17:32:05 <SumitNaiksatam> #startmeeting Networking Advanced Services
17:32:05 <openstack> Meeting started Wed Jun 11 17:32:05 2014 UTC and is due to finish in 60 minutes.  The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:32:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:32:08 <openstack> The meeting name has been set to 'networking_advanced_services'
17:32:13 <SumitNaiksatam> #info agenda: https://wiki.openstack.org/wiki/Meetings/AdvancedServices
17:33:05 <SumitNaiksatam> for the last couple of weeks we have started to track the priority blueprints in this meeting
17:33:15 <SumitNaiksatam> #link https://wiki.openstack.org/wiki/Neutron/AdvancedServices/JunoPlan
17:33:37 <SumitNaiksatam> before we get into each blueprint, anything anyone wants to bring up at outset?
17:34:25 <SumitNaiksatam> i am little concerned that our specs are still in review at the end of Juno 1
17:34:25 <s3wong> SumitNaiksatam: why back to work so early :-) ?
17:34:31 <SumitNaiksatam> s3wong: ha :-)
17:34:42 <SumitNaiksatam> i will get back to real work soon
17:34:52 <SumitNaiksatam> this is just a break ;-)
17:35:03 <pgpus> Last week few were approved I thought
17:35:06 <SumitNaiksatam> lets discuss the review logistics in the open discussion
17:35:20 <SumitNaiksatam> pgpus: not sure which ones you are referring ot
17:35:24 <SumitNaiksatam> *to
17:35:57 <SumitNaiksatam> enikanorov: there?
17:36:03 <pgpus> Mot sure me too but of the 5 we had 2 or 3 were in approved state
17:36:31 <SumitNaiksatam> pgpus: only the general Juno plan is approved
17:36:31 <s3wong> pgpus: I think only SumitNaiksatam 's umbrella bp was approved
17:36:37 <SumitNaiksatam> s3wong: yeah
17:36:39 <Kanzhe> hi
17:36:46 <SumitNaiksatam> enikanorov: doesnt seem to be on
17:36:48 <gduan> Hi
17:37:00 <SumitNaiksatam> but Kanzhe is popped in at the right time :-)
17:37:07 <SumitNaiksatam> #topic Service base definition and Insertion
17:37:10 <Kanzhe> Welcome back, SumitNaiksatam
17:37:11 <s3wong> good :-)
17:37:19 <SumitNaiksatam> Kanzhe: thanks :-)
17:37:20 <pgpus> Ok I don't see any of the 5 approved, may be was referring to some other Blue prints
17:37:21 <SumitNaiksatam> #link https://review.openstack.org/#/c/93128
17:37:28 <SumitNaiksatam> pgpus: ok
17:37:45 <SumitNaiksatam> Kanzhe: there were a few pending comments on the blueprint
17:37:51 <s3wong> pgpus: this one is approved #link https://review.openstack.org/#/c/92200
17:38:05 <SumitNaiksatam> Kanzhe: are there any issues that you would want to bring up for discussion here?
17:38:18 <s3wong> SumitNaiksatam: yeah, I saw that you -1 it
17:38:19 <Kanzhe> SumitNaiksatam: Yes, I will address the comments later today.
17:38:21 <SumitNaiksatam> Kanzhe: or you can take care of the review comments?
17:38:28 <SumitNaiksatam> s3wong: yeah i did
17:39:03 <SumitNaiksatam> s3wong: i can work with you guys though, i will try not be the bottleneck :-)
17:39:17 <SridarK> Kanzhe: s3wong: I also reviewed today - have some minor clarifications
17:39:21 <s3wong> SumitNaiksatam: sure, good
17:39:30 <s3wong> SridarK: yes, I too noticed you have -1 it :-)
17:39:35 <Kanzhe> SridarK: thanks.
17:39:36 <SumitNaiksatam> so i would like to poll the other reviewers who have assigned themselves
17:39:54 <pgpus> I had seen some blue print using protocol:port being used of service insertion with firewall as opposed to L2/L3 insertion we were looking at, so are there multiple blue prints to this topic?\
17:40:10 <SumitNaiksatam> pgpus: link?
17:40:17 <SridarK> s3wong: overall looks good nothing negative here (no pun intended)
17:40:46 <s3wong> SridarK: it's OK :-) . We will address your concerns
17:40:55 <pgpus> I will send u later as it was on a diff system so later that
17:41:00 <SumitNaiksatam> LouisF: are you there?
17:41:07 <LouisF> yes
17:41:15 <SumitNaiksatam> dont see regxboi marios here
17:41:27 <SumitNaiksatam> LouisF: you signed up to review, have you reviewed?
17:41:35 <LouisF> will do so
17:41:50 <SumitNaiksatam> i am looking at the list at the top of:
17:41:54 <SumitNaiksatam> #link https://wiki.openstack.org/wiki/Neutron/AdvancedServices/JunoPlan
17:42:06 <LouisF> yes i am commited to review
17:42:18 <SumitNaiksatam> LouisF: thanks, hopefully soon :-)
17:42:21 <s3wong> Kanzhe will update soon, so we would love for everyone to review once the latest one is posted
17:42:30 <SumitNaiksatam> LouisF: we are already missing our first milestone
17:42:39 <SumitNaiksatam> is ivar here?
17:42:51 <SumitNaiksatam> probably not
17:43:01 <SumitNaiksatam> ok so i went through the assigned reviewers
17:43:15 <SumitNaiksatam> i dont want to start pinging the cores until we have consensus in the team here
17:43:28 <SumitNaiksatam> anyone has major issues with this spec?
17:43:43 <SumitNaiksatam> or its just that we havent reviewed it carefully?
17:43:56 <s3wong> SumitNaiksatam: I am going to San Antonio next week to work with LBaaS team for them to conform to this as they revamp their APIs
17:44:06 <SumitNaiksatam> s3wong: nice
17:44:12 <s3wong> So hopefully our own team has reached a consensus by then :-)
17:44:23 <SumitNaiksatam> s3wong: absolutely
17:44:45 <Kanzhe> s3wong: SumitNaiksatam It would be great to put an internal target for review feedbacks.
17:45:05 <SumitNaiksatam> Kanzhe: our target to get this approved was today
17:45:18 <s3wong> Kanzhe: great idea. LouisF, SridarK?
17:45:19 <SumitNaiksatam> Kanzhe: or if not approved at least to understand why it is not a being approved
17:45:22 <Kanzhe> SumitNaiksatam: great! :-)
17:45:33 <LouisF> will review today
17:45:40 <SumitNaiksatam> Kanzhe: unfortuntaley we dont yet have enough reviews even within the sub-team here
17:45:51 <SridarK> s3wong: yes agree
17:45:55 <SumitNaiksatam> is kevinbenton here?
17:46:22 <kevinbenton> yes
17:46:31 <Kanzhe> SumitNaiksatam: agreed. I broadcasted a plea on the mailing list, but was silently ignore.
17:46:43 <SumitNaiksatam> Kanzhe: mailing list will not help
17:46:48 <banix> I will review by Friday.
17:46:58 <SumitNaiksatam> we have to work within our team here first
17:47:09 <SumitNaiksatam> banix: thanks, i was just going to put you on the spot ;-)
17:47:11 <gduan> I will review the BP too.
17:47:19 <SumitNaiksatam> kevinbenton: you have volunteered to review
17:47:26 <SumitNaiksatam> kevinbenton: are you happy with this spec?
17:47:30 <SumitNaiksatam> gduan: thanks
17:47:39 <pgpus> Is trhere anything that needs to be updated to design https://docs.google.com/document/d/1fmCWpCxAN4g5txmCJVmBDt02GYew2kvyRsh0Wl3YF2U/edit?pli=1#
17:47:42 <SumitNaiksatam> cgoncalves: you also signed up, any comments?
17:47:47 <s3wong> Kanzhe: update by tonight then :-) ?
17:48:06 <Kanzhe> s3wong: yes.
17:48:23 <SumitNaiksatam> kevinbenton: i know you are part of the design team, but i would like to see a +1 if you dont have any issues :-)
17:48:25 <kevinbenton> SumitNaiksatam: I haven’t looked at the latest one, I will review the next upload
17:48:31 <SumitNaiksatam> kevinbenton: ok thanks
17:48:35 <cgoncalves> SumitNaiksatam: sure, will do
17:48:39 <SumitNaiksatam> pgpus: dont look at that document
17:48:41 <Kanzhe> pgpus: I don't see any design change yet.
17:48:52 <SumitNaiksatam> pgpus: #link https://review.openstack.org/#/c/93128
17:49:13 <s3wong> pgpus: yeah, please ignore the document, and focus on the gerrit spec reivew
17:49:56 <SumitNaiksatam> #action LouisF kevinbenton gduan banix cgoncalves regxboi marios ivar to review https://review.openstack.org/#/c/93128 by end of week
17:49:56 <pgpus> ok thanks
17:50:13 <SumitNaiksatam> hemanthravi: can you review as well?
17:51:08 <SumitNaiksatam> Kanzhe s3wong: i would request you to please return the review favor with the other blueprints
17:51:24 <hemanthravi> SumitNaiksatam: yes travelling this week will do it by mon
17:51:30 <SumitNaiksatam> Kanzhe s3wong: i know you guys are terribly busy, but review begets review ;-)
17:51:47 <s3wong> SumitNaiksatam: absolutely. Ready to review the other four bps for the team
17:51:47 <SumitNaiksatam> hemanthravi: thanks, can you please add yourself to the reviewers (or I can do that ;-))
17:52:01 <cgoncalves> SumitNaiksatam: as owner of that ^ google doc, should one (you?) add a warning message to it saying that doc is deprecated and pointing to the right URL (blueprint URL)?
17:52:05 <Kanzhe> SumitNaiksatam: which one?
17:52:16 <hemanthravi> SumitNaiksatam: this for 93128 right?
17:52:20 <SumitNaiksatam> cgoncalves: good suggestion, i will do it, my bad
17:52:28 <SumitNaiksatam> hemanthravi: correct
17:52:35 <SumitNaiksatam> Kanzhe: all the others
17:52:39 <hemanthravi> SumitNaiksatam: i'll do that
17:52:45 <s3wong> Kanzhe: all the other four below ours #link https://wiki.openstack.org/wiki/Neutron/AdvancedServices/JunoPlan
17:52:55 <SumitNaiksatam> Kanzhe: at least the flavors, traffic steering and chaining
17:53:01 <SumitNaiksatam> enikanorov: there?
17:53:06 <Cathy_> I will review the blueprints too, service chaining specifically
17:53:13 <Kanzhe> SumitNaiksatam: Kanzhe yes. Will do it by the weekend.
17:53:18 <s3wong> Kanzhe: Oh, and flavor also...
17:53:27 <SumitNaiksatam> Cathy_: nice, thanks much in advance
17:53:53 <SumitNaiksatam> Cathy_: if you feel comfortable please add yourself to the reviewers list: https://wiki.openstack.org/wiki/Neutron/AdvancedServices/JunoPlan
17:54:02 <Cathy_> sure, will do
17:54:03 <SumitNaiksatam> Cathy_: that way I can hound you ;-P
17:54:36 <SumitNaiksatam> ok it doesnt seem enikanorov is still not around
17:54:44 <SumitNaiksatam> #topic Traffic steering
17:54:52 <SumitNaiksatam> #link https://review.openstack.org/#/c/92477
17:55:06 <SumitNaiksatam> cgoncalves: are you planning a new patch?
17:55:47 <cgoncalves> SumitNaiksatam: we'd like first to get input from banix et al. on Joao's last comments
17:56:05 <SumitNaiksatam> cgoncalves: ok please go ahead, we can have the discussion now and resolve it if possible
17:56:08 <SumitNaiksatam> banix: ?
17:56:21 <banix> cgoncalves: will do by end of day (night) today
17:56:30 <cgoncalves> banix: thanks
17:56:38 <enikanorov> SumitNaiksatam: sorry, i'm late
17:56:55 <s3wong> SumitNaiksatam: now we can go to flavor!
17:56:55 <enikanorov> i'll give an update when you give me a timeslot
17:57:08 <cgoncalves> all: note that I've submitted three patches for reviewing but marked as WIP
17:57:11 <cgoncalves> #link https://review.openstack.org/#/q/topic:bp/traffic-steering-abstraction,n,z
17:57:12 <SumitNaiksatam> enikanorov: no worries, we can do flavors next, people are getting restless without the customary start with flavors discussion ;-P
17:57:24 <enikanorov> :)
17:57:25 <SumitNaiksatam> cgoncalves: nice
17:57:36 <s3wong> cgoncalves: OK - still need to review your spec; sorry for the delay
17:57:38 <cgoncalves> marked as WIP because BP has not yet been approved
17:58:09 <SumitNaiksatam> cgoncalves: thats the righ approach, very much appreciate the process you are following
17:58:17 <cgoncalves> more codebase will follow once more reviewing is given
17:58:30 <SumitNaiksatam> cgoncalves: which other reviewer do you want to check with?
17:58:49 <SumitNaiksatam> *reviewers
17:59:05 <cgoncalves> SumitNaiksatam: I'm not sure it's the best approach to follow, though, but I will follow a similar codebase as ML2 and GP. how does that sound to you all?
17:59:16 <pgpus> I can provided I get some help on Gerrit from one of you?
17:59:18 <SumitNaiksatam> cgoncalves: yeah good
17:59:26 <SumitNaiksatam> pgpus: nice, much appreciated
17:59:44 <SumitNaiksatam> is Youcef_ here?
17:59:44 <cgoncalves> SumitNaiksatam: Ryan Moats as he brought some comments too
17:59:57 <pgpus> OK  I will review the 3 of the listed one and work with cgoncalves
18:00:03 <SumitNaiksatam> cgoncalves: yes, but i dont see ryan here today
18:00:07 <cgoncalves> pgpus: thanks!
18:00:27 <SumitNaiksatam> Youcef_ had comments on both this and the insertion bp
18:00:34 <SumitNaiksatam> *good comments
18:00:37 <banix> cgoncalves: ryan is out the rest of the week but let me go through your comments
18:00:40 <cgoncalves> SumitNaiksatam: yes. better take this discussion to gerrit
18:00:44 <SumitNaiksatam> i will hound him :-)
18:00:51 <SumitNaiksatam> cgoncalves: ok
18:01:05 <cgoncalves> banix: thanks ;)
18:01:12 <SumitNaiksatam> anyone else want to bring up an technical issue with the traffic steering blueprint?
18:01:17 <banix> cgoncalves: np
18:01:19 <SumitNaiksatam> *any
18:01:37 <cgoncalves> I think there may be some other folks interested in this BP
18:01:47 <cgoncalves> not sure, though, if any of them are here
18:01:56 <s3wong> cgoncalves: outside of this subteam?
18:02:18 <s3wong> cgoncalves: the NFV folks, perhaps?
18:02:22 <cgoncalves> s3wong: yes. NFV team
18:02:39 <cgoncalves> s3wong: https://wiki.openstack.org/wiki/Meetings/NFV
18:02:51 <cgoncalves> s3wong: search for "steering", for instance
18:03:03 <SumitNaiksatam> #action banix pgpus regxboi Kanzhe s3wong kevinbenton hemanthravi LouisF  to review https://review.openstack.org/#/c/92477 and provide feedback by end of week
18:03:07 <SumitNaiksatam> cgoncalves: yeah
18:03:14 <SumitNaiksatam> cgoncalves: did you attend the NFV meeting?
18:03:26 <s3wong> cgoncalves: you may want to add your bp on the wiki page here #link https://wiki.openstack.org/wiki/Meetings/NFV
18:03:29 <SumitNaiksatam> or anyone else?
18:03:35 <banix> SumitNaiksatam: liking these action items already :)
18:03:35 <s3wong> SumitNaiksatam: I have
18:03:42 <cgoncalves> I also just got a contact from someone else not in these teams asking for more details
18:03:47 <banix> SumitNaiksatam: sounds like a good way of tracking
18:03:49 * SumitNaiksatam hides for cover :-P
18:03:49 <cgoncalves> SumitNaiksatam: I did
18:04:04 <cgoncalves> s3wong: it is listed there. check the bottom of second table
18:04:06 <SumitNaiksatam> cgoncalves: ok good
18:04:08 <s3wong> cgoncalves: that's good. Ask them to join as reviewers :-)
18:04:16 <SumitNaiksatam> s3wong: nice one
18:04:38 <SumitNaiksatam> banix: i am happy being the bad guy here
18:04:45 <cgoncalves> s3wong: sure!
18:04:45 <s3wong> cgoncalves: got it, didn't scroll all down enough. Sorry
18:05:18 <SumitNaiksatam> cgoncalves: we will have a quick NFV update later, perhaps you can do that
18:05:22 <SumitNaiksatam> ok flavors
18:05:30 <SumitNaiksatam> #topic Flavors
18:05:32 <banix> SumitNaiksatam: no i think this is a good way of tracking things; when one have an action item you pay more attention; we all want to do the reviews but things get pushed around wrt priority; so i like the approach
18:05:49 <SumitNaiksatam> banix: exactly
18:05:53 <cgoncalves> SumitNaiksatam: later this meeting or?
18:05:58 <SumitNaiksatam> #link https://review.openstack.org/#/c/90070
18:06:06 <SumitNaiksatam> cgoncalves: yes later in the agenda
18:06:10 <SumitNaiksatam> enikanorov: ?
18:06:17 <enikanorov> here
18:06:22 <enikanorov> ok
18:06:29 <SumitNaiksatam> enikanorov: is there another patch coming?
18:06:40 <enikanorov> SumitNaiksatam: you mean spec?
18:06:46 <SumitNaiksatam> enikanorov: yeah
18:07:02 <enikanorov> yes, i think the only remaining question is about tags format
18:07:06 <SumitNaiksatam> enikanorov: i meant patch set, sorry
18:07:28 <enikanorov> my understanding was that with many different requirements it might be more flexible to hae it in a form of string
18:07:34 <SumitNaiksatam> enikanorov: so who is blocking that?
18:07:55 <enikanorov> there was a couple of questions along the way about that
18:08:05 <enikanorov> some asked if we need additional model for Tag
18:08:06 <SumitNaiksatam> enikanorov: ok
18:08:15 <enikanorov> and add Tags one by one to Flavor
18:08:24 <enikanorov> but I think it's too complex to be usable
18:08:33 <SumitNaiksatam> enikanorov: i made the comment about there being more structure
18:08:45 <SumitNaiksatam> nati_ueno: there?
18:08:50 <enikanorov> but technically, what kind of structure it could be?
18:09:18 <garyduan> enikanorov: Does the spec define how "supported capabilities" are inputed?
18:09:27 <SumitNaiksatam> enikanorov: i was just thinking something more than a one single flat string for all the tags and valures
18:09:31 <SumitNaiksatam> *values
18:09:45 <enikanorov> SumitNaiksatam: so what is that 'something'?
18:09:47 <SumitNaiksatam> enikanorov: i will not block this
18:10:00 <SumitNaiksatam> enikanorov: something similar to the dict that i was suggesting in the comments
18:10:08 <pgpus> Diff services have diff capability so how do we structure flavor, more likely string format with key values should be Ok
18:10:11 <enikanorov> garyduan: "cap_name:cap_value,cap_name2:cap_value2"
18:10:52 <pgpus> That spound perfect provided we know service instace can use them with correct interpretation
18:10:59 <garyduan> enikanorov: what I mean is there is a predefined set of allowed tags that driver can use
18:11:18 <SumitNaiksatam> garyduan: my earlier suggesting was along similar lines
18:11:31 <enikanorov> garyduan: I would be glad if someone could help me with defining that
18:11:53 <SumitNaiksatam> enikanorov: is there a possibility that two different drivers can have the same capability name and be using them in different ways?
18:11:57 <enikanorov> although i'm not sure it shoul be necessarily in flavor API implementation
18:12:27 <enikanorov> SumitNaiksatam: that should be avoided. Capabilites are user facing so they create expectations
18:12:34 <pgpus> OK unfortunately unlike nova flavor we do not have fixed mem storage like common absractions fully similar across services
18:12:43 <enikanorov> expectations of consistency
18:12:48 <SumitNaiksatam> pgpus: yes that is the issue
18:13:02 <enikanorov> since user doesn't know what implementation he gets, it should be consistent across drivers
18:13:09 <SumitNaiksatam> enikanorov: but without being prescriptive, it is difficult to be consistent
18:13:22 <garyduan> enikanorov: what these tags are can be figured out later for each services
18:13:24 <enikanorov> SumitNaiksatam: please explain?
18:13:28 <SumitNaiksatam> enikanorov: so who ensures the consistency?
18:13:47 <garyduan> enikanorov: but where to define them. Are they hard-coded in Neutron?
18:13:48 <pgpus> Atleast service_type is common across all falvors
18:13:49 <enikanorov> I think deployers/cloud admins should ensure it
18:14:20 <SumitNaiksatam> enikanorov: but the deployers are different from the entities who develop the drivers
18:14:34 <SumitNaiksatam> enikanorov: and we are saying that the drivers express their capabilities
18:14:47 <pgpus> If Provider has a different concept from same service_type they should be able to override the service_type\
18:14:55 <enikanorov> yes, they're different. Those who maintain service should ensure that certain feature works similarly in all drivers
18:15:18 <enikanorov> also, driver authors should be verbose in defining capabilities supported by their driver
18:15:18 <garyduan> I think what Sumit means is
18:15:32 <garyduan> Operator A may want to have tag X, Y, Z
18:15:44 <garyduan> and B wants tag S, T, W
18:16:20 <garyduan> as vendor driver, which tag set should it expose?
18:16:45 <enikanorov> the tag names should be hardcoded unless it's vendor-specific
18:16:55 <SumitNaiksatam> enikanorov: so essentially, you seem to be saying that at the time of reviewing a driver, we as reviewers should ensure consistency?
18:17:09 <SumitNaiksatam> * driver code
18:17:19 <pgpus> Every Service_Type should have atleast two or three standard attributes like for for firewall igress, egress and l2 or l3 should be minmum just for exaple sake
18:17:23 <s3wong> enikanorov: so should the hardcoded tag names be part of the spec/API/DB?
18:17:45 <enikanorov> yes, I think every driver should implement some feature X such that user would have same experience with it
18:17:45 <garyduan> s3wong: my question too.
18:17:50 <pgpus> So tags S=A has A,B,C ... anything after that
18:17:57 <enikanorov> and we as reviewers should ensure that
18:18:06 <pgpus> Service-B has, A,B,C.. anything after that
18:18:13 <SumitNaiksatam> enikanorov: in that case, can i suggest an evolutionary approach
18:18:20 <pgpus> both S-A & S-B being two implementations of FW
18:18:37 <enikanorov> SumitNaiksatam: please explain?
18:18:48 <SumitNaiksatam> enikanorov: how about we define a module where we note these capability names as they are populated by drivers?
18:18:59 <pgpus> same with LB that may have S-A A,B,C,D ... and any more
18:19:02 <enikanorov> SumitNaiksatam: yes, that will work for us
18:19:09 <SumitNaiksatam> enikanorov: since the problem now seems to be to able to identify these tags at the outset
18:19:13 <pgpus> and S-B A,B.C,D and any thing more
18:19:37 <pgpus> but for a give Service A,B,C or A,BC,D must be common minimum
18:19:40 <SumitNaiksatam> enikanorov: every new tag name should be added to this capabilites module (like a common constants module)
18:19:58 <enikanorov> SumitNaiksatam: yes, something like that
18:20:23 <s3wong> SumitNaiksatam: enikanorov: OK - that makes sense, separate framework from service-type-specific attributes
18:20:24 <SumitNaiksatam> enikanorov: that way we have the tag (names) all defined in one place, which evolve over time, but we ensure that at least people are aware of what is being used
18:20:42 <enikanorov> SumitNaiksatam: agree
18:20:48 <garyduan> SumitNaiksatam: these are hardcoded "common tags or specific to service types" tag, right?
18:20:52 <SumitNaiksatam> enikanorov: we should also ensure that these tag names are documented when they are populated
18:20:53 <pgpus> OK that looks good so you define a set of constant which are part of standards for that specifc service and rest are optional
18:21:02 <SumitNaiksatam> pgpus: yes
18:21:13 <enikanorov> SumitNaiksatam: ok
18:21:27 <SumitNaiksatam> enikanorov: to help us maintain consistency
18:21:29 <garyduan> agree
18:21:35 <SridarK> SumitNaiksatam: enikanorov so there is a review process to get a new tag added ?
18:21:35 <pgpus> I am ok with that, that will work
18:21:38 <nati_ueno> Hi!
18:21:41 <enikanorov> #action enikanorov to add notes on tags consistency to blueprint spec
18:21:57 <enikanorov> SridarK: common sense? :)
18:22:02 <SumitNaiksatam> SridarK: that would be part of the review for the driver, enikanorov right?
18:22:10 <enikanorov> SumitNaiksatam: sure
18:22:13 <SumitNaiksatam> nati_ueno: hi, wanted to put you on the spot :-)
18:22:21 <nati_ueno> SumitNaiksatam: Show time!
18:22:26 <SumitNaiksatam> nati_ueno: did you get a chance to look at the flavors spec?
18:22:27 <SridarK> enikanorov: SumitNaiksatam yes that will help excess proliferation.
18:22:34 <s3wong> SridarK: I would imagine who ever add their service's flavor support would have to define tags and have them reviewed
18:22:37 <SumitNaiksatam> nati_ueno: #link https://review.openstack.org/#/c/90070
18:22:52 <nati_ueno> SumitNaiksatam: I didn't read the latest yet. please let me have a look
18:23:02 <SridarK> s3wong: yes a sort of IANA allocation will keep things sane
18:23:13 <SumitNaiksatam> nati_ueno: thanks, this is blocker for lots of services’ stuff
18:23:25 <SumitNaiksatam> SridarK: nice analogy :-)
18:23:27 <s3wong> SridarK: in here we have Neutron core-dev to serve the role :-)
18:23:34 <enikanorov> let nati_ueno put another -1 so i could resolve his comments as well :)
18:23:49 <SridarK> :-)
18:24:06 <SumitNaiksatam> nati_ueno: enikanorov is challenging you to put a +2 i think :-)
18:24:07 <enikanorov> So with regards to implementation that is under way, I'd like to put everything that is not yet fully decided - out of the patch
18:24:20 <enikanorov> so it only will consist of API and db part (+UTs of course)
18:24:27 <nati_ueno> enikanorov: he he that challenge is accepted
18:24:30 <enikanorov> SumitNaiksatam: eventually, yes :)
18:24:32 <s3wong> enikanorov: makes sense
18:25:05 <SumitNaiksatam> enikanorov: have we identified people who will work on each service?
18:25:35 <enikanorov> SumitNaiksatam: no, i don't think so. garyduan for fwaas, ... ?
18:25:53 <s3wong> SumitNaiksatam: enikanorov himself for LBaaS? That is, once the dust is settled there :-)
18:25:54 <enikanorov> i'll gladly help with integration
18:26:04 <enikanorov> it's a bit early to say about lbaas
18:26:08 <SumitNaiksatam> #agreed nati_ueno accepts enikanorov challenge to +2 flavors patch https://review.openstack.org/#/c/90070
18:26:10 <enikanorov> after the code sprint - may be!
18:26:10 <garyduan> enikanorov: please also explain in the spec how vendor should expose "specific to vendor" tags
18:26:20 <garyduan> enikanorov: anything not hardcoded?
18:26:35 <SumitNaiksatam> garyduan: are you doing fwaas?
18:26:36 <s3wong> enikanorov: so optimistic :-)
18:26:39 <garyduan> yes
18:26:43 <SumitNaiksatam> garyduan: nice
18:26:49 <enikanorov> garyduan: vendor exposes to admin, admin decides wether to put those caps into flavors
18:26:54 <SumitNaiksatam> nati_ueno: what about vpnaas?
18:27:20 <garyduan> enikanorov: ok.
18:27:45 <nati_ueno> SumitNaiksatam: I think there is nothing special for vpnaas
18:27:51 <SumitNaiksatam> nati_ueno: ok
18:27:52 <nati_ueno> accoding to the flavor part
18:28:11 <nati_ueno> I feel there is more and more vendor paramters in the vpnaas, but it is another issue
18:28:32 <nati_ueno> may be we should bind flavor and such extended parater for validation ..
18:28:32 <SumitNaiksatam> ok as a team can we agree to resolve all issues with the flavors spec by next week; i would like to see a bunch of +1s by that time
18:28:47 <SumitNaiksatam> and i would like to see some +1s to my comment now ;-P
18:29:00 <rkukura> SumitNaiksatam: I’ll be sure to review that one
18:29:00 <s3wong> SumitNaiksatam: +1
18:29:08 <SumitNaiksatam> rkukura: thanks!
18:29:28 <SridarK> SumitNaiksatam: +1
18:29:34 <garyduan> SumitNaiksatam: +1
18:29:38 <Kanzhe> +1
18:29:41 <vinay_yadhav> +1
18:29:42 <banix> sounds good
18:29:45 <enikanorov> +1!
18:30:16 <s3wong> SumitNaiksatam: time's up for the meeting, BTW
18:30:26 <SumitNaiksatam> #action rkukura s3wong SridarK garyduan Kanzhe vinay_yadhav banix nati_ueno SumitNaiksatam to review flavors patch and resolve issues before next weeks meeting
18:30:35 <cgoncalves> SumitNaiksatam: +1 (just because I'm scare of being haunted by you)
18:30:35 <SumitNaiksatam> s3wong: yes we are running a little behind
18:30:40 <SumitNaiksatam> cgoncalves: ha
18:30:47 <SumitNaiksatam> cgoncalves: you managed to escpat
18:31:00 <SumitNaiksatam> few more minutes if people are willing to stay
18:31:08 * cgoncalves does the chicken dance
18:31:13 <SumitNaiksatam> cgoncalves: ha
18:31:16 <vinay_yadhav> sumit: TaaS
18:31:17 <pgpus> Sumit Thanks will follow with cg
18:31:21 <SumitNaiksatam> vinay_yadhav: yes
18:31:37 <SumitNaiksatam> #topic Tap Service
18:31:40 <s3wong> SumitNaiksatam: let's skip updates on NFV and serviceVM for this week then
18:31:43 <vinay_yadhav> we have resolved some comment from previous patch and have got some +1
18:31:56 <SumitNaiksatam> #link https://review.openstack.org/#/c/96149/
18:32:01 <SumitNaiksatam> s3wong: yes
18:32:02 <garyduan> vinay_yadhav: thanks for initiating the spec
18:32:23 <s3wong> vinay_yadhav: nice
18:32:25 <garyduan> vinay_yadhav: have you considered to support one-arm type of service as well
18:32:28 <SumitNaiksatam> vinay_yadhav: nice work, i do see +1s
18:32:32 <vinay_yadhav> We would like to see more reviews so that we can get the spec accepted
18:32:45 <SumitNaiksatam> vinay_yadhav: sorry, i did not get a chance to review the latest
18:32:45 <garyduan> vinay_yadhav: besides mirroring
18:32:57 <SumitNaiksatam> vinay_yadhav: any blockers at this point?
18:33:15 <vinay_yadhav> garyduan: not as of yet
18:33:21 <vinay_yadhav> i dont see any
18:33:56 <Kanzhe> vinay_yadhav: I will review your spec this weekend.
18:34:03 <SumitNaiksatam> Kanzhe: thanks
18:34:06 <vinay_yadhav> Kanzhe: Thanks
18:34:23 <SumitNaiksatam> #action SumitNaiksatam Kanzhe garyduan to review https://review.openstack.org/#/c/96149/
18:34:42 <vinay_yadhav> cool thanx
18:34:50 <anil_rao> Thanks
18:35:00 <garyduan> vinay_yadhav: I will add comments on the spec
18:35:08 <SumitNaiksatam> i dont see mandeep, so i will skip service chaining, please respond to the review: https://review.openstack.org/#/c/93524
18:35:18 <vinay_yadhav> garyduan: thanx
18:35:19 <s3wong> vinay_yadhav: I will review as well (actually I already added myself as reviewer anyway)
18:35:26 <SumitNaiksatam> #topic Open Discussion
18:35:37 <vinay_yadhav> s3wong: sure
18:35:49 <SumitNaiksatam> anythin anyone wants to bring up?
18:36:14 <SumitNaiksatam> i am really hoping that we can get consensus within the team to +1 the specs by next week
18:36:15 <cgoncalves> SumitNaiksatam: I think jmsoares has something
18:36:25 <SumitNaiksatam> we can accordingly go to the core reviewers
18:36:31 <SumitNaiksatam> jmsoares: sure, please go ahead
18:36:43 * SumitNaiksatam appreciates everyone staying longer
18:36:58 <jmsoares> about the NFV meeting:  apart from logistic stuff, the discussion was focused on 1) Gap analysis (functional), 2) NFV use-cases, and 3) Workload analysis (performance).
18:37:03 * SumitNaiksatam and thanks fwaas team for always being patient
18:37:18 <SumitNaiksatam> jmsoares: great, thanks for that update
18:37:19 <jmsoares> 1) What ETSI NFV is defining that doesn't (currently) align with OpenStack. Some members that are in ETSI will try to bring some of the most relevant (to OpenStack) draft documents public.
18:37:28 <jmsoares> 2) Do a gap analysis focused on NFV use-cases.
18:37:31 <SumitNaiksatam> jmsoares: any action items for us?
18:37:39 <jmsoares> this is all I remember :)
18:37:57 <SumitNaiksatam> jmsoares: sure, any immediate action items for us?
18:38:00 <jmsoares> not really.
18:38:03 <s3wong> SumitNaiksatam: currently some NFV team people (including me) will map NFV requirements to the list of BPs
18:38:05 <SridarK> SumitNaiksatam: no worries
18:38:06 <SumitNaiksatam> jmsoares: ok
18:38:14 <s3wong> SumitNaiksatam: I will make sure our BPs get prominently featured :-)
18:38:23 <SumitNaiksatam> s3wong: great, thanks :-)
18:38:31 <SumitNaiksatam> SridarK: thanks
18:38:37 <SumitNaiksatam> ok anything else that we missed?
18:38:38 <jmsoares> s3wong: exatcly...that's the main action point in the group now.
18:39:14 <SumitNaiksatam> alrighty, lets call it a wrap
18:39:22 <s3wong> SumitNaiksatam: +1
18:39:27 <SumitNaiksatam> please review the action items after the meeting, and act on them :-P
18:39:30 <banix> bye everybody
18:39:33 <SumitNaiksatam> thanks all!
18:39:34 <SumitNaiksatam> bye
18:39:36 <s3wong> thanks!
18:39:39 <SumitNaiksatam> #endmeeting