Reference documentation
Basic knowledge of computer network core
computer network performance index
Rate: Transmission data rate. Unit b/s kb/s Mb/s
Bandwidth: Maximum transfer rate.
Throughput: The amount of data passing through a certain network channel or interface per unit time, in units of b/kb/Mb, etc.
Latency:
Sending delay (the time it takes for a host or router to send data).
Propagation delay (the time it takes for an electromagnetic wave to travel in a channel).
processing delay
Queue delay.
Delay-bandwidth product = propagation delay * bandwidth
Round trip time RTT
Utilization
channel utilization
network utilization
computer network architecture
OSI seven-layer model
Protocol: Rules/standards/conventions established for the exchange of data in a network
The benefits of layering:
- Each layer is independent. Break down the big problem into several independent small problems
- Good flexibility. As long as the interface remains unchanged. The internal implementation can be modified without affecting the upper and lower layers
- Structurally easy to separate
- Easy to implement and maintain
- promote standardization
Five layers of TCP/IP protocol
Application Layer
Transport Layer
Network Layer
Data Link Layer
Physical Layer
physical layer
- The main task of the physical layer is to determine the characteristics related to the transmission media interface
- Mechanical properties
- Electrical Characteristics
- Features
- Process characteristics
- information interaction
- Simplex communication
- half duplex
- full duplex
- Channel Multiplexing Technology
- frequency division multiplexing
- time division multiplexing
- wavelength division multiplexing
- code division multiplexing
data link layer
Data link: In addition to a physical line representing the physical layer, it also includes some communication protocols that control data transmission. The combination of the two is a data link.
Network adapter: The data link layer protocol is generally implemented by the network adapter. It realizes the functions of the data link layer and the physical layer. Its main function is to realize the communication between the computer and the external local area network.
Protocol data unit of the data link layer: frame
Point-to-point data link communication steps (point-to-point data links, data links using broadcast channels are the two main data links):
- Node A encapsulates the IP datagram called by the network layer of the node into a frame by adding the header and the tail.
- Node A sends the encapsulated frame to Node B through the physical layer link.
- When node B checks that the received frame is error-free, it will be handed over to the network layer of node B. Otherwise, the frame will be discarded.
There are three basic problems to be solved by the data link layer protocol:
- Encapsulation into framing: frame start service SOH terminator EOT
- Transparent transmission: place the transmission text with SOH or EOT. It causes an error to start or an error to end
- Error Detection: CRC Cyclic Redundancy Detection is widely used
PPP point-to-point protocol: is the most used protocol at the data link layer
Provide unreliable datagram services. Because the data link layer does not necessarily provide more functions than the IP protocol of the network layer. Therefore, the PPP protocol does not need error correction and does not require sequence numbers. It does not require flow control. Simple is the primary requirement of the PPP protocol and biggest feature.
Ethernet: The most widely used LAN specification today. Uses CSMA/CD technology. And runs at 10M/s over various cables.
CSMA/CD Protocol: Carrier Sense Multipoint Access/Collision Detection
CSMA/CD core points:
- Multipoint access: Many computers are interconnected to a bus in a multipoint access manner, and only one computer can occupy the bus to transmit data at a time
- Carrier Sense: Each station (computer) must constantly detect whether the channel is transmitting data, and can obtain the right to send if it is not occupied
- Collision detection: listen while sending, if it is detected that there are two stations on the bus transmitting data at the same time, stop the transmission immediately
MAC address: The physical address of the Ethernet
Bridge: It can forward and filter frames, can connect different physical layers, MAC sublayers and different Ethernets, and can extend Ethernet at the data link layer, but the disadvantage is that it increases the delay
Hub: can forward bit streams, work at the physical layer, and extend Ethernet at the physical layer
Network layer
- IP address classification and subnet mask
- Address Resolution Protocol ARP: resolves IP addresses of hosts and routers to MAC addresses
- ARP mechanism: Each host has an ARP notification cache. There is a mapping table from the IP address to the MAC address of each host/router in the local area network. And this mapping table is updated frequently.
- International Control Message Protocol ICMP: Provides host or router query conditions/message errors/abnormal conditions
- ICMP message type: 1. Error report message 2. Query message
- ICMP application: use the ping command. At the application layer, over the transport layer, directly use the ICMP protocol to send back request and reply messages. Test connectivity between hosts.
transport layer
Although the IP layer sends the packet data to the destination host.
But strictly speaking, the communication between two hosts in a computer network is actually the communication between the application processes on the two hosts. The endpoint of the communication is not the host but the application process on the host. The network layer provides logical communication between hosts. The transport layer provides between ports logical communication.
UDP
TCP
application layer
DNS: Domain Name System
DNS domain name resolution: resolve domain names into corresponding IP addresses
DNS name server type:
1. Root Domain Name Server
2. Top-level domain name server
3. Authority Domain Name Server
4. Local Domain Name Server
DNS server domain name resolution process:
1. The user enters the domain name of the website to be accessed in the browser. If the operating system checks that the mapping relationship of this domain name is cached in the local hosts file, it will be called directly to complete the domain name resolution.
2. If there is no hosts file, the browser will Request resolution from the local DNS. If the mapping relationship is cached, return the result and complete the resolution.
3. If the local DNS does not have it, send the request to the RootDNS (root DNS server). The root DNS server will tell the local server to query the website for authorization. DNS server. That is, send the IP address of the website authorized DNS server to the local DNS server (the website authorized DNS server is the top-level, authoritative domain name server).
4. The website authorized DNS server sends the IP address obtained by resolution back to the local DNS. Local DNS caches the mapping relationship and sends the IP address back to the user.
5. After the browser gets the IP address, it sends an HTTP request to it.HTTP: Hypertext Transfer Protocol. It is a stateless/connectionless protocol. But it relies on TCP for data transfer, which is stateful/connected.
HTTP must know and must know - summary of common interview questions