08:00:48 <licanwei> #startmeeting Watcher
08:00:49 <openstack> Meeting started Wed May 22 08:00:48 2019 UTC and is due to finish in 60 minutes.  The chair is licanwei. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:00:50 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
08:00:52 <openstack> The meeting name has been set to 'watcher'
08:01:03 <Dantalion> Hello
08:01:30 <licanwei> Dantalion: hi
08:01:36 <chenke> Dantalion: how are you
08:01:47 <Dantalion> chenke: I am good thanks for asking
08:02:35 <chenke> i am glad to be here to talking about watcher.
08:02:51 <licanwei> #topic review action items
08:03:23 <licanwei> https://review.opendev.org/659099 Spec for Grafana proxy datasource needs review
08:04:14 <chenke> ok
08:04:29 <Dantalion> Tokens do not expire in Grafana, I will post this on the spec after meeting
08:04:33 <licanwei> chenke: good~
08:04:54 <licanwei> Dantalion: thanks
08:06:47 <licanwei> Dantalion: does Watcher depend on some Grafana libs?
08:07:09 <Dantalion> No Grafana exposes rest-api
08:07:26 <licanwei> Dantalion: OK~
08:07:34 <licanwei> next
08:07:44 <licanwei> https://review.opendev.org/659008 Spec for support-keystoneclient-option.rst needs review
08:08:34 <licanwei> I'll review this spec
08:08:57 <licanwei> https://review.opendev.org/656448 Add Placement helper needs review
08:10:03 <chenke> licanwei: i will review it after this meetting
08:10:22 <licanwei> chenke: thanks
08:10:46 <licanwei> https://review.opendev.org/656622 formal datasource interface implementation needs review
08:11:27 <chenke> i had reviewed this commit "formal datasource interface implementation "
08:11:50 <chenke> Overall, it looks good to me.
08:12:29 <licanwei> Dantalion: I'll review it again,
08:12:49 <Dantalion> Okaj thanks
08:13:11 <licanwei> it's too many code~
08:13:37 <licanwei> https://review.opendev.org/660132 remove unused cliutils.py file as osc_lib is used needs review
08:14:55 <licanwei> review +1
08:14:56 <chenke> Does this file cliutils.py never be used before?
08:15:37 <licanwei> https://review.opendev.org/660295 Remove redundant key 'hostname' needs review
08:16:40 <licanwei> review +1
08:17:08 <licanwei> https://review.opendev.org/#/c/657944/ Audit API supports new force option needs review
08:18:21 <chenke> ok, i will review it .
08:18:43 <chenke> the docs about force option i had reviewed.
08:20:12 <licanwei> Dantalion: your comment :'this might lead to a wrong actionplan if the data models are different?
08:20:50 <licanwei> Dantalion: I don't understand your mean
08:21:30 <Dantalion> licanwei: in approved spec it is specified that forcing new audit while one is ongoing might have consequences for actionplan
08:21:50 <chenke> i am also confused with this.
08:22:24 <Dantalion> last sentence of problem description: https://specs.openstack.org/openstack/watcher-specs/specs/train/approved/add-force-field-to-audit.html
08:23:33 <Dantalion> I think we should add the same warning about wrong actionplan to the release note
08:24:18 <chenke> i agree it.
08:24:28 <licanwei> Dantalion: yes,If the audit has the same datamodel ,maybe creating wrong actionplan
08:24:46 <licanwei> Dantalion: I'll update the releasenote
08:24:52 <licanwei> Dantalion: thanks
08:24:58 <Dantalion> licanwei: Yes thank you
08:25:30 <licanwei> https://review.opendev.org/657211 Add force field to Audit needs review
08:27:38 <Dantalion> This looks good to me, nit: is to do boolean type same as other openstack services but the wfield will already ensure a default value so it is not mandatory
08:29:30 <chenke> Before the actionplan start, datamodel has been changed. I think this question need to be discussed.
08:30:04 <licanwei> Dantalion: object field type?
08:30:21 <chenke> i will read the code after our meeting.
08:33:13 <licanwei> next
08:33:22 <licanwei> https://review.opendev.org/658571 Cleanup doc dependencies needs review
08:34:29 <licanwei> review +1
08:34:54 <licanwei> https://review.opendev.org/#/c/659688/ Optimize NovaClusterDataModelCollector.add_instance_node needs review
08:35:47 <licanwei> review +1
08:35:55 <Dantalion> I will look at this one
08:36:38 <chenke> I am looking at this commit.
08:37:26 <licanwei> #topic Discussion
08:37:39 <chenke> this patch is releated to this :https://review.opendev.org/#/c/659886/
08:37:46 <licanwei> whether we need to rename the field 'human_id' of the data model?
08:38:26 <Dantalion> So human_id and hostname are becoming one in nova?
08:38:50 <chenke> novaclient add this option human_id
08:39:35 <licanwei> Dantalion: no hostname in data model
08:40:28 <chenke> but if we use cmd "nova show server_id" ,there is not "human_id"
08:40:44 <licanwei> maybe 'name' is better than human_id
08:40:48 <chenke> only have "name" option
08:41:11 <Dantalion> I think data model names should reflect nova names
08:41:29 <Dantalion> so name and hostname instead of humand_id
08:41:57 <Dantalion> name and hostname can be different right?
08:41:58 <chenke> yes.
08:42:03 <chenke> no
08:42:29 <Dantalion> if name is always equal to hostname than maybe just have one datamodel attribute not two
08:42:58 <chenke> yes
08:43:35 <Dantalion> For instances the name is usually not a hostname right?
08:43:37 <licanwei> Dantalion: agree
08:43:57 <chenke> so i want to delete hostname, because it maybe confused with the name of the node.
08:44:45 <Dantalion> I think naming the attribute 'name' and removing the second attribute is the good option
08:44:54 <chenke> now , in this patch, i use human_id to get instance.name. https://review.opendev.org/#/c/660011/4/watcher/decision_engine/strategy/strategies/base.py
08:45:32 <chenke> I agree.
08:45:58 <licanwei> Dantalion: we should use name or hostname instead of human_id
08:46:23 <Dantalion> hostname might give impression it is FQDN while it is not that is why I think name is better
08:46:54 <chenke> what's the mean about FQDN?
08:47:02 <Dantalion> Fully Qualified Domain Name
08:47:07 <licanwei> Dantalion: agree
08:47:10 <Dantalion> example: google.com
08:47:53 <chenke> oh, i understand it.
08:49:04 <licanwei> I'm no other question
08:49:18 <Dantalion> I will try split up formal datasource
08:49:25 <Dantalion> but will have lots of Depends-On
08:50:09 <chenke> i think future patch can split it.
08:50:16 <chenke> now it's ok.
08:53:22 <licanwei> thank you all ,i'll end the meeting
08:53:28 <chenke> I have no other questions.
08:53:38 <Dantalion> thank you until next time
08:53:40 <Dantalion> :)
08:53:48 <chenke> licanwei: thanks too.
08:54:08 <chenke> Dantalion: bye~
08:54:12 <licanwei> Dantalion: see you :)
08:54:17 <chenke> licanwei: bye~
08:54:19 <licanwei> #endmeeting