16:03:05 <sridhar_ram> #startmeeting tacker
16:03:07 <openstack> Meeting started Tue Jun 28 16:03:05 2016 UTC and is due to finish in 60 minutes.  The chair is sridhar_ram. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:03:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:03:10 <openstack> The meeting name has been set to 'tacker'
16:03:22 <sridhar_ram> #topic Roll Call
16:03:30 <KanagarajM> hi
16:03:32 <sridhar_ram> who is here ?
16:03:38 <vishwanathj> o/
16:03:41 <s3wong> o/
16:03:41 <janki_> o/
16:03:50 <trozet> o/
16:03:57 <sripriya> o/
16:03:58 <tung_doan> hi all
16:04:13 <sridhar_ram> hello all !
16:04:25 <tbh> o/
16:04:26 <sridhar_ram> let's start..
16:04:31 <sridhar_ram> #topic Agenda
16:04:44 <sridhar_ram> #link https://wiki.openstack.org/wiki/Meetings/Tacker
16:05:33 <sridhar_ram> Also, given tung_doan is here i would like to get few words from his side on alarm-monitor
16:05:51 <tung_doan> sridhar_ram: ok
16:05:51 <sridhar_ram> #topic Annoucements
16:06:38 <sridhar_ram> #info Tacker client version 0.4.0 (newton)  is released
16:06:41 <sridhar_ram> #link http://lists.openstack.org/pipermail/openstack-announce/2016-June/001259.html
16:07:05 <sridhar_ram> We have been making regular releases off stable branches. This is the first time we are starting to make a tackerclient release off master.
16:07:22 <KanagarajM> nice !
16:07:30 <sridhar_ram> will add tackerclient to global-requirements for us and other projects (like Mistral) to use
16:07:51 <sridhar_ram> FYI, https://review.openstack.org/#/c/334641/
16:07:54 <sripriya> cool!
16:07:55 <sridhar_ram> KanagarajM: thanks!
16:07:59 <vishwanathj> cool
16:08:07 <sridhar_ram> yes, finally!
16:08:44 <sridhar_ram> OpenStack Barcelona Summit talk submission deadline is fast approaching
16:09:11 <sridhar_ram> You folks should try to put a good set of submissions on Tacker!
16:09:21 <sridhar_ram> s/a/a lot of/
16:09:36 <KanagarajM> sridhar_ram, i'm thinking of the Scaling could be nice topic :)
16:09:50 <KanagarajM> with alarm in place
16:09:50 <sridhar_ram> KanagarajM: of course !
16:09:55 <sridhar_ram> JULY 13, 2016 AT 11:59PM PDT (JULY 14 6:59 UTC) IS THE DEADLINE TO SUBMIT A TALK
16:10:20 <sridhar_ram> Please watchout for local UTC time... and don't make it last minute :)
16:10:30 <tung_doan> KanagarajM: scaling + monitoring is impossible? I try to do it
16:10:33 <sridhar_ram> moving on...
16:10:56 <sridhar_ram> #topic OPNFV Summit Update
16:11:16 <sridhar_ram> Videos are available at Videos - https://www.youtube.com/channel/UC3EjXLJbub0tPFpnI3vEmYg
16:11:34 <sridhar_ram> Tacker design summit session is available at https://www.youtube.com/watch?v=DSaQth_nJHI
16:12:00 <sridhar_ram> In short, lot of NFV MANO discussions and sessions..
16:12:17 <sridhar_ram> Open-O, OpenBaton and Tacker was discussed quite a bit
16:12:56 <sridhar_ram> Two demos using Tacker + VNFFFG where shown in the demo floor
16:13:09 <sridhar_ram> Kudos to trozet efforts !!
16:13:26 <trozet> thanks :)
16:13:35 <vishwanathj> +1
16:13:38 <tbh> +1
16:13:40 <sripriya> great job trozet!
16:13:41 <KanagarajM> good to see the tacker values :)
16:13:53 <sridhar_ram> we also need to engage more with various OPNFV Projects
16:14:21 <sridhar_ram> FuncTest is of immediate importance... we need Tacker tests running in OPNFV Scenario testbed
16:14:31 <bobh> sridhar_ram: sorry late
16:14:55 <sridhar_ram> We also have an oppurtunities to engage with lots of other projects like VNF Event Stream (VES), Domino, ...
16:15:05 <sridhar_ram> bobh: howdy! we are just warming up
16:15:20 <sridhar_ram> There is also a main Tacker talk session, still waiting for the youtube link...
16:15:25 <sridhar_ram> will pass it onto the team...
16:15:29 <sridhar_ram> when i get it
16:15:54 <sridhar_ram> meanwhile, if some you want to work on Tacker on the OPNFV project do let me know.. i will connect you with the right people
16:16:02 <sridhar_ram> *some of you
16:16:17 <sridhar_ram> any questions ?
16:16:33 <sridhar_ram> moving on..
16:16:46 <sridhar_ram> #topic Alarm-Monitoring
16:17:02 * sridhar_ram apologizes for the slight change in agenda order
16:17:15 <sridhar_ram> tung_doan: can you please give an update where this spec stands ?
16:17:37 <sridhar_ram> reviewers of alarm-spec : please pitch in :)
16:18:05 <tung_doan> sridhar_ram: I already updated spec last week: https://review.openstack.org/#/c/306562/14/specs/newton/alarm-based-monitoring-driver.rst
16:19:09 <tung_doan> sridhar_ram: i wtill wait for reviewers. Also, I updated CLI cmmand  in L146. Let see.
16:19:48 <sridhar_ram> tung_doan: one area i've some concern is, this spec refers to some scaling related things.. wouldn't that be an overlap with what is in the scaling spec ?
16:20:24 <tung_doan> sridhar_ram: absolutely not.
16:20:56 <sridhar_ram> ah, perhaps i was looking at an older version.. my bad
16:21:12 <sridhar_ram> alright, cool then..
16:21:30 <tung_doan> sridhar_ram: IMO, scaling will be done once triggers sent by Ceilometer. it totally different scenario
16:21:48 <sridhar_ram> bobh: as usual i'm going to bug you to review this  :)
16:21:58 <sridhar_ram> tung_doan: perfect!
16:22:03 <tung_doan> sridhar_ram:  currently, I am going with this
16:22:20 <bobh> sridhar_ram: I figured :-)
16:22:34 <sridhar_ram> tung_doan: what happens after an alarm from ceilometer is totally upto what is specified in the TOSCA template..
16:23:02 <sridhar_ram> tung_doan: .. at some point in the future an action might involving sending an SMS - who knows
16:23:22 <sridhar_ram> *involve
16:23:34 <sridhar_ram> tung_doan: thanks for the update
16:23:40 <tung_doan> sridhar_ram:  regrading to notification like SMS, I already think about it
16:23:59 <sridhar_ram> tung_doan: ha, ha.. that is not futuristic enough :)
16:24:12 <sripriya> tung_doan: will you be creating a new abstract driver for the alarm framework?
16:24:45 <tung_doan> sridhar_ram:  right
16:25:56 <sridhar_ram> tung_doan: did u catch srirpriya's questions ?
16:26:31 <tung_doan> sridhar_ram:  what question? sridhar :)
16:26:54 <sridhar_ram> nevermind, let's take it to #tacker after the meeting
16:27:02 <sridhar_ram> need to get back to the rest of the agenda
16:27:07 <sridhar_ram> tung_doan: thanks for the update..
16:27:22 <sridhar_ram> i hope we can land alarm and scaling spec soon
16:27:27 <sridhar_ram> and move on to code-reviews
16:27:38 <tung_doan> sridhar_ram:  please take some minitues to review my specs.. Thanks
16:28:03 <sridhar_ram> tung_doan: bobh is on it ;-)
16:28:15 <sridhar_ram> #topic VNFFG --> VNFM --> networking-sfc interaction
16:28:36 <sridhar_ram> s3wong: trozet: that was a good discussion in the channel on this topic..
16:28:48 <sridhar_ram> I've capture this based on that...
16:29:00 <sridhar_ram> #link http://paste.openstack.org/show/523666/
16:29:36 <sridhar_ram> FYI, for the folks interested in the VNF-FFG workflow... please take a look at the ^^ diagram
16:29:56 <sridhar_ram> trozet: s3wong: does this flow sounds about right ?
16:30:23 <s3wong> sridhar_ram: looks good
16:30:40 <trozet> what is the vnffg abstract driver?
16:30:52 <trozet> oh nvm
16:30:53 <s3wong> trozet: abstract class for all drivers
16:30:55 <trozet> i forgot about hte new driver
16:31:07 <s3wong> trozet: common interface for all drivers
16:31:16 <trozet> shouldnt VNFM also have an abstract driver?
16:31:31 <KanagarajM> looks reasonable ! also should we look to typing the descriptor in the catalog http://paste.openstack.org/show/523666/
16:32:20 <KanagarajM> typo ^^ https://review.openstack.org/#/c/333216/1/tacker/extensions/nfvo.py
16:32:23 <s3wong> trozet: it really should :-)
16:32:38 <sridhar_ram> trozet: i didn't get the question.. we do have abstract infra drivers.. https://github.com/openstack/tacker/blob/master/tacker/vm/infra_drivers/abstract_driver.py
16:33:11 <trozet> maybe the diagram just doesn't show it
16:33:18 <s3wong> sridhar_ram: I think trozet was talking about the diagram
16:34:30 <sridhar_ram> s3wong: trozet: do you mean a VNFM abstract driver for NFVO to call ?
16:34:32 <sripriya> trozet: are you referring to the heat infra_driver?
16:34:42 * sridhar_ram is still puzzled
16:34:54 <trozet> there is just a direct line from VNFM to heat
16:35:13 <trozet> i would htink there would be abstraction there, since there is abstraction between VNFFG to it's VIM type
16:35:18 <s3wong> sridhar_ram: in the diagram, between VNFM plugin and Heat infra driver, there should be an infra abstract driver layer
16:35:21 <trozet> i don't mean to hold up the meeting
16:35:56 <sridhar_ram> trozet: ah, yes, indeed.. there is an infra abstract driver as i poiinted out earlier
16:36:06 <sridhar_ram> trozet: no worries...
16:36:26 <sridhar_ram> trozet: you never know where some wrong assumptions
16:36:55 <sridhar_ram> trozet: s3wong: i'd like to get some clarity on this get_vnf_details() ...
16:36:57 <sripriya> sridhar_ram: s3wong: trozet: are we introducing a new abstract driver for vnffg or will it merge with existing abstract_vim_driver?
16:37:39 <s3wong> sripriya: probably new, as the driver is specific to VNFFG
16:37:46 <s3wong> sridhar_ram: yes
16:37:48 <trozet> sripriya: i'm not sure yet because I haven't looked at the VIM driver
16:37:55 <sridhar_ram> sripriya: i'd suggest we leave this separate and no tie w/ abstract_vim_driver
16:38:13 <sripriya> s3wong: sridhar_ram: ok
16:38:46 <sridhar_ram> get_vnf_details() in this flow, it should be any special for VNFFG .. i think it should be a generic enhancement
16:39:21 <sripriya> +1
16:39:21 <sridhar_ram> GET /vnfs/<uuid> with some flag indicating "details" should return more information about the VNF
16:39:49 <s3wong> sridhar_ram: +1
16:39:50 <sridhar_ram> all the heat resources created for the VNF - VDUs, CPs,
16:40:07 <sridhar_ram> trozet: what do you think ?
16:40:35 <KanagarajM> sridhar_ram, +1 it would be nice enhancement.
16:40:37 <sridhar_ram> can we make that a generic enhancement and VNFFG patchsets will have dependecny on that
16:40:59 * sridhar_ram my typing malaise is back :(
16:41:09 <sripriya> sridhar_ram: something like GET /vnfs/uuid/detail ..
16:41:17 <trozet> sridhar_ram: I think it should be generic, and return VNFD logical mapping names to VNFD instance components
16:42:05 <sridhar_ram> sripriya: that would work, or even a flag in the json request body.. no hard opinion here
16:42:37 <s3wong> sridhar_ram: who is going to implement the API?  :-)
16:42:44 <sripriya> sridhar_ram: sure
16:42:46 <sridhar_ram> trozet: sounds good.. we can use this to do 'tacker vnf-show --details <vnf-id>'
16:43:39 <sridhar_ram> trozet: do you have some code already in place to get such vnf-details ?
16:43:51 <sridhar_ram> trozet: do you need some help from the community on this peice
16:43:53 <sridhar_ram> ?
16:44:05 <trozet> sridhar_ram: no code.  If someone wants to implement that it would be helpful
16:44:40 <sridhar_ram> any volunteers ?
16:44:48 <sridhar_ram> janki_: i remember you were interested in pitching in for VNFFG .. is this something you can help ?
16:45:12 <janki_> sridhar_ram: sure, was just typing that only
16:45:28 <sridhar_ram> janki_: cool, thanks...
16:45:32 <janki_> :)
16:45:48 <sridhar_ram> janki_: please keep in mind this will be blocking dependency for trozet
16:46:04 <sridhar_ram> janki_: so, keep trozet happy :)
16:46:07 <janki_> Ok, will get right to it
16:46:20 <sridhar_ram> janki_: thanks!
16:46:43 <sridhar_ram> i hope that diagram can be used in some vnffg-defref
16:47:02 <sridhar_ram> s3wong: trozet: janki_ : thanks!
16:47:05 <sridhar_ram> moving on..
16:47:17 <s3wong> janki_: thanks!
16:47:23 <sridhar_ram> #topic Grooming Midcycle Meetup topics
16:47:26 <KanagarajM> sridhar_ram, should we start to typing the descriptor in the catalog, seems it got missed over the discussions :)
16:47:39 <sridhar_ram> #undo
16:47:40 <openstack> Removing item from minutes: <ircmeeting.items.Topic object at 0x7f2b35c4da10>
16:47:48 * sridhar_ram oops
16:47:55 <janki_> s3wong: my pleasure
16:48:01 <sridhar_ram> #topic VNFFG
16:48:20 <sridhar_ram> KanagarajM: can you please elaborate ?
16:49:10 <KanagarajM> sridhar_ram, we now bring anothe descriptor for the VNFFGD.
16:49:29 <sridhar_ram> KanagarajM: yes, that's correct
16:49:39 <KanagarajM> sridhar_ram, so instead of modelling it as new resource in REST, should be update the existing VNFD to be generic enough to hold the all kind of descriptor
16:50:20 <sridhar_ram> KanagarajM: no, VNFD is for VNFs... VNFFGD descriptor flows covering multiple VNFs..
16:50:30 <sridhar_ram> KanagarajM: they don't logically map
16:50:48 <KanagarajM> sridhar_ram, i am not telling to directly on the VNFD,
16:51:40 <KanagarajM> sridhar_ram, instead make a VNFD as generic enough to hold VNFD, VNFFGD
16:52:29 <sridhar_ram> hmm... VNFFGD is actually close to NSD descriptor than VNFD
16:53:08 <janki_> I say we keep the descriptors separate
16:53:19 <janki_> a single descriptor would be long
16:53:25 <sridhar_ram> these descriptors are something we are preserving from ETSI architecture... i don't think we should overload VNFD to also include FFGD..
16:53:29 <janki_> it might become difficult for new comers to grap it
16:53:37 <sridhar_ram> the later refers to multiple VNFs
16:53:43 <janki_> given almost similar short forms for the terms
16:53:49 <KanagarajM> sridhar_ram, by using another attributed called 'type' this would make catalog simple
16:53:49 <KanagarajM> https://review.openstack.org/#/c/333216/1/tacker/extensions/nfvo.py i tried to comment the same
16:54:09 <KanagarajM> sridhar_ram, no... some thing like as below
16:54:43 <sridhar_ram> KanagarajM: sure.. we need to take up the topic of "NFV Catalog" evolution as a topic in our midcycle meetup
16:54:57 <KanagarajM> sridhar_ram, sure.
16:55:09 <sridhar_ram> lets take this offline in the channel.. or continue in the next week's call..
16:55:26 <KanagarajM> sridhar_ram, yeah.
16:55:49 <sridhar_ram> #topic Midcycle Meetup
16:55:58 <sridhar_ram> #link https://etherpad.openstack.org/p/tacker-newton-midcycle
16:56:13 <sridhar_ram> Please enter your topics of interest..
16:57:10 <sridhar_ram> Also, please respond to the following Doodle poll to pick a date in the last weekl of July..
16:57:24 <sridhar_ram> picking a time will be a challenge
16:58:01 <sridhar_ram> i've suggested a time that seems equally painful for everyone (ex US east coast)
16:58:17 <sridhar_ram> please pick upto 2 slots
16:59:08 <sridhar_ram> if you have other time slots in mind or if this suggested slot doesn't work.. please give a shout out..
16:59:23 <sridhar_ram> #topic Open Discussion
16:59:34 <sridhar_ram> thanks everyone for joining today...
16:59:48 <sridhar_ram> that all the time we have
16:59:56 <sridhar_ram> bye
17:00:06 <s3wong> bye
17:00:08 <sripriya> bye
17:00:14 <sridhar_ram> #endmeeting