Alberto,

Our tests showed  very promising results for public cloud Pub/Sub (DHT based). 
We got about 120ms performance out of these systems.
Given that these systems are designed to mover large chunk of data, it is very 
reasonable to assume that the performance for these DHT based mapping systems 
can be further improved for 5G.

Arashmid

From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of Alberto Rodriguez-Natal
Sent: Friday, March 16, 2018 4:20 AM
To: dmm@ietf.org
Cc: l...@ietf.org list <l...@ietf.org>
Subject: [DMM] Mapping System scalability

Hi all,

In the DMM call this week, some people asked about the scalability of the 
Mapping System. The LISP community has delivered different solutions to address 
that challenge over the years. Here are some pointers to different Mapping 
System implementations, I'm sure that the folks at the LISP WG list can provide 
even more references.

LISP Delegated Database Tree (LISP DDT) [1] follows a DNS-like structure and 
it's designed to be deployed at Internet scale. For an academic work on a 
DNS-based Mapping System take a look at [2].

Designs leveraging DHTs have also been considered to enable high scalabilty, 
see for instance [3][4]. For a survey on different Mapping System options see 
[5] (I wasn't able to find an open version of this article unfortunately).

For those interested in current research efforts, recent works include 
decentralized designs [6] or even Blockchain-based approaches [7].

Hope this helps,
Alberto

[1] https://tools.ietf.org/html/rfc8111
[2] http://people.ac.upc.edu/ljakab/2010-jakab-jsac-lisp-tree.pdf
[3] https://tools.ietf.org/html/draft-cheng-lisp-shdht-04
[4] 
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.139.735&rep=rep1&type=pdf
[5] http://ieeexplore.ieee.org/document/6422285/
[6] https://tools.ietf.org/html/draft-farinacci-lisp-decent-00
[7] 
https://datatracker.ietf.org/meeting/97/materials/slides-97-lisp-blockchain-based-mapping-system/
_______________________________________________
dmm mailing list
dmm@ietf.org
https://www.ietf.org/mailman/listinfo/dmm

Reply via email to