Transcript Pb Statement for a Common Network Address Mapping System

ID EnAbled networkS:
Pb Statement for a Common
Network Address Mapping System
Draft-padma-Ideas-problem-statement-00
Dino Farinacci lispers.net [email protected]
Dave Meyer Brocade [email protected]
David Lake UoSurrey [email protected]
Tom Herbert Facebook [email protected]
Michael Menthe UoTuebingen [email protected]
Dipenkar (Ray) Raychaudhuri Rutgers University
[email protected]
Julius Mueller ATT [email protected]
Presenter: Padma Pillay-Esnault Huawei [email protected]
Motivation

New requirements ….
◦ Over the last 10 years the network usage has evolved with
the explosive scale of mobile devices and IoT (expected to
be several billions by year 2025)
◦ ID protocols provide dissociation between the location/ID
and are good candidates solutions for
 Session continuity for All types of mobility – IP mobility, celullar
mobility …
 IoT – Reachability of devices
 Heterogeneous networks
◦ Today the ID based protocols have been deployed mostly in
the DC space and SD-WAN
◦ The large scale use across multiple administrative domains
require deployment of a global Network Mapping System
Problem Statement

Need for a standardized single Common Control
Plane
◦ There are many data plane ID solutions ( HIP, LISP, ILNP, ILA)
◦ Competing solutions operate in silos
◦ While it makes sense to have different data plane solutions as
they may have very different objectives

It would be beneficial to have a standardized
common Control plane
◦ Common infrastructure may facilitate cross-platform
◦ Remove impediment to deployment – same information in
different MS
◦ Newer ID-aware applications may leverage an existing infra
Problem Statement

Flexible, Open and Efficient Mapping System
Interfaces
◦ If there is no standard to accessing the Network Mapping
System, there will be more divergence as we go forward
◦ The mapping system must have a flexible, open and
extensible for novel usage of ID and mapping types
◦ Multiple examples of lack of interoperability
 VM mobility in the data center.
 VXLAN data-plane but each vendor has their own
SDN-like control-plane.
 SD-WAN space. Use of IPsec for data-plane and
proprietary control-planes with their own SaS service
to operate it.
Problem Statement

Identifier Structure and Life Span
◦ Currently no guidance on allocation
◦ agreed upon ID format and scope may facilitate interoperability
◦ Siloed communication

Confidentiality
◦ Access to mapping system may reveal the location of ID

Security
◦ Scanning tools may reveal information that is used later for a
cyber attack based on vulnerabilities.
◦ Access to MS

Automatic Bootstrapping
◦ A must for the scale of M2M, IOT communication
◦ ZTP – highly desirable
◦ Particularly pertinent for Industrial Internet
Network Mapping System (NMS)
Requirements
o
Mapping Responsibility
o Can be regional ( per AS, regional authority …)
o Owner of ID
o Considering most things do not move outside of country leverage real life scenarios
o
Distribution and Redundancy
o Cannot be single point of failure
o Must de distributed
o A collection of NMS collaborating
o
Massively scalable
o Able to handle a lot of changes and be very fast
o Aggregation?
o Caches?
o Hierarchy
Network Mapping System (NMS)
Requirements
o Mapping
System Security
oSecured access, lookup, add …
oConfidentiality
oCrypto signed
oPolicy
o Flexibility
for Future Apps
oOne standardize interface, metadata
o Smart, dynamic
– autonomic
oMachine Learning
Further discussions




Side Meeting
Date: Thursday, 17th November 2016
Time : 6:00 - 7:30pm
Venue: Studio2
Mailing List: IDEAS
List address: [email protected]
Archive: https://mailarchive.ietf.org/arch/search/?email_li
st=ideas
 To subscribe: https://www.ietf.org/mailman/listinfo/ideas




Areas: RTG, OPS
Agenda

Padma Pillay-Esnault (Huawei) - Introduction on problem statement for
IDEAS (10 mins)

Tom Herbert ( Facebook) - The ILA protocol and NMS (10 mins)

Dino Farinacci ( Lispers) - LISP Mapping system, How it works? (10 mins)

Gerry Forster ( UoSurrey) - ETSi NGP: GTP, Mobility & Flat 5G
Architecture( 15 mins)

Fabio Maino(Cisco) - Deployment experience of Mapping Systems ( 10
mins)

Dave Meyer (UoOregon/Brocade) - Machine Learning and Network
Mapping System ( 15 Mins)

A. Cabellos, J Vilanova & F Maino (UoCatalunya, Ecole P. Lausanne, Cisco) – A
Blockchain-based Mapping System (15 mins)
What’s next?
Invite LISP WG and larger community to
IDEAS Side Meeting
o Several areas for work
o
o
o
o
Allocation and Format of ID
On what this should be based?
Leverage ID indirection properties for security?
…
Ops – Need more data
o
o
o
o
Deploy a network mapping system internal
Interoperations beween NMS
Best Practices?
Questions?