18:01:47 <SumitNaiksatam> #startmeeting networking_policy
18:01:48 <openstack> Meeting started Thu Dec 11 18:01:47 2014 UTC and is due to finish in 60 minutes.  The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:01:49 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:01:52 <openstack> The meeting name has been set to 'networking_policy'
18:02:11 <SumitNaiksatam> #info agneda https://wiki.openstack.org/wiki/Meetings/Neutron_Group_Policy#Dec_11th.2C_2014
18:02:17 <SumitNaiksatam> #topic Bugs
18:02:43 <SumitNaiksatam> we have one critical
18:02:50 <SumitNaiksatam> #link https://bugs.launchpad.net/group-based-policy/+bug/1397721
18:02:58 <SumitNaiksatam> ivar-lazzaro: thats on your plate
18:03:02 <SumitNaiksatam> hemanthravi: hi
18:03:20 <SumitNaiksatam> ivar-lazzaro: i know you are swamped with other stuff
18:03:26 <hemanthravi> hi
18:03:28 <SumitNaiksatam> Youcef: hi!
18:03:35 <LouisF> hi
18:03:52 <SumitNaiksatam> LouisF: hi!
18:03:56 <SumitNaiksatam> is mageshgv here?
18:03:56 <ivar-lazzaro> SumitNaiksatam: this is actually solved by https://review.openstack.org/#/c/140201/
18:04:07 <mageshgv> SumitNaiksatam:hi
18:04:15 <ivar-lazzaro> SumitNaiksatam: not sure why the lp isn't updated
18:04:23 <SumitNaiksatam> ivar-lazzaro: awesome
18:04:28 <SumitNaiksatam> mageshgv: great thanks for joining
18:04:53 <SumitNaiksatam> ivar-lazzaro: i think you it might just a need some teaking in the commit message
18:05:24 <SumitNaiksatam> mageshgv: you seem to be making good progress with the ton of “high” priority bugs on your plate
18:05:55 <SumitNaiksatam> mageshgv: any blockers for you at this point?
18:06:22 <mageshgv> SumitNaiksatam: yes, making good progress, no blockers :)
18:06:31 <SumitNaiksatam> mageshgv: great!
18:06:59 <SumitNaiksatam> ivar-lazzaro: i did not notice, you add a second critical yesterday: #link https://bugs.launchpad.net/group-based-policy/+bug/1401347
18:07:18 <SumitNaiksatam> its a critical if there is not workaround
18:07:20 <ivar-lazzaro> yes, that one also fixed
18:07:28 <SumitNaiksatam> but anyway you have the fix
18:07:31 <ivar-lazzaro> there is not
18:07:42 <ivar-lazzaro> once the port is deleted the PT leaks forever
18:07:44 <SumitNaiksatam> minor pep8 issue
18:07:51 <SumitNaiksatam> ivar-lazzaro: okay got it
18:08:03 <SumitNaiksatam> ivar-lazzaro: i will track this as soon as you fix the pep8
18:08:16 <SumitNaiksatam> #topic Features merged in the last week
18:08:52 <SumitNaiksatam> Resources sharing (with shared attributed) : #link https://review.openstack.org/#/c/134692/
18:09:26 <SumitNaiksatam> this helps us to enable a bunch of use cases, most importantly, external connectivity (which is in review right now)
18:09:51 <SumitNaiksatam> Extensions loading from drivers: #link https://review.openstack.org/#/c/138249/
18:10:05 <SumitNaiksatam> banix: this ^^^ should make you happy :-)
18:10:37 <banix> SumitNaiksatam: that’s right. thanks.
18:10:57 <SumitNaiksatam> with the above change the GBP model becomes extensible for individual drivers if they decide to do so
18:11:07 <SumitNaiksatam> decide -> choose
18:11:16 <SumitNaiksatam> LouisF: perhaps you were also looking for the above change
18:11:28 <SumitNaiksatam> unfortunately it has landed a little late in the cycle
18:11:47 <SumitNaiksatam> but it is there, so we can start planning on how we can use it
18:11:53 <SumitNaiksatam> thanks rkukura for that
18:11:59 <LouisF> SumitNaiksatam: yes
18:12:14 <igordcard> SumitNaiksatam, hi
18:12:14 <SumitNaiksatam> Policy Rule Updates: #link https://review.openstack.org/138338
18:12:26 <SumitNaiksatam> thanks to s3wong for that
18:12:37 <SumitNaiksatam> igordcard: hi, go ahead
18:13:03 <SumitNaiksatam> s3wong: i just finished thanking you!
18:13:09 <s3wong> sorry, late
18:13:18 <SumitNaiksatam> np
18:13:20 <s3wong> bad weather to get to office
18:13:27 <SumitNaiksatam> yeah, its pretty out there
18:13:47 <SumitNaiksatam> apparently parts of SF dont have power anymore
18:14:06 <SumitNaiksatam> juice up your laptops!! :-P
18:14:17 <s3wong> would have been a good excuse to leave office if it happens here too :-)
18:14:23 <banix> did the temp drop below 50 for you guys? ;)
18:14:29 <SumitNaiksatam> s3wong: :-)
18:14:43 <s3wong> banix: actually, I don't think so, I think we are still @ around 60 :-)
18:14:45 <SumitNaiksatam> banix: okay okay…not as bad as guys
18:14:55 <SumitNaiksatam> banix: but we at least have something to whine about now :-P
18:14:59 <s3wong> banix: it is freezing out here :-)
18:15:11 <SumitNaiksatam> s3wong: lol
18:15:18 <SumitNaiksatam> okay so those were the three main features in the last week
18:15:32 <SumitNaiksatam> any questions on those?
18:16:05 <SumitNaiksatam> #topic Pending reviews (features)
18:16:25 <SumitNaiksatam> External connectivity #link https://review.openstack.org/#/q/status:open+project:stackforge/group-based-policy+branch:master+topic:bp/external-connectivity,n,z
18:16:28 <SumitNaiksatam> oops
18:16:39 <SumitNaiksatam> well may be it was the right link
18:16:59 <SumitNaiksatam> ivar-lazzaro is at it again! ;-)
18:17:18 <SumitNaiksatam> on a more serious note, this is great work in a short period of time
18:17:30 <mageshgv> SumitNaiksatam: the link doesnt show any open patch
18:17:37 <SumitNaiksatam> mageshgv: ooops
18:17:42 <ivar-lazzaro> I've given some manual testing to the patches yesterday, using implicit shared default ES
18:18:01 <SumitNaiksatam> ivar-lazzaro: okay
18:18:08 <SumitNaiksatam> perhaps ivar-lazzaro changed the topic on those patches
18:18:22 <ivar-lazzaro> everything went fine for very simple cases, I invite you all to give a try to find any bug asap :)
18:18:34 <SumitNaiksatam> so starting with this patch: #link https://review.openstack.org/#/c/137267/
18:18:59 <ivar-lazzaro> yeah now something different is in the topic field, sorry about that
18:19:02 <SumitNaiksatam> ivar-lazzaro: it will be instructive for the team if you can create a short wiki page, and outline how you have tested this for the neutron mapping
18:19:34 <SumitNaiksatam> ivar-lazzaro: i can create a landing page for you and you can fill it up per your convenience
18:19:35 <ivar-lazzaro> SumitNaiksatam: sure, I'll try to get some time to do so
18:19:43 <SumitNaiksatam> ivar-lazzaro: thanks
18:20:04 <SumitNaiksatam> this is critical functionality for this release, so please please review if you get a chance (and also to try it out)
18:20:27 <SumitNaiksatam> any questions for ivar-lazzaro on this?
18:20:47 <SumitNaiksatam> does everyone understand what ivar-lazzaro is trying to do here?
18:21:03 <SumitNaiksatam> ivar-lazzaro: perhaps a quick one line summary to level set?
18:21:12 <SumitNaiksatam> sorry to put you on the spot
18:21:17 <ivar-lazzaro> np
18:21:28 <LouisF> quick summary?
18:21:45 <ivar-lazzaro> 3 new objects have been added to the API: External Segment (ES), External Policy (EP), Nat Pool (NP)
18:22:10 <ivar-lazzaro> together, they describe how a given L3_Policy can connect to the external world
18:22:27 <ivar-lazzaro> The ES physically represents the external network/subnet
18:22:50 <ivar-lazzaro> the EP is basically an External EPG, that can provide and consume contracts on behalf of a given portion of the external world
18:22:50 <SumitNaiksatam> ivar-lazzaro: okay
18:23:09 <ivar-lazzaro> and the Nat Pool (not used in this iteration) represents the floating IPs pool
18:23:23 <SumitNaiksatam> ivar-lazzaro: so EP is a “special” EPG, right?
18:23:28 <ivar-lazzaro> yes
18:23:37 <SumitNaiksatam> ivar-lazzaro: how does it differ from EPG?
18:23:40 <ivar-lazzaro> It exists on a given External Segment
18:23:47 <SumitNaiksatam> errr…PTG
18:23:47 <ivar-lazzaro> and can provide and consume contracts
18:23:56 <ivar-lazzaro> (PRS)
18:24:04 <ivar-lazzaro> the difference is that it doesn't have endpoints
18:24:12 <SumitNaiksatam> ivar-lazzaro: ah there you go
18:24:36 <SumitNaiksatam> ivar-lazzaro: also ES and EP can be CUD only by admin?
18:24:40 <ivar-lazzaro> It defines the policy for the external segment, in particular for its routes
18:24:46 <SumitNaiksatam> ivar-lazzaro: i guess also NAT pool
18:25:09 <ivar-lazzaro> that's correct. Everything of course depends on the policy.json
18:25:25 <ivar-lazzaro> so the role access can vary dynamically
18:25:44 <ivar-lazzaro> but yeah, that's the typical use case
18:26:28 <ivar-lazzaro> any questions?
18:26:42 <SumitNaiksatam> ivar-lazzaro: and in the context of the neutron mapping, what we should be looking for is whether the external network gets created properly
18:26:58 <LouisF> ivar-lazzaro: how is the ES defined? what attributes does it have?
18:26:59 <SumitNaiksatam> ivar-lazzaro: and the router’s external gateway is set properly?
18:27:01 <ivar-lazzaro> yes, and also shared properly
18:27:21 <mageshgv> ivar-lazzaro: one quick question. Does the external segment and external policy help in modelling only the external network segment the openstack uses or does it also have a way to represent anything outside the cloud
18:27:37 <ivar-lazzaro> LouisF: #link https://github.com/stackforge/group-based-policy-specs/blob/master/specs/juno/external-connectivity.rst :)
18:27:44 <LouisF> ivar-lazzaro: thx
18:28:23 <ivar-lazzaro> mageshgv: in this first iteration, especially for the RMD, it's tied on what Neutron can represent
18:28:38 <mageshgv> ivar-lazzaro: alright
18:29:02 <ivar-lazzaro> mageshgv: however, the idea is to make the ES very very simple, so that each driver can extend it
18:29:26 <ivar-lazzaro> mageshgv: in order to represent the outside of the cloud however they need to
18:30:16 <ivar-lazzaro> mageshgv: in a way the ES is meant to be extended in order to be used properly. The plan is to add more core attributes once we realize what can be defined as "common" requirement for the external world
18:31:06 <SumitNaiksatam> ivar-lazzaro: can you provide a little more context on what “more” core attributes you anticipate
18:31:14 <mageshgv> ivar-lazzaro: thanks. that helps. wanted to make sure we have that in mind :)
18:31:18 <SumitNaiksatam> trying to get a feel for which direction we are going
18:31:39 <ivar-lazzaro> so far the core attributes are *cidr* and *port_address_translation*
18:32:06 <ivar-lazzaro> I feel that each driver will have to describe every L2 related attribute
18:32:20 <ivar-lazzaro> like encap, physnet (in Neutron's terminology) or encap_value
18:32:56 <ivar-lazzaro> however it's not clear now what is the common denominator
18:33:13 <SumitNaiksatam> ivar-lazzaro: okay makes sense
18:33:55 <SumitNaiksatam> ivar-lazzaro: i was just going to say that as long as you (as a vendor) were already mapping to the physical network in Neutron, you should be able to do it here as well
18:34:48 <SumitNaiksatam> any more questions on this?
18:35:39 <SumitNaiksatam> mageshgv: you are working on (1) hierarchical PRS (2) Policy Action updates
18:36:28 <SumitNaiksatam> Support Hierarchial Policy Rule Set for Redirects: #link https://review.openstack.org/140286
18:36:43 <mageshgv> SumitNaiksatam: yes, (1) is ready for initial review, (2) is completed
18:36:52 <SumitNaiksatam> Handle Policy Action value update(Service Chain Spec update) #link https://review.openstack.org/138338
18:37:28 <SumitNaiksatam> mageshgv: thats awesome, having a tough time keeping pace with the flurry of patches!
18:37:58 <SumitNaiksatam> any questions for mageshgv on those two?
18:38:32 <SumitNaiksatam> #topic New Policy drivers
18:38:39 <SumitNaiksatam> ODL: #link https://review.openstack.org/#/c/138556/
18:38:44 <SumitNaiksatam> this is coming along nicely
18:39:11 <SumitNaiksatam> i think this one is the first patch in the series: #link https://review.openstack.org/#/c/138235/3
18:39:43 <SumitNaiksatam> we have yi and yapeng working on the OpenStack side, and keith and thomas bachman working on the ODL side
18:39:44 <s3wong> SumitNaiksatam: finally out of WIP mode?
18:39:58 <SumitNaiksatam> s3wong: first one is, second patch is not
18:40:07 <SumitNaiksatam> s3wong: but the testing is going well
18:40:18 <s3wong> SumitNaiksatam: OK
18:40:23 <SumitNaiksatam> i mean there is end-to-end intergration tested
18:40:40 <SumitNaiksatam> with respect to EPs being registered on the ODL side
18:40:55 <SumitNaiksatam> we have not tested the datapath yet
18:41:10 <SumitNaiksatam> LouisF: you probably work with yapeng?
18:41:21 <SumitNaiksatam> is ronak here?
18:41:44 <s3wong> SumitNaiksatam: I think yapeng and LouisF are in different BUs and different countries :-)
18:42:36 <SumitNaiksatam> s3wong: yeah i know yapeng is in Canada
18:42:40 <SumitNaiksatam> #topic Packaging
18:43:00 <SumitNaiksatam> rkukura does not seem to be here (he is probably on his flight back from the neutron mid-summit)
18:43:10 <SumitNaiksatam> the update there is that this is proceeding well
18:43:30 <SumitNaiksatam> rkukura’s patch to fedora was approved for the GBP server side package
18:43:52 <SumitNaiksatam> so its a matter of now keeping it updated with the release version
18:43:56 <LouisF> SumitNaiksatam: yes
18:44:05 <SumitNaiksatam> i belive he has this done for the client as well
18:44:14 <SumitNaiksatam> and the heat and horizon packages are in the works
18:44:22 <SumitNaiksatam> #topic Open Discussion
18:44:43 <SumitNaiksatam> sorry we have been skipping the discussion on the kilo specs
18:45:04 <SumitNaiksatam> but we have too many things on our plate to finish by dec 19th, hence the delay on that discussion
18:45:25 <SumitNaiksatam> we should be able to focus our discussion on those once we reach the release point
18:45:38 <SumitNaiksatam> LouisF: thanks for being patient on that, and apologies for not reviewing your specs
18:46:24 <SumitNaiksatam> anything else we want to discuss today?
18:46:37 <SumitNaiksatam> btw, the update on the Horizon front is that its shaping up well
18:46:51 <SumitNaiksatam> there is one critical issue with creating service chain specs
18:47:09 <SumitNaiksatam> (and another regression on creating the policy classifier)
18:47:17 <SumitNaiksatam> but other than that its mostly functional
18:47:40 <LouisF> SumitNaiksatam: np
18:47:58 <SumitNaiksatam> anything else?
18:48:30 <SumitNaiksatam> ivar-lazzaro and mageshgv thanks for the herculean effort!
18:48:43 <SumitNaiksatam> alrighty, !until next week
18:48:49 <SumitNaiksatam> thanks for attending
18:48:51 <SumitNaiksatam> bye
18:48:53 <s3wong> bye
18:48:54 <mageshgv> bye
18:48:57 <banix> bye
18:48:57 <ivar-lazzaro> thanks, bye!
18:49:00 <SumitNaiksatam> #endmeeting