Comments on: Video of Mobile VMs Demo

By: Deshawn Pelletier

Deshawn Pelletier — Mon, 06 Feb 2012 17:54:45 +0000

Thank you pertaining to sharing that fantastic subject material on your web site. I discovered it on the search engines. I may check to come back if you post much more aricles.

By: தங்கமணிஅருண்

தங்கமணிஅருண் — Mon, 07 Mar 2011 01:17:42 +0000

Dear Reaz,

Sooner or later the game will changes. So wait and wach.

By: Reza F.

Reza F. — Sat, 22 May 2010 21:34:28 +0000

Quite promising. There are 3 questions:

1. Will switch and router vendors support OpenFlow in commercial products, as it can break their current business models?

2. Will OpenFlow complement (more flexibility) or replace TRILL (standard), when the latter is supported in switches?

3. Will the vendors of Flow collection and analysis tools adopt OpenFlow?

Thanks,

By: Yiannis

Yiannis — Tue, 29 Dec 2009 21:54:10 +0000

Hi David,

This is not supposed to be the correct way to do virtual machine mobility or client-server mobility for gaming. The goal of the demo is to highlight the flexibility one could have to manage a similar network. It’s up to the administrator (researcher, network owner, you name it) to decide the algorithm that fits your needs for load-balancing, cost etc. Regarding security, the control of the network takes place using OpenFlow through SSL encryption, and ignorant/malicious users shouldn’t be able to install a bad route.

Thanks,
Yiannis

By: David B

David B — Tue, 29 Dec 2009 19:22:47 +0000

I don’t get it. It seems like you are adding control over elements that you probably do not know anything about except ping times and names.

This seems very dangerous. Just because you can see the basic topology doesn’t mean that those links are equal and that you should use them. They could be congested, nearly congested, low latency, high jitter, unsecured, costly in terms of dollars or have other problems.

What stops malicious or ignorant people from just picking the most circuitous route possible?