The origin and development of Ethernet, you will understand after reading this!

Basics

1. About Ethernet

Ethernet was developed by the Palo Alto Research Center (PARC), a division of Xerox, in the mid-1970s. Xerox first invented a 2Mbps Ethernet, and later worked with Intel and DEC to develop a 10Mbps Ethernet, commonly known as (Ethernet II or Ethernet DIX). Later, IEEE standardized Ethernet as IEEE 802.3 through the 802 Committee (802 Comitee). It is very similar to Ethernet II.

[[345002]]

In TCP/IP, the encapsulation format of IP datagrams in Ethernet is defined by RFC 894, and the encapsulation format of IP datagrams in IEEE802.3 networks is defined by RFC1042. The most commonly used encapsulation format today is the format defined by RFC894, usually called Ethernet II or Ethernet DIX.

2. Managing MAC table

 show mac address-table
 clear mac address-table

Bind a mac address to an interface

 Switch(config)# mac address-table static machine's mac interface vlan vlan number

To cancel use no mac addres-static ....

Ethernet Data Link Layer

In Ethernet, different media access methods are provided for different duplex modes:

• In half-duplex mode, the CSMA/CD access method is used.
• In full-duplex mode, data can be sent and received directly without having to determine the busy or idle state of the link in advance.

Half-duplex and full-duplex are concepts of the physical layer, while providing different access methods for the duplex mode of the physical layer is a concept of the data link layer. This forms an important feature of Ethernet: the data link layer and the physical layer are related.

Since the physical layer and data link layer of Ethernet are related, a specific data link layer needs to be provided for access to different working modes of the physical layer, which brings some inconvenience to the design and application.

For this reason, some organizations and manufacturers have proposed to further divide the data link layer into the logical link control sublayer (LLC) and the media access control sublayer (MAC). In this way, different physical layers correspond to different MAC sublayers, and the LLC sublayer can be completely independent. As shown in Figure 1-4.

1. MAC sublayer

The MAC sublayer is responsible for the following tasks:

• Provides access to physical links.
• Link-level site identification: Identify each site on the network at the data link layer.
• That is to say, a site address, namely the MAC address, is reserved at this level to identify a unique site on the network.
• Link-level data transmission: Receives data from the LLC sublayer, attaches the MAC address and control information, and sends the data to the physical link; provides verification and other functions in this process.

The MAC sublayer is related to the physical layer, that is, different physical layers have different MAC sublayers for access. In Ethernet, there are mainly two types of MAC:

• Half-duplex MAC: The physical layer operates in half-duplex mode to provide access.
• Full-duplex MAC: The physical layer operation mode is full-duplex to provide access.

Both types of MAC are integrated into the network card. When the network card is initialized, automatic negotiation is generally performed. The operating mode is determined based on the result of the automatic negotiation, and then the corresponding access MAC is selected based on the operating mode.

MAC Address

The MAC address is burned into the ROM of the network interface controller (NIC).

The high bit is the individual/group bit. When its value is 0, it can be considered that this address is actually the MAC address of the device. When its value is 1, it can be considered that this address represents the broadcast address or multicast address in Ethernet, or the broadcast address or functional address in TR and FDDI. The next bit is the G/L bit (also called U/L, where U stands for global). When this bit is set to 0, it represents a globally managed address (assigned by IEEE). When this bit is 1, it represents an address that is locally managed (like in DECnet). Ethernet has always used globally unique addresses.

2. Ethernet frame format

PS: In Ethernet, due to the existence of conflicts, if two hosts on the shared medium send frames at the same time, conflicts will occur.

In Ethernet, the minimum frame length is 64 bytes.

PS: Currently, the Ethernet frames we use are basically Ethernet II frames.

IEEE802.3 frames can be divided into the following categories according to the values ​​of the DSAP and SSAP fields:

(1) When both DSAP and SSAP are set to the specific value 0xff, the 802.3 frame becomes a Netware ETHERNET frame, which is used to carry NetWare type data.

(2) When both DSAP and SSAP are set to the specific value 0xaa, the 802.3 frame becomes an ETHERNET_SNAP frame.

• ETHERNET_SNAP frames can be used to transport multiple protocols. Therefore, SNAP can be seen as an extension that allows manufacturers to create their own Ethernet transport protocols.
• The ETHERNET_SNAP standard was developed by the IEEE802.1 committee to ensure interoperability between IEEE802.3 LANs and Ethernet.

(3) Other values ​​of DSAP and SSAP are all pure IEEE802.3 frames.

3. LLC sublayer

In the previous introduction, the frame structure formed by the MAC sublayer was mentioned, including IEEE802.3 frames and ETHERNET_II frames.

In the ETHERNET_II frame, the Type field distinguishes the upper layer protocol. At this time, there is no need to implement the LLC sublayer, and only a MAC sublayer is included. In addition to defining traditional link layer services, the LC sublayer in the IEEE802.3 frame also adds some other useful features. These features are provided by the DSAP, SSAP and Control fields.

For example, there are three types of point-to-point transmission services:

• Connectionless packet transfer service: Current Ethernet implementations use this service.
• Connection-oriented reliable data transmission service: establish a connection in advance and then transmit data, and the reliability of data is guaranteed during the transmission process.
• Connectionless data transmission service with confirmation: This type of data transmission service does not require establishing a connection, but it adds a confirmation mechanism in data transmission, which greatly increases reliability.

The following example illustrates the application of SSAP and DSAP. Assume that terminal system A and terminal system B want to use a connection-oriented reliable data transmission service. The following process will occur:

• A sends a data frame to B, requesting to establish a connection-oriented reliable connection.
• After receiving the message, B determines whether its resources are sufficient (i.e., whether too many connections have been established). If so, it returns a confirmation message containing the SAP value identifying the connection.
• After receiving the response, A knows that B has established a connection with itself locally. A also opens a SAP value to represent the connection and sends a confirmation to B. The connection is established.
• A's LLC sublayer encapsulates the data it wants to transmit, where the DSAP byte is filled with the SAP returned by B and the SSAP byte is filled with the SAP it has created, and then sends it to the MAC sublayer.
• A's MAC sublayer adds the MAC address and LENGTH field and sends it to the data link.
• After receiving the data frame, the MAC sublayer of B submits it to the LLC sublayer. The LLC sublayer determines the connection to which the data frame belongs based on the DSAP field.
• B performs corresponding checks and confirmations according to the type of connection, and only sends it to the upper layer after passing these checks and confirmations.
• After the data transmission is completed, A sends a data frame to B to tell B to disconnect the connection and the communication ends.