Principles of automatic Ethernet negotiation and related network tests

The principle of automatic Ethernet negotiation and related network test automatic negotiation are one of the most controversial concepts in Ethernet. Is it good or bad? This depends on the situation. When you use automatic negotiation, the supply, driver version, and administrator selection are all considerations. Www.2cto.com
What is automatic negotiation? Article 802.3 of the 28th standard defines the automatic negotiation function as follows: it allows a device to advertise the way in which it runs to a device at the far end of the link, and detects the corresponding running mode of the remote notification. The purpose of Automatic negotiation is to provide an information exchange method for the two devices that share a link, and automatically configure them to work under the optimal capability. Literally, automatic negotiation is a way to achieve the maximum possible transmission rate between two devices. It allows devices to "discuss" possible transfer rates in one way, and then select the best rate acceptable to both parties. They use a token called a fast link pulse to exchange notifications about their respective transmission capabilities. Slave allows the Peer to know the transmission capability of the source end. When the secondary node is switched, the two sites use the following best method to detect the two sites based on their priorities from high to low. 1000BASE-T full duplex 1000BASE-T100BASE-T2 full duplex 100BASE-TX full duplex 100BASE-T2100BASE-T4100BASE-TX10BASE-T full duplex 10BASE-T for example, A and B are automatically negotiating, and A has 10/100/1000 full half duplex capability, however, B only has 10/100 full half duplex capability, so the maximum link capability of both sides is 100, full duplex. Once both parties negotiate automatically, the link will run under the optimal capability supported by both parties. What problems does www.2cto.com automatically discuss? Most of the problems related to automatic negotiation are due to the fact that one party does not work in the Automatic negotiation mode. When one site works in the Automatic negotiation mode and the other party does not, only one Party sends a fast link pulse. The other party has already set a specific rate and duplex mode, so that it will not negotiate with the peer end. He has been forcibly set, so he will not consider the working method of his connection end.

Because the sites that are forcibly set do not tell the sites that are being negotiated their own rate and single-duplex mode, the sites that are automatically negotiated must decide the appropriate rate and single-duplex mode to match the peer end, this is called Parallel detection. The communication rate can be identified by the link pulse from the peer end of the negotiation site monitoring. 10,100 and 1000Mbs Ethernet use different signal methods, so the negotiation site can identify the working rate of the Peer end. However, full half duplex is another thing. Because the sites that are forcibly set are not negotiated, there is no way for the sites to know which duplex method the sites work in. In order to avoid full half duplex mismatch, the negotiated site must use the same rate as the forced site according to the 802.3 standard, but work in half duplex mode. Regardless of the speed (except 10 gig), half duplex is the default Ethernet mode. In many cases, full half-duplex mismatch occurs. For both ends of www.2cto.com to reach the full duplex mode, either both ends are automatically negotiated or both ends are forcibly set.

Do not enable automatic negotiation at one end and forcibly set at the other end. This causes a duplex mismatch. We recommend that you leave the settings at both ends in Automatic negotiation to reduce human errors. How can I know the full half duplex mismatch problem? Duplex mismatch may cause conflicts on the link because one end can be sent and received at the same time, while the other end cannot. Full-duplex sites cannot work well with half-duplex sites. Conflicts may occur on the half-duplex site side. If the affected port is retransmitted, the poor throughput and the high number of errors may damage the link performance. When you use OptiView of fluke network, an error is displayed on the switch port.