18:00:17 <bh526r> #startmeeting gluon
18:00:18 <openstack> Meeting started Wed Aug 16 18:00:17 2017 UTC and is due to finish in 60 minutes.  The chair is bh526r. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:21 <openstack> The meeting name has been set to 'gluon'
18:00:31 <bh526r> #topic Roll Call
18:00:38 <bh526r> #info Bin Hu
18:00:41 <jinli> #info JinLi
18:00:46 <jinli> Hi Bin
18:00:49 <bh526r> Hi Jin
18:01:29 <bh526r> looks like only you and I. so let's quickly discuss those patches
18:01:37 <jinli> sure
18:01:42 <bh526r> #topic Pike Status
18:02:06 <bh526r> #info Bin cut the stable/pike branch, and tagged 1.0.5
18:02:22 <bh526r> #info There are 4 outstanding patches
18:02:29 <pcarver> hi
18:02:40 <jinli> Hi paul
18:02:56 <bh526r> Hi Paul
18:03:06 <bh526r> #info Patch 1: Gluon object-level authorization
18:03:22 <bh526r> #link https://review.openstack.org/#/c/482752/
18:03:34 <bh526r> #info Patch 2: Policies in yaml
18:03:39 <krenczewski> #info Kamil Renczewski
18:03:50 <bh526r> #link https://review.openstack.org/#/c/486254/
18:03:54 <bh526r> Hi Kamil
18:04:00 <jinli> Hi Kamil
18:04:02 <krenczewski> Hi
18:05:07 <bh526r> #info Bin, Jin and Tin discussed offline, and concluded that for Gluon, because the APIs are generated from YAML, putting policies in YAML is a more flexible way for Gluon
18:05:32 <bh526r> #info Default policies will be in base.yaml.
18:06:28 <bh526r> #info Policies for new services (new YAML) will be defined in the same YAML. In this case, service-specific policies defined in YAML will be used
18:06:53 <bh526r> #info If new YAML doesn't contain service-specific policies, default policies in base.YAML will be used
18:07:45 <bh526r> #info Jin will finalize the coding in above 2 patches this week based on the way to go
18:08:11 <bh526r> #info Patch 3: WIP: Resolve warnings in Sphinx documentation
18:08:20 <bh526r> #link https://review.openstack.org/#/c/491544/
18:08:39 <bh526r> #info This patch seems completed.
18:09:15 <bh526r> @jinli, I saw your comments and new patchset uploaded by Omar.
18:09:21 <bh526r> Do you have more comments?
18:09:55 <jinli> I think this patch should be ok, I just have some questions.
18:10:00 <jinli> let me discuss with Omar
18:10:16 <jinli> then we can merge it later this week
18:10:33 <bh526r> sounds good to me. So I will wait for your signal
18:10:50 <jinli> Tin reviewed it and gave a +1, Tin is the one who discovered this issue
18:11:15 <jinli> Tin gave +1 in patch 5
18:11:53 <bh526r> I see. Now it is patchset 6.
18:12:18 <jinli> I am glad we fix this issue. it causes random errors and fail jenkin once a while
18:12:34 <bh526r> Can you discuss with Omar, and check with Tin regarding your questions on patchset 6?
18:12:42 <jinli> sure,
18:12:48 <bh526r> If you are good, we can merge it later this week
18:12:54 <jinli> ok
18:13:35 <bh526r> #info Jin has some comments. Jin will discuss with Omar and double check with Tin.
18:14:00 <bh526r> #info If everything is fine, it will be approved. Bin will take action to merge it.
18:14:29 <bh526r> #info Patch 4: The para into "add_model()“ should be changed to model in function ”build_api()“
18:14:59 <bh526r> #link https://review.openstack.org/#/c/492916/
18:15:19 <bh526r> any comments on this patch?
18:16:04 <jinli> that load_model_for_service(service) is a singleton, so it doesn't make that much difference to replace it with model.
18:16:39 <jinli> the patch is ok, less confusing
18:17:01 <bh526r> great
18:17:43 <bh526r> If no more comments, I suggest we approve it.
18:19:28 <bh526r> #info Group approved it. Bin will take the action to merge it in master and cherrypick to stable/pike
18:19:48 <bh526r> I think that's all outstanding patches.
18:21:29 <bh526r> Kamil, do you feel we can do a PoC in Sydney with Contrail really working with ODL?
18:21:47 <krenczewski> I think yes
18:22:01 <bh526r> Great.
18:22:22 <bh526r> Georg is still on family leave.
18:22:25 <krenczewski> I didn't discuss details about Sydney summin with Sukhdev yet
18:22:45 <krenczewski> So we need to agree on scenario first
18:22:55 <bh526r> I will draft the demo scenario, and share with you
18:23:03 <krenczewski> OK, great
18:23:16 <bh526r> Then we can see which are doable, and which are stretch
18:24:27 <bh526r> Input from Georg is also important once he comes back
18:25:17 <bh526r> Let me put together the scenario first, and we can discuss it further.
18:25:28 <krenczewski> Sure
18:25:39 <bh526r> #topic AOB
18:25:45 <krenczewski> We're working on same lab as previously?
18:26:00 <bh526r> #info Group discussed a potential PoC/demo in Sydney summit
18:26:21 <bh526r> Yes, the same lab is working as I checked last week.
18:26:31 <bh526r> In Germany
18:26:40 <krenczewski> I will need to re-deploy Contrail
18:27:05 <krenczewski> We're using now R4.0 branch, previously was R3.1
18:27:23 <bh526r> Surely, because this time, we want to see the API calls go through Contrail, and through ODL
18:27:45 <bh526r> And we also need to re-deploy Gluon once we have stable Pike version
18:29:03 <bh526r> #info We need to:
18:29:18 <bh526r> #info 1. Complete Gluon Pike stable version
18:29:26 <bh526r> #info 2. Re-deploy Gluon
18:29:49 <bh526r> #info 3. Re-deploy Contrail (4.0)
18:30:09 <bh526r> #info 4. Double check Lab's availability
18:30:24 <bh526r> #info 5. perhaps re-deploy ODL and ONOS
18:30:57 <bh526r> #info 6. Ensure API call flows through Contrail, ODL and ONOS
18:31:22 <bh526r> #info Bin will draft the demo scenarion first, and share with everyone for further discussion
18:31:56 <bh526r> anything else?
18:32:05 <krenczewski> I am good
18:32:19 <bh526r> great and thank you.
18:32:44 <bh526r> If nothing else, we can adjourn our meeting for now
18:33:01 <bh526r> #info Meeting adjourned
18:33:06 <bh526r> #endmeeting