Comments for Comfortably Geek http://lalith.in Lalith Suresh Wed, 23 May 2012 19:48:03 +0000 hourly 1 http://wordpress.com/ Comment on Chasing the Thesis Carrot by lalithsuresh http://lalith.in/2012/05/23/chasing-the-thesis-carrot/#comment-788 Wed, 23 May 2012 19:48:03 +0000 http://lalith.in/?p=1551#comment-788 Yeah I do. It’s called hope!

]]> Comment on Chasing the Thesis Carrot by wmalik http://lalith.in/2012/05/23/chasing-the-thesis-carrot/#comment-787 Wed, 23 May 2012 19:46:15 +0000 http://lalith.in/?p=1551#comment-787 Dude, you waste a whole minute every day on Google+! Seeeeeriously?

]]> Comment on HOWTO: Getting started with ns-3-click – Part I by lalithsuresh http://lalith.in/2011/05/22/howto-getting-started-with-ns-3-click-part-i/#comment-768 Mon, 30 Apr 2012 16:16:10 +0000 http://lalith.in/?p=925#comment-768 Thanks for pointing that out. It used to work at the time I wrote it. Fixed.

]]> Comment on HOWTO: Getting started with ns-3-click – Part I by Andrew Bolster (@Bolster) http://lalith.in/2011/05/22/howto-getting-started-with-ns-3-click-part-i/#comment-767 Mon, 30 Apr 2012 16:08:05 +0000 http://lalith.in/?p=925#comment-767 Just FYI,the git clone that you used is incorrect (well, it’s not working for me).
Instead of the http link, the fresh git link works git://read.cs.ucla.edu/git/click

]]> Comment on Fun with TCP CUBIC by amitsingchandele http://lalith.in/2012/02/15/fun-with-tcp-cubic/#comment-746 Tue, 10 Apr 2012 18:19:07 +0000 http://lalith.in/?p=1456#comment-746 Can you please tell me how did you generated Congestion window graph for TCP Cubic?

]]> Comment on Network Operating System? by lalithsuresh http://lalith.in/2012/03/16/network-operating-system/#comment-725 Sun, 01 Apr 2012 11:18:41 +0000 http://lalith.in/?p=1524#comment-725 This is what my architecture seems to have boiled down to pretty much. It looks something like this:

| App1 | App2 |
——————–
Master App
——————–
Controller
——————–
network of switches

The master app exposes a set of primitives and interfaces (for statistics collection, for instance) to the ‘upper’ applications, which operate solely based on those. Of course, I have no clue on how to generalise this, because all applications in my system including the master are doing something very scenario specific and are doing more than changing flow tables (which I guess is making it more complicated :) ).

]]> Comment on Network Operating System? by md1clv http://lalith.in/2012/03/16/network-operating-system/#comment-724 Sun, 01 Apr 2012 11:07:14 +0000 http://lalith.in/?p=1524#comment-724 It seems to me that there will need to be an arbitrator application which talks to the controller, and the other applications will talk to the arbitrator instead of directly to the controller.

The arbitrator’s job would be to make sure that the “right” switching decision gets passed to the controller when there are multiple conflicting options.

]]> Comment on Network Operating System? by Jason Edelman (@jedelman8) http://lalith.in/2012/03/16/network-operating-system/#comment-721 Fri, 30 Mar 2012 21:38:38 +0000 http://lalith.in/?p=1524#comment-721 I’ll admit, I read your post earlier from the road, but now re-reading it, what Rexford was working on was making changes to a distributed network to multiple nodes, not changes from different applications to the same controller. However, there may be similarities.

]]> Comment on Network Operating System? by kyleforster http://lalith.in/2012/03/16/network-operating-system/#comment-720 Fri, 30 Mar 2012 20:07:25 +0000 http://lalith.in/?p=1524#comment-720 > I was under the impression that a network operator using a particular controller could choose between different 3rd party OpenFlow applications to handle different complexities with the network

Also worth taking a peek at FlowVisor for some inspiration — http://conferences.sigcomm.org/sigcomm/2009/demos/sigcomm-pd-2009-final65.pdf

]]> Comment on Network Operating System? by lalithsuresh http://lalith.in/2012/03/16/network-operating-system/#comment-718 Fri, 30 Mar 2012 18:49:18 +0000 http://lalith.in/?p=1524#comment-718 Thanks for clarifying that. Is there any work underway on any of those directions?

]]>