14:00:18 <slaweq> #startmeeting neutron_drivers
14:00:19 <openstack> Meeting started Fri Nov 29 14:00:18 2019 UTC and is due to finish in 60 minutes.  The chair is slaweq. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:22 <slaweq> hi
14:00:23 <openstack> The meeting name has been set to 'neutron_drivers'
14:00:32 <davidsha> o/
14:00:56 <ralonsoh> hi
14:00:58 <slaweq> welcome davidsha, lets wait few minutes for quorum
14:01:01 <slaweq> hi ralonsoh
14:01:09 <davidsha> kk
14:02:31 <slaweq> I know that ppl from US are off today but I hope amotoki and yamamoto will join us soon
14:02:45 <amotoki> hi
14:03:03 <slaweq> hi amotoki
14:06:33 <amotoki> sorry for a bit late. Asthma started to hit me again during the Shanghai summit and I was forced to take off a couple days this week :(  hopefully I can recover soon....
14:06:47 <slaweq> amotoki: no problem
14:06:56 <slaweq> I hope You will feel better soon
14:07:06 <slaweq> and we are still waiting for quorum
14:07:19 <slaweq> but it seems that yamamoto will not join today
14:07:23 <amotoki> yeah, this week is the thanks giving week
14:07:38 <slaweq> so as mlavalle and haleyb are off for sure we may not have quorum today
14:07:52 <slaweq> so I will not speak about new rfe today
14:08:07 <slaweq> but as davidsha is here, maybe we can speak a bit about his topic
14:08:13 <slaweq> what do You think?
14:08:16 <amotoki> sounds good
14:08:19 <ralonsoh> +1
14:08:24 <davidsha> kk, thanks!
14:08:27 <slaweq> #topic On Demand agenda
14:08:46 <slaweq> so davidsha please go on :)
14:08:49 <davidsha> https://review.opendev.org/#/c/678865/ I'm working on integration between Neutron Clssifier and Neutron QoS, specifically thes DSCP marking rule
14:09:37 <davidsha> I've submitted a spec discussing the changes to the API, from feedback I need to expand on the changes to the DB schema
14:11:18 <davidsha> Part of this spec also talks about moving the Neutron Classifier service plugin and DB models into the Neutron repo, this was the only solution I could see for creating foreign keys to the classification objects from the QoS Rule objects.
14:11:50 <slaweq> davidsha: yes, I read this spec recently and this part is also something what I wanted to raise on drivers meeting
14:12:03 <ralonsoh> (^^ that's the key point: the DB changes needed in Neutron, specially in dscp)
14:12:47 <davidsha> It's also needed because Neutron classifier's DB migration happens when other out of tree DB migration happen, after Neutrons, but QoS's DB migrations happen at the same time as Neutrons.
14:13:17 <davidsha> So they can't point to the tables because the migration haven't happened yet.
14:14:14 <amotoki> the classifier stuff would be a common base once we adopt the neutron classifier, so it sounds reasonable to have thme in the neutron repo at a glance.
14:14:34 <ralonsoh> I agree with amotoki
14:14:36 <amotoki> one question is how neutron stadium project can consume them.
14:14:51 <slaweq> amotoki: yes, it can be also used by e.g. security group rules in the future IIRC
14:15:03 <davidsha> I have a PoC patch showing how DSCP is consuming it, 2 secs
14:15:21 <amotoki> if the only reason is FFK in db models, what happens if stadium projects would like to use NC models.
14:15:25 <amotoki> NC = neutron classifiers
14:15:35 <davidsha> https://review.opendev.org/#/c/636333/
14:15:38 <ralonsoh> slaweq, as far as I remember, the API will be the QoS one
14:16:02 <davidsha> The extension API consumes the classification API resources
14:16:47 <amotoki> In the API level, I don't see a problem as long as the change is consistent.
14:16:51 <davidsha> So you could reuse the same classification for QoS and Security Groups and SFC etc.
14:18:35 <ralonsoh> (btw, the API https://review.opendev.org/#/c/670049/)
14:18:59 <slaweq> amotoki: I think that if other stadium project will want to use NC models, it will need to depends on Neutron and than Neutron db models will be created first so stadium will be able to use them e.g. in own FK
14:20:02 <davidsha> slaweq: does that Repo already exist?
14:20:06 <amotoki> slaweq: ah. I see. I read the discussion again and noticed that the original problem davidsha raised is about db migration
14:20:35 <slaweq> davidsha: which repo?
14:21:18 <davidsha> A Neutron db models repo, or is it just Neutron-lib?
14:21:42 <slaweq> davidsha: neutron db models are in neutron repo
14:22:50 <davidsha> kk, misunderstood!
14:23:37 <ralonsoh> so, summarizing QoS meetings discussions
14:23:57 <ralonsoh> 1) are the DB changes ok? this is going to modify the dscp rule
14:24:03 <ralonsoh> 2) is the API correct?
14:24:14 <ralonsoh> 3) is the new data model and logic correct?
14:24:49 <ralonsoh> all related to classification resources and groups
14:25:19 <ralonsoh> (IMO, all three are ok for me)
14:25:27 <amotoki> I think 2 and 3 can be discussed in the review.
14:25:32 <ralonsoh> sure
14:25:43 <ralonsoh> and 1 also, in the spec
14:25:53 <amotoki> re 1, it sounds reasonble to me to have DB models in the neutron repo.
14:26:01 <ralonsoh> fantastic!
14:26:09 <slaweq> amotoki++
14:26:14 <davidsha> Cool!
14:26:19 <slaweq> lets discuss that in gerrit comments
14:27:50 <davidsha> Would it be possible to get the rfe approved then?
14:28:21 <slaweq> davidsha: do You have link to the rfe?
14:28:52 <davidsha> https://bugs.launchpad.net/neutron/+bug/1476527
14:28:52 <openstack> Launchpad bug 1476527 in neutron "[RFE] Add common classifier resource" [Wishlist,Triaged] - Assigned to Igor D.C. (igordcard)
14:29:04 <slaweq> this rfe is approved already
14:29:07 <slaweq> :)
14:29:12 <slaweq> but we need to merge spec now
14:29:40 <slaweq> next week I will ask mlavalle, yamamoto and haleyb also to review this spec
14:30:10 <amotoki> it applies to me too :)
14:30:14 <davidsha> kk, ack!
14:30:35 <slaweq> ok, thx for discussing about that
14:30:46 <ralonsoh> thanks!
14:30:54 <slaweq> my goal was mostly to talk about potentially rehoming service plugin to be in-tree plugin
14:30:55 <davidsha> Thanks!
14:30:55 <amotoki> thanks
14:31:04 <slaweq> and we talked about it a bit :)
14:31:10 <slaweq> so thanks for attending
14:31:15 <slaweq> ahh, wait
14:31:27 <slaweq> maybe You have anything else You would like to talk about today?
14:31:32 <slaweq> :)
14:31:37 <ralonsoh> no
14:31:41 <amotoki> nothing from me
14:31:52 <davidsha> I'm good, thanks
14:31:58 <slaweq> ok, so thx for attending
14:32:03 <slaweq> and have a great weekend
14:32:08 <slaweq> #endmeeting