18:00:22 <bh526r> #startmeeting gluon 18:00:23 <openstack> Meeting started Wed Nov 30 18:00:22 2016 UTC and is due to finish in 60 minutes. The chair is bh526r. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:27 <openstack> The meeting name has been set to 'gluon' 18:00:36 <bh526r> hello guys 18:00:42 <krenczewski> Hi 18:00:44 <jinli> Hi all 18:00:50 <jinli> #info JinLi 18:01:33 <bh526r> #topic Roll Call 18:01:39 <bh526r> thank you Jin 18:01:44 <bh526r> #info Bin Hu 18:01:48 <krenczewski> #info Kamil Renczewski 18:01:53 <bh526r> Hi Kamil 18:02:04 <krenczewski> Sorry I couldn't attend yesterday 18:02:53 <bh526r> No problem 18:03:09 <bh526r> #topic Admin Update 18:03:33 <bh526r> #info No admin update from my side. 18:03:49 <bh526r> #topic Gluon Tasks 18:04:26 <bh526r> #info Our task list for Ocata release is in our wiki: https://wiki.openstack.org/wiki/Gluon/Tasks-Ocata 18:05:08 <bh526r> #info Original discussion etherpad during Design Summit in Barcelona is at: https://etherpad.openstack.org/p/ocata-gluon-work-plan 18:05:43 <bh526r> #info The etherpad during joint Nova-Neutron session is at: https://etherpad.openstack.org/p/ocata-nova-neutron-session 18:06:15 <bh526r> #info All of the work is along with the task list 18:06:33 <bh526r> #topic Status Update 18:07:13 <bh526r> #info There have been about 9 outstanding patches in review since our last meeting 18:07:54 <bh526r> #info Now we are progressing in 7 areas 18:09:12 <bh526r> #info Task "Service Binding Model Proposal" based on approve BP https://blueprints.launchpad.net/python-gluon/+spec/port-service-binding-model, tracked by PM tool https://jira.opnfv.org/browse/NETREADY-39. 18:09:48 <bh526r> #info There were 3 patches for this task: 18:09:57 <bh526r> #info Patch 1: https://review.openstack.org/#/c/392250/ 18:10:19 <bh526r> #info Patch 2: https://review.openstack.org/#/c/395738/ 18:11:30 <bh526r> #info There were many comments in review. 18:11:54 <bh526r> #info Patch authors Georg (392250) and Tom (395738) took those comments, and jointly submitted a new patch 18:12:20 <bh526r> #info Patch 3 https://review.openstack.org/#/c/400012/ jointly by authors of 392250 and 395738 18:13:18 <bh526r> #info 400012 was approved and merged on Monday. 392250 and 395738 were abandoned 18:14:30 <bh526r> #info Patch 400012 will serve as the baseline of service binding model, and more work is expected, such as formal definition of parameters etc, as I indicated in review approval 18:15:43 <bh526r> #info Task "Authentication and Access Control" based on approved BP https://blueprints.launchpad.net/python-gluon/+spec/gluon-auth. 18:16:14 <bh526r> #info it is tracked by 2 JIRAs: https://jira.opnfv.org/browse/NETREADY-47 and https://jira.opnfv.org/browse/NETREADY-48 18:16:32 <bh526r> #info Kamal submitted a patch https://review.openstack.org/#/c/398577/ 18:17:27 <bh526r> #info It is under review, and there have been comments as of this morning 18:18:01 <bh526r> #info It is expected that Kamal will take into consideration those comments, and submit a new patchset for further review 18:19:13 <bh526r> #info Task "Support of Multiple YAMLs / Proton APIs" is based on approved BP https://blueprints.launchpad.net/python-gluon/+spec/multiple-api-support, and tracked by JIRA https://jira.opnfv.org/browse/NETREADY-58. 18:20:01 <bh526r> #info The code patch is https://review.openstack.org/#/c/399051/ 18:20:44 <bh526r> #info The code was approved and merged on Monday 18:22:14 <bh526r> #info Task "OpenContrail's Mechanism Driver" is based on approved BP https://blueprints.launchpad.net/python-gluon/+spec/gluon-contrail-integration 18:22:14 <bh526r> #info Kamil submitted a patch "https://review.openstack.org/#/c/402071/" 18:23:01 <bh526r> #info There was a comment regarding it is a mechanism driver, and the code might not belong to Gluon repo 18:24:14 <bh526r> Kamil, since you are here, I think the question is that: because it is a mechanism driver, when it is deployed with OpenStack, you need to modify neutron.conf, right? 18:25:19 <krenczewski> If openstack is configured with ML2 plugin enabled, then I need to add an entry point for this mechanism driver 18:25:29 <krenczewski> and modify ml2.ini 18:25:59 <krenczewski> I think than I don't need to modify neutron.conf, but I am not 100% sure right now 18:26:04 <krenczewski> I need to double check this 18:26:51 <bh526r> OK, I see. So at least you need to modify ml2.ini, and possibly neutron.conf 18:28:05 <bh526r> In Gluon's circumstance, we extended core ML2 plugin, and modify neutron.conf so that core plugin's entry point is Gluon's Extended ML2 Plugin. 18:28:09 <krenczewski> I need to modify: /etc/neutron/plugins/ml2/ml2_conf.ini file 18:28:57 <krenczewski> and also add entry to /opt/stack/neutron/neutron.egg-info/entry_points.txt 18:29:07 <bh526r> I see. which line / config parameter in ml2_conf.ini? 18:29:11 <krenczewski> I don't need to modify neutron.conf 18:29:21 <bh526r> Great. 18:30:05 <krenczewski> In section [ml2] I need to add my mechnosm driver to option "mechanism_drivers" 18:30:18 <bh526r> I see. 18:30:50 <jinli> Bin please have someone take a look at 393935 18:31:13 <krenczewski> And of course after this I need to restart Neutron service 18:31:15 <jinli> I make new patch to fix pep8 issue 18:31:38 <bh526r> So when integrating with current Gluon, your change in ml2_conf.ini should transparently work with Gluon's change in neutron.conf, right? 18:32:02 <bh526r> I assume your current deployment testing is in this way, right? 18:32:20 <krenczewski> yes 18:32:28 <bh526r> Jin, we will talk about testing after we finish this Contrail topic 18:32:40 <bh526r> Great, Kamil. 18:33:33 <bh526r> So we don't need to worry about and maintain Contrail's mechanism driver code, and those 2 configuration changes will work together just as-is. 18:33:50 <krenczewski> yes 18:34:05 <bh526r> I am very happy to hear this. 18:34:56 <bh526r> Under this circumstance, do you feel we need to merge Contrail's mechanism driver code to Gluon repo? Or is it better to maintain it in Contrail's own repo? 18:35:25 <krenczewski> I am not sure. 18:35:47 <krenczewski> I think that Nachi should answer this question 18:35:52 <bh526r> Then it is will be deployment / installer's work to fetch those repos respectively. 18:36:28 <krenczewski> I think that this is possible 18:36:41 <bh526r> Sounds good. Let me ask Nachi, and also get input from Ian (devstack installer), Georg (Apex installer), Niko (Fuel installer) etc. 18:37:11 <bh526r> Then based on input from all of them, we can find the best way. 18:37:39 <krenczewski> OK, great. 18:39:01 <bh526r> #info In order to use Contrail's Mechanism Driver in OpenStack, we need to modify: /etc/neutron/plugins/ml2/ml2_conf.ini file 18:39:54 <bh526r> #info and also add entry to /opt/stack/neutron/neutron.egg-info/entry_points.txt 18:40:46 <bh526r> #info In section [ml2] we need to add my mechnosm driver to option "mechanism_drivers" 18:41:19 <bh526r> #info But we don't need to modify neutron.conf for Contrail's Mechanism Driver 18:42:17 <bh526r> #info On the other hand, we need to modify neutron.conf so that it points to Gluon's Extended ML2 Plugin which subclasses/wraps original core ML2 Plugin code 18:43:04 <bh526r> #info So Gluon's Extended ML2 Plugin and Contrail's Mechanism Driver are independent 18:44:18 <bh526r> #info Making both changes in neutron.conf for Gluon and changes in ml2_conf.ini and entry_points.txt will make Gluon work with Contrail through original Neutron's ML2 Plugin 18:45:13 <bh526r> #info In another word, those 2 configurations change will work together seamlessly and transparently as is 18:45:55 <bh526r> #info This has been verified by Kamil in his deployment of Contrail with Gluon 18:47:43 <bh526r> #info Regarding which repo Contrail's Mechanism Driver belongs to, Bin will communicate with Nachi, and installer assignees (Ian of devstack, Georg of Apex and Niko of Fuel) to get their input, and find the best way of hosting Contrail Mechanism Driver code as well as considering installer support 18:48:30 <bh526r> Kamil, anything else you want to add to my summary above? 18:49:19 <krenczewski> I think that is all. 18:49:33 <bh526r> Thank you Kamil 18:49:47 <bh526r> #info Task "https://jira.opnfv.org/browse/NETREADY-50" is tracked by JIRA https://jira.opnfv.org/browse/NETREADY-50 18:50:00 <krenczewski> Thanks 18:50:12 <bh526r> #info Sorry, Task "Testing" is tracked by JIRA https://jira.opnfv.org/browse/NETREADY-50 18:50:52 <bh526r> #info Patch https://review.openstack.org/#/c/388821/ and patch https://review.openstack.org/#/c/393935/ 18:51:01 <bh526r> Jin? 18:51:11 <jinli> yes 18:51:21 <jinli> sorry, didn't know you are talking with Kamil 18:52:13 <bh526r> #info Jin has fixed PEP8 issue in patch https://review.openstack.org/#/c/393935/ 18:54:42 <bh526r> Can Ian and Paul take a look at this patch https://review.openstack.org/#/c/393935/ again at your earliest convenience? 18:55:24 <bh526r> Jin, anything else from you? 18:56:08 <jinli> no, that's it 18:56:11 <jinli> thanks 18:56:24 <bh526r> OK. 18:56:59 <bh526r> #info There is another patch https://review.openstack.org/#/c/395031/ of developer documentation based on code 18:57:41 <bh526r> #info Because the code has been changed rapidly, Darla will update it 18:57:55 <bh526r> #topic AOB 18:58:07 <bh526r> Anything else from anyone? 18:58:34 <bh526r> If nothing else, let's adjourn the meeting 18:58:39 <bh526r> #info Noe 18:58:45 <bh526r> #info None 18:58:52 <bh526r> #info Meeting adjourned 18:59:08 <bh526r> Thank you Kamil, Jin and everyone 18:59:14 <bh526r> #endmeeting