| *** zzxwill has joined #senlin | 01:14 | |
| openstackgerrit | Merged openstack/senlin: Set API version request on request objects https://review.openstack.org/296407 | 01:23 |
|---|---|---|
| *** xuhaiwei has joined #senlin | 01:33 | |
| *** yanyanhu has joined #senlin | 01:43 | |
| *** Qiming has joined #senlin | 01:52 | |
| *** elynn has joined #senlin | 02:11 | |
| *** zzxwill has quit IRC | 02:14 | |
| *** zzxwill has joined #senlin | 02:16 | |
| elynn | morning senlin | 02:26 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Add global API version check for micronversion https://review.openstack.org/298081 | 02:37 |
| yanyanhu | morning, elynn | 02:52 |
| openstackgerrit | Merged openstack/senlin: Make version resource a subclass of Controller https://review.openstack.org/296492 | 02:54 |
| elynn | Hi Qiming | 03:03 |
| Qiming | yes, elynn? | 03:03 |
| elynn | Saw your comments on tempest plugin patch, | 03:03 |
| elynn | do you mean we will expose action id in cluster/node create API? | 03:04 |
| elynn | in the future? | 03:04 |
| Qiming | I am suggesting we do this in a consistent way | 03:04 |
| Qiming | either explicitly exposing or implicitly exposing it in body should be okay | 03:05 |
| Qiming | my concern is about the inconsistency | 03:05 |
| Qiming | sometimes we are throwing away all other fields, leaving only the action id | 03:06 |
| Qiming | sometimes, we don't | 03:06 |
| elynn | ok, so actually we should change location pointed to action id in creation if it returns 202, right? | 03:06 |
| Qiming | yep | 03:07 |
| Qiming | we'd better strive to make the caller's life simpler | 03:07 |
| Qiming | for example, when you are invoking a (senlin) api, you will first check if the status code is 202 | 03:08 |
| Qiming | and you will check the 'location' field in the body if the code is 202 | 03:08 |
| elynn | yes, I think you are right, let me modify the codes in senlin-engine first. | 03:09 |
| Qiming | I don't think we have anything wrong in senlin-engine | 03:09 |
| Qiming | we are already returning whatever needed from senlin-api | 03:09 |
| Qiming | the point is about how would you check it from sdk and from client side | 03:10 |
| elynn | now the location pointed to cluster resource itself I think. | 03:10 |
| Qiming | which api are you talking about? | 03:10 |
| elynn | let me check | 03:10 |
| elynn | https://github.com/openstack/senlin/blob/master/senlin/api/openstack/v1/clusters.py#L180 | 03:11 |
| elynn | https://github.com/openstack/senlin/blob/master/senlin/api/openstack/v1/nodes.py#L120 | 03:12 |
| elynn | These two . | 03:12 |
| Qiming | okay, that one is a mistack, it should reference the action id | 03:12 |
| elynn | yes, What I mean is to change these two places :) | 03:13 |
| *** zzxwill has quit IRC | 03:13 | |
| Qiming | ok | 03:13 |
| elynn | I will prepare a patch for it :) then modify the tempest patch. | 03:14 |
| Qiming | great | 03:14 |
| *** zzxwill has joined #senlin | 03:14 | |
| elynn | And I think I also encounter db problems when running tempest tests, it still there... | 03:15 |
| Qiming | db problems? | 03:17 |
| elynn | something like transaction locked | 03:18 |
| Qiming | well, file a bug if you can reproduce it | 03:18 |
| elynn | Can't find the logs right now, will paste it later if it happens. | 03:19 |
| *** zzxwill has quit IRC | 03:52 | |
| *** yuanying has quit IRC | 03:53 | |
| *** elynn has quit IRC | 04:07 | |
| *** zzxwill has joined #senlin | 04:22 | |
| *** zzxwill has quit IRC | 04:40 | |
| *** elynn has joined #senlin | 04:50 | |
| *** elynn has quit IRC | 04:54 | |
| *** elynn has joined #senlin | 04:54 | |
| *** yuanying has joined #senlin | 04:56 | |
| *** zzxwill has joined #senlin | 05:09 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Rework senlin docs based on CLI changes https://review.openstack.org/297552 | 05:16 |
| *** yuanying has quit IRC | 05:40 | |
| *** elynn has quit IRC | 05:41 | |
| *** yuanying has joined #senlin | 05:48 | |
| *** elynn has joined #senlin | 05:49 | |
| *** elynn has quit IRC | 05:54 | |
| *** elynn has joined #senlin | 05:54 | |
| *** elynn has quit IRC | 06:09 | |
| *** elynn has joined #senlin | 06:12 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Add API version header in responses https://review.openstack.org/298108 | 06:12 |
| *** elynn has quit IRC | 06:16 | |
| *** elynn has joined #senlin | 06:16 | |
| openstackgerrit | Merged openstack/senlin: Add versioned method support https://review.openstack.org/297639 | 06:20 |
| yanyanhu | hi, elynn, around? | 06:22 |
| *** elynn has quit IRC | 06:25 | |
| yanyanhu | hi, Qiming, around? | 06:31 |
| yanyanhu | Is defining sdk profile the only way to specify parameters like region, service name when creating a senlinclient object? | 06:32 |
| *** elynn has joined #senlin | 06:38 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Add API version header in responses https://review.openstack.org/298108 | 06:41 |
| *** elynn has quit IRC | 06:42 | |
| *** elynn has joined #senlin | 06:43 | |
| yanyanhu | hi, elynn, around? | 06:48 |
| *** gongysh has joined #senlin | 06:53 | |
| *** elynn has quit IRC | 06:53 | |
| *** Liuqing has joined #senlin | 06:59 | |
| *** elynn has joined #senlin | 07:00 | |
| *** elynn has quit IRC | 07:05 | |
| *** elynn has joined #senlin | 07:06 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Refactor API version range support https://review.openstack.org/298129 | 07:08 |
| Qiming | lixinhui, there? | 07:10 |
| lixinhui | Yes, Qiming | 07:11 |
| Qiming | got a question about lbaas | 07:12 |
| Qiming | is it a hard requirement for us to use octavia if we use lbaas v2? | 07:12 |
| lixinhui | no it is not | 07:13 |
| Qiming | okay | 07:13 |
| Qiming | it is optional, right? | 07:13 |
| lixinhui | yes, only one of the backend | 07:13 |
| lixinhui | of lbaas | 07:13 |
| Qiming | okay | 07:13 |
| Qiming | is it fair to say that our preliminary experience with octavia is not good? | 07:14 |
| lixinhui | no | 07:14 |
| lixinhui | I wanna give fair feedback | 07:15 |
| lixinhui | just the bugs are still there | 07:15 |
| Qiming | it is just an learning curve and some common mistakes made by new users? | 07:15 |
| lixinhui | I think its configurations are quiet easy by following their sample | 07:17 |
| lixinhui | but diagnosis is hard | 07:17 |
| Qiming | or .. we can say, we spent about two weeks on integrating octavia into the whole picture but we gave up eventually because there are some bugs to be fixed | 07:17 |
| lixinhui | when suffering from network problem | 07:17 |
| lixinhui | yes | 07:17 |
| lixinhui | that is the fact | 07:17 |
| Qiming | okay | 07:17 |
| Qiming | don't want to exaggerate, just share our experience with others | 07:18 |
| lixinhui | Okay, I see | 07:18 |
| Qiming | I do believe creating an "invisible" vm and doing a lot of things "secretly" will make it a painful process to debug | 07:20 |
| lixinhui | indeed it is | 07:20 |
| Qiming | okay | 07:20 |
| Qiming | thanks | 07:20 |
| lixinhui | np | 07:21 |
| yanyanhu | hi, Qiming, around? | 07:22 |
| yanyanhu | I'm now trying to add senlinclient plugin for rally | 07:23 |
| yanyanhu | but I found the only way to specify region_name when creating senlinclient is to creating a sdk profile to wrap the region_name and pass it | 07:24 |
| Qiming | okay? | 07:24 |
| yanyanhu | that means we need to add sdk requirement to rally which could not be appropriate | 07:25 |
| yanyanhu | http://git.openstack.org/cgit/openstack/python-senlinclient/tree/senlinclient/v1/client.py#n18 | 07:25 |
| yanyanhu | this line | 07:25 |
| yanyanhu | I'm thinking maybe we can handle the region_name kwargs and create sdk profile in senlinclient | 07:26 |
| Qiming | yes | 07:26 |
| yanyanhu | to hide this detail from enduser or other services who invoke senlinclient | 07:26 |
| Qiming | need to check with elynn | 07:27 |
| Qiming | how is 'preferences' parameter used in Heat | 07:27 |
| yanyanhu | yes, heat plugin has the same issue | 07:27 |
| Qiming | we got to be consistent about the model | 07:27 |
| yanyanhu | actually, it isn't used now... | 07:27 |
| yanyanhu | I think | 07:27 |
| yanyanhu | in heat client plugin for senlin | 07:27 |
| yanyanhu | let me check it again | 07:27 |
| Qiming | okay, better double check | 07:28 |
| yanyanhu | http://git.openstack.org/cgit/openstack/heat/tree/heat/engine/clients/os/senlin.py#n39 | 07:28 |
| yanyanhu | looks so | 07:28 |
| yanyanhu | so it looks like a common requirement | 07:29 |
| Qiming | okay | 07:29 |
| yanyanhu | will talk with elynn to see how to handle it | 07:29 |
| yanyanhu | to meet requirement from both heat and rally side | 07:29 |
| Qiming | yes | 07:29 |
| Qiming | senlinclient should check if preferences is specified | 07:30 |
| Qiming | and it should ALSO check if 'region_name' is provided in the kwargs | 07:30 |
| yanyanhu | yes | 07:31 |
| yanyanhu | will open a bug for it | 07:31 |
| yanyanhu | does this count new feature or just a bug? | 07:32 |
| Qiming | file a bug would be okay | 07:32 |
| yanyanhu | ok | 07:33 |
| *** EldonZhao has joined #senlin | 07:34 | |
| elynn | Hi, i'm back, yanyanhu so you suspect that senlin plugin can 't work in multi region env? | 07:39 |
| yanyanhu | it can, just not convenient for enduser | 07:39 |
| yanyanhu | to define the region_name parameter inside sdk profile | 07:39 |
| EldonZhao | I got the same question about muti-region scene. | 07:40 |
| EldonZhao | my openstack server has two regions | 07:40 |
| yanyanhu | as Qiming said, we should handle both cases of profile or kwarg region_name is provided | 07:40 |
| yanyanhu | when creating senlinclient object | 07:40 |
| yanyanhu | profile/preference | 07:41 |
| elynn | let me check.. | 07:41 |
| elynn | Is there any docs about how sdk works with region_name? | 07:42 |
| yanyanhu | http://git.openstack.org/cgit/openstack/python-senlinclient/tree/senlinclient/v1/client.py#n18 | 07:42 |
| yanyanhu | here | 07:42 |
| yanyanhu | you can refer to the docstring of Connection class | 07:42 |
| yanyanhu | I guess no detail document there... | 07:42 |
| yanyanhu | let me find the code for u | 07:42 |
| yanyanhu | http://git.openstack.org/cgit/openstack/python-openstacksdk/tree/openstack/connection.py#n148 | 07:43 |
| yanyanhu | here | 07:43 |
| elynn | hmm... seems it's bug in senlin plugin, I haven't setup the client correctly. | 07:45 |
| elynn | Then senlin dashboard has the same problem... | 07:46 |
| yanyanhu | elynn, I think the current implementation works | 07:46 |
| yanyanhu | just could induce extra dependency for sdk if user to define region_name | 07:47 |
| yanyanhu | need to define | 07:47 |
| yanyanhu | will try to propose a patch for it. | 07:47 |
| elynn | Why do we need t o depend on sdk? | 07:48 |
| elynn | you mean heat need to depend on idk? | 07:48 |
| elynn | sdk | 07:48 |
| yanyanhu | if you want to specify region_name when creating senlinclient, you have to define a profile object to wrap it | 07:48 |
| Qiming | either you explicitly pass a region_name in kwargs (thus we need to fix senlinclient) | 07:48 |
| Qiming | or you have to construct a openstack.preferences object and pass it to the senlinclient.v1.client constructor | 07:49 |
| yanyanhu | exactly | 07:49 |
| Qiming | I believe the former one is more user friendly | 07:49 |
| elynn | From the string in sdk, I think we just need to pass a parameter 'profile' include region_name in it is ok? | 07:49 |
| yanyanhu | I think adding a handling here will work | 07:49 |
| yanyanhu | http://git.openstack.org/cgit/openstack/python-senlinclient/tree/senlinclient/v1/client.py#n18 | 07:49 |
| Qiming | opentack profile is not a dict, it is an object | 07:49 |
| Qiming | that is an arguable design | 07:50 |
| elynn | oh... | 07:50 |
| elynn | Then I think yanyanhu's solution is workable, maybe we can expose region_name as parameter and wrap it in senlinclient. | 07:51 |
| yanyanhu | hi, elynn, no need to add extra parameter I think, just handle region_name kwarg if user provides it | 07:53 |
| elynn | ok | 07:54 |
| yanyanhu | not a difficult fix : ) | 07:54 |
| yanyanhu | just for compatibility | 07:54 |
| elynn | sounds ok for me | 07:55 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Add API version header in responses https://review.openstack.org/298108 | 08:08 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Refactor API version range support https://review.openstack.org/298129 | 08:13 |
| *** yuanying has quit IRC | 08:31 | |
| *** yuanying has joined #senlin | 08:31 | |
| *** yuanying has quit IRC | 08:32 | |
| *** yuanying has joined #senlin | 08:39 | |
| *** EldonZhao has quit IRC | 08:40 | |
| *** zzxwill has quit IRC | 08:45 | |
| yanyanhu | network is broken again, ft | 08:55 |
| yanyanhu | can't do any test... | 08:55 |
| *** zzxwill has joined #senlin | 08:56 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Rework senlin docs based on CLI changes https://review.openstack.org/297552 | 08:56 |
| *** zzxwill has quit IRC | 08:57 | |
| openstackgerrit | Yanyan Hu proposed openstack/python-senlinclient: Support more parameters for senlinclient creation https://review.openstack.org/298147 | 08:59 |
| openstackgerrit | Yanyan Hu proposed openstack/python-senlinclient: Support more parameters for senlinclient creation https://review.openstack.org/298147 | 09:03 |
| *** zzxwill has joined #senlin | 09:05 | |
| *** EldonZhao has joined #senlin | 09:05 | |
| openstackgerrit | Merged openstack/senlin: Add global API version check for micronversion https://review.openstack.org/298081 | 09:05 |
| *** EldonZhao has left #senlin | 09:05 | |
| *** zzxwill has quit IRC | 09:06 | |
| *** EldonZhao has joined #senlin | 09:06 | |
| *** yuanying has quit IRC | 09:16 | |
| *** Qiming has quit IRC | 09:17 | |
| *** Qiming has joined #senlin | 09:17 | |
| *** shu-mutou has joined #senlin | 09:18 | |
| *** EldonZhao has quit IRC | 09:34 | |
| *** EldonZhao has joined #senlin | 09:35 | |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Modify location to point to action https://review.openstack.org/298175 | 09:36 |
| Qiming | elynn, I'm +2 ing your patch above | 09:44 |
| Qiming | but please understand this is because we do have a flexibility here: http://git.openstack.org/cgit/openstack/api-wg/tree/guidelines/http.rst#n105 | 09:44 |
| Qiming | when asynchronously creating a resource, we can point the location to either the resource or the "status resource" ('action' in our case) | 09:45 |
| elynn | ok | 09:48 |
| elynn | so you mean my commit message misguide others? | 09:48 |
| Qiming | elynn, please also propose a patch to api doc here: | 09:49 |
| Qiming | http://git.openstack.org/cgit/openstack/api-site/tree/api-ref/src/wadls/clustering-api/src/v1/wadl/clustering-api.wadl#n212 | 09:49 |
| Qiming | line 212 should be changed to "&actionLocationParameter;" | 09:49 |
| elynn | oh, yes, will propose patch for api doc. | 09:49 |
| openstackgerrit | Yanyan Hu proposed openstack/python-senlinclient: Support more parameters for senlinclient creation https://review.openstack.org/298147 | 09:55 |
| openstackgerrit | Yanyan Hu proposed openstack/python-senlinclient: Support more parameters for senlinclient creation https://review.openstack.org/298147 | 10:02 |
| *** yanyanhu has quit IRC | 10:14 | |
| *** elynn has quit IRC | 10:18 | |
| *** zzxwill has joined #senlin | 10:19 | |
| *** Qiming has quit IRC | 10:24 | |
| *** Liuqing has quit IRC | 10:25 | |
| *** shu-mutou is now known as shu-mutou-AFK | 10:40 | |
| *** Liuqing has joined #senlin | 10:51 | |
| *** Liuqing_ has joined #senlin | 11:15 | |
| *** Liuqing has quit IRC | 11:16 | |
| *** Qiming has joined #senlin | 11:30 | |
| *** zzxwill has quit IRC | 11:47 | |
| *** shu-mutou-AFK has quit IRC | 11:56 | |
| *** Liuqing_ has quit IRC | 12:02 | |
| *** zzxwill has joined #senlin | 12:17 | |
| *** zzxwill has quit IRC | 12:29 | |
| *** EldonZhao has quit IRC | 12:34 | |
| *** zzxwill has joined #senlin | 12:42 | |
| openstackgerrit | Merged openstack/python-senlinclient: Support more parameters for senlinclient creation https://review.openstack.org/298147 | 13:08 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Modify location to point to action https://review.openstack.org/298175 | 13:19 |
| *** zzxwill has quit IRC | 13:36 | |
| *** openstackgerrit has quit IRC | 13:48 | |
| *** openstackgerrit has joined #senlin | 13:49 | |
| *** zzxwill has joined #senlin | 13:59 | |
| openstackgerrit | Merged openstack/senlin: Add API version header in responses https://review.openstack.org/298108 | 14:48 |
| -openstackstatus- NOTICE: Gerrit is restarting on review.openstack.org in an attempt to address an issue reading an object from the ec2-api repository | 15:22 | |
| openstackgerrit | Merged openstack/senlin: Rework senlin docs based on CLI changes https://review.openstack.org/297552 | 15:25 |
| *** Qiming has quit IRC | 15:29 | |
| *** zzxwill has quit IRC | 16:13 | |
| *** cschulz has quit IRC | 21:14 | |
| *** cschulz has joined #senlin | 21:23 | |
| *** yuanying has joined #senlin | 23:11 | |
| *** Qiming has joined #senlin | 23:29 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!