|
|||||||
![]() |
|
|
LinkBack | Thread Tools | Search this Thread | Rate Thread | Display Modes |
|
|
#1 |
|
Member (13 bit)
|
A buddy of mine and I were talking, and we got upon a topic both of us had no clue upone.
When a client requests a domain name, such as http://www.pcmech.com, how does the DNS system work? We found two possibilities that seem to work:
We'd appriciate someone to tell us the correct way things are done. |
|
|
|
|
|
#2 |
|
Premium Member
Join Date: Jun 1999
Posts: 9,231
|
Hi Dok:
The answer is (2) There are 2 kinds of lookups actually iterative & recursive.. In iterative: The client's DNS request, say http://www.institution.ernet.in [a UDP packet by gethostbyname() ]if not found on the cached DNS tables of the ISP, is then sent by the ISP to the root server of the domain in question - .in. The Root server simply sends the address of the server [lets call it 'ABC'] that has the SOA for the domain. The ISP's server now with this information asks 'ABC' for the information pertaining to the .ernet and is responded to with a new server address.. and so on. Remember the tree system of DNS is bottom up as opposed to a filesystem. In terms of a recursive lookup, the resolver actually sends recursive queries. The queried name server then has the obligation to respond with the answer to that query or with an error code. The name server cannot refer the resolver to another name server. In case the queried name server is not authoritative for the requested data, it has to resolve the query itself, either recursively or iteratively. Current implementations resolve the query iteratively and do not pass the query to another server. Hope this is of some use... Cheers |
|
|
|
|
|
#3 |
|
Member (13 bit)
|
I was right!...kinda. I guess he didn't think that each ISP would want a so called "HUGE" database to deal with.
Thanks for the info EX! |
|
|
|
![]() |
| Bookmarks |
| Thread Tools | Search this Thread |
| Display Modes | Rate This Thread |
|
|