Switch Management and Operating Concepts

Layer 2 MAC Destination Address (DA)

Layer 2 MAC Source Address and Destination Address (XOR of SA and DA)

Layer 3 IP Source Address (SIP)

Layer 3 IP Destination Address (DIP)

Layer 3 IP Source Address and IP Destination Address (XOR of SIP and DIP)

Remember the following guidelines when creating a link aggregation group:

The ports used in a group must all be of the same media type.

The ports used for each group must all be on the same switch.

The ports in a group must be contiguous.

Ports can only be assigned to one link aggregation group.

Use the same link aggregation mode (Trunk or LACP) on both ends of the link.

None of the ports in a group can be configured as a mirror source port or a mirror target port.

All of the ports in a group must be treated as a whole when added to, or deleted from, a VLAN.

STP (Spanning Tree Protocol) will use the port parameters of the master port in the calculation of port cost and in determining the state of the link aggregation group. The following formula is used to calculate the path cost:

group path cost = (path cost of master port) minus (number of ports in the group)

STP treats all ports in a link aggregation group as a single port and will block the entire group if it is a redundant link.

Data transmitted to a specific host (destination address) will always be transmitted over the same port in the group. This allows packets in a data stream to arrive in the same order they were sent.

The configuration of the lowest numbered port in the group becomes the configuration for all of the ports in the aggregation group. This port is called the master port of the group, and all configuration options-including the VLAN configuration-that are applied to the master port are applied to the entire link aggregation group.

Load sharing is automatically applied to the links in the link aggregation group, and a link failure within the group causes the network traffic to be directed to the remaining links in the group. The default load sharing algorithm is based on the source IP address, but options to use the destination IP address, source MAC address, destination MAC address, or a combination of them is selectable.

Switches or servers that use a load-balancing scheme that sends the packets of a host-to-host data stream over multiple ports cannot use this scheme in creating a link aggregation connection with the ZT 8101 switch.

The link aggregation group(s) should be configured prior to connecting any cable between the switches to avoid creating a data loop. Before removing a link aggregation group, you should disconnect all link aggregation cables or disable the all ports in the link aggregation group to avoid creating a data loop.

When using a load sharing algorithm based on layer 3 IP addresses, dynamic rebalancing of a reestablished failed link in the link aggregation group will not occur. The load sharing algorithm will be applied to any new IP addresses learned after the link is reestablished, but the exiting learned IP addresses will not be rebalanced over the entire link aggregation group.

34

Intel® NetStructure™ ZT 8101 10/100 Ethernet Switch User’s Manual

Page 34
Image 34
Intel ZT 8101 10/100 user manual Switch Management and Operating Concepts

ZT 8101 10/100 specifications

The Intel ZT 8101 10/100 is a highly regarded network interface controller designed for efficient data communication in both home and enterprise environments. This versatile chip offers robust support for various networking technologies, making it suitable for a wide range of applications.

One of the most significant features of the Intel ZT 8101 is its capability to operate at both 10 Mbps and 100 Mbps, allowing for seamless integration into existing networks. This dual-speed functionality ensures that users can enjoy the benefits of faster data transfer rates while still maintaining compatibility with legacy hardware. The device automatically detects the network speed, facilitating a plug-and-play experience that minimizes user intervention.

The Intel ZT 8101 utilizes advanced features such as full-duplex support, which enables simultaneous data transmission and reception. This capability significantly enhances network efficiency and maximizes throughput, making it ideal for environments with high data traffic. Moreover, the chip employs sophisticated packet processing algorithms to prioritize data, reducing latency and ensuring smoother communication.

In terms of power efficiency, the Intel ZT 8101 is designed to consume minimal power, making it a suitable choice for energy-conscious applications. Its low power consumption allows for more efficient operations, contributing to overall system stability and longevity. Additionally, it incorporates power management features that can dynamically adjust the power usage based on network demand.

Another notable characteristic is the integration of hardware-based flow control, which helps prevent data packet loss during high-utilization periods. This capability is essential for maintaining the integrity of data transmission in busy network environments. The chip’s robust error detection and correction mechanisms further enhance data reliability, minimizing the chances of transmission errors.

The Intel ZT 8101 supports various network standards, including IEEE 802.3 and IEEE 802.3u, ensuring compatibility with a wide range of Ethernet devices. Its flexibility makes it an excellent choice for network upgrades, expansions, or new installations, allowing users to tailor their network infrastructure according to specific needs.

In summary, the Intel ZT 8101 10/100 network interface controller stands out for its dual-speed support, energy efficiency, and advanced networking features. Whether for home users looking to improve their network performance or businesses seeking reliable data communication solutions, the ZT 8101 represents a compelling choice that combines technology, reliability, and efficiency. As network demands continue to evolve, this Intel chip remains a fundamental component in many networking scenarios.