| *** Kevin_Zheng has joined #openstack-nimble | 00:55 | |
| *** l4yerffeJ has quit IRC | 01:09 | |
| *** l4yerffeJ has joined #openstack-nimble | 01:09 | |
| zhenguo | shaohe_feng, luyao: are you around? | 01:59 |
|---|---|---|
| shaohe_feng | zhenguo: Hi | 01:59 |
| shaohe_feng | zhenguo: o\ | 01:59 |
| zhenguo | shaohe_feng: hi, we will hold a meeting in 1 min | 02:00 |
| *** Jeffrey4l has joined #openstack-nimble | 02:00 | |
| zhenguo | #startmeeting nimble | 02:01 |
| openstack | Meeting started Thu Dec 22 02:01:15 2016 UTC and is due to finish in 60 minutes. The chair is zhenguo. Information about MeetBot at http://wiki.debian.org/MeetBot. | 02:01 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 02:01 |
| openstack | The meeting name has been set to 'nimble' | 02:01 |
| zhenguo | o/ | 02:01 |
| shaohe_feng | zhenguo: Notice you work at 12:00 last night | 02:01 |
| zhenguo | shaohe_feng: hah | 02:01 |
| zhenguo | anyone else here? | 02:02 |
| zhenguo | liusheng: are you around? | 02:03 |
| luyao | zhenguo, Hi | 02:03 |
| zhenguo | luyao: o/ | 02:04 |
| zhenguo | ok, as always, the agenda: | 02:04 |
| liusheng | o/ | 02:04 |
| zhenguo | liusheng: o/ | 02:04 |
| zhenguo | #link https://wiki.openstack.org/wiki/Meetings/Nimble#Agenda_for_next_meeting | 02:04 |
| zhenguo | #topic announcements and reminders | 02:05 |
| *** _liusheng has joined #openstack-nimble | 02:05 | |
| zhenguo | The taskflow work for create instance has been done | 02:05 |
| _liusheng | zhenguo: awesome :) | 02:05 |
| zhenguo | please help to test it more | 02:05 |
| Kevin_Zheng | awesome | 02:05 |
| *** liusheng has quit IRC | 02:05 | |
| zhenguo | We are preparing for the renaming stuff | 02:06 |
| zhenguo | Mogan... | 02:06 |
| _liusheng | do we have check the name with legal people ? | 02:07 |
| zhenguo | yes, and we need to pay :( | 02:07 |
| _liusheng | they said money is not a problem, lol | 02:08 |
| zhenguo | _liusheng: lol, yes money is not a problem | 02:08 |
| zhenguo | not much of announcement, anyone else have a thing here? | 02:08 |
| zhenguo | oh, as always the contribution report: | 02:09 |
| *** yuntongjin has joined #openstack-nimble | 02:09 | |
| zhenguo | #link http://stackalytics.com/report/contribution/nimble/90 | 02:09 |
| zhenguo | #link http://stackalytics.com/report/contribution/python-nimbleclient/90 | 02:09 |
| zhenguo | thanks _liusheng for taking over the client work | 02:10 |
| _liusheng | zhenguo: np :) | 02:10 |
| zhenguo | _liusheng: really appreciate | 02:11 |
| zhenguo | ok, anything else? | 02:11 |
| _liusheng | zhenguo: it is my duty :) | 02:11 |
| zhenguo | _liusheng: :) | 02:11 |
| zhenguo | ok, let's move on | 02:11 |
| zhenguo | #topic task tracking | 02:11 |
| zhenguo | #link https://etherpad.openstack.org/p/nimble-task | 02:12 |
| luyao | zhenguo, I almost finished my thesis proposal | 02:12 |
| luyao | zhenguo, I will come back on nimble ,:) | 02:12 |
| zhenguo | luyao: oh, good news! | 02:12 |
| _liusheng | luyao: good news, welcome :) | 02:12 |
| zhenguo | thanks shaohe_feng for adding new tasks there | 02:13 |
| zhenguo | I want to talk about the priority of these tasks | 02:13 |
| zhenguo | I'm now refactoring the deleting instance codes | 02:14 |
| shaohe_feng | nice | 02:14 |
| shaohe_feng | delete shoulbe be high | 02:14 |
| zhenguo | yes, and as luyao is coming back, we have 4 developers now :P | 02:15 |
| zhenguo | so I would like everyone focues on one task | 02:15 |
| zhenguo | shaohe_feng: seems configdrive is a big task | 02:17 |
| zhenguo | keypair, password, injected file, all depend on that | 02:17 |
| shaohe_feng | zhenguo: yes, but I have learn the mechanism | 02:18 |
| zhenguo | shahe_feng: thanks | 02:19 |
| shaohe_feng | zhenguo: physical work, haha | 02:19 |
| zhenguo | shaohe_feng: maybe you can work with luyao for the whole thing | 02:19 |
| zhenguo | shaohe_feng: including the keypair and injected_file work | 02:19 |
| shaohe_feng | zhenguo: good suggestion | 02:20 |
| zhenguo | after finish the deleting refactor, I will focuses on floating ip support, and will dig out why our tempest is broken | 02:21 |
| _liusheng | zhenguo: thanks for that | 02:21 |
| _liusheng | zhenguo: so what's my task :) | 02:22 |
| zhenguo | _liusheng: hah, not sure whether I can dig out. | 02:22 |
| zhenguo | _liusheng: I think we have quotas, notification, ... | 02:23 |
| zhenguo | which is the high priorities | 02:23 |
| _liusheng | zhenguo: oh, yes, I can pick the notification task, also quotas | 02:23 |
| zhenguo | _liusheng: thanks | 02:23 |
| zhenguo | besides those, I think we need to support multi instance create | 02:24 |
| _liusheng | I like the task :) | 02:24 |
| zhenguo | hah | 02:24 |
| zhenguo | it's a basic functionality | 02:25 |
| _liusheng | zhenguo: yes | 02:25 |
| zhenguo | I think we should support that ASAP | 02:25 |
| shaohe_feng | _liusheng: expert on that field. | 02:25 |
| _liusheng | zhenguo: no :) | 02:25 |
| zhenguo | shaohe_feng: hah, indeed | 02:25 |
| zhenguo | ok, so everyone has a task now! | 02:26 |
| zhenguo | anyone else have athing here? | 02:27 |
| zhenguo | ok, what's next | 02:28 |
| zhenguo | #topic open discussion | 02:28 |
| zhenguo | anything want to discuss here? | 02:29 |
| zhenguo | ok, thanks you all for joinging the meeting! | 02:30 |
| _liusheng | thanks zhenguo | 02:30 |
| zhenguo | #endmeeting | 02:30 |
| openstack | Meeting ended Thu Dec 22 02:30:46 2016 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 02:30 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/nimble/2016/nimble.2016-12-22-02.01.html | 02:30 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/nimble/2016/nimble.2016-12-22-02.01.txt | 02:30 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/nimble/2016/nimble.2016-12-22-02.01.log.html | 02:30 |
| shaohe_feng | zhenguo: so the meeting is finished? I can go to get Medicare card of my son | 02:32 |
| shaohe_feng | zhenguo: will come back in half hour. | 02:33 |
| zhenguo | shaohe_feng: yes, it's finished | 02:33 |
| zhenguo | luyao: When you are back, you can test whether the new taskflow has meet the requirements of what you expected before. | 02:36 |
| *** liusheng has joined #openstack-nimble | 02:44 | |
| * zhenguo brb | 03:03 | |
| *** liusheng has quit IRC | 03:25 | |
| *** liusheng has joined #openstack-nimble | 03:25 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Add node NotFound to messaging expected_exceptions https://review.openstack.org/413915 | 05:36 |
| *** _liusheng has quit IRC | 05:57 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Refactor delete_instance https://review.openstack.org/413677 | 06:18 |
| *** openstackgerrit has quit IRC | 06:48 | |
| zhenguo | shaohe_feng, liusheng: wrt https://review.openstack.org/#/c/413677/ , do you think we need to change to rpc cast from call? | 06:55 |
| liusheng | zhenguo: does the deletion in ironic will take a long time ? | 06:56 |
| zhenguo | liusheng: yes, but we don't need to wait, just want to confirm it moves to a unprovison state, | 06:57 |
| zhenguo | liusheng: this is copied from nova | 06:57 |
| liusheng | zhenguo: hmm, I am ok with it, but I agree with shaohe's review comment | 06:58 |
| zhenguo | liusheng: yes, need to change the return code to Accepted | 06:59 |
| liusheng | zhenguo: yes | 07:00 |
| zhenguo | liusheng: thanks | 07:01 |
| * zhenguo brb | 07:04 | |
| *** yuntongjin has quit IRC | 07:47 | |
| *** yuntongjin has joined #openstack-nimble | 07:50 | |
| zhenguo | liusheng, shaohe_feng: after a second think, maybe we should introduce deleted to instnace table, wdyt? | 08:03 |
| liusheng | zhenguo: why we need that | 08:04 |
| zhenguo | liusheng: maybe users want to list the instances they used last month, or mybe charging system need that? | 08:05 |
| zhenguo | liusheng: I know there are many users complain the increasing instance table as the deleted, but seems we can't avoid that | 08:09 |
| liusheng | zhenguo: I don't mind that if we really need | 08:11 |
| zhenguo | liusheng, shaohe_feng: I browsed the nova codes about terminate instance, before they casting the request to compute, it has set the instance to deleted status, | 08:11 |
| liusheng | zhenguo: just copy from nova ? | 08:11 |
| zhenguo | liusheng: not sure, but seems nova is mature | 08:12 |
| zhenguo | liusheng: so although nova use cast, but it's return code also use 204 | 08:12 |
| zhenguo | liusheng: they don't care about the resouces cleaning, if the db access successfuly to change instance to deleted, they assume it succeed to delete the instance | 08:13 |
| zhenguo | liusheng: but for us, there's no deleted, so we need to delete it from db, then we can't to cleaning other resouces as we don't know what should be cleaned. | 08:14 |
| liusheng | zhenguo: yes, so there will maybe also a feature that admin user can query delete instances | 08:14 |
| zhenguo | liusheng: yes | 08:15 |
| liusheng | zhenguo: sounds reasonable | 08:15 |
| zhenguo | liusheng: so we need to just add deleted for instances or all tables :P | 08:15 |
| liusheng | zhenguo: yes, the *deleted* field wil be int num ? | 08:16 |
| zhenguo | liusheng: yes, seems just 0/1 | 08:16 |
| zhenguo | liusheng: not sure | 08:17 |
| liusheng | zhenguo: but seems in Nova it is numble that will automatically increas | 08:17 |
| liusheng | zhenguo: if a resource has been deleted, the num is 0 | 08:17 |
| zhenguo | liusheng: maybe need some nova guys to explain the 'deleted' for us | 08:19 |
| liusheng | zhenguo: lol | 08:19 |
| liusheng | zhenguo: for us, the Nova guys is zhenyu | 08:19 |
| liusheng | zhenguo: I know a little Nova | 08:19 |
| zhenguo | liusheng: hah | 08:20 |
| liusheng | zhenguo: btw, could you help to take a look: https://review.openstack.org/#/c/411729/ | 08:20 |
| liusheng | zhenguo: and https://review.openstack.org/#/c/412326/ | 08:21 |
| liusheng | zhenguo: I just want to change *instance* --> server | 08:21 |
| zhenguo | liusheng: sure | 08:21 |
| liusheng | zhenguo: since I like use cli to test some functionalities, lol | 08:21 |
| zhenguo | liusheng: ditto, lol | 08:21 |
| zhenguo | liusheng: just change to server instead of 'openstack baremetal server' ? | 08:22 |
| liusheng | zhenguo: *instance* --> *server* and *type* --> *flavor* | 08:22 |
| liusheng | zhenguo: what do you think ? | 08:23 |
| zhenguo | liusheng: totally agree | 08:23 |
| zhenguo | liusheng: how about the 'compute' | 08:23 |
| Kevin_Zheng | liusheng is right | 08:23 |
| Kevin_Zheng | he teached me a leason | 08:23 |
| zhenguo | Kevin_Zheng: hah | 08:24 |
| liusheng | zhenguo: we use "baremetal flavor" and "baremetal server" | 08:24 |
| zhenguo | liusheng: excellent | 08:24 |
| zhenguo | Kevin_Zheng: we want to add a deleted filed to instance like nova | 08:24 |
| Kevin_Zheng | cool | 08:27 |
| liusheng | Kevin_Zheng, zhenguo we have copied most of Nova, lol | 08:28 |
| zhenguo | liusheng: hah | 08:28 |
| zhenguo | liusheng: we need to stand on the top of Nova | 08:28 |
| Kevin_Zheng | looks like only you two are ATC | 08:31 |
| liusheng | Kevin_Zheng: we have big customer | 08:33 |
| zhenguo | Kevin_Zheng: yes, do you want to come back, lol | 08:41 |
| Kevin_Zheng | you know we have some high priority job to do :) | 08:42 |
| zhenguo | hah | 08:42 |
| Kevin_Zheng | lol | 08:42 |
| shaohe_feng | zhenguo: what is ATC | 08:52 |
| zhenguo | shaohe_feng: active technical contributor | 08:53 |
| *** openstackgerrit has joined #openstack-nimble | 08:57 | |
| openstackgerrit | ShaoHe Feng proposed openstack/python-nimbleclient: Add support for setting power state for instance https://review.openstack.org/411729 | 08:57 |
| shaohe_feng | zhenguo: Yes, got it. You two are high active | 08:58 |
| shaohe_feng | zhenguo: liusheng: we will add a deleted field? | 08:58 |
| zhenguo | shaohe_feng: hah, you are also a ATC, and luyao will come back soon | 08:58 |
| liusheng | shaohe_feng: zhenguo proposed that | 08:58 |
| zhenguo | shaohe_feng, liusheng: yes, I want to add deleted | 08:59 |
| zhenguo | shaohe_feng, liusheng: then we just set the instance to deleted before casting the request to engine, and we don't need to change the API retrun status | 08:59 |
| shaohe_feng | zhenguo: so for the user, it is an sync call? | 09:02 |
| zhenguo | shaohe_feng: user just care about whether the instance is in deleted status, lol | 09:02 |
| zhenguo | shaohe_feng: we set the status before we do rpc cast | 09:03 |
| zhenguo | shaohe_feng: as long as we successfully set the instance status it succeed. | 09:03 |
| shaohe_feng | zhenguo: OK. | 09:04 |
| shaohe_feng | sounds good. | 09:04 |
| zhenguo | shaohe_feng: thanks, I will refactor it again | 09:09 |
| shaohe_feng | zhenguo: waiting for the new patch | 09:10 |
| zhenguo | shaohe_feng: I will try to sent out by today | 09:10 |
| shaohe_feng | do not need rush\ | 09:12 |
| shaohe_feng | zhenguo: you need a rest. | 09:12 |
| zhenguo | shaohe_feng: hah | 09:12 |
| shaohe_feng | zhenguo: you stay up last night | 09:12 |
| zhenguo | shaohe_feng: np, we need to be quickly | 09:13 |
| zhenguo | shaohe_feng: there's really a long list in the task tracking etherpad, before O release, I want to make nimble usable | 09:14 |
| *** yuntongjin has quit IRC | 09:14 | |
| *** yuntongjin has joined #openstack-nimble | 09:15 | |
| zhenguo | in fact, we even can't compare with nova now, wdyt? | 09:15 |
| shaohe_feng | zhenguo: yes. we need our characteristic | 09:18 |
| zhenguo | shaohe_feng: yes, so we need to finish the basic functionalities first, and make it usable | 09:19 |
| shaohe_feng | zhenguo: yes, basic functionalities and stable is the most important | 09:20 |
| zhenguo | shaohe_feng: sure, but if you only have that, how to compare with nova | 09:20 |
| shaohe_feng | zhenguo: you know, nova has make may improvement, but it still many defect。 | 09:21 |
| shaohe_feng | zhenguo: such as some advance features, live migration | 09:22 |
| shaohe_feng | zhenguo: still very bad. | 09:22 |
| zhenguo | shaohe_feng: yes, but we don't care about that | 09:22 |
| zhenguo | shaohe_feng: we just need to compare the bare metal related features | 09:23 |
| shaohe_feng | zhenguo: Yes, I just want to say, it need time to make a project perfect. | 09:24 |
| shaohe_feng | zhenguo: so you need rest on demand. ") | 09:25 |
| zhenguo | shaohe_feng: hah | 09:25 |
| zhenguo | shaohe_feng: O is a very import release for us | 09:25 |
| zhenguo | s/import/important | 09:26 |
| zhenguo | shaohe_feng: maybe someone will test it, I don't want it that bad. | 09:26 |
| shaohe_feng | zhenguo: OK. I will test it. | 09:28 |
| zhenguo | shaohe_feng: thanks | 09:28 |
| openstackgerrit | Merged openstack/python-nimbleclient: Add support for setting power state for instance https://review.openstack.org/411729 | 09:29 |
| zhenguo | liusheng: all client related patches have been landed. | 09:29 |
| openstackgerrit | Merged openstack/python-nimbleclient: Support retuning detailed info of "instance list" command https://review.openstack.org/412326 | 09:30 |
| liusheng | zhenguo: thanks | 09:34 |
| zhenguo | liusheng: np | 09:34 |
| zhenguo | liusheng: when creating instance, after specify --nic, I want to replace uuid with net-id like nova | 09:35 |
| zhenguo | liusheng: seems --nic uuid=xxx it's a bit confused, wdyt? | 09:35 |
| zhenguo | liusheng: as we need to support net-id, port-id.. | 09:36 |
| liusheng | zhenguo: agree | 09:36 |
| zhenguo | liusheng: thanks, but no rush, seems we need to refactor the networks on nimble side first. | 09:41 |
| liusheng | zhenguo: yes, currently we only support net-id, right ? | 09:41 |
| zhenguo | liusheng: yes, | 09:42 |
| * zhenguo brb | 09:48 | |
| *** yuntongjin has quit IRC | 11:20 | |
| zhenguo | liusheng: wrt nova's deleted, why it's a automatically increased? | 11:33 |
| *** yuntongjin has joined #openstack-nimble | 11:33 | |
| liusheng | zhenguo: I was a wrong, the deleted items' *deleted* is automatically increased, I said opposite :) | 11:35 |
| zhenguo | liusheng: lol | 11:35 |
| zhenguo | liusheng: so it's a boolen in object and a integer in db | 11:36 |
| liusheng | zhenguo: I think the main reason is because the deleted instance can release some unique constraint attributes which need to be unique in non-deleted items | 11:36 |
| liusheng | zhenguo: yes, the non-deleted items' *deleted* is 0 | 11:37 |
| zhenguo | liusheng: yes | 11:37 |
| liusheng | zhenguo: e.g. the nova's flavor, the id is unique, but when a flavor was deleted. the id can be reused | 11:38 |
| liusheng | zhenguo: so the *id* and *deleted* combination can be a unqiue key | 11:39 |
| liusheng | zhenguo: I am also not sure, I cannot remember clearlly, lol | 11:39 |
| zhenguo | liusheng: yes, it is | 11:39 |
| zhenguo | liusheng: do we have a key like that? | 11:41 |
| zhenguo | liusheng: seems not, as in instance the unique constraint is uuid which is not allowed to be specified by users | 11:42 |
| liusheng | zhenguo: maybe in future | 11:42 |
| liusheng | zhenguo: yes, currently the uuid of instance is the only uqniue of instance | 11:42 |
| liusheng | zhenguo: in some scenarios, in users' perspective, some attributes maybe uqniue in the alive resources, but for db, when the resources were deleted, it should be constrained | 11:44 |
| zhenguo | liusheng: yes, | 11:45 |
| zhenguo | liusheng: another question, do you know the difference between deleted_at and terminated-at | 11:45 |
| liusheng | zhenguo: oh, I remembered, if we support unique name in future, that will need that, hah | 11:45 |
| zhenguo | liusheng: hah, | 11:46 |
| liusheng | zhenguo: but it also can be controlled by code logical | 11:46 |
| liusheng | zhenguo: nova have both the attributes ? | 11:46 |
| zhenguo | liusheng: yes, but hope we will never need to change to support unique name again | 11:47 |
| zhenguo | liusheng: yes | 11:47 |
| zhenguo | liusheng: and I find another attribute disable_terminate which seems useful. | 11:47 |
| zhenguo | liusheng: there are really many things we need to learn from nova, we should not against it. | 11:48 |
| liusheng | zhenguo: the terminated_at indicate the time when the dom is destroyed in libvirt | 11:48 |
| liusheng | zhenguo: the deleted_at indicate the time when the record is deleted in db | 11:49 |
| liusheng | zhenguo: hah | 11:49 |
| zhenguo | liusheng: aha, thanks, so seems we only need deleted now | 11:49 |
| liusheng | zhenguo: yes, for auditing it is really useful | 11:50 |
| zhenguo | liusheng: yes, and seems it's not that hard to implement it | 11:50 |
| liusheng | zhenguo: lol | 11:51 |
| zhenguo | liusheng: hah, we need to keep learning from nova when we don't know what to do... | 11:52 |
| openstackgerrit | liusheng proposed openstack/nimble: Add instance CURD related functional tests https://review.openstack.org/398068 | 11:57 |
| *** yuntongjin has quit IRC | 12:10 | |
| *** yuntongjin1 has joined #openstack-nimble | 12:10 | |
| *** yuntongjin1 has quit IRC | 14:38 | |
| zhenguo | liusheng: are you still around? | 15:08 |
| zhenguo | liusheng: why did you remove id column from instance table? | 15:08 |
| *** yuntongjin has joined #openstack-nimble | 15:14 | |
| *** l4yerffeJ_ has joined #openstack-nimble | 15:33 | |
| *** l4yerffeJ has quit IRC | 15:33 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Revert "Use uuid as the primary key" https://review.openstack.org/414215 | 16:13 |
| zhenguo | liusheng: I reverted the patch to set uuid as primary key, why we don't use id as primary key? and soft_delete seems depending on id column. | 16:15 |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Add node NotFound to messaging expected_exceptions https://review.openstack.org/413915 | 16:19 |
| *** Jeffrey4l has quit IRC | 16:46 | |
| *** Jeffrey4l has joined #openstack-nimble | 17:07 | |
| *** Jeffrey4l has quit IRC | 17:15 | |
| *** Jeffrey4l has joined #openstack-nimble | 17:15 | |
| *** Jeffrey4l has quit IRC | 17:23 | |
| *** Jeffrey4l has joined #openstack-nimble | 17:23 | |
| *** l4yerffeJ_ has quit IRC | 17:29 | |
| *** l4yerffeJ_ has joined #openstack-nimble | 17:29 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/nimble: Updated from global requirements https://review.openstack.org/414271 | 18:52 |
| *** yuntongjin has quit IRC | 20:33 | |
| *** yuntongjin has joined #openstack-nimble | 20:34 | |
| *** l4yerffeJ has joined #openstack-nimble | 21:32 | |
| *** l4yerffeJ_ has quit IRC | 21:35 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
| *** Jeffrey4l has joined #openstack-nimble | 21:48 | |
| *** l4yerffeJ has quit IRC | 23:44 | |
| *** l4yerffeJ has joined #openstack-nimble | 23:45 | |
| *** yuntongjin has quit IRC | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!