Architecture of network application in computer network 13--

Contents of this page

1. Introduction to three architectures for network applications

2. Client/server architecture

3.P2P structure

4. Hybrid structure

1. Introduction to three architectures for network applications

There are three main architectures for network applications:

A. Client/server structure =client-server (c/s)

B. Peer-to-peer structure =peer-to-peer

C. Mixed structure =hybrid

Well, let's take a look at these three types of structures in detail below.

2. Client/server architecture

First look at the client/server structure, this structure is not difficult for everyone to understand.

It works by requesting a service from the client to the server, and the server feeds back the results of the service to the client. We usually play the lol,web is the use of this C/s structure.

It requires a high level of server because you don't want to have a server problem when you're in lol. The requirements of the client is not high, after all, you do not want to play the game when you turn off the computer directly, the impact on others is not big.

What exactly does it have to do with the server and the client's requirements? Let's wait and see:

A. Requirements for servers: to provide permanently accessible addresses/domain names (the ever-changing domain name is hard to remember)

takes 24 hours to service all the time (so players can lol anytime, anywhere)

Scalability (scalability with a large number of servers, which is necessary when the user is massively increased)

B. Requirements for clients: able to communicate with the server, using the services provided by the server (if this is not possible, there is no way, never a single Loler)

can intermittently access the network (come to, want to walk on, 5 missing 1 so amputated opposite)

You can use dynamic IP(it doesn't matter, no matter how things change, nobody cares)

does not communicate directly with other clients (yes!) Have to put my big server in the eye)

Finally, we have a picture for your reference:

3.P2P structure

Can the client and the client communicate with each other? No, not in front of the said, this is one of the basic requirements of the client. But that is in the C/s system structure, now has changed, is peer-to structure! What's the peer? Peer is equality and freedom. In-Peer architecture: There is no forever online server , each server, can be a client, each client, can also be a server. If you want to do it, leave your job. Any end system/node can communicate directly . nodes can be intermittent access to the network (rest and Rest), can also change the IP address (free relocation, no need to change the account).

The peer structure is really so bull? Yes, there is a picture for proof:

So why do you need another network application architecture? In a word, the structure is complex and difficult to manage . Even so, due to its high degree of shrinkage , it is still the eyes of the bright pearl.

4. Mixed structure say gold no can't pure, nobody is perfect, everything has two sides, C/s and peer network structure is so. We should take the essence of it and discard its dross. Mix them and breed super rice.Mixed StructureThat's how it's done. Crossing, and let me come to you:

Mixing, how to mix? Peer management is difficult, then we will be mixed with C/s centralized management. The server pressure in C/s is too high, we will be mixed with the p in the peer to reduce its burden. That's what Napster did.

Who's Napster?

It's ignorant. Napster is ..... (Please own Baidu).

Let's superficial understanding first (whoever it is!) ) and look at how it works:

Its file transfer uses the peer-to structure, the file search uses the C/s structure (centralized). That is, the peer in the p, what you want, what you have, all tell me the server, the one who has what you want, I call it immediately to you, this person who wants you, you immediately pass this thing to it. That is, each node is obliged to register its own content with the central server , each node has the right to submit a query request to the central server to find the content of interest .

Oh yeah! Finished. I hope everyone can understand.

Architecture of network application in computer network 13--