Talk:Web Science/Part1: Foundations of the web/Domain Name System/Delegating DNS queries

Check your understanding clarification
4. How can the DNS system be manipulated (e.g. by governments) to block (censor) web sites? I think it is real censorship for most of the users. Using the same logic I can say that even blocking by IP is not real censorship (like in Russia) because you can always use VPN, proxy, etc. in order to avoid any blocking and visit whatever you want.
 * This is no real censorship since the websites can still be accessed via their IP addresses.

Question number 2 and option "most likely B = C". My former ISP used about 2 weeks DNS cache. I had a lot of problems as a web-sites administrator because of that fact =) --oleamm (discuss • contribs) 04:28, 4 November 2013 (UTC)

Information from the video
Delegating DNS queries

You see domains everywhere. Let’s see “testing.eng.xyz.com” that has such a graph: “com”-> “xyz”-> “eng”-> “testing”. Basically all his nodes represent the domain like we have whole com-domain name that covers everything below it. Or for “asmpx.l.google.com” we have “com”-> “google”-> “l”-> “aspmx” and here we see a google.com-domain name (“google”-> “l”-> “aspmx”) which surrounds everything underneath google. In the top of Domain tree we have root zone that contains itself “''. (root)''”and the top level Domains (.com or .edu or .uk or .net). Below that is the .com-zone that contains itself and all the .com-domains like google, yahoo, amazon, youtube and the example “xyz”. Below that we have .xyz-zone that knows about the divisions like “eng” (“com”-> “xyz”-> “eng”-> “testing”). And there is an “eng”(for engineering)-zone. Each example knows about itself and items right below it.

How DNS queries work?

We have “mobile.dev.eng.xyz.com” and “com”-> “xyz”-> “eng”-> “dev” -> “mobile” for it. Let’s imagine that someone is trying to reach the point “mobile”. The query is going to come into the top in the root-zone first. The root name server that knows about itself and the top level domains is going to look at the query and see that it knows about “com”. So the query gets path down to the com-zone. The name servise for the com-domain look at the query and sees that it should go to xyz.com-domain. So it hands off the query to the next zone down. Then he sees that he should go to “eng”. So again the query is handed down to the eng.xyz.com-zone and from there it goes to dev.eng.xyz.com-zone and the answer is found. This answer is then sent back to the computer that made a request.

But for example, someone is looking for “aspmx” that is in “com”-> “google”-> “l”-> “aspmx”. Once again the query comes into the top in the root-zone first then it is handed to com-zone. Than it goes to the google.com-zone and the answer is found right here. So we send back the answr --Jane Kruch (discuss • contribs) 21:37, 6 November 2013 (UTC)