Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  1. x axis - average Round Trip Time from SLAC to country in msec, we want low RTTs i.e. to the left (but depends on distance)
  2. yaxis - throughput in kilobits we want large values (towards the top)
  3. Each bubble is a country. Moving the mouse over the bubble identifies the country.
    Bubble size represents population size
  4. Color = region of the world
  5. Seen from SLAC to countries of the world
  6. Aggregated by year

Slider

  1. Allows Notice the slider at the bottom. Moving this right and left allows moving forward and backward in time from 1998 onwards
  2. Notice how RTT reduces with time as one goes the light blue bubble representing China, moves to the left with time, that is the RTT reduced  as they shifted from satellite connections to terrestrial connections, and one gets also installed faster links.
  3. Also Notice the throughput also increases with time, this is represented by the bubbles moving upwards.
  4. AS time moves forwards Also note we start to monitor sites in more and more bubbles appear as more countries are monitored
  5. Can also free run
  6. Moving the mouse over the region causes countries in the region to blink
  7. There is a free run feature that automatically flies through time, and shows all the effects of time on network performance for the countries, just like in the movies.

Change axis type

  1. Change There are options to change an axis from Lin to Loga Linear to a  Logarithmic scale. These options are labelled Log and Lin
  2. To see the data more clearlywe clearly we can use the Log scale feature. Choosing the Log scale enables us to look at data covering a wide range of values from a few tens of thousands of kbits/sec ro a few to tens of kbitkbits/sec. So clicking on log on the y axis we see the points are more evenly spread in throughput and the same ios true for the x axis. We can also perform the same spreading using the Log scale on the x-axis for the Round Trip Time 
  3. Note the correlation between throughput and Round Trip Time
  4. As exepcted This is as expected since throughput rougly roughly scales as 1/Round trip the inverse of the Round Trip Time

Identify countries

  1. Move mouse over bubble to identify the country
  2. Move mouse over region and hosts in region blink
  3. Notice Africa in botton right hand quadrant of high Round Trip Time and low throughput
  4. Track Country via pull down or by clicking on them, say China & India

...