15:00:51 <serverascode> #startmeeting operators_telco_nfv
15:00:52 <openstack> Meeting started Wed Dec 14 15:00:51 2016 UTC and is due to finish in 60 minutes.  The chair is serverascode. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:53 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:57 <openstack> The meeting name has been set to 'operators_telco_nfv'
15:01:07 <serverascode> #topic roll call / opening
15:01:23 <serverascode> hello! anyone here for the ops telecom nfv meeting? :)
15:01:52 <GeraldK> o/
15:02:18 <serverascode> hi GeraldK, it might just be you and I, we shall see
15:02:30 <serverascode> I'll give it a minute or so
15:02:32 <GeraldK> hi serverascode. let's see
15:03:24 <serverascode> I know Adrien said he would not be here today
15:04:38 <serverascode> ok I think it is just you and I
15:05:07 <GeraldK> okay. np.
15:05:35 <GeraldK> shall we still go through the agena?
15:05:43 * PerfectChaos waves
15:05:47 <serverascode> #topic Creating a project repository in openstack git
15:06:01 <serverascode> hi PerfectChaos :)
15:06:32 <serverascode> ok, so for the first topic I was writing up some things about a "generic nfv platform" and realized what I was probably doing was writing a "spec" in openstack terms
15:06:48 <serverascode> or a "blueprint", as I haven't really figured out the difference
15:06:59 <serverascode> some working groups/teams have a repo in git
15:07:03 <serverascode> let me find one...
15:07:17 <GeraldK> yes. e.g. PWG
15:07:28 <serverascode> #link  https://git.openstack.org/cgit/?q=wg
15:07:35 <serverascode> that will show a couple of working group repos
15:08:02 <serverascode> should we setup a repository for our projects?
15:08:20 <serverascode> even if it is just for the docs on the specs, ie. what we are actually doing
15:08:53 <GeraldK> as soon as we have something more stable, yes, why not.
15:09:31 <serverascode> ok, I think so too, I'll at least start the process of creating one
15:09:38 <GeraldK> easier to find than an etherpad page
15:09:48 <PerfectChaos> It potentially helps us, and I don't think we lose out in any way for having one, so sure.
15:09:51 <GeraldK> and we can have the reviews / comments on it.
15:09:59 <serverascode> yeah I agree
15:10:02 * PerfectChaos nods
15:10:19 <serverascode> #action serverascode setup an openstack git repo for this group
15:10:21 <PerfectChaos> There's a lot of infrastructure with gerrit and all that it gives us access to for reviewing spec type things
15:10:53 <serverascode> cool, good to hear you are both in favor, as some people don't like the enforced structure of the openstack gerrit/git system :)
15:10:57 <GeraldK> #me agrees
15:11:22 <serverascode> (some people would be me, but I can see the value in doing it)
15:11:56 <serverascode> ok if I call it something like "otn-wg" ? as in operators telecom nfv working group?
15:12:25 <GeraldK> see my last agenda item on a quick discussion "wg" vs "team"
15:12:51 <serverascode> right ok we'll cover it then, great
15:13:00 <serverascode> next topic?
15:13:10 <serverascode> I should say moving to next agenda item
15:13:28 <GeraldK> okay
15:13:32 <serverascode> #topic upgrades project proposal
15:13:48 <serverascode> all yours GeraldK :)
15:13:54 <GeraldK> #link https://etherpad.openstack.org/p/ops-telco-nfv-upgrades initial draft
15:14:09 <GeraldK> i have drafted some points around the upgrade topic
15:14:28 <GeraldK> i noticed there had also been a session in BCN on the topic
15:14:56 <serverascode> great, yeah there is a lot of information left in the etherpads of previous summits and such
15:15:24 <GeraldK> i will try to collect more info
15:15:39 <GeraldK> one issue on our side is the problem with skipping releases which is what we do so far. see line 12
15:16:22 <serverascode> right, that is interesting as I believe most openstack projects now expect not to jump versions
15:16:26 <PerfectChaos> What's the proposal here, for this group?
15:16:34 <GeraldK> i believe getting more feedback / input from Telcos would be great for this discussion
15:16:44 <PerfectChaos> Is this just discussing these various upgrade-related points and information-gathering?
15:16:55 <GeraldK> I'd like to see pain points of others
15:17:49 <GeraldK> if several Telcos have similar pain points we can try to see how we can approach the openstack services to address it.
15:18:08 * PerfectChaos nods
15:18:27 <GeraldK> maybe, also other have found a good solution on that already
15:18:32 <serverascode> so do you see it as a requirements gathering project at first? to talk to other telelcoms and find out what they would like to be able to do?
15:18:50 <serverascode> or I should say what issues they are having with upgrades
15:19:04 <GeraldK> yes, maybe that would be a necessary first step.
15:19:36 <GeraldK> e.g. create a survey on this topic
15:19:44 <serverascode> do you know if the previous telecom wg did a usecase on upgrades?
15:19:50 <serverascode> or "user story"
15:20:07 <serverascode> (I'm taking a look right now)
15:20:20 <PerfectChaos> Yes, I was about to ask whether we should be conducting a survey for these information gathering sorts of projects
15:20:31 <GeraldK> definitely, it had been discussed in a meeting as far as I remember
15:21:11 <serverascode> #link https://git.openstack.org/cgit/openstack/openstack-user-stories/tree/user-stories/proposed
15:21:20 <serverascode> not sure if that is actually all the user stories, b/c there is not many there
15:21:39 <serverascode> oh there is a rolling upgrades story
15:21:42 <serverascode> #link https://git.openstack.org/cgit/openstack/openstack-user-stories/tree/user-stories/proposed/rolling-upgrades.rst
15:22:16 <serverascode> I'm slightly hesitant to work on user-stories, but if there is not a "jump version" story that might be interesting
15:22:32 <serverascode> or other requirements you mention
15:23:01 <GeraldK> that is the user stories from PWG
15:23:16 <GeraldK> #link https://review.openstack.org/gitweb?p=openstack/telcowg-usecases.git;a=tree;f=usecases;h=f4004434a7ca606ff6eca3d71e26a2b5a7b978e3;hb=HEAD TelcoWG use cases
15:24:00 <serverascode> I think those got pushed into the pwg once the old telco wg closed
15:24:11 <GeraldK> Rolling upgrades is part of the upgrading discussion, but just one part
15:24:25 <serverascode> I agree regarding rolling upgrades being only one part
15:24:43 <GeraldK> creating a user story in PWG could be a second step after the survey
15:25:08 <serverascode> ok that sounds doable to me :)
15:25:31 <serverascode> so are we agreeing that some kind of survey would be the next step?
15:25:37 <GeraldK> +1
15:25:41 <PerfectChaos> Yup~
15:26:01 <GeraldK> someone having experiences how to do such surveys?
15:26:04 <PerfectChaos> Since getting people to come in here and talk about their problems seems to be difficult.
15:26:04 <serverascode> #agreed Some kind of telecom upgrade survey would be the next step
15:26:35 <serverascode> I know that the openstack foundation does surveys
15:26:55 <serverascode> I know that some people also get somewhat upset when random surveys are sent to the openstack lits
15:26:58 <serverascode> *list
15:27:10 <GeraldK> okay. once we have a set of questions ready, we can reach out to them how to best do it
15:27:19 <PerfectChaos> I'll be honest, last time I attempted to conduct any sort of survey was well over a decade ago and I was still in school at the time.
15:27:47 <PerfectChaos> Does feel like asking the foundation for advice would be a useful step if none of us happen to be super-confident in the art of surveying
15:27:56 <GeraldK> alternatively, have a session at the Forum?
15:28:49 <serverascode> I agree that it would be best to check with the foundation first on surveys just to make sure everything is proper
15:28:52 <serverascode> I can do that
15:28:58 <PerfectChaos> That could work too, though also has going through the foundation as a necessary step, so we could discuss that possibility with then when we do that
15:28:58 <serverascode> unless someone else wants to
15:29:04 <GeraldK> thanks serverascode
15:29:05 <PerfectChaos> *them
15:29:15 <serverascode> #action serverascode check on surveys with foundation
15:29:31 <serverascode> also, I agree a session at the next summit would be a great idea
15:30:04 <PerfectChaos> Thanks.
15:30:15 <serverascode> ok, so is our next step to write some questions for the survey?
15:30:26 <GeraldK> maybe a slot in the OTN session is sufficient
15:30:41 <GeraldK> serverascode: yes
15:31:00 <serverascode> (just looking at irc bot commands)
15:31:09 <GeraldK> put an action on me. will not be able to work on before Jan
15:31:23 <serverascode> ok
15:31:40 <serverascode> #action GeraldK to begin developing survey questions
15:32:04 <serverascode> #agreed Either a session or part of a session on telecom upgrades at next summit would be beneficial
15:32:21 <serverascode> anything else on this topic today?
15:32:44 <GeraldK> not from my side
15:33:01 <PerfectChaos> On the coming up with questions front - put it on an etherpad so that I and others can potentially comment/help out?
15:33:20 <GeraldK> hi DaSchab
15:33:51 <GeraldK> PerfectChaos: https://etherpad.openstack.org/p/ops-telco-nfv-upgrades
15:34:14 <serverascode> #topic project proposal "Generic OpenStack NFV Platform"
15:34:16 <PerfectChaos> Ah, yes, putting it on that pad works
15:34:34 <DaSchab> hi
15:34:40 <serverascode> #link https://etherpad.openstack.org/p/OTN-nfvi-project
15:35:00 <serverascode> hi DaShab :)
15:35:18 <serverascode> ok, I started writing up what I thought would be a good start on a "Generic OpenStack NFV Platform"
15:35:31 <serverascode> but the first part was using the phrase "Generic OpenStack NFV Platform"
15:35:58 <serverascode> I think NFV Platform makes sense as a term? I had previously kept using "VIM" or "NFVi" interchangibly
15:36:30 <serverascode> do people think that "Generic OpenStack NFV Platform" makes sense as a general term, and as a project?
15:36:40 <DaSchab> it makes sense. it's more than just a VIM
15:36:54 <GeraldK> +1
15:37:21 <serverascode> I guess it's a lot like how OPNFV uses platform in their name
15:37:56 <serverascode> ok, so after I realized I should use the word platform, I started to put together some parts of the platform
15:38:07 <serverascode> which is when I realized I was writing more of a spec
15:38:30 <serverascode> that I was listing out pieces of the platform, and requirements and such
15:38:35 <GeraldK> i like the approach
15:39:25 <serverascode> yeah and then I started a spreadsheet, but I wasn't sure that was the right format either
15:39:52 <serverascode> My feeling is the next step is to put it into a doc in an openstack git repo and start detailing parts of the platform
15:39:56 <GeraldK> how would this be different from the OPNFV Platform?
15:40:40 <serverascode> GeraldK that is a good question, it would be similar, and in fact we would use some of the OPNFV components
15:40:54 <GeraldK> if we want to run tests we would need to setup the platform not just write a spec. which resources would we use?
15:41:23 <serverascode> right, I was thinking the OSIC would be the first option for actual resources
15:41:33 <GeraldK> my proposal: write a spec. see how much we align with OPNFV. propose a new scenario for OPNFV. propose test cases to OPNFV
15:41:42 <serverascode> #link https://osic.org/
15:41:55 <serverascode> ok, that approach would work for me
15:41:59 <GeraldK> OPNFV already has all the environment and test framework setup
15:42:30 <serverascode> ok, part of this is I'm really hoping to bridge the openstack and opnfv communities
15:42:30 <GeraldK> i guess they would be happy to have some resources to develop further test cases or extend the set of scenarios that they have
15:43:32 <serverascode> ok, so does a first step of writing a very basic spec, and finding people in both the openstack and opnfv communities to comment on it make sense?
15:43:40 <GeraldK> OPNFV also has a few VNF services they are using in their tests, like a simple vIMS
15:43:53 <GeraldK> serverascode: +1
15:44:01 <serverascode> right, yeah can't test without VNFs :) which are actually hard to come by
15:44:28 <serverascode> anyone else have any thoughts/comments/critiques?
15:44:31 <GeraldK> if we have some agreement in this direction I can reach out to some OPNFV folks to comment on it
15:45:09 <serverascode> if no other comments I'll suggest we agree on the basic start of this project?
15:45:18 <GeraldK> what I would like to see a bit better is what we want to achieve/produce/test
15:45:48 <GeraldK> "functional tests" IMHO is a bit too generic
15:45:51 <DaSchab> sorry that I joined your discussion at this late stage, but what is your intention to achieve with this paper and the alignment with OPNV
15:46:23 <DaSchab> I missed the last weeks since the summit and now want to get onboard, sorry for nasty questions
15:46:29 <serverascode> no worries :)
15:46:52 <serverascode> my intention is to help openstack operators who are charged with running NFV environments
15:46:53 <GeraldK> DaSchab: on the alignment with OPNFV, I would rather say utilize synergies, is that we should avoid doing the same work twice
15:47:13 <serverascode> in a way I'd like to bridge the opnfv and openstack communities work around an nfv platform
15:47:42 <serverascode> and to start, create a very basic, generic nfv reference platform (which may already exist in opnfv's docs)
15:47:51 <DaSchab> understand
15:48:01 <GeraldK> DaSchab: the topic had been raised in the Summit that people would like to run performance tests and would need a reference platform for it
15:48:16 <DaSchab> I know, I'm also interested inthis topic
15:48:39 <serverascode> ok, I'm going to say we are agreed in this direction for now so we can move onto the next topic :)
15:49:01 <GeraldK> serverascode: any actions or agrees ?
15:49:29 <serverascode> #agreed We agree that an initial attempt at a "Generic OpenStack NFV Platform" by writing a spec and gathering comments on it
15:49:45 <serverascode> hopefully that is an ok start, and we can make changes later
15:49:59 <GeraldK> +1
15:50:06 <serverascode> #action serverascode create an inital doc in the repo
15:50:17 <serverascode> #topic Next meeting
15:50:24 <GeraldK> 10 min left
15:50:39 <serverascode> #info Last Dec meeting cancelled, next meeting not until 2nd week of January
15:50:49 <serverascode> #info We meet on even weeks
15:51:08 <serverascode> #topic "Working Group" vs "Team"
15:51:25 <GeraldK> there had been a similar discussion in PWG whether they should rename to team as according OpenStack definition a wg is limited in time, whereas a team is an onoging effort
15:51:26 <serverascode> GeraldK?
15:52:00 <GeraldK> I am not saying we should change name, but maybe having a discussion on it would be good before setting up the reop
15:52:05 <serverascode> right, my impression is that we are a functional team as opposed to a working group
15:52:08 <GeraldK> s/reop/repo/
15:52:15 <serverascode> by I'm confused by new groups being created taht don't seem like working groups
15:52:19 <serverascode> like the arch-wg
15:52:34 <serverascode> so I don't really know what to do myself
15:52:43 <PerfectChaos> Is this just because that distinction isn't clearly documented somewhere obvious?
15:52:53 <GeraldK> I don't have a strong position on it
15:53:16 <serverascode> #link https://wiki.openstack.org/wiki/Governance/Foundation/UserCommittee
15:53:24 <serverascode> I think that is the list of working groups and such
15:53:31 <GeraldK> PWG has decided -altough they are more like a team- not to change to team right now as the term PWG is widely known
15:54:32 <GeraldK> or we keep a neutral repo name like ops-telco
15:54:35 <PerfectChaos> Heh, we don't appear to currently be either
15:54:49 <serverascode> ah, ok, yes I agree on a neutron repo name
15:55:17 <serverascode> this is something to try to figure out there as it is confusing
15:55:33 <serverascode> whether we are a working group or team, based on the user commitee definitions
15:55:38 <serverascode> I can email the user commitee list
15:55:42 <serverascode> and ask
15:55:57 <PerfectChaos> I guess if we had to pick one and there is something somewhere documenting that WGs are time-limited whilst teams are on-going, I'd prefer team
15:56:14 <serverascode> I agree that we are a team as opposed to wg
15:56:42 <DaSchab> +1
15:56:54 <serverascode> perhaps lets leave this til Jan so we have time for the last topic?
15:57:00 <GeraldK> okay
15:57:01 <serverascode> though we only have 3 minutes
15:57:10 <GeraldK> i can be quick
15:57:19 <serverascode> #agree table wg vs team to jan
15:57:41 <serverascode> #topic telco related ux persona
15:57:43 <GeraldK> #link http://docs.openstack.org/contributor-guide/ux-ui-guidelines/ux-personas.html OpenStack UX personas
15:58:07 <GeraldK> UX has defined personas that should be used for description of user stories, etc.
15:58:18 <GeraldK> there is not Telco, so I have started drafting one
15:58:24 <serverascode> ah, interesting
15:58:32 <GeraldK> #link https://etherpad.openstack.org/p/osux-persona-nocengineer NOC engineer
15:58:37 <GeraldK> feedback welcome
15:59:00 <GeraldK> we may also need to draft a model company and other personas in the Telco
15:59:31 <GeraldK> just drop your feedback in the Etherpad page
15:59:45 <serverascode> ok, cool, thanks GeraldK, that is something that is needed I think
15:59:51 <serverascode> I'm going to have to end the meeting
15:59:58 <serverascode> thanks everyone for all your input!
16:00:01 <DaSchab> do we need an application owner?
16:00:03 <zhubingbing> ;)
16:00:11 <serverascode> #endmeeting