|
Home > Archive > CCNA > March 2001 > OSI question
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
|
|
| dmaftei 2001-03-14, 5:10 pm |
| Hey folks,
I saw this question in the Network+ forum some time ago. What's your answer?!
What layer of the OSI model is responsible for Name Resolution?
a. DataLink
b. Network
c. Application
d. Transport
Cheers! | |
|
| network, dns
Thats my pick 
Azam | |
|
| I would agree that DNS and WINS are at the Network Layer of the OSI Model because names have to use IP addresses. If you were doing MAC addresses, that would be at the Data Link Layer.
tristinjbg, MCP+I, MCSE | |
|
| Name resolution: generally the process of associating a name with a network location. I go with network also,
Bean | |
|
| In my opinion, should'nt DNS (port 53) be in the application layer just the same as telnet (port 23)? Any thoughts people ?
Nico | |
|
| hehe.. I think dmaftei has all you guys
the correct answer is the Transport Layer
(reflects my thoughts only)
Explaination:
The functions of the Transport Layer can best be described using an airline analogy. Suppose you are flying your children to Grandmas house unaccompanied. The Data Link Layer would be the planes making the connecting flights. A small fee will insure that ground attendants or Network Layer will get your kids from the first flight to their connecting flight. The Transport Layer will call Grandma to let her know they are coming and what their luggage looks like, and will expect a call from Grandma when they have arrived safe and sound. | |
|
| Interesting Analogy, Imran. It will help me to remember better
cheers,
Nico | |
|
| quote:
Originally posted by nico
In my opinion, should'nt DNS (port 53) be in the application layer just the same as telnet (port 23)? Any thoughts people ?
Nico
I agree with you Nico.
Actually this is an example of a case where it helps to UNDERSTAND the OSI model rather than memorizing it. Name resolution packets are encapsulated within transport layer segments. This is true for ordinary Internet host names (usually resolved with DNS) as well as NetBIOS names (usually resolved with WINS in the MS TCP/IP world). You could argue that the layer above the transport layer in the OSI model is the session layer. In TCP/IP world the session layer and presentation layer are (usually) not separate layers but part of the application layer.
Even if we are talking pure ISO OSI name resolution data has to be encapsulated within presentation layer messages so that all parties have an agreed upon representation of the name as well as its address. So we still arrive at the application layer as the answer to dmafteis question.
Your comments and corrections are welcomed. ESPECIALLY if you disagree.
Terje | |
|
| quote:
Originally posted by nico
In my opinion, should'nt DNS (port 53) be in the application layer just the same as telnet (port 23)? Any thoughts people ?
I think nslookup would be in the application layer, but DNS as a protocol would live futher down like in the transport or session layer. Since session isn't an option, I'll go with transport.
MadChef | |
|
| quote:
Originally posted by MadChef
I think nslookup would be in the application layer,
The OSI model is a conceptual model for arranging PROTOCOLS. "nslookup" is not a protocol at all, it's a program (or application if you want).
Important: The application layer of the OSI model is for application layer protocols, not for applications. X.400 is a protocol for exchanging mail. It belongs on the application layer of the OSI model. The actual mail program that happends to use X.400 does not belong in the application layer of OSI nor anywhere else in OSI for that matter.
quote:
but DNS as a protocol would live futher down like in the transport or session layer. Since session isn't an option, I'll go with transport.
DNS is encapsulated in UDP (sometimes TCP). UDP and TCP are generally considered transport layer protocols so DNS should be above that.
Feel free to disagree. I appreciate your opinion.
Terje | |
|
| This is straight from Cisco's mouth, take it however you want. These protocols are in a section on application layer.
-----------------------------------------------
The application layer supports addressing protocols and network management. It also has protocols for file transfer, e-mail, and remote login.
DNS (Domain Name System) is a system used in the Internet for translating names of domains and their publicly advertised network nodes into addresses.
WINS (Windows Internet Naming Service) is a Microsoft-developed standard for Microsoft Windows NT that automatically associates NT workstations with Internet domain names.
HOSTS is a file created by network administrators and maintained on servers. They are used to provide static mapping between IP addresses and computer names.
POP3 (Post Office Protocol) is an Internet standard for storing e-mail on a mail server until you can access it and download it to your computer. It allows users to receive mail from their inboxes using various levels of security.
SMTP (Simple Mail Transport Protocol) governs the transmission of e-mail over computer networks. It does not provide support for transmission of data other than plain text.
SNMP (Simple Network Management Protocol) is a protocol that provides a means to monitor and control network devices, and to manage configurations, statistics collection, performance and security.
FTP (File Transfer Protocol) is a reliable connection-oriented service that uses TCP to transfer files between systems that support FTP. It supports bi-directional binary file and ASCII file transfers.
TFTP (Trivial File Transfer Protocol) is a connectionless unreliable service that uses UDP to transfer files between systems that support the TFTP. It is useful in some LANs because it operates faster than FTP in a stable environment.
HTTP (Hypertext Transfer Protocol) is the Internet standard that supports the exchange of information on the World Wide Web, as well as on internal networks. It supports many different file types, including text, graphic, sound, and video. It defines the process by which Web browsers originate requests for information to send to Web servers. - | |
|
| quote:
Originally posted by Terje
The OSI model is a conceptual model for arranging PROTOCOLS. "nslookup" is not a protocol at all, it's a program (or application if you want).
That was not a particularly well thought out example on my part. My bad. Caught again!
Anyway, to further strengthen the confusion, protocols.com and decodes.com both list DNS a session layer protocol.
http://www.protocols.com/pbook/tcpip.htm
MadChef | |
|
| dmaftei:
Can you please tell us the correct answer | |
|
| Now look what ya did, Dan!
IMO, name lookup is one of those protocols that does not directly map to only one layer. For example, during my CCNA studies, RIP was always defined as a network layer protocol according to Cisco.
Come to find out, it spans multiple upper layers as well. (Ignore this if you're studying for CCNA). The same holds true for a router, it has functions which span all layers, but is generally considered to be a network layer device.
By this definition (what is the protocol's or device's main function) and if you want to map a protocol or device to a single layer , then name lookup would more reside at the network layer since it's main function is the resolution of IP addresses.
This definition would make a reverse lookup an application layer protocol...or even a layer 8 protocol.
Food for thought. | |
|
| Its got to be Application. Otherwise my books are wrong. | |
|
| Application.
They both utilize tcp/ip and udp/ip so you can rule out layers 3 and 4. | |
|
| First let me thank you all for your replies. It happened the way I expected: people expressing their opinions, willing to discuss the issue, to accept others' ideas... That's why you, guys, are such a great bunch!
Now, there is a short story about this question. A while ago (one of my "bored" days ) I was wandering around, and I saw this question in the Network+ forum (if you're curious to see that thread, here it is: http://www.examnotes.net/forums/sho...?threadid=14014). It sort of bothered me that the question had eight answers, all of them transport!! I believe the right answer is application, and I dared to post an answer saying so. I was told to "firmly place foot in mouth now"... Whew!! This kind of pissed me off, so I abandoned the matter. For no apparent reason, a couple of days ago I remembered this thing, and I thought to ask you about it (for the reasons shown in the first paragraph ).
imran1430
I don't know if there is A CORRECT answer to this question. Apparently most people (myself included) believe name resolution belongs to the Application layer. I am also aware of the places MadChef mentioned (protocols and decodes), that place name resolution at the Session layer. However, I do NOT think that name resolution has anything to do with the Network or Transport layers. My final answer: application.
firechicken
I like what I did! Starting an intelligent discussion with intelligent people is something I enjoy!
Again thanks all for your answers!
Cheers! | |
|
| dmaftei,
It's members like you and the other veterans that keeps me comming back. I'm more confident in my posts because of you guys/gals because of what I have learned here.
Thanks always for the help,
Bean | |
|
| quote:
Originally posted by dmaftei
imran1430
I don't know if there is A CORRECT answer to this question. Apparently most people (myself included) believe name resolution belongs to the Application layer. I am also aware of the places MadChef mentioned (protocols and decodes), that place name resolution at the Session layer. However, I do NOT think that name resolution has anything to do with the Network or Transport layers. My final answer: application.
IS tere anyways you can explain what the anaology is talking about, Im confused now !  | |
|
| Well,, guess I need to study some more on the OSI model.
Da_n, just when I thought I had it.
"by failing to prepare, you are preparing to fail" | |
|
| quote:
Originally posted by Terje
Even if we are talking pure ISO OSI name resolution data has to be encapsulated within presentation layer messages so that all parties have an agreed upon representation of the name as well as its address.
I would say Terje hit a home run here! It's perfectly feasible for a little-endian client using ASCII (Windows) to send a DNS query to a big-endian server using EBCDIC (Aix ?). Both little-endian to/from big-endian and ASCII to/from EBCDIC conversions take place at the Presentation layer. Therefore name resolution lives at the Application layer.
Thank you, Terje, for settling this issue! |
|
|
|
|