Summary of the packet structure under the TCP/IP protocol cluster

I. Hierarchy of TCP/IP protocol clusters:
 
II. Specific protocols include:
 
Note: 1. ARP and RARP are used for switching between physical addresses and network addresses. They should belong to the physical link layer. However, since ARP and RARP both have their own frame types in Ethernet frames, they are drawn between Ethernet drivers and IP addresses. 2. ICMP and IGMP are usually used to query network management, that is, they are considered part of the IP layer. But they are all encapsulated in IP message transmission, and the IP protocol type fields have their own fields, so they are painted between the IP address and the transmission layer. 3. The port number is enclosed in the brackets of the service layer, and the protocol field type is included in the brackets of other layers.
Iii. packet structure (1) Ethernet CSMA/CD network 802.3. Token Bus 802.4 and Token Ring bus 802.5 are the standard sets published by the 802 committee, of which 802.2 define the common characteristics of the three. Currently, Ethernet is basically used, and CSMA/CD is used for access. However, 802.3 and 802.2 (RFC1042) define two different frame structures than Ethernet (RFC894. Here, only the frame Encapsulation Format of Ethernet is shown:
 
Note: The data part must contain at least 46 bytes. If the data part is not enough, You need to insert the padding bytes.
(2) ARP and RARP
Note: 1. the ARP frame type is 0x0806, and RARP is 0x0835. 2. If the hardware type is 1, it indicates the ethernet address (MAC). If the protocol type is 0x0800, it indicates that the Protocol address type is IP. 3. When the hardware type is 1 and the protocol type is 0x0800, the length of the hardware address and Protocol address is 6 or 4 bytes, that is, 48 or 32 bits, respectively. 4. The operation types can be ARP request (1), ARP response (2), RARP request (3), and RARP response (4 ). 5. Note that the last four and the first two have duplicate data.
(3) ICMP
Note: 1. Different Types and codes are combined to explain different packets. 2. The second 4-byte message has different contents. For example, the Ping program is an identifier and a serial number.
(4) IGMP
Note: 1. IGMP is a 2.32-bit Internet Group address used to identify different groups.
(5) IP
Note: 1. version Description: IPv4 and IPv62. service type: for example, minimum consumption, maximum throughput, maximum reliability, and minimum cost. the total length is up to 65535, but MTU and host requirements are limited. The total length includes the header and data 4. ID number identifies each packet sent by the host. there are three flag spaces, including whether to split (the first is not used, the second is not part identification, and the third is whether there are subsequent parts ). The Slice offset is used to calculate the data location of the slice. 6. TTL indicates the maximum number of routers that a datagram can pass. 7. The protocol is used to describe the protocol type encapsulated in an IP address.
(6) TCP
Note: 1. serial number and validation serial number are used to provide confirmation and retransmission 2. because there are optional options, there is a header length of 3. URG identification emergency pointer, ACK identification confirmation, PSH identification needs to immediately pass the data packet to the upper layer, RST identification reset, FIN identification ends when waving. 4. window size is used for throttling (Sliding Window Mechanism) 5. Emergency pointer pointing to the location of emergency data
(7) UDP
Note: 1. The length of a data packet includes the header and data. 2. Check is the same as IP, TCP, IP, ICMP, and IGMP, but not CRC (the Ethernet frame is CRC)
4. Summary these packets are available in the TCP/IP explanation. record them as a deep point of memory.