[Ndn-interest] [EXT] NDN Research Issue, Name-Lookup in NFD

Mobin Ranjbar mobinranjbar at hotmail.com
Fri Apr 17 04:20:37 PDT 2020


Millions of lookups per second is not huge for a NoSQL databases. In Big Data situation, We will talk about far more beyond billions. I have developed NFD with Redis integration for Content-Store, results looked nice.


Mobin Ranjbar
Big Data Engineer
Founder and Head or Board, Farafekr Technology Ltd.<http://www.farafekr.co>
Founder, Hadoop.ir<https://www.hadoop.ir>
Founder, BigDataWatcher<http://www.bigdatawatcher.com>
Co-Founder, Iran Startups<http://www.irstartups.com>
Editor in Chief, Techly.co<http://www.techly.co>
Linkedin<http://linkedin.com/in/mobinranjbar> | Twitter<https://twitter.com/MobinRanjbar> | Skype: mobinranjbar
From: Ndn-interest <ndn-interest-bounces at lists.cs.ucla.edu> on behalf of Junxiao Shi <shijunxiao at email.arizona.edu>
Sent: April 16, 2020 10:31 PM
To: avakili <avakili at ut.ac.ir>
Cc: ndn-interest at lists.cs.ucla.edu <ndn-interest at lists.cs.ucla.edu>
Subject: Re: [Ndn-interest] [EXT] NDN Research Issue, Name-Lookup in NFD

Hi avakili

NDN design principle requires an Interest with prefix name to match a Date with longer name. “Exact match only” is fundamentally not NDN.

If you are thinking about adopting an existing implementation of NoSQL database (LevelDB and similar), it’s unlikely to deliver sufficient performance at millions of lookups per second.

Yours, Junxiao

On Thu, Apr 16, 2020 at 07:00 avakili <avakili at ut.ac.ir<mailto:avakili at ut.ac.ir>> wrote:

External Email

Dear Respected NDN team,

Greetings. I am Abdul Rahman Vakili from Afghanistan. I am doing my PhD studies in Tehran University in Iran in IT field.

I am very interested in NDN network after reading about and want o focus on a part of it.

I had a look on NDN name lookup already done  (Exact matching) with Trie-Based, Bloom filter and Tabel hashing.

I have had a try with similarity matching techniques such as Levenshtein, Fuzzywuzzy and Fuzzy.

I am thinking of a model of using Key-Value Based store solutions using NoSQL databases in order to handle part of searching records (CS/PIT/FIB) of NFD routing tables. This way may have better scalability and better prefix matching management.

Hope to get a good feedback from your side and make my PhD research going on NDN and later help Afghanistan to join your cool Testbed network. I will appreciate in case you advise any other research issue in NDN are too.

Best Regards

Abdul Rahman Vakili

CS Lecturer, IT manager

Herat University, Afghanistan.

PhD Student in Tehran University, Iran

Ndn-interest mailing list
Ndn-interest at lists.cs.ucla.edu<mailto:Ndn-interest at lists.cs.ucla.edu>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndn-interest/attachments/20200417/7916db49/attachment.html>

More information about the Ndn-interest mailing list