Sprint Demo 5.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:0445cede127e8d3e21f68898fdd4bc3d387796a691e2dd971648a97f12ffefda"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Name Based Routing Project"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Addressable Problem:\n"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "-Reduce the effects of DDoS attack\n",
      "\n",
      "-Speed up content delivery\n",
      "\n",
      "-And open up new markets by routing with named strings instead of DNS and IP."
     ]
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "What's new/different?"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Akamai's solution:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "FastTCP \n",
      "-- improves traditional TCP to increase throughput and control congestion.\n",
      "-- maintain optimal send rate to not exceed network bandwidth, reducing loss rate. "
     ]
    },
    {
     "cell_type": "heading",
     "level": 6,
     "metadata": {},
     "source": [
      "\"[FastTCP] can only help if there is excess available bandwidth that loss-based TCP cannot take advantage of\" - Akamai White Paper"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Our solution:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "Improve Traditional IP by using Name Based Routing techniques \n",
      "    (does not require new or additional hardware)\n",
      "    \n",
      "\n",
      "    \n",
      "-Create a parent/child hierachy relationship between network routers\n",
      "-Any arbitrary string can be registered as a Domain name\n",
      "-Allows for delivery of requests from multiple sources\n"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Our Hypothesis:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "-We believes it will beat standard IP for DDoS attacks\n",
      "\n",
      "-It will deliver content faster than the best current content delivery method (Akamai-style content delivery)"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Setup:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "- Probability p of a request arising at any leaf node (1/total machine count)\n",
      "\n",
      "- 2 Queues:\n",
      "    -- The forward queue, which queues up requests, \n",
      "    -- The return queue, which queues up return requests.\n",
      "\n",
      "- Heat Map \n",
      "    -- Combines both queues into one for display purposes only.    \n",
      "\n",
      "- Ticks \n",
      "    -- (Used as unit of time)\n",
      "    -- We don't want to use CPU time.\n",
      "\n"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "Progress:"
     ]
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": [
      "    Past:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "-Graphical User Interface \n",
      "    Illustrates differences betweeb Name Based Routing and traditional IP \n",
      "    -- requests 'video'\n",
      "    -- DDoS     'video'\n",
      "    \n",
      "\n",
      "-Discrete Event Simulator \n",
      "    -- collected stats on Round Trip Time for a level 4 Name Based Network (1400 machines) vs IP (1400 machines)"
     ]
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": [
      "Current:"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": [
      "-Graphical User Interface\n",
      "    -- NB performance while supporting delivery from multiple sources 'video'\n",
      "    -- IP performance under a spike in requests for particular content 'video'\n",
      "    \n",
      "-Discreate Event Simulator\n",
      "    -- Simulation that collects Rount Trip Time to fulfill a request for IP and NB for:\n",
      "        --- 4 levels (1,400 machines)\n",
      "        --- 5 levels (16,807 machines)\n",
      "        --- 6 levels (117,648 machines)"
     ]
    },
    {
     "cell_type": "raw",
     "metadata": {},
     "source": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from graphs import *\n",
      "%matplotlib inline\n",
      "rtt_means()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEZCAYAAAB8culNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiRJkiRJkiRJkqRG8b/32rPibwGz1gAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1077354d0>"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "average_difference()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEACAYAAABCl1qQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nR++eAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x104023f90>"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}