Baoyan postgraduate entrance examination interview - Network

Ashore some midstream 985, The important knowledge points summarized below ：

1. Functions of computer network ：

① data communication ： File transfer 、 E-mail, etc

② Resource sharing ： Data sharing 、 Software sharing, etc

③ distributed processing ： Assign tasks to other computers on the network to reduce the burden

⑤ Load balancing

④ Improve reliability ： They are interchangeable

TCP/IP Five layer structure and corresponding protocols 、 effect ：

The physical layer	（ Transmission unit ： The bit ） Transparently transmit bitstreams over physical media		Two theorems , Channel multiplexing technology
The link layer ：	（ Transmission unit ： frame ） Package into frames , Transparent transmission , Error control , flow control , Provide services for network layer	SDLC,HDLC,PPP,STP	The sliding window ,GBN Deal with the SR agreement ;ALOHA Department agreement ,CSMA Department agreement ,MAC frame
The network layer	（ Transmission unit ： The datagram ） Datagram service , Heterogeneous network interconnection 、 Routing and forwarding 、 Congestion control	IP,ARP（IP-MAC）DHCP（ Dynamic allocation IP）	Routing and algorithms 、 agreement ; Network forwarding process ;NAT; Subnet partition ; Multicast ; Datagram and virtual circuit , Three exchanges
Transport layer	（ Transmission unit ： Message segment ）① Logical communication between different processes ② Provide a port for the application layer to deliver the data of the application process down ③ Reuse and reuse technology , Error detection technology ④ Two protocols are provided TCP,UDP	TCP,UDP	Three handshakes 、 Four handshakes 、 flow control 、 Congestion control 、
application layer		FTP,Http

What are the relay systems ？

① The physical layer ： Repeater 、 A hub

② Data link layer ： bridge 、 Switch

③ The network layer ： Router

④ The network layer goes up ： gateway

ISO,OSI More session layer and presentation layer

TCP、IP The physical layer , The data link layer is merged into the network interface layer

Code element , channel , Interaction mode of communication parties , Two theorems

Network address translation NAT： It refers to the conversion of private network address to public network address

Switching mode of communication network ：

① Circuit switching ： Connection required 、 The data transfer 、 Connection release

② Message switching ： No connection needed , Store and forward transmission mode is adopted , Dynamic distribution circuit .

③ Packet switching ： Solve the problem of large messages in message exchange , The length of the group is fixed , Simplified storage management .

How media access is controlled

Static channel division and medium control

① Frequency division multiplexing

② Time division multiplexing

③ Bo Fen takes

④ Code points : Each chip is equivalent to a vector , The chips sent by each station are orthogonal to each other .

6. Dynamic partition of media access control ：

① ALOHA： Don't test , Send at random , There are conflicts , Random resend —— Throughput is very low

② Timeslot ALOHA ： It can only be sent at the beginning of each time slot —— Throughput ratio ALOHA Twice as big

③ CSMA agreement ： Before sending, listen to whether the public channel is sending , Send when the channel is found to be idle （ Whether there is data on the road is unknown ）

④ CSMA/CD : carrier sense 、 collision detection - Listen before sending , Send and listen , If no collision is detected after the contention period, it is determined that no collision has occurred . If... Is detected , Then send a congestion signal to let all users know and use the backoff algorithm , Wait a random time to return to step 2.

⑤ CSMA/CA : If the channel is detected to be idle, wait DIFS after , Send the whole frame of data ;

（ From the second frame ） Otherwise, execute the backoff algorithm , If a busy channel is detected, the backoff timer remains unchanged , Otherwise the countdown , Until 0 , Send the whole frame , After receiving the confirmation frame, the second frame will be sent .

Flow control mode and transmission mechanism

Stop waiting ： Send a frame , Send the next frame after receiving the response

Sliding window mechanism ： Send window ： The range of frames that can be sent , Slide a frame forward without receiving a confirmation frame

Receiving window ： The range of windows that can be received , Every time a frame is received, the window slides forward and sends a confirmation frame

Reliable transmission mechanism ：

Stop waiting for agreement ： Over time retransmission

back off N Frame protocol ： The sender can send multiple frames simultaneously and continuously , After the receiver detects the ordered frames and makes them correct , Return a confirmation frame （ The current maximum number of frames ）, When the sender receives the acknowledgement frame , The window slides to the next frame of the maximum numbered frame .

If a group of frames is wrong , The group is discarded , Send a confirmation frame , The content is the number of the currently confirmed maximum number frame

Choose retransmission protocol （SR）： Single frame confirmation , The receiving window becomes larger , Frames in the window are welcome , It doesn't require order , If in front of the window N If all frames have been received, they can be delivered to the network layer and the window can be slid .

Send window ： The elements in the window can be sent , If the element in the window is sent and timed by the timer , If an acknowledgement frame is not received within the time limit, the frame will be retransmitted . front N Each element receives a confirmation frame , Windows can slide .

bridge 、 A hub 、 Switch

A hub ：

bridge ：

The data processing object is a frame , It has the ability of addressing and routing , according to mac The destination address of the frame forwards and filters the frame .

The same conflict domain can become a network segment

advantage ： It can interconnect different types of LANs , Expand the scope of Physics , Different physical layers can be used

shortcoming ： It is suitable for LAN with few users , Otherwise, it will occur Broadcast storm （ Too much broadcast information causes network congestion , Intra segment broadcast ）

Transparent bridge ：

Plug and play , Self learning algorithm

​

Source routing bridge ： When sending frames , The detailed best routing information （ Routing is minimal 、 Shortest time ） Put it at the beginning of the frame .

Ethernet switch ：– Multi interface bridge , The same principle as the bridge , There are multiple interfaces , That is, you can connect multiple conflict domains .

cut-through switches

Store and forward switch

NAT, HYPERNET , Multicast , Socket

Subnet mask and subnet partition ：

Subnet partition ： network number + The two-level address of the host number is not flexible enough , Utilization is very low

Partition subnet knowledge will IP The host number part of the address is divided , Can be expressed as {< network number >,< Subnet number >,< Host number >} In the third part of , The first two parts can be collectively referred to as network prefixes

HYPERNET ：

Reduce the number of routes , Reduce the routing burden .

ARP Address resolution protocol ：IP Address to MAC Address mapping

Dynamic Host Configuration Protocol DHCP： Dynamically assign to the host IP Address , It's the application layer protocol , be based on UDP

Internet control message protocol ICMP： Improve IP The chance of a successful datagram delivery

Multicast ：

Similar to broadcasting , Multicast data is sent only once , When the transmission path changes, the data is copied and then forwarded , Multicast the last time

Multicast address 224.0.0.0 - 239.255.255.255

Socket ：

IP Part is used to distinguish different hosts , The port number is used to distinguish different processes

Network layer forwarding packet flow

1. From the destination host IP Address D Get the network address N,

If network N Connect directly to this router , Then the datagram is delivered directly to the destination host D, This is called direct delivery .

If the destination address in the routing table is D Specific host routes for , Then the datagram is sent to the next hop router specified in the routing table ;

Otherwise, the table has reached N The routing of the network , Then send the datagram to the next hop router specified in the routing table ;

If there is a default route in the routing table , The datagram is sent to the default router specified in the routing table .

Otherwise, it is delivered to the default route , If there is no default route, the forwarding packet will report an error .

[ Failed to transfer the external chain picture , The origin station may have anti-theft chain mechanism , It is suggested to save the pictures and upload them directly (img-CrEKiZrU-1643375281769)(mindmaster_image/1e484639018d7785a262e22f567ccbc4.png)]

​

Three handshakes , Four handshakes

Three handshakes established , Four handshakes released

SYN Flooding attack

wait for 2MSL The reason for time ：

TCP And UDP

UDP

No connection needed , Unreliable transmission ;

The message segment setting length is moderate , No need to slice

The head length is short , Low overhead

no congestion control , It will not affect the sending efficiency of the host

Support one-to-one , One to many , For one more , Many to many interactive communication

UDP test ： Consider a message segment as a number of segments composed of 16 A message segment consisting of bit substrings , Sum by binary inverse code operation , If the result of the summation is 1 Right （ Insufficient even digit complement 0）

TCP

Point to point , Need to establish a connection , reliable , Byte stream oriented , Full duplex communication

why TCP reliable ,UDP unreliable

TCP There are three handshakes to establish a connection , Four waves close the connection mechanism . In addition, there are sliding windows and congestion control algorithms . The most important thing is to retain the mechanism of timeout retransmission . There is also verification for each message , Ensure the reliability of each message .

UDP Datagram oriented connectionless , The datagram goes out , You don't keep data backups . Only in the IP Add checksum multiplexing to the header of datagram . UDP There is no concept of server and client . UDP If the message is too long, give it to IP Cut into small pieces , If a message is discarded .

TCP Why it takes three times instead of two ： Prevent the invalid request message segment from being suddenly transmitted to the server B, So there's a mistake

** Why three handshakes when connecting , It's four handshakes when it's closed ？**https://www.cnblogs.com/bj-mr-li/p/11106390.html

answer ： Because when Server End receipt Client Terminal SYN After connecting the request message , It can be sent directly SYN+ACK message . among ACK Messages are used to answer ,SYN Messages are used to synchronize . But when you close the connection , When Server End receipt FIN When the message , It's not likely to shut down immediately SOCKET, So you can only reply one first ACK message , tell Client End ,“ You sent it FIN I received the message ”. Only when I Server All messages are sent , I can send FIN message , So we can't send . So it takes four steps to shake hands .

Why? TIME_WAIT Status needs to pass 2MSL( Maximum segment lifetime ) To return to CLOSE state ？

stay Client Send the last ACK reply , But it's time to ACK May be lost **.Server If not received ACK, Will be sent repeatedly FIN fragment .** therefore Client Can't close now , It has to confirm Server Received this ACK.Client Will send out ACK Then go to TIME_WAIT state .Client A timer will be set , wait for 2MSL Time for . If it is received again within that time FIN, that Client Will be reissued ACK And wait again 2MSL.

None of the following ： sketch SYN attack

SYN Attack and use TCP Protocol defect , By sending a large number of semi connection requests , Occupy half connection queue , cost CPU And memory resources .

How to optimize ：

1. To shorten the SYN Timeout Time
2. Record IP**, If you receive a continuous IP Duplication SYN message , From this IP All packets from the address will be discarded .**

Flow control and congestion control ：

TCP flow control

When the two sides are communicating , The rate of the sender is not necessarily the same as that of the receiver , If the sender's transmission rate is too fast , It will cause the receiver to be unable to handle , At this time, the receiver can only store the data that cannot be processed in the cache （ Out of order packets are also stored in the cache ）.

The following figure shows how to use the cumulative confirmation algorithm ：

[ Failed to transfer the external chain picture , The origin station may have anti-theft chain mechanism , It is suggested to save the pictures and upload them directly (img-TwdfTVQy-1643375281771)(mindmaster_image/183406-762447.png)]

TCP Congestion control ：

Slow start and hard plug to avoid ：

Congestion window from 1 Start exponential growth to the threshold , Post addition growth , If there is a network jam （ That is, the confirmation is not received on time ） The threshold is less current window （ Not the original threshold ） Half of ,

Congestion window from 1 Continue to do the above actions .

Fast retransmission and fast recovery ：

redundancy ACK

Whenever an out of order segment larger than the expected sequence number arrives , Send a redundancy ACK, Indicates the sequence number of the next expected byte .

The sender has sent 1,2,3,4,5 Message segment
The receiver receives 1, Return to 1 The confirmation of ( The confirmation number is 2 The first byte of )
The receiver receives 3, Still return to 1 The confirmation of ( The confirmation number is 2 The first byte of )
The receiver receives 4, Still return to 1 The confirmation of ( The confirmation number is 2 The first byte of )
The receiver receives 5, Still return to 1 The confirmation of ( The confirmation number is 2 The first byte of )

The sender received 3 For message segments 1 Redundancy of ACK—— Think 2 newspaper The text segment is missing , Retransmission 2 Message segment No （ The fast retransmission ）

Fast retransmission ： The sender continuously receives 3 Redundancy ACK, It's a packet loss （ At this time, there is no need to control whether the arrival time is timed out ） , Packet loss means that the network is likely to be congested , To prevent or prevent network congestion , A fast retransmission is required + Fast recovery algorithm , Algorithm is as follows ：
The threshold value is reduced to current window Half the size , Then the window size is added .

DNS And http,ftp

1.P2p Model ：

Not many hosts have the function of uploading and downloading, which reduces the pressure of the service , Multiple customers and can share documents directly .

2.DNS：

[ Failed to transfer the external chain picture , The origin station may have anti-theft chain mechanism , It is suggested to save the pictures and upload them directly (img-FYmUwmPh-1643375281773)(mindmaster_image/172831-854679.png)]

[ Failed to transfer the external chain picture , The origin station may have anti-theft chain mechanism , It is suggested to save the pictures and upload them directly (img-NsO88mEw-1643375281773)(mindmaster_image/183644-380898.png)]

3.FTP：

FTP use Customer / The server works The way ,TCP Reliable transmission

A main process , Responsible for accepting new requests ; In addition, there is the nogango series process , Responsible for handling individual requests . The working steps are as follows ：

① Open the control port 21 , Enable client processes to connect to

② Wait for the client process to send a request

③ Start slave process control client request , The master and slave processes are concurrent , The slave process terminates after processing

④ Back to the waiting state , Continue to accept requests from other customer processes

Http characteristic ：

Reliable transmission 、 No state （ User visits more than once , The response was the same ）

Persistent and non persistent connections

The order in which a page is accessed

The order in which a page is accessed ：

① Browser analysis URL

② Browser direction DNS Domain name server requests resolution IP Address

③ DNS It is concluded that IP

④ Browser pass IP Address to find the server and request a link （80 port ）tcp

⑤ Send a specific message after the connection is established http request

⑥ The server passes http The response sends the file to the browser

⑦ Release tcp

⑧ The browser parses the file and displays it to the user

http And https The difference between

http All transmitted content is clear text , And neither the client nor the server can verify each other's identity . https Having security ssl Encrypted transport protocol , Encryption uses symmetric encryption , https The agreement needs to reach ca Apply for a certificate , Generally there are few free certificates , You need to pay a fee .

sketch TLS/SSL, HTTP, HTTPS The relationship between

SSL Its full name is Secure Sockets Layer That is, the condom layer , His successor is TLSTransport Layer Security Transport layer security protocol , Are used to provide security support for data communication at the transport layer .

Can be HTTPS The agreement is simply understood as HTTP agreement ＋TLS/SSL

https Connection process

1. The browser sends the supported encryption algorithm information to the server
2. The server selects a set of encryption algorithms supported by the browser , Send back to browser in the form of certificate
3. client (SSL/TLS) Analyze the certificate and verify the validity of the certificate , Generate symmetric encryption key , We call this key client key, That is, the client key , Use the public key of the server to encrypt the client key asymmetrically .
4. The client will initiate HTTPS Second of HTTP request , Send the encrypted client symmetric key to the server
5. After the server receives the ciphertext from the client , Will use their own private key for asymmetric decryption , The decrypted plaintext is the client key , Then use client key to encrypt data symmetrically , So the data becomes ciphertext .
6. The server sends the encrypted ciphertext to the client
7. The client receives the ciphertext sent by the server , Use client key to decrypt it symmetrically , Get the data sent by the server . such HTTPS Second of HTTP End of request , Whole HTTPS Transmission complete

After entering a web address in the browser , What exactly happened

1. Conduct DNS Parsing operation , according to DNS The result of parsing can be found on the server IP Address

2. adopt ip Addressing and arp, Find the server , And use three handshakes to establish TCP Connect

3. Browser generation HTTP message , send out HTTP request , Wait for the server to respond

4. Server processing request , And return to browser

5. according to HTTP Open long connection or not , Conduct TCP The process of waving

6. The browser renders the page according to the received static resources