*** catintheroof has quit IRC | 00:29 | |
*** catintheroof has joined #openstack-tricircle | 00:30 | |
*** catintheroof has quit IRC | 00:30 | |
*** dongfeng has joined #openstack-tricircle | 00:32 | |
*** catintheroof has joined #openstack-tricircle | 00:42 | |
*** Yipei has joined #openstack-tricircle | 01:20 | |
*** gongysh has joined #openstack-tricircle | 02:16 | |
*** joehuang has joined #openstack-tricircle | 03:00 | |
dongfeng | hello, zhiyuan, for how to deal with the job delete and job redo, should them be written to the spec? | 03:43 |
---|---|---|
dongfeng | here is the simple description of the delete and redo work. the link is:http://paste.openstack.org/show/602149/ | 03:48 |
*** catintheroof has quit IRC | 03:57 | |
*** catintheroof has joined #openstack-tricircle | 03:58 | |
*** catintheroof has quit IRC | 03:59 | |
joehuang | to Dongfeng, it's good to add it to the doc. | 06:17 |
dongfeng | hello, joe, zhiyaun's suggestion: One suggestion, shall we change url from /v1.0/job to /v1.0/jobs? Since we are using plural form for pod API and routing API | 06:31 |
dongfeng | to joe, I will add the description of delete and redo work in doc. It seems that zhiyuan's ieda is good, I plan to change /v1.0/job to /v1.0/jobs. | 06:34 |
joehuang | yes | 06:57 |
dongfeng | helli, zhiyuan,one question: if the project_id is the one to creates and runs the job, it is not necessary to specify it when create job. | 06:57 |
dongfeng | the job creation code:http://paste.openstack.org/show/602186/ | 07:00 |
dongfeng | we directly extract project_id from context. | 07:00 |
joehuang | not from context, the job creation is mostly called by admin, so the project_id in context is admin project, but admin can create a job for specified project(non-admin), so need to provide it in the creation body | 07:12 |
dongfeng | if admin create a job for tenant A, then the project_id of the job is A' project_id? | 07:15 |
dongfeng | in this Admin API, project_id should be specified by user. but in other place, project_id is from the context, is it correct? | 07:19 |
joehuang | no, job is not visible for general user | 07:31 |
joehuang | Admin will create a job in case of something wrong, it's for maintainance purpose | 07:32 |
dongfeng | currently when create a job, only job type and resource_id is provided, so if project_id is added to job, where to provide project_id parameter? Many places should be changed. | 07:33 |
Yipei | @joehuang, have you received my message yesterday? | 07:36 |
Yipei | i failed creating lb via local neutron | 07:36 |
Yipei | mcreate failed (client error): The server could not comply with the request since it is either malformed or otherwise incorrect. | 07:38 |
joehuang | what's error in the log | 07:39 |
Yipei | http://paste.openstack.org/show/602044/ | 07:39 |
Yipei | line 83 (client error): The server could not comply with the request since it is either malformed or otherwise incorrect. | 07:40 |
Yipei | i compare the log of local neutron with that of central neutron in the link | 07:41 |
joehuang | Because local Neutron can not find the subnet by name, the local plugin can only query the subnet through id, if id is not found locally, try to get it from Central server | 07:42 |
Yipei | i see, so if i replace the name with id, it may work | 07:43 |
joehuang | currently network/port query from local Nova is supported, so not sure subnet query by ID will work or not. Function is there, you can have a try, and trace the code. If it doen | 07:46 |
joehuang | if it doesn't work, you may have to add some code | 07:46 |
Yipei | local Nova? | 07:47 |
Yipei | i tried, and it fails | 07:47 |
joehuang | I just discussed with ZHiyuan, | 07:54 |
joehuang | If there is no network( not found in local neutron), just query subnet, then query from Central neutron use this subnet id, and return the content to the caller in local Neutron plugin | 07:56 |
joehuang | you have to add some code | 07:56 |
Yipei | got it | 07:57 |
Yipei | i checked the databases, there do exist an entry of the queried subnet | 08:02 |
Yipei | and the get_subnet function is implemented in the local_plugin | 08:02 |
*** dongfeng has quit IRC | 08:05 | |
*** dongfeng1 has joined #openstack-tricircle | 08:05 | |
*** dongfeng1 is now known as dongfeng | 08:05 | |
dongfeng | if project_id is added to job, for Admin API, we can receive the parameter from user, for other job creator, I still don't know the source of the project_id. apart from the Admin API, the project_id of context isn't the project_id of job? | 08:43 |
*** dongfeng has quit IRC | 09:41 | |
*** dongfeng has joined #openstack-tricircle | 09:41 | |
*** joehuang has quit IRC | 09:47 | |
*** gongysh has quit IRC | 10:26 | |
*** gongysh has joined #openstack-tricircle | 11:41 | |
*** catintheroof has joined #openstack-tricircle | 12:34 | |
*** Yipei has quit IRC | 13:07 | |
*** gongysh has quit IRC | 13:32 | |
*** catintheroof has quit IRC | 15:00 | |
*** catintheroof has joined #openstack-tricircle | 15:01 | |
*** catintheroof has quit IRC | 17:07 | |
*** catintheroof has joined #openstack-tricircle | 17:08 | |
*** catintheroof has quit IRC | 22:23 | |
*** Kevin_Zheng has quit IRC | 22:53 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!