| *** zhangyang has quit IRC | 00:11 | |
| *** wanghao has joined #openstack-mogan | 00:43 | |
| *** little has joined #openstack-mogan | 00:53 | |
| *** yuntongjin has joined #openstack-mogan | 01:25 | |
| *** kevinz has joined #openstack-mogan | 01:28 | |
| *** zhangjl has joined #openstack-mogan | 01:47 | |
| zhenguo | morning mogan | 01:50 |
|---|---|---|
| zhenguo | Gerrit downtime on Thursday 2017-01-12 at 20:00 UTC, we will finish renaming then, lol | 01:51 |
| zhangjl | good news | 01:51 |
| *** zhangyang has joined #openstack-mogan | 02:07 | |
| * zhenguo away for lunch | 03:56 | |
| zhenguo | hi all, do you know what's the difference between state and status? | 06:14 |
| zhenguo | liusheng: hi, after a second this maybe it's better to introduce a state machine for instance state, leveraging check_instance_state like nova will be not that clear, and nova also plan to introduce sm, as we are new we can do that at beginning | 06:23 |
| zhenguo | second think | 06:23 |
| liusheng | zhenguo: yes, agree, not everything of Nova is reasonalbe | 06:24 |
| zhenguo | liusheng: only a state machine for instance sate, which can control states transitions. power_state and task_state just show the status | 06:24 |
| zhenguo | liusheng: hah | 06:24 |
| zhenguo | liusheng: I still a bit confused about status and state | 06:24 |
| liusheng | zhenguo: just need to ping Kevin_Zheng | 06:25 |
| liusheng | zhenguo: he is not in workplace now | 06:25 |
| zhenguo | liusheng: ok | 06:26 |
| zhenguo | liusheng: ironic has a provision_state, target_provision_state, power_state, target_power_state, just these | 06:27 |
| liusheng | zhenguo: so many.. | 06:27 |
| liusheng | zhenguo: lol | 06:27 |
| zhenguo | liusheng: indeed, hah | 06:27 |
| zhenguo | liusheng: we will have a power_state, task_state and state | 06:28 |
| zhenguo | liusheng: task_state will include the ing status of power_state and state like powering_on, spawning. | 06:29 |
| liusheng | zhenguo: I am thinking about how about merged tast_state and state ? | 06:29 |
| zhenguo | liusheng: you mean only have a state? | 06:30 |
| liusheng | zhenguo: state will includes middle state and fixed state | 06:30 |
| zhenguo | liusheng: yes | 06:30 |
| liusheng | zhenguo: yes, just a primitive thought | 06:31 |
| liusheng | zhenguo: the state can be active, stopped, booting, stopping | 06:31 |
| zhenguo | liusheng: seems ok, as I find some duplicate with task_state and state | 06:31 |
| Kevin_Zheng | whats up | 06:31 |
| zhenguo | hey | 06:31 |
| zhenguo | we are discussing about task_state and state | 06:32 |
| liusheng | Kevin_Zheng: we need a Nova expert now, lol | 06:32 |
| Kevin_Zheng | one sec | 06:32 |
| zhenguo | ok | 06:32 |
| zhenguo | liusheng: seems it's more clear to just using one state, hah | 06:36 |
| zhenguo | liusheng: we can use a state maching to control the states transition | 06:37 |
| Kevin_Zheng | http://git.openstack.org/cgit/openstack/nova/tree/nova/api/openstack/common.py#n46 | 06:37 |
| Kevin_Zheng | you can check this out | 06:37 |
| Kevin_Zheng | it is a mapping between vm_state+task_state and Status | 06:38 |
| liusheng | zhenguo: yes there isn't overlap between persistent state and processing state | 06:38 |
| zhenguo | too messy | 06:38 |
| zhenguo | Kevin_Zheng: we plan to just use one instance state, wdyt? | 06:39 |
| liusheng | zhenguo: may we need to designe a map firstly to show state transition | 06:40 |
| zhenguo | liusheng: yes, that could be done with the state machine defination | 06:40 |
| zhenguo | s/defination/definition | 06:41 |
| zhenguo | we don't need to show that specific states like nova task_state | 06:41 |
| liusheng | e.g. when a Nova instance is in 'building' state, the task state maybe in scheduling, networking, spawing. but seems building is easy to understood by end users. | 06:45 |
| Kevin_Zheng | I think it is ok, because for baremetal instances | 06:48 |
| Kevin_Zheng | power_state means status? | 06:48 |
| zhenguo | power_state will be power on or power off | 06:51 |
| zhenguo | liusheng: yes, and nova task_state always just present in 1 sec or less | 06:52 |
| liusheng | zhenguo: yes | 06:54 |
| zhenguo | and when building instance it's really rare to see all task states, so I doubt whether we need that, and we have a task flow now | 06:54 |
| zhenguo | task flow can be useful to track specific task states transition | 06:55 |
| liusheng | zhenguo: we need to design all the state transitio to see how many state/tast_state we will have | 06:58 |
| zhenguo | liusheng: ok, will draft a patch soon | 06:58 |
| liusheng | zhenguo: coool! | 06:59 |
| zhenguo | liusheng: within a state machine, hah | 06:59 |
| *** yuntongjin has quit IRC | 07:20 | |
| *** yuntongjin has joined #openstack-mogan | 07:20 | |
| *** kevinz has quit IRC | 07:40 | |
| *** kevinz has joined #openstack-mogan | 07:41 | |
| zhenguo | liusheng: after repo renamed, seems we need to upload all inprogress patches to the new repo | 07:46 |
| liusheng | zhenguo: ok, the gerrit won't handle the patches automatically ? | 07:47 |
| zhenguo | liusheng: not sure, seems they can, according to the email Existing reviews, project watches, etc, for these projects will all be carried over. | 07:48 |
| zhenguo | liusheng: they will do this rename tomorrow, hah | 07:49 |
| liusheng | zhenguo: so let's wait! | 07:49 |
| zhenguo | liusheng: hah | 07:49 |
| zhenguo | liusheng: powering on will go to active state and powering of will be stopped, right? | 07:51 |
| liusheng | zhenguo: seems not exactly, | 07:53 |
| zhenguo | liusheng: I find in nova active means vm is running | 07:54 |
| liusheng | zhenguo: I remember there is other state after instance end powering off | 07:56 |
| zhenguo | liusheng: I find a STOPPED state for vm | 07:57 |
| zhenguo | liusheng: maybe we should also introduce a maintenance state, which will block other state transiation | 07:58 |
| liusheng | zhenguo: you mean two set of states ? | 08:00 |
| zhenguo | liusheng: no, just one staet | 08:00 |
| liusheng | zhenguo: which one state ? | 08:00 |
| zhenguo | liusheng: just instance state, I mean it can be set to maintenance | 08:01 |
| liusheng | zhenguo: oh, I see, you mean like ironic | 08:01 |
| zhenguo | liusheng: yes, as baremtal is not like vm, sometimes we need to do some maintenance | 08:01 |
| *** yuntongjin has quit IRC | 08:02 | |
| little | In nova, the status is the combination of vm_state and task_state, you can read the _STATE_MAP defination in nova :\nova\api\openstack\common.py | 08:18 |
| little | zhenguo: | 08:19 |
| zhenguo | little: thanks, let me check | 08:19 |
| little | I think the status is readable for human | 08:21 |
| little | OK | 08:21 |
| zhenguo | little: you mean we should keep the current status field? | 08:21 |
| little | yes, I think we can learn from nova | 08:24 |
| zhenguo | little: seems OK | 08:24 |
| zhenguo | liusheng: so we don't need to replace status with state, wdyt? | 08:24 |
| zhenguo | as nova will transfer it to status in the response body | 08:25 |
| zhenguo | nova's state management seems too complex | 08:25 |
| little | yes, a bit | 08:25 |
| little | we can simple it | 08:26 |
| little | status and task_state | 08:26 |
| zhenguo | little: we plan to get rid of task_state | 08:26 |
| zhenguo | little: only leave one state/status field | 08:27 |
| little | what is the defference of state and status? | 08:27 |
| zhenguo | as status is a combination of vm_state and task_state, so we can just use a status | 08:27 |
| zhenguo | not sure the difference | 08:28 |
| zhenguo | but we can keep to use status | 08:28 |
| zhenguo | follow nova way | 08:28 |
| little | yes, we no need to use combination of vm_state and task_state, just use status | 08:29 |
| little | this is simple | 08:29 |
| zhenguo | yes, and it's clear | 08:29 |
| zhenguo | we will introduce a state machine to control the states transiation as well | 08:29 |
| little | Om | 08:30 |
| little | OK | 08:30 |
| zhenguo | liusheng, shaohe_feng, wanghao, zhangjl: wdyt? | 08:30 |
| zhangjl | zhenguo: IMO, same as nova maybe is better | 08:31 |
| *** l4yerffeJ has joined #openstack-mogan | 08:31 | |
| zhenguo | zhangjl: but nova's seems not design that well | 08:31 |
| zhenguo | zhangjl: and they also have a plan to introduce state machine | 08:32 |
| zhangjl | zhenguo:ok | 08:32 |
| zhenguo | zhangjl: I will sent a draft patch by today, you can help review it | 08:33 |
| zhangjl | zhenguo:gj | 08:33 |
| zhangjl | maybe, this is helpful #link:http://docs.openstack.org/developer/ironic/dev/states.html | 08:35 |
| zhenguo | zhangjl: yes, will borrow the fsm from ironic, hah | 08:35 |
| zhangjl | zhenguo:that`s good | 08:36 |
| wanghao | zhenguo: agree with keeping only one status in mogan, since we can change the status to show what task_state it is. | 08:36 |
| zhenguo | wanghao: yes, and we have a taskflow to track the specific states transitions in log | 08:37 |
| zhenguo | zhangjl: btw, for the api doc, we plan to setup a http server ourselves. | 08:37 |
| zhangjl | zhenguo:good news | 08:38 |
| zhenguo | zhangjl: hah | 08:38 |
| wanghao | zhenguo: yes, and also support to introduce the maintance status as god sheng said, that's helpful to baremetal host. | 08:39 |
| zhenguo | wanghao: ok, thanks | 08:39 |
| wanghao | zhenguo: Cinder also has this status FYI. | 08:39 |
| little | zhenguo: where is the spec docs? | 08:39 |
| *** l4yerffeJ has quit IRC | 08:40 | |
| zhenguo | little: currently not, after renaming tomorrow, I will initialize the mogan-specs | 08:40 |
| little | zhenguo: good | 08:40 |
| zhenguo | wanghao: you mean volume has a maintenance status? | 08:40 |
| *** l4yerffeJ has joined #openstack-mogan | 08:40 | |
| wanghao | zhenguo: yes | 08:42 |
| zhenguo | wanghao: hah, ok will add that. | 08:43 |
| liusheng | wanghao: I am not god sheng, lol | 08:43 |
| zhenguo | lol | 08:44 |
| wanghao | liusheng: yes, you're | 08:44 |
| zhenguo | wanghao: agree | 08:44 |
| zhangjl | wanghao:+1 | 08:44 |
| wanghao | zhenguo: In migrate_volume, user can lock volume, that means user can't operate this volume in mirgrating process, the volume will be maintenance. | 08:44 |
| * liusheng have to escape | 08:44 | |
| wanghao | haha | 08:44 |
| little | I just mistook 'god sheng' to 'gou sheng', sorry | 08:47 |
| zhangjl | little:hahaha | 08:48 |
| liusheng | little: :( | 08:48 |
| zhenguo | wanghao: you mean when users lock the volume it will go to maintenance state? | 08:48 |
| zhenguo | wanghao: we should also support lock instance | 08:48 |
| zhenguo | I find nova also has a option to disable instance terminate | 08:49 |
| zhenguo | little: lol | 08:50 |
| liusheng | zhenguo: which one ? | 08:50 |
| zhenguo | liusheng: instance has a filed named disable_terminate | 08:51 |
| zhenguo | liusheng: it will first check this when deleting instance | 08:51 |
| zhangjl | zhenguo:shall we lock instance to disable terminate instead of add a filed? | 08:52 |
| wanghao | zhenguo: yes, in cinder, when user uses migrate operation, he has a option choice to lock volume. | 08:53 |
| liusheng | zhenguo: Kevin told me seems the attribute is useless.. | 08:53 |
| zhenguo | zhangjl: need to dig | 08:53 |
| zhenguo | wanghao: ok, thanks | 08:53 |
| liusheng | zhenguo: it is always False, and not way to set it to true | 08:53 |
| liusheng | s/not way/no way | 08:53 |
| zhenguo | liusheng: hah, if so we can just lock the instance to disable terminate | 08:53 |
| zhenguo | liusheng: and it's more generic | 08:53 |
| wanghao | zhenguo: for mogan, I agree it's a lock API to change instance to maintance. | 08:54 |
| zhenguo | zhangjl: yes, seems that's better | 08:54 |
| zhenguo | wanghao: ok | 08:54 |
| wanghao | zhenguo: Like vmware, in Vcenter, user can set host to maintance. | 08:54 |
| zhenguo | zhangjl: if you are interested in locking instance, maybe you can take this task, hah | 08:54 |
| zhangjl | zhenguo:let me try | 08:55 |
| liusheng | zhenguo: vmware expert! | 08:55 |
| zhenguo | wanghao: yes, we should support maintenance | 08:55 |
| liusheng | zhangjl: wanghao | 08:55 |
| zhenguo | hah | 08:55 |
| zhenguo | zhangjl: thanks | 08:55 |
| wanghao | :) | 08:56 |
| zhenguo | hah, ok so things become clear, we will keep to use status and not introduce task_state, need to add a state machine to control states transitions | 08:58 |
| *** l4yerffeJ has quit IRC | 09:01 | |
| *** l4yerffeJ has joined #openstack-mogan | 09:01 | |
| zhenguo | zhangjl: for locking instance, you can first adding a new task here https://etherpad.openstack.org/p/nimble-task , and maybe register a bp for that | 09:07 |
| zhangjl | zhenguo:ok | 09:08 |
| zhenguo | zhangjl: thanks | 09:08 |
| zhangjl | zhenguo:np | 09:08 |
| *** wanghao has quit IRC | 09:28 | |
| *** l4yerffeJ has quit IRC | 09:47 | |
| *** l4yerffeJ has joined #openstack-mogan | 09:47 | |
| *** l4yerffeJ has quit IRC | 10:08 | |
| *** l4yerffeJ has joined #openstack-mogan | 10:08 | |
| *** zhangjl has quit IRC | 10:13 | |
| *** little has quit IRC | 10:16 | |
| *** kevinz has quit IRC | 10:27 | |
| zhenguo | seems cinder doesn't have a state_file, just using hard code for status | 11:32 |
| *** kevinz has joined #openstack-mogan | 12:06 | |
| *** yuntongjin has joined #openstack-mogan | 12:31 | |
| *** kevinz has quit IRC | 12:31 | |
| *** kevinz has joined #openstack-mogan | 12:32 | |
| *** _liusheng has joined #openstack-mogan | 12:38 | |
| zhenguo | FYI, http://docs.openstack.org/developer/nova/vmstates.html | 12:42 |
| *** kevinz has quit IRC | 12:49 | |
| *** kevinz has joined #openstack-mogan | 12:50 | |
| Kevin_Zheng | Still working on states | 13:03 |
| *** little has joined #openstack-mogan | 13:25 | |
| *** little has quit IRC | 13:42 | |
| *** little has joined #openstack-mogan | 13:43 | |
| *** Jeffrey4l_ has quit IRC | 13:45 | |
| *** little has quit IRC | 13:47 | |
| *** yuntongjin has quit IRC | 13:51 | |
| *** little has joined #openstack-mogan | 13:55 | |
| *** little has quit IRC | 14:11 | |
| *** little has joined #openstack-mogan | 14:12 | |
| *** little has quit IRC | 14:16 | |
| *** kevinz has quit IRC | 14:31 | |
| *** _liusheng has quit IRC | 14:47 | |
| *** little has joined #openstack-mogan | 14:49 | |
| *** little has quit IRC | 14:54 | |
| *** little has joined #openstack-mogan | 14:56 | |
| *** little has quit IRC | 15:13 | |
| *** little has joined #openstack-mogan | 15:13 | |
| *** little has quit IRC | 15:18 | |
| *** Jeffrey4l has joined #openstack-mogan | 15:40 | |
| *** little has joined #openstack-mogan | 15:54 | |
| *** little has quit IRC | 16:11 | |
| *** little has joined #openstack-mogan | 16:11 | |
| *** little has quit IRC | 16:15 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Introduce a state machine for instance status https://review.openstack.org/419024 | 16:30 |
| *** little has joined #openstack-mogan | 16:46 | |
| *** little has quit IRC | 16:51 | |
| *** little has joined #openstack-mogan | 16:55 | |
| *** little has quit IRC | 17:12 | |
| *** wxy has quit IRC | 17:14 | |
| *** wxy has joined #openstack-mogan | 19:12 | |
| *** l4yerffeJ has quit IRC | 20:08 | |
| *** l4yerffeJ has joined #openstack-mogan | 20:09 | |
| *** l4yerffeJ_ has joined #openstack-mogan | 21:34 | |
| *** Jeffrey4l_ has joined #openstack-mogan | 21:34 | |
| *** l4yerffeJ has quit IRC | 21:35 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!