Some quick data to mull over
Check this out: http://pastebin.com/maae99d7
It might not mean so much to you, so let me boil it down and do some analysis on what they mean. Quick a note on the DNS servers I’m pinging:
- 8.8.8.8 is Google public DNS’ primary DNS server address.
- 24.143.124.6 is my local provider’s primary DNS server address (or so I’m told via DHCP)
Now, given the few queries I performed, it looks like my local DNS provider is winning in terms of latency for every query, save for the query where it’s obvious Google already had a cached answer the first time around (for amazon.co.jp). It also looks like most non-cached responses are returned in the 100 ms range and cached responses are returned in the 50 ms range. This seemed to be fairly consistent.
Those are pretty good numbers, but they don’t beat my local DNS provider. I’m not convinced there is too much value added here in the short term, but in the long run the value added would be an improvement of Google’s products, which trickle down to a better end user experience for me (and everyone else).
I’m going to do a more indepth analysis when I get some more free time.
(As a note, dig by default uses UDP packets to query the servers. Google says they support TCP as well, so those numbers would probably have an overall higher latency).
