4
a url
in the message, we can find the answer is stitched together
http://gaia.cs.umass.edu/cs453/index.html
Uniform Resource Locator, commonly known as web addresses
standard format follows
the protocol type: [// server address [: port number]] [/ resource-level UNIX file path] filename [# fragment ID] [query?]
b http version
1.1
in fact, http addition, only version 1.0
c continuously connected
with a field Connection: keep-alive
D ip address
no ip address, ip because something is the network layer
and the application layer protocol is http, the packet does not contain the ip address
E
Mozilla / 5.0
on why this thing can refer to this article very interesting
http://www.cnblogs.com/ifantastic/p/3481231.html
Different browsers have different ways to resolve the page, so the server needs to know the type of browser in order to facilitate the transmission of the corresponding version of the response page
5
a can, find 200 indicates success, back data 2008.3.7
In addition, there are other status codes, begins with the first number determines the right or wrong reasons
| status code | content |
|---|---|
| 1** | Information, the server receives the request, the requester needs to continue operation |
| 2** | Successful operation has been successfully received and processed |
| 3** | Redirection, further action is required to complete the request |
| 4** | Client error, the request contains a syntax error or could not fulfill the request |
| 5** | Server Error The server error has occurred during the processing of the request |
404 is our most common server can not find the page that is requested address is wrong
Last-Modified b: Sat, 10, DEC2005 and 18:27:46
c Content_Length: 3874
d <Doc, (http whenever files are the beginning)!
agree persistent connection, keep-Alive (also see http1.1 version)
7、
2RTT0+RTT1+RTT2+……+RTTn
This study is a multi-layer DNS server, such as this case,
click on the link (such as amazon), the browser will first ask to access the root server
root server returns a com DNE servers address consuming RTT1
then visit comDNS servers, returns Amazon.com ... ...
Ip address needs to be found so RTT1 + RTT2 ......
Then add the established tcp connection and sends a request message to 2RTT0
26
ab are possible
reasons: Because I was so dry (funny)
Real reason:
A, in the P2P transmission, the transmission range continues to expand in order to ensure the update, there is a rule: Every so often random hair a chunk to a user
so long as Bob in a large enough number of users of P2P network, you You will receive all the documents needed to spell chunk needed
b, Sure, a computer will do, not to mention more than one computer?
In fact, better results, each computer can download different chunk, then pieces together, faster
Supplementary questions
1, a brief description of the characteristics of the three architecture application layer
| structure | C/S | P2P | mixing |
|---|---|---|---|
| Outline | The client connection to a central server | Point to point transmission | LAN P2P, the entire Internet on C / S |
| advantage | Permanent service, domain stability | Any node communication, fast, highly scalable, | Combined advantages of both |
| Shortcoming | Flow, easy to block the server, down | Difficult to manage, difficult to maintain copyright | To avoid the disadvantages of both |
2. What is a socket, the user agent and Web caching? Description of the relationship between the user process and the socket
Sockets: TCP with the host IP address with the port number as the endpoint of the TCP connection, the endpoint is a socket
User-agent: a special string head, so that the server can identify the operating system version used by customers, CPU, browser, etc.
web cache: for temporary storage of web pages, a technique to reduce server latency
Socket socket is for a process to communicate with each other and other processes channels and methods
3. What is P2P file sharing? Brief characteristics of the contents of three methods of positioning
p2p file sharing: p2p software allows users to access media files
1) centralized directory
How it works:
1. Note: When the peer starts, its IP address and can share content notification server
2, the query: User queries need to share objects
3, access to content:
4, Update: When the peer access to new objects when or deleting objects, notice the directory server update
Question: single points of failure, performance bottlenecks, copyright infringement, reliability (transmission dispersion, orientation concentration)
2) query flooding
3) level of network coverage
