Inventory Migration Use Case

Meeting 08 Jun 2020 

ATTENDEES - Fred Feisullin  Vivian Pressley  Andy Mayer  Joanne Liu Rudel  

James Forsyth  Linda Horn Marge Hillis  Ryan Hallahan Scott Blandford  William Reehil   Zu Qiang (Ericsson)

NOTES:

MEETING June 8 2020

PROBLEM STATEMENT: Inventory Migration (?)

Carriers, has multiple databases

Some could be domain specific (xport) & wireline vs wireless etc

A&AI unifying, 1-stop-shop for inventory & topo data w/ orchestration – configuration

Maintaining descriptions of inventory & future things such as state

Needs to be some kind of migration from these SP’s databases into A&AI

Trying to integration brown-field inventory systems, existing tools for prov 7 activation

Things specific to wireless !wireline. PNFs have own databases how much do we move over?

How to integration w/ seamless e2e service orchestration

An intermediate state, a mapping between existing DBs into an ETL

Or representation “a hop” to a single-A&AI database, schema TBD.

Solution -  Create an overlay, start from scratch

WR – there isn’t something to automate this process, internally U/C to map to schema.

If the ask is to automate the process, that would need to be a new req.

FF – your approach is to do it on a case-by-case basis based on user stories.

WR – when the U/Cs come to us, we have people write those into user-stories and fit them into the model based on the centralized inventory system; if there needs to be mapping from old to new, the mapping work is done manually.

FF – if you are doing this manually; is there some template we can build from? Maybe vendors give a model or schema. One form of automation.

WR – The framework does support multiple schemas, take framework and redefine the schema.

Is that global enough;

FF – give use case to vendor to develop schema. Good to have some initial starting point. So orchestrator could understand. Anytime there are new solutions by release, new features added.

JLR – SP domain, wireline/RAN,

FF – any 5G RAN domain that is needed to deliver the service, xport, VRAN, 5G Core, LTE, FH-BH, service orchestration & assurance E2E. Routers. Capacity planning need representation in inventory. Legacy inventory systems. Not compatible with A&AI need to move information there into A&AI. Information there that needs to be transferred over, based on use cases. This is just an example.

VP – is this database coming from a DB, any kind of data, and somehow create A&AI schema and populate A&AI??

FF – limit scope to big inventory & topo DBs used by SPs today, one impacted by migration. Not small internal database, not small product family, larger vendors.

JLR – maybe choose a Legacy inventory system.

VP – maybe a PoC, and going through steps of going through this exercise.

FF – in inventory & topo DBs working w/ DCAE, SO etc flows to be supported would start with this representation.

JLR – A&AI does have topo information (limited)

FF – need to find another contact. Need to find user stories & requirements, formats of database, identify info in those DBs.

BC – Timeline?

FF – the need is there now. Being handled manually now. And that isn’t scalable in the long-run

JLR – If we have a module developed in ONAP, an application that addresses this problem,

FF – whether we use the arch to drive vendor; or open source – if a structure is defined and it is standardized is the goal instead of a 1-off process. As ONAP gains traction solving this problem will help adoption. If this is a common problem solve together.

WR – I can help w/ A&AI architecture.

FF – is this doable?

JF – we get to a point where carriers will adopt, and find we needed this xyz. This is probably a year out, before they know they need this.

FF – this is timely now, before it becomes a big problem

JF – brought to the community as a use case or feature enhancement. Importing Legacy inventory systems data, dicey in the open source world. Maybe needs to be more generic. From a resource perspective, if we were to frame it as a U/C take to ONAP community, needs a sponsor, needs ONAP partners to contribute resources to get this work done. If VZW, ATT, BC, Orange could collaborate chances its gets done to meet all carriers is a possibility.

FF – agreed. We shouldn’t use a vendor product name, talk about particular type of schemas & data in DBs.

JF – “Legacy inventory systems”.  Interest in forming a team around this topic, some responses, category of they don’t know that they need it, so not on top of their mind, not a pressing concern.

FF – consolidating legacy inventory systems is a large problem, to automate this task is daunting. Need to show business use case and need.

JF – present at Virtual DDF.

JLR – In 5G ONAP platform, C&PS would provide topo info in a 5G RAN network (in R8); for your situation, this would be a brand new database. To address your use case, these legacy inventory systems, which hold config information for xNFs. Could we use heat templates into SDC. We have a new network, physical attributes … what about SPs with legacy inventory systems,  you want to link that data into ONAP right? Do the greenfield first?

FF – Greenfield and brown field aspect, orchestration of LTE, focus on the problem then figure out how to do it. You have existing systems, overlays OSS and BSS. Problem, situation, impact of problems, Status quo, next steps. Sell it to communicate the problem and its impact.

BC – Formulation, Scheduling, Resourcing, Socialize (DDF)

JF – if want anything to happen you have to start with a use case presentation to U/C S/C, in the DDF you can have discussions about an idea, are other partners have experience with it. The DDF is a place where it can be socialized. A few slides. Anyone who is interested will be at the DDF, it is a good forum to have discussion. Does ONAP, do carriers care about this as a need as they implement, and give people a forum where they can participate. This is a big problem that requires multiple carriers to participate and seek to solve it. If AT&T is just going to do this, there isn’t enough resources, it is early in the process need to gauge community interest and buy-in to move forward.

Recording