HN Explorer

7 comments

[ 0.19 ms ] story [ 339 ms ] thread
carbocation
I ran this against my ISP's DNS as cached by my router, Google's new 8.8.8.8, Level3's 4.2.2.1 and 4.2.2.2, and OpenDNS.

It performed 200 tests in 1 run, using my 100,000 Firefox bookmarks as its data source in 'weighted' mode.

My results were as follows: "In this test, Your current primary DNS server [my router's cache of Comcast's DNS] is 29% Faster than Google Public DNS".

I am in Boston. In ranked order of performance: My router's cache of Comcast (48.28ms), Google's 8.8.8.8 (62.49ms), OpenDNS (68.14ms), Level3's 4.2.2.1 (94.37ms), Level3's 4.2.2.2 (113.51ms)

*Update - I clarified that I tested my router's cache of Comcast's DNS, not Comcast's DNS servers themselves.

besologic
200 tests, 1 run. ~33K Firefox items. Weighted.

I was surprised to find my ISP (Cox, San Diego) performed pretty favorably.

Cox (92.73 ms), UltraDNS (100.95 ms), OpenDNS (117.07 ms), Google Public DNS (197.01 ms)

jf
I was also surprised to see that my IPS's DNS servers performed so well against all other services.
dfranke
It shouldn't be surprising. Your ISP's server serves enough customers that most popular sites will be cached most of the time, and you're likely to have a shorter ping time to your ISP's server than to any other.
lazyant
Slightly off-topic but when I first got my MacBook about two years ago the networking was awfully slow compared to my Windows or Linux boxes.

I suspected a DNS problem and after installing a local DNS cache server it improved dramatically, so a local cache-only/forwarder DNS server is another solution for a faster networking experience.

dschobel
Cool that it considers automatic domain redirection an error:

www.google.com. hijacked (google.navigation.opendns.com.) NXDOMAIN Hijacking

jamesk2
I'm getting very different results when I run these tests. At prime time 7 pm to 9 pm pacific time on weekdays is different than sundays at noon.