlin | 06:24 | |
| *** ChanServ changes topic to "test message" | 06:26 | |
| *** Shijia_ has joined #senlin | 06:27 | |
| xuhaiwei_ | hi, welcome to senlin zhenguo | 06:27 |
| zhenguo | o/ | 06:28 |
| *** Shijia has quit IRC | 06:28 | |
| *** ChanServ changes topic to "IRCLog: http://eavesdrop.openstack.org/irclogs/%23senlin/ | Bugs: bugs.launchpad.net/senlin | Review: https://review.openstack.org/#/q/project:stackforge/senlin,n,z" | 06:28 | |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: Check cluster size constraint before doing node join/leave https://review.openstack.org/215475 | 06:42 |
| *** elynn has quit IRC | 06:47 | |
| openstackgerrit | xu-haiwei proposed stackforge/senlin: Handle exception in neutron_v2 module https://review.openstack.org/215051 | 07:03 |
| *** xuhaiwei_ has quit IRC | 07:10 | |
| Qiming | hi, Yanyan | 07:23 |
| Qiming | I'm looking at the functional tests | 07:24 |
| Yanyan | hi | 07:24 |
| Qiming | still wondering if the fake API client, API response classes are actually needed | 07:24 |
| openstackgerrit | Merged stackforge/senlin: Correct description about marker option in getting_started doc https://review.openstack.org/215418 | 07:25 |
| Yanyan | you mean maybe we don't need to fake the return value of those interfaces in like nova driver? | 07:26 |
| Qiming | in Senlin, we are not making any HTTP requests directly, all our interactions with other services are done thru sdk | 07:26 |
| *** kun_huang has left #senlin | 07:26 | |
| Qiming | we fake the nova driver's directly, that would suffice? | 07:26 |
| Yanyan | nova driver's? | 07:27 |
| Yanyan | you mean sdk's function call? | 07:27 |
| Qiming | this file: https://review.openstack.org/#/c/214090/12/senlin/tests/functional/common/client.py | 07:28 |
| *** Shijia_ has quit IRC | 07:29 | |
| Yanyan | yes? | 07:29 |
| *** Shijia has joined #senlin | 07:29 | |
| Yanyan | hi, Qiming, this client if for the test code talking with senlin service | 07:32 |
| Qiming | where is the code? | 07:32 |
| Yanyan | you mean the code of test case? | 07:33 |
| Yanyan | something like this https://review.openstack.org/#/c/214090/12/senlin/tests/functional/test_profile.py | 07:34 |
| Qiming | senlin server code | 07:34 |
| *** xuhaiwei has joined #senlin | 07:34 | |
| Yanyan | oh, it is installed by devstack using the code from our git repo | 07:34 |
| Yanyan | it will be located in /opt/stack/new/senlin/ | 07:35 |
| Yanyan | by default | 07:35 |
| Qiming | I mean, in senlin server code, we are not talking to senlin via HTTP request/response, right? | 07:35 |
| Qiming | even for functional tests, we can still mock a cluster driver | 07:36 |
| Yanyan | cluster driver? the interface implemented in sdk for clustering service? | 07:37 |
| Qiming | yes | 07:37 |
| Qiming | we could have had all senlin resources committed to sdk, if not for their lock for release | 07:37 |
| Yanyan | you mean we can use sdk rather than API client to talk with senlin | 07:37 |
| Yanyan | when doing functional test | 07:38 |
| Qiming | API client makes not sense in our context | 07:38 |
| Yanyan | oh, if so, that will be the best | 07:38 |
| Qiming | if we are not using sdk, the story could be different | 07:38 |
| Yanyan | we don't need to handle all those API related things by ourself | 07:38 |
| Qiming | for functional tests, we don't need API client, because we can use faked driver for senlin, just like faked driver for nova | 07:39 |
| Qiming | for integration test, when that day comes, we could have committed all Senlin resources to SDK | 07:39 |
| Qiming | the raw API client stuff is still useless, right? | 07:39 |
| Yanyan | umm, but since sdk doesn't provide enough support for senlin, how can we send those requests to senlin service without both API and senlinclient | 07:41 |
| Yanyan | sure, after sdk is ready, we don't need the APIclient anymore | 07:41 |
| Yanyan | since we need real interaction between test case and senlin API and engine | 07:42 |
| Qiming | okay, understand | 07:43 |
| *** Shijia has quit IRC | 07:43 | |
| Yanyan | since API client is now the startpoint of a test progress :) | 07:43 |
| *** Shijia has joined #senlin | 07:44 | |
| Qiming | I see | 07:44 |
| Qiming | we really need to push the SDK guys | 07:47 |
| Yanyan | absolutely... | 07:47 |
| openstackgerrit | xu-haiwei proposed stackforge/senlin: Handle exception in neutron_v2 module https://review.openstack.org/215051 | 07:47 |
| Qiming | we still have action, build_info, event, policy, webhook, cluster_policy and trigger to be pushed to SDK | 07:48 |
| Yanyan | yes | 07:48 |
| *** Shijia has quit IRC | 07:48 | |
| Yanyan | but seems those guys are very busy recently... | 07:48 |
| Yanyan | I found bunch of bugs in sdk are still in open status | 07:49 |
| Qiming | dont feel it that way, not sure what they are busying at | 07:49 |
| Qiming | patches from TH and BC were not that much recently | 07:49 |
| Yanyan | hope they can finish their current work soon and start to accept new resource implementation | 07:50 |
| Yanyan | yes | 07:50 |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: Ensure policy enforcement level is less equal than 100 https://review.openstack.org/215501 | 07:53 |
| *** Shijia has joined #senlin | 07:54 | |
| Yanyan | hi, Qiming, I have tried to propose some fix for those issues Chris mentioned in his test result and marked them in the updated version of test document | 07:54 |
| Yanyan | will send you guys the updated one later | 07:54 |
| Qiming | okay, good | 07:54 |
| *** Shijia has quit IRC | 07:58 | |
| xuhaiwei | Chris is mathspanda and also LinPeiyu? | 07:59 |
| mathspanda | no, he is not me. | 08:00 |
| mathspanda | he is a german intern. | 08:00 |
| xuhaiwei | ok, thanks | 08:00 |
| mathspanda | hi, please look at this paste: http://paste.openstack.org/show/423203/ | 08:01 |
| *** mathspanda_ has joined #senlin | 08:05 | |
| Qiming | yes, start_time and end_time are not datetime data type | 08:06 |
| Qiming | they are currently used to count how many seconds taken to do an action | 08:06 |
| Qiming | also, you are not supposed to do this stuff in each individual action subclass, unless absolutely necessary | 08:07 |
| Qiming | all concrete action types use the same logic for execution flow and status update | 08:08 |
| Qiming | line 46 in that file and line 65 are all wrong | 08:09 |
| mathspanda_ | ok.i understand. | 08:10 |
| *** mathspanda_ has quit IRC | 08:12 | |
| *** Shijia has joined #senlin | 08:14 | |
| *** Qiming has quit IRC | 08:14 | |
| *** Shijia has quit IRC | 08:19 | |
| openstackgerrit | xu-haiwei proposed stackforge/senlin: Handle exception in neutron_v2 module https://review.openstack.org/215051 | 08:55 |
| *** Shijia has joined #senlin | 08:59 | |
| *** xuhaiwei has quit IRC | 09:01 | |
| *** Shijia has quit IRC | 09:03 | |
| *** LiuWei has quit IRC | 09:16 | |
| *** mathspanda has quit IRC | 09:32 | |
| *** Yanyan has quit IRC | 09:43 | |
| *** Shijia has joined #senlin | 09:49 | |
| *** Shijia has quit IRC | 09:53 | |
| *** Shijia has joined #senlin | 10:09 | |
| *** Shijia has quit IRC | 10:13 | |
| *** Shijia has joined #senlin | 10:34 | |
| *** Shijia has quit IRC | 10:38 | |
| *** Shijia has joined #senlin | 11:04 | |
| *** Shijia has quit IRC | 11:08 | |
| *** Shijia has joined #senlin | 11:15 | |
| *** Shijia has quit IRC | 11:58 | |
| *** Shijia has joined #senlin | 11:59 | |
| *** Shijia has quit IRC | 12:03 | |
| *** Shijia has joined #senlin | 12:07 | |
| *** Shijia has quit IRC | 13:06 | |
| *** Shijia has joined #senlin | 13:07 | |
| *** Shijia has quit IRC | 13:16 | |
| *** Shijia has joined #senlin | 13:37 | |
| *** Shijia has quit IRC | 13:47 | |
| *** Shijia has joined #senlin | 13:47 | |
| *** Shijia has quit IRC | 13:52 | |
| *** Qiming has joined #senlin | 14:02 | |
| *** Shijia has joined #senlin | 14:17 | |
| *** Shijia has quit IRC | 14:22 | |
| *** Qiming has quit IRC | 14:24 | |
| *** Shijia has joined #senlin | 15:12 | |
| *** Shijia has quit IRC | 15:16 | |
| *** Shijia has joined #senlin | 15:25 | |
| *** Shijia has quit IRC | 15:36 | |
| *** Shijia has joined #senlin | 15:36 | |
| *** Shijia has quit IRC | 15:40 | |
| *** Qiming has joined #senlin | 15:48 | |
| *** Qiming has quit IRC | 16:18 | |
| *** Shijia has joined #senlin | 16:31 | |
| *** Shijia has quit IRC | 16:36 | |
| *** Shijia has joined #senlin | 16:51 | |
| *** Shijia has quit IRC | 16:56 | |
| *** Shijia has joined #senlin | 16:57 | |
| *** Shijia has quit IRC | 17:01 | |
| *** Shijia has joined #senlin | 17:21 | |
| *** Shijia has quit IRC | 17:26 | |
| *** Shijia has joined #senlin | 17:37 | |
| *** Shijia has quit IRC | 17:41 | |
| *** Shijia has joined #senlin | 18:07 | |
| *** Shijia has quit IRC | 18:12 | |
| *** Shijia has joined #senlin | 18:16 | |
| *** Shijia has quit IRC | 18:21 | |
| *** Shijia has joined #senlin | 18:22 | |
| *** Shijia has quit IRC | 18:26 | |
| *** Shijia has joined #senlin | 18:31 | |
| *** Shijia has quit IRC | 18:36 | |
| *** Shijia has joined #senlin | 18:46 | |
| *** Shijia has quit IRC | 18:51 | |
| *** Shijia has joined #senlin | 19:07 | |
| *** Shijia has quit IRC | 19:11 | |
| *** Shijia has joined #senlin | 19:37 | |
| *** Shijia has quit IRC | 19:42 | |
| *** Shijia has joined #senlin | 19:46 | |
| *** Shijia has quit IRC | 19:51 | |
| *** Shijia has joined #senlin | 20:07 | |
| *** Shijia has quit IRC | 20:11 | |
| *** Shijia has joined #senlin | 20:36 | |
| *** Shijia has quit IRC | 20:40 | |
| *** Shijia has joined #senlin | 21:11 | |
| *** Shijia has quit IRC | 21:16 | |
| *** Shijia has joined #senlin | 21:31 | |
| *** Shijia has quit IRC | 21:36 | |
| *** jruano has joined #senlin | 21:56 | |
| *** Shijia has joined #senlin | 21:57 | |
| *** Shijia has quit IRC | 22:02 | |
| *** Shijia has joined #senlin | 22:32 | |
| *** Shijia has quit IRC | 22:36 | |
| *** Shijia has joined #senlin | 23:02 | |
| *** Shijia has quit IRC | 23:06 | |
| *** Shijia has joined #senlin | 23:27 | |
| *** Shijia has quit IRC | 23:31 | |
| *** Shijia has joined #senlin | 23:32 | |
| *** jruano has quit IRC | 23:37 | |
| *** Shijia has quit IRC | 23:41 | |
| *** jruano has joined #senlin | 23:55 | |
| *** Shijia has joined #senlin | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!