| *** wanghao has joined #openstack-mogan | 00:37 | |
| *** little has joined #openstack-mogan | 01:05 | |
| *** little has quit IRC | 01:20 | |
| *** RuiChen has quit IRC | 01:23 | |
| *** RuiChen has joined #openstack-mogan | 01:26 | |
| zhenguo | morning mogan! | 02:04 |
|---|---|---|
| zhenguo | shaohe_feng, liusheng: are you around? | 02:06 |
| *** little has joined #openstack-mogan | 02:11 | |
| *** liusheng has joined #openstack-mogan | 02:32 | |
| *** yuntongjin has joined #openstack-mogan | 02:32 | |
| *** yuntongjin has quit IRC | 02:38 | |
| *** yuntongjin has joined #openstack-mogan | 02:40 | |
| *** RuiChen has quit IRC | 02:42 | |
| *** RuiChen has joined #openstack-mogan | 02:42 | |
| openstackgerrit | liusheng proposed openstack/nimble: Don't need to wait fixed time after deletion in tempest https://review.openstack.org/417724 | 02:49 |
| *** RuiChen has quit IRC | 02:58 | |
| *** RuiChen has joined #openstack-mogan | 03:02 | |
| *** kevinz has joined #openstack-mogan | 03:13 | |
| *** Jeffrey4l_ has quit IRC | 03:18 | |
| *** yuntongjin has quit IRC | 03:26 | |
| openstackgerrit | liusheng proposed openstack/nimble: Don't need to wait fixed time after deletion in tempest https://review.openstack.org/417724 | 03:50 |
| openstackgerrit | liusheng proposed openstack/nimble: Versioned notification: Add instance action notification base https://review.openstack.org/417335 | 04:07 |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Move power target check before fetching instance from db https://review.openstack.org/417740 | 04:25 |
| *** yuntongjin has joined #openstack-mogan | 04:50 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: [WIP] Add rebuild instance support https://review.openstack.org/417746 | 05:13 |
| *** wanghao_ has joined #openstack-mogan | 06:05 | |
| *** wanghao has quit IRC | 06:07 | |
| *** little has quit IRC | 06:17 | |
| *** little has joined #openstack-mogan | 06:17 | |
| openstackgerrit | liusheng proposed openstack/nimble: Versioned notification add support for instance.create.start https://review.openstack.org/417755 | 06:34 |
| *** zhangjl has joined #openstack-mogan | 06:44 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: [WIP] Support Floating IPs association/dissociation https://review.openstack.org/417760 | 06:50 |
| zhenguo | shaohe_feng: are you around？ | 06:50 |
| zhenguo | hi all, before doing further work, I would like you all to help review https://review.openstack.org/417746 and https://review.openstack.org/417760 to see our API defination which affects the following APIs. | 06:53 |
| zhenguo | as we don't follow nova's way to define REST APIs like instances/instance_ident/action, hope to get suggestions from all of you, thanks! | 06:55 |
| * zhenguo brb | 06:56 | |
| *** little has quit IRC | 07:04 | |
| *** little has joined #openstack-mogan | 07:04 | |
| zhangjl | hey, all | 07:44 |
| zhangjl | my devstack cannot work | 07:44 |
| zhangjl | /opt/stack/mogan/devstack/plugin.sh: line 106: /usr/local/bin/mogan-dbsync: No such file or directory | 07:44 |
| shaohe_feng | zhangjl: yes. | 07:44 |
| shaohe_feng | zhenguo: yes | 07:45 |
| shaohe_feng | zhenguo: we need to support rootwrap in mogan. | 07:45 |
| shaohe_feng | zhenguo: I have seen, most project support rootwrap | 07:46 |
| zhangjl | agree | 07:46 |
| shaohe_feng | zhenguo: for we will not allow a user can do everything by "sudo" | 07:46 |
| zhenguo | zhangjl, shaohe_feng: hi | 07:47 |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Remove target_power_state https://review.openstack.org/417769 | 07:47 |
| zhangjl | zhenguo:hi | 07:47 |
| zhenguo | zhangjl: is this your first time to rebuild the devastack env after renaming? | 07:48 |
| zhangjl | yes | 07:48 |
| zhenguo | zhangjl: it works well here, and we have a tempest job to build it for every patch | 07:48 |
| zhenguo | zhangjl: so maybe you need to deep clean the env and rebuild again | 07:48 |
| zhangjl | zhenguo:so strange. ok, let me try | 07:49 |
| zhenguo | shaohe_feng: yes, agree | 07:49 |
| zhenguo | zhangjl: yes, it's weird, I just clean up the python lib and rebuild again then it's ok | 07:49 |
| zhangjl | zhenguo:ok, i see | 07:49 |
| zhenguo | shaohe_feng: hi, I would like to introduce a new provision_state/task_state to instance to tack our task | 07:51 |
| zhenguo | s/tack/track | 07:52 |
| zhenguo | shaohe_feng: and will save power_state to db instead of issuing a request to ironic every time, although this need db sync with ironic | 07:53 |
| zhenguo | shaohe_feng: I find a power_state_sync periodic task also have other benefits, like when the node power goes down for some reason, it will power it gain to keep consistent with the instance | 07:55 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server start' to set a server power on https://review.openstack.org/417772 | 08:03 |
| zhenguo | hi every task owner, please remember to set status on the etherpad https://etherpad.openstack.org/p/nimble-task , or please feel to ping me if needed, thanks! | 08:06 |
| zhenguo | s/set/update | 08:06 |
| zhenguo | ls | 08:06 |
| shaohe_feng | zhenguo: do you means the power_state_sync periodic task will also power on/off the node? | 08:14 |
| zhenguo | shaohe_feng: yes | 08:14 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server start' to set a server power on https://review.openstack.org/417772 | 08:15 |
| zhenguo | shaohe_feng: like when the instance is in power on, you manually power off the node, it will power it again to be consistent with instance | 08:15 |
| zhenguo | shaohe_feng: and I think saving a power state to db seems resonable, I don't quite like the current way to get it from ironic every time | 08:16 |
| *** openstackgerrit has quit IRC | 08:18 | |
| zhenguo | shaohe_feng, liusheng: for API defination, we will try to follow the REST instruction, instance will have two sub resources, states and networks. | 08:21 |
| zhenguo | all networks related actions will under networks like instances/instance_uuid/networks/floatingip, instances/instance_uuid/networks/ports, and GET /instances/instance_uuid/networks will return all network info | 08:22 |
| zhenguo | all states related actions will under states like instances/instance_uuid/power which we currently do | 08:23 |
| *** openstackgerrit has joined #openstack-mogan | 08:23 | |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server stop' to set a server power off https://review.openstack.org/417778 | 08:23 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server reboot' to set a server power reboot https://review.openstack.org/417779 | 08:23 |
| zhenguo | and a new provision action added like instances/instance_uuid/provision, then rebuild/rescure and all other instance related operations will do like this | 08:24 |
| zhenguo | and GET instances/instance_uuid/states will retrun {'power_state': xx, 'task_state'}, task_state will be introduced later | 08:24 |
| zhenguo | hi all, wdyt? do you think nova's instances/instance_uuid/action pattern is better or our way is ok? | 08:26 |
| liusheng | zhenguo: so if we only GET instances/{instance-uuid}, it won't return these fields ? | 08:27 |
| zhenguo | liusheng: no, it will retrun needed fields including power_state and network_info | 08:27 |
| liusheng | zhenguo: I mean return the power_state, network_info | 08:27 |
| liusheng | zhenguo: yes | 08:27 |
| liusheng | zhenguo: I am afraid that will brings many APIs | 08:28 |
| zhenguo | liusheng: just two more, GET states/networks | 08:29 |
| zhenguo | liusheng: but if users find it's not useful, we can not provide them | 08:29 |
| liusheng | zhenguo: only GET ? or PUT ? | 08:29 |
| zhenguo | liusheng: GET | 08:30 |
| zhenguo | liusheng: states/power and states/provision will be PUT | 08:30 |
| zhenguo | liusheng: you need to add a json field, so not that APIs | 08:30 |
| zhenguo | not that many APIs I mean | 08:31 |
| zhenguo | liusheng: nova only has one API for management instances/instance_uuid/action, hah | 08:31 |
| liusheng | zhenguo: lol, but seems Nova don't have a specific API to query a server's power state | 08:33 |
| zhenguo | liusheng: yes, I rememeber at the begining of this mogan, shaohe_feng do many reserches about REST API defination, finally he said nova's API is bad | 08:33 |
| liusheng | zhenguo: hah | 08:34 |
| liusheng | zhenguo: we need a expert | 08:34 |
| zhenguo | liusheng: hah, yes | 08:34 |
| zhenguo | liusheng: in fact, the best API defination should be that when reading the verb and URL you should know what you will do with the resource | 08:36 |
| zhenguo | liusheng: like POST/PUT/DELETE /instances | 08:37 |
| liusheng | zhenguo: eys | 08:37 |
| liusheng | zhenguo: yes | 08:37 |
| shaohe_feng | liusheng: nova one api can do many thing. | 08:40 |
| zhenguo | shaohe_feng: hah | 08:41 |
| shaohe_feng | liusheng: the redhat's API define follow API pattern. | 08:42 |
| zhenguo | liusheng: just FYI, I noticed Ironic also add versioned object notification suport recently, maybe we can leverage some notifications from Ironic, like we can catch some notifications from ironic, not sure if it is possible | 08:43 |
| shaohe_feng | liusheng: nova is rackspace pattern. | 08:44 |
| shaohe_feng | liusheng: redhat's API follow RESTFUL pattern. | 08:44 |
| shaohe_feng | liusheng: not sure we should compatible with nova | 08:45 |
| liusheng | zhenguo: hmm, we focus on different things | 08:45 |
| zhenguo | liusheng: ok | 08:45 |
| liusheng | shaohe_feng: thanks for the info! | 08:45 |
| shaohe_feng | zhenguo: liusheng: such as the up layer, for example, will effect heat with different API style? | 08:47 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server reboot' to set a server power reboot https://review.openstack.org/417779 | 08:48 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server stop' to set a server power off https://review.openstack.org/417778 | 08:48 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refator: use 'server start' to set a server power on https://review.openstack.org/417772 | 08:48 |
| zhenguo | shaohe_feng: seems not, different API do different job is better, in heat we will not share the same thing with nova, so will not impact it | 08:49 |
| liusheng | shaohe_feng: you mean mention this with Heat core ? | 08:49 |
| shaohe_feng | zhenguo: liusheng: Good. | 08:52 |
| liusheng | zhenguo, shaohe_feng heat calls clients of other projects | 08:53 |
| zhenguo | shaohe_feng: hi, If you don't have objection for this https://review.openstack.org/#/c/417769/ , I will merge it directly :P | 08:58 |
| shaohe_feng | zhenguo: let me check it first. | 09:01 |
| zhenguo | shaohe_feng: ok | 09:01 |
| shaohe_feng | zhenguo: It is reasonable to remove target_power_state. it is not a native attr of node states | 09:04 |
| zhenguo | shaohe_feng: hah, thanks, will introduce a task_state later | 09:04 |
| * zhenguo brb | 09:06 | |
| *** yuntongjin has quit IRC | 09:20 | |
| *** wanghao_ has quit IRC | 09:27 | |
| *** little has quit IRC | 09:28 | |
| zhenguo | hi all, please feel free to create/update the BP https://blueprints.launchpad.net/mogan you are working on, seems the owner in task tracking etherpad is inconsistent with launchpad. | 09:28 |
| *** little has joined #openstack-mogan | 09:28 | |
| *** little has quit IRC | 09:49 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Run sync_node_resources periodic task immediately https://review.openstack.org/417803 | 09:51 |
| openstackgerrit | liusheng proposed openstack/nimble: Remove dict compatibility or versioned object https://review.openstack.org/417808 | 10:10 |
| *** kevinz has quit IRC | 10:16 | |
| *** zhangjl has quit IRC | 10:16 | |
| shaohe_feng | zhenguo: liusheng: how do I enable openstack cli autocomplete | 10:20 |
| liusheng | shaohe_feng: type "openstack" and into interactive modelm, does it work ? | 10:34 |
| *** wanghao has joined #openstack-mogan | 10:46 | |
| zhenguo | shaohe_feng: wrt https://review.openstack.org/#/c/417308/ , I get your comments, you mean we don't need the looping call to wait the power transition finished? | 11:23 |
| *** wanghao has quit IRC | 11:32 | |
| *** kevinz has joined #openstack-mogan | 12:16 | |
| *** kevinz has quit IRC | 13:49 | |
| *** kevinz has joined #openstack-mogan | 13:50 | |
| *** kevinz has quit IRC | 14:15 | |
| *** Jeffrey4l_ has joined #openstack-mogan | 15:35 | |
| openstackgerrit | Merged openstack/python-nimbleclient: Enable tests coverage report when run tox -ecover https://review.openstack.org/416393 | 15:58 |
| *** Jeffrey4l_ has quit IRC | 16:22 | |
| *** Jeffrey4l_ has joined #openstack-mogan | 16:24 | |
| *** l4yerffeJ has joined #openstack-mogan | 16:24 | |
| *** Jeffrey4l__ has joined #openstack-mogan | 16:35 | |
| *** l4yerffeJ has quit IRC | 16:38 | |
| *** Jeffrey4l_ has quit IRC | 16:39 | |
| *** l4yerffeJ has joined #openstack-mogan | 16:43 | |
| *** l4yerffeJ has quit IRC | 16:44 | |
| *** Jeffrey4l__ has quit IRC | 16:51 | |
| openstackgerrit | Merged openstack/python-nimbleclient: The --nic parameter should be required https://review.openstack.org/417259 | 17:16 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refator: use 'server start' to set a server power on https://review.openstack.org/417772 | 17:27 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refator: use 'server stop' to set a server power off https://review.openstack.org/417778 | 17:29 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refator: use 'server reboot' to set a server power reboot https://review.openstack.org/417779 | 17:29 |
| openstackgerrit | Merged openstack/python-nimbleclient: Remove the --long parameter when listing flavors https://review.openstack.org/416404 | 17:33 |
| *** liusheng has quit IRC | 17:40 | |
| *** liusheng has joined #openstack-mogan | 17:41 | |
| *** liusheng has quit IRC | 17:47 | |
| *** liusheng has joined #openstack-mogan | 17:47 | |
| *** Jeffrey4l_ has joined #openstack-mogan | 18:23 | |
| *** Jeffrey4l__ has joined #openstack-mogan | 21:34 | |
| *** Jeffrey4l_ has quit IRC | 21:35 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!