| *** l4yerffeJ has quit IRC | 00:28 | |
| *** l4yerffeJ has joined #openstack-nimble | 00:28 | |
| zhenguo | morning nimble! | 01:48 |
|---|---|---|
| openstackgerrit | Merged openstack/nimble: Updated from global requirements https://review.openstack.org/414271 | 02:10 |
| shaohe_feng | zhenguo: morning | 02:56 |
| zhenguo | shaohe_feng: o/ | 02:56 |
| liusheng | morning | 02:56 |
| zhenguo | liusheng: o/ | 02:57 |
| shaohe_feng | zhenguo: I'm in a internal meeting. | 03:00 |
| shaohe_feng | zhenguo: It's for vGPU. intel's vGPU has been in upstreamed | 03:01 |
| shaohe_feng | zhenguo: kernal upstream | 03:01 |
| zhenguo | shaohe_feng: cool, | 03:01 |
| zhenguo | shaohe_feng: i have some internal stuff to deal with, will ping you back, | 03:01 |
| * zhenguo brb | 03:01 | |
| shaohe_feng | zhenguo: allo the patch, more than 20 thousand lines. | 03:02 |
| shaohe_feng | zhenguo: al of these patches are done by Chinese. | 03:02 |
| *** yuntongjin has joined #openstack-nimble | 03:47 | |
| *** l4yerffeJ has quit IRC | 03:57 | |
| *** yuntongjin has quit IRC | 04:35 | |
| *** kevinz has joined #openstack-nimble | 05:02 | |
| *** yuntongjin has joined #openstack-nimble | 05:53 | |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part1) https://review.openstack.org/414448 | 07:06 |
| *** yuntongjin has quit IRC | 07:17 | |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part1) https://review.openstack.org/414448 | 07:34 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part2) https://review.openstack.org/414464 | 07:34 |
| *** yuntongjin has joined #openstack-nimble | 07:42 | |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part2) https://review.openstack.org/414464 | 07:52 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part2) https://review.openstack.org/414464 | 07:59 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part1) https://review.openstack.org/414448 | 07:59 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part3) https://review.openstack.org/414472 | 08:15 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part4) https://review.openstack.org/414481 | 08:49 |
| *** yuntongjin has quit IRC | 09:03 | |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part4) https://review.openstack.org/414481 | 09:23 |
| openstackgerrit | liusheng proposed openstack/python-nimbleclient: Refactor: change *instance* to *server* (part4) https://review.openstack.org/414481 | 09:34 |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Refactor to soft delete instance https://review.openstack.org/414517 | 09:40 |
| zhenguo | liusheng: hi, I add instance.id back | 09:41 |
| liusheng | zhenguo: why need instance.id | 09:41 |
| zhenguo | liusheng: as soft_delete() will set instance.deleted=instance.id | 09:42 |
| zhenguo | liusheng: id is a internal column, need by soft_delete() method | 09:42 |
| liusheng | zhenguo: oh, I see | 09:43 |
| liusheng | zhenguo: it is only used internallly, right ? | 09:44 |
| zhenguo | liusheng: yes, it will never expose to end users | 09:44 |
| liusheng | zhenguo: ok, the patch I committed is not only remove the instance.id and set uuid as primary key | 09:45 |
| zhenguo | liusheng: why uuid as primary key | 09:45 |
| zhenguo | liusheng: id should be primary key and uuid is unique constraints | 09:45 |
| liusheng | zhenguo: I mean I have fix some other issues in the patch, not only use uuid as primary key | 09:46 |
| zhenguo | liusheng: ok | 09:46 |
| zhenguo | liusheng: I find a issue when using 'instance list' | 09:47 |
| zhenguo | liusheng: there are are Instance Type and Networks, but it's empty | 09:47 |
| liusheng | zhenguo: I have found one, not sure same, lol | 09:47 |
| liusheng | zhenguo: yes | 09:47 |
| zhenguo | liusheng: haha | 09:47 |
| liusheng | zhenguo: oh, it is not same issue | 09:48 |
| liusheng | zhenguo: you mean the two columns are empty ? | 09:48 |
| zhenguo | liusheng: yes, but if I use instance show, it will present | 09:49 |
| liusheng | zhenguo: I have fix it, https://review.openstack.org/#/c/412326/ | 09:49 |
| liusheng | zhenguo: do you have update your client | 09:49 |
| liusheng | https://bugs.launchpad.net/python-nimbleclient/+bug/1651021 | 09:50 |
| openstack | Launchpad bug 1651021 in python-nimbleclient "The instance list command don't return instance_type, networks" [Undecided,Fix released] - Assigned to Liusheng (liusheng) | 09:50 |
| zhenguo | liusheng: oh, seems not, will update soon | 09:50 |
| liusheng | zhenguo: another issue I found, the "instance list" command will return emty some times, but some times it will return the instance list | 09:50 |
| liusheng | zhenguo: it is werid | 09:51 |
| liusheng | zhenguo: something wrong with nimble-api side | 09:51 |
| zhenguo | liusheng: you mean empty instance list or some fields ? | 09:51 |
| liusheng | root@liusheng:/opt/stack/python-nimbleclient# openstack baremetal server list | 09:51 |
| liusheng | zhenguo: it return empty | 09:51 |
| liusheng | zhenguo: but actually I have a instance | 09:52 |
| zhenguo | liusheng: is there some error in api log | 09:52 |
| liusheng | zhenguo: no | 09:53 |
| liusheng | zhenguo: I have checked in db | 09:53 |
| zhenguo | liusheng: really wired | 09:53 |
| liusheng | zhenguo: some times it is ok | 09:53 |
| zhenguo | s/wired/weird | 09:53 |
| zhenguo | liusheng: oh, just realized, you have finished the rename work? | 09:56 |
| liusheng | zhenguo: yes | 09:56 |
| liusheng | zhenguo: the 4 patches | 09:56 |
| zhenguo | liusheng: nice | 09:56 |
| zhenguo | liusheng: I will have a presentation in tomorrow's OpenStack Meetup, maybe I can show our new CLI, hah | 09:57 |
| liusheng | zhenguo: cool | 09:57 |
| liusheng | zhenguo: so may need to fix the issue I said today, lol | 09:58 |
| zhenguo | liusheng: yes, hah | 09:58 |
| liusheng | zhenguo: I will fix it tonight | 09:59 |
| zhenguo | liusheng: really appreciate | 09:59 |
| liusheng | zhenguo: np :) | 09:59 |
| zhenguo | shaohe_feng: will you go to tomorrow's meetup? | 10:00 |
| zhenguo | shaohe_feng: seems in Intel | 10:00 |
| liusheng | zhenguo: seems I don't update the mimble code in my env | 10:02 |
| liusheng | zhenguo: maybe it is not an issue, hah | 10:02 |
| zhenguo | liusheng: lol | 10:02 |
| liusheng | zhenguo: I cannot reproduce now :( | 10:02 |
| zhenguo | liusheng: hah, so i'ts not a issue if you can't produce | 10:03 |
| liusheng | zhenguo: openstack baremetal server list --detailed, this command can show the flavor, networks | 10:03 |
| liusheng | zhenguo: openstack baremetal server list only uuid, name, status | 10:03 |
| zhenguo | liusheng: ok, I'm updating my client now, | 10:04 |
| liusheng | zhenguo: they are GET /instances and GET /instances/detail in nimble-api side | 10:04 |
| zhenguo | liusheng: yes, and detail will get some information from ironic | 10:05 |
| zhenguo | liusheng: but I don't quite like the current way | 10:05 |
| liusheng | zhenguo: what you prefer | 10:06 |
| zhenguo | liusehng: I would like to cache the node information instead of issuing ironic API call then there's a request comming | 10:06 |
| zhenguo | s/then/when | 10:06 |
| zhenguo | liusheng: we should try to reduce users' requests time, | 10:07 |
| liusheng | zhenguo: agree! | 10:08 |
| zhenguo | liusheng: hah | 10:08 |
| zhenguo | liusheng: oh, after updated, I see the instance type and network info column disappeared, thanks! | 10:10 |
| liusheng | zhenguo: hah | 10:10 |
| liusheng | zhenguo: I need to go home :) | 10:11 |
| * zhenguo away for dinner | 10:11 | |
| zhenguo | liusheng: see you | 10:11 |
| liusheng | zhenguo: enjoy | 10:11 |
| zhenguo | liusheng: thanks | 10:11 |
| *** kevinz has quit IRC | 10:24 | |
| *** l4yerffeJ has joined #openstack-nimble | 11:33 | |
| openstackgerrit | Zhenguo Niu proposed openstack/nimble: Refactor to soft delete instance https://review.openstack.org/414517 | 11:51 |
| openstackgerrit | Merged openstack/nimble: Refactor to soft delete instance https://review.openstack.org/414517 | 13:04 |
| shaohe_feng | zhenguo: I will go to the meeting up. | 15:12 |
| zhenguo | shaohe_feng: hah, see you there | 15:13 |
| shaohe_feng | zhenguo: yes. just see your message | 15:13 |
| shaohe_feng | zhenguo: I'm out of office this afternoon | 15:14 |
| zhenguo | shaohe_feng: hah, why are you still online | 15:14 |
| shaohe_feng | zhenguo: I'n need to by a house for my car. | 15:14 |
| zhenguo | shaohe_feng: in fact, tomorrow is a work day for us | 15:14 |
| shaohe_feng | zhenguo: yes. | 15:15 |
| zhenguo | shaohe_feng: I men for huawei guys | 15:15 |
| shaohe_feng | zhenguo: oh, got it | 15:15 |
| shaohe_feng | zhenguo: every the last satudays | 15:15 |
| zhenguo | shaohe_feng: yeah, hah | 15:15 |
| shaohe_feng | zhenguo: so you deserved high salary. | 15:16 |
| zhenguo | shaohe_feng: we will get double pay for the work day like tomorrow | 15:16 |
| zhenguo | extra work da | 15:17 |
| shaohe_feng | zhenguo: I'm not buzy at this time every day. | 15:17 |
| shaohe_feng | zhenguo: My son go to sleep, so I can do some work. haha | 15:18 |
| zhenguo | shaohe_feng: hah | 15:18 |
| shaohe_feng | zhenguo: double pay. Huawei is a conscientious company | 15:19 |
| zhenguo | shaohe_feng: lol | 15:20 |
| openstackgerrit | Merged openstack/nimble: Add node NotFound to messaging expected_exceptions https://review.openstack.org/413915 | 15:20 |
| zhenguo | shaohe_feng: I have changed to use soft delete now | 15:22 |
| shaohe_feng | zhenguo: so the soft delete means does not delete the DB? | 15:22 |
| zhenguo | shaohe_feng: yes, seems soft delete is a common use, not only for openstack | 15:23 |
| shaohe_feng | zhenguo: nice. | 15:24 |
| shaohe_feng | zhenguo: do we support hard delete? :) | 15:24 |
| zhenguo | shaohe_feng: no | 15:25 |
| zhenguo | liusheng: thanks for the cool CLI, I'm really like to type like that, lol | 15:27 |
| shaohe_feng | zhenguo: OK. A good design means meet usrs' requirements | 15:28 |
| zhenguo | shaohe_feng: sure, hah | 15:29 |
| zhenguo | shaohe_feng: I also learned some other thing from nova while reading their soft delete codes | 15:30 |
| zhenguo | shaohe_feng: they have a terminate_disabled field, which will disable user to delete the instance | 15:31 |
| shaohe_feng | zhenguo: why they are disable user to delete the instance. | 15:32 |
| shaohe_feng | zhenguo: is it useful for nimble? | 15:32 |
| zhenguo | shaohe_feng: it's useful | 15:33 |
| zhenguo | shaohe_feng: we may have many users in one tenant, | 15:33 |
| shaohe_feng | zhenguo: for example, some scenarios, such as we do not allow delete when building? | 15:33 |
| zhenguo | shaohe_feng: yes | 15:33 |
| zhenguo | shaohe_feng: no, not exactly | 15:34 |
| shaohe_feng | zhenguo: so many users in one tenant, what will happen? | 15:34 |
| zhenguo | shaohe_feng: I mean the instance should never be deleted if that attribute is set | 15:34 |
| zhenguo | shaohe_feng: they will see all instances in the tenant, | 15:35 |
| zhenguo | shaohe_feng: but admin may not want end users delete some important instances | 15:35 |
| shaohe_feng | zhenguo: so they set the instance as terminate_disabled as true? | 15:36 |
| zhenguo | shaohe_feng: I think yes, but not sure | 15:38 |
| shaohe_feng | zhenguo: only admin can decide what are important instances, but any principle to make the decision? | 15:39 |
| shaohe_feng | zhenguo: So I think the "important" is not absolute | 15:40 |
| zhenguo | shaohe_feng: we are software, the principle should be set by human users, we just need to provide API | 15:40 |
| shaohe_feng | zhenguo: they can become un-important on day, so admin can also set terminate_disabled as false. :) | 15:41 |
| zhenguo | shaohe_feng: yes, we provide the API, they can do what they want, hah | 15:42 |
| shaohe_feng | zhenguo: we can seen the “套路” is add a field in DB. :) | 15:43 |
| zhenguo | shaohe_feng: lol | 15:43 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refactor: change *instance* to *server* (part1) https://review.openstack.org/414448 | 15:50 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refactor: change *instance* to *server* (part2) https://review.openstack.org/414464 | 15:51 |
| shaohe_feng | zhenguo: Administrator password for the instance can also be in configdrive | 15:52 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refactor: change *instance* to *server* (part3) https://review.openstack.org/414472 | 15:52 |
| openstackgerrit | Merged openstack/python-nimbleclient: Refactor: change *instance* to *server* (part4) https://review.openstack.org/414481 | 15:52 |
| zhenguo | shaohe_feng: yes | 15:53 |
| shaohe_feng | zhenguo: generate a configdrive shoulde be a task? | 15:59 |
| shaohe_feng | zhenguo: or just as a setp in the do_node_deploy routine? | 16:00 |
| openstackgerrit | Zhenguo Niu proposed openstack/python-nimbleclient: Set --power-state to be a required argument https://review.openstack.org/414620 | 16:00 |
| zhenguo | shaohe_feng: seems like it is in node deploy | 16:01 |
| shaohe_feng | zhenguo: OK. | 16:02 |
| zhenguo | shaohe_feng: you can pass the configdrive to ironic via provision API, then it will store it in swfit | 16:02 |
| shaohe_feng | zhenguo: yes. that's why ironic need swift. hah | 16:03 |
| zhenguo | shaohe_feng: yes, as ironic python agent cant get authorized from keystone, so they need swift to provide a tempurl | 16:04 |
| zhenguo | shaohe_feng: I have to go to bed now, see you tomorrow | 16:07 |
| shaohe_feng | zhenguo: see you. | 16:07 |
| * zhenguo away | 16:08 | |
| *** l4yerffeJ_ has joined #openstack-nimble | 21:32 | |
| *** l4yerffeJ_ has quit IRC | 21:32 | |
| *** l4yerffeJ_ has joined #openstack-nimble | 21:33 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
| *** l4yerffeJ has quit IRC | 21:35 | |
| *** Jeffrey4l has joined #openstack-nimble | 21:48 | |
| *** l4yerffeJ_ has quit IRC | 23:12 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!