Improving the responsiveness of websites with R

July 14, 2009
139 Views

Philip Jacob wanted to improve the responsiveness of the website for StyleFeeder.com customers, and narrowed the problem down to DNS lookup performance. Using R, he collected data on DNS lookup timings for 3 providers (and a “control”) to compare the results. When plotted as a graph (using R, of course), the result is clear: Philip summarizes his conclusions thus: But the real news here is that Dynect demolished the other participants in this test with a 37ms response time, less than half that of the first runner up. It’s not everyday that you can chop 40-70ms off of your mean…

Philip Jacob wanted to improve the responsiveness of the website for StyleFeeder.com customers, and narrowed the problem down to DNS lookup performance. Using R, he collected data on DNS lookup timings for 3 providers (and a “control”) to compare the results. When plotted as a graph (using R, of course), the result is clear:

Dns

Philip summarizes his conclusions thus:

But the real news here is that Dynect demolished the other participants in this test with a 37ms response time, less than half that of the first runner up.  It’s not everyday that you can chop 40-70ms off of your mean response time, so when those opportunities arise, it’s definitely worthwhile.  Companies like StyleFeeder that have tons of new users who don’t have these DNS entries in their resolver caches will definitely benefit from the speedup.

Stylefeeder Tech Blog: DNS Performance Redux

Link to original post