I. Working Principle and Application of Bridges
Bridges were widely used in business since the beginning of 1980s. Why use a bridge? The reasons are as follows:
(1) Compared with physical layer devices, the bridge can divide the conflict domains to reduce the possibility of conflicts in the network and improve the overall network performance.
(2) The bridge can extend the maximum distance of the LAN. If no bridge is used, the maximum transmission distance of the LAN is 2.5 kilometers. A bridge can even allow the network to span any long distance. For example, you can connect two network segments through a pair of bridges and a point-to-point lease serial line, or a point-to-point link connection through a pair of bridges and a lease satellite channel. A bridge is a network interconnection device in the data link layer (L2) of the OSI reference model, also known as a data link layer device. Therefore, the bridge can connect different network segments of two physical layers. For example, the physical layer transmission medium of one network segment is a coaxial cable, and the physical layer transmission medium of the other network segment is unshielded twisted pair wires. Of course, the network layer (Layer 3) and above of the two network segments must be the same, because the bridge cannot perform protocol conversion. The earliest bridge had only two ports, and now there are multiple ports of the bridge. Hosts in the network segment connected by the bridge can send data frames to each other through the bridge. The header of the data frame contains the source physical address and the destination physical address. The bridge can analyze the physical address and perform corresponding * operations on the data frame. Different CIDR blocks connected by the bridge are the same broadcast domain, and each CIDR block is an independent conflict domain. The bridge can allow communication in different conflicting regions to be performed simultaneously. Therefore, when designing network bridging, you can consider connecting hosts that frequently communicate with each other to the same network segment to improve network performance. In addition, because the bridge needs to store, forward, and analyze data frames, it also increases the latency of the entire network.
Ii. Bridge Type
1. Transparent Bridge)
Transparent bridges are currently the most widely used bridges in the market. They are mainly used to connect Ethernet networks. They can also be used to connect the licensing ring network and FDDI (Optical Fiber Distributed Interface) networks. The transparent bridge verifies the received data frame. If an error is found, the frame is discarded. Transparent bridges transmit data between network segments through learning, forwarding, filtering, and flooding. A transparent bridge can detect the data frames transmitted on each connected network segment and record the relationship between the corresponding physical address and the port in the table (called forwarding table, forwarding information base) (This process is called Learning ). If the destination address of the data frame is in the forwarding table, the frame is forwarded to the corresponding port based on the forwarding table (this process is called forwarding ). If the destination address and Source Address of the data frame are in the same port, the frame is not forwarded (discarded) (this process is called filtering ). If the destination address of the data frame is not in the forwarding table, it is forwarded to all ports except the source port (this process is called flood ).
2. Source Route Bridge)
Source Routing bridges are mainly used to connect the ring network and FDDI network. When a host in a ring communicates with another host in a ring, if the forwarding table does not have route information for the host to the destination, the source host must first determine the route to the destination. The procedure is as follows:
(1) The source host first sends a local test frame containing the target address. If the test frame is returned, the destination host is in the local ring.
(2) If the test frame is not returned, it indicates that the destination host is not in the local ring. For IBM's System Network Architecture (SNA), each source routing Bridge forwards the test frame to all ports and finally to the destination host.
(3) Each time the destination host receives a test frame, it returns the original path to the source host.
(4) normally, the source host only receives the first returned test frame and uses its path as the optimal path. This frame contains the starting ring and bridge number passing through from the source host to the destination host. The data frame sent by the source host to the destination host carries the complete routing information. Then, the net bridge forwards the data to the destination host.
3. Source Route transparent Bridge)
Source Routing transparent bridges are mainly used in environments where Ethernet and the wildcard Ring Network coexist. This type of bridge provides both the active and transparent Bridge functions. The bridge function can be provided for the Ethernet and the Ring Plate respectively. However, you must note that the source routing transparent bridge cannot bridge the Ethernet and the ring-based network, that is, it cannot forward data frames between the Ethernet and the ring-based network.
4. Source Route translation Bridge)
The source route translation bridge is also used for the coexistence of Ethernet and the gateway. However, unlike the source routing transparent Bridge, this bridge integrates the source routing Bridge and transparent bridge algorithm, and can provide the bridge function between the Ethernet and the ring. That is to say, the Ethernet frame from Ethernet to the ring net will be converted to the ring net format when it passes through the bridge, and vice versa. The specific conversion is completed by the Logical Link Control (LLC) Sub-layer of the data link layer.
The bridge works at the data link layer, connects two LANs, and forwards Frames Based on the MAC address. It can be seen as a "low-layer Router" (the router works at the network layer, based on Network addresses such as IP addresses ).
A remote bridge connects two remote LANs through a normally slow link (such as a telephone line). For a local bridge, the performance is important, and for a remote bridge, it is more important to run normally over long distances.
Comparison between bridges and Routers
The net bridge does not know the information of the high-level protocol in the forwarding frame, which enables it to process IP addresses, IPX, and other protocols at the same time. It also provides a network without routing protocols (such as netbeui) segment function.
Because routers process network-layer data, they are more likely to interconnect different data link layers, such as the licensing ring CIDR blocks and Ethernet segments. Bridges are usually more difficult to control than routers. Protocols such as IP addresses have complex routing protocols, making it easy for network administrators to manage routes; protocols such as IP addresses also provide more information about how to segment networks (even if their addresses provide such information ). The bridge only uses the MAC address and physical topology. Therefore, bridges are generally suitable for small and simple networks.
Ii. Reasons for use
Many organizations have multiple local networks and want to connect them. One organization has multiple local networks for the following six reasons:
First, many university departments or company departments have their own LAN, which is mainly used to connect their own personal computers, workstations and servers. Because different systems (or departments) work in different ways, different LAN is selected. These systems (or departments) need to communicate with each other sooner or later, and thus require a bridge.
Secondly, a unit is geographically dispersed and far apart. Instead of installing a coaxial cable network in all locations, it is better to build a LAN in each location, connect with a bridge and an infrared link, and the cost may be lower.
3rd, it may be necessary to divide a logical single LAN into multiple LAN to adjust the load. For example, multiple local networks connected by bridges have a set of workstations and their own file servers. Therefore, most of the communication is limited to a single local area network, reducing the burden on the backbone network.
4th. In some cases, there is no problem with a single LAN in terms of load, but the physical distance between the machines with the farthest distance is too far (for example, more than 802.3 defined as 2.5 km ). Even if the cable is not laid properly, the network will still not work properly due to the long back-and-forth delay. The only way is to segment the LAN and place a bridge between each segment. By using bridges, you can increase the total physical distance of your work.
5th, reliability issues. In a single LAN, a defective node constantly outputs useless information flows, which seriously damages the LAN operation. A bridge can be set to a key part in a LAN, just like a fire door in a building, to prevent the entire system from being damaged due to a single node malfunction.
6th. The bridge helps ensure security and confidentiality. Most LAN interfaces have a promiscuousmode in which the computer receives all frames, including those that are not sent to the host. If multiple bridges are configured in the network and important information that does not need to be forwarded is intercepted with caution, the network can be separated to prevent information theft.
Iii. compatibility issues
Some may naively think that a bridge from a 802 LAN to another 802 LAN is very simple, but it is not. In the nine combinations of 802. X to 802. Y, each has its own special problem to solve. Before discussing these special issues, let's take a look at the common problems these bridges face.
First, various LAN adopt different frame formats. This incompatibility is not caused by technical reasons, but simply by companies that support three standards (Xerox, GM, and IBM), no one is willing to change the standards they support. The result is that the format of frames to be copied between different local networks needs to be rearranged, which takes up the CPU time and recalculates the checksum, in addition, errors that cannot be detected due to a bridge storage error may occur.
The second problem is that the interconnected LAN does not have to run at the same data transmission rate. When a fast LAN sends a long continuous frame to a slow LAN, the speed at which the bridge processes the frame is slower than that at which the frame enters. The bridge must use a buffer to store frames that cannot be processed, and be careful to exhaust the memory. Even a 802.4 to 10 Mbit/s 802.3 Bridge of 10 Mbit/s also has this problem to some extent. Because 802.3 of the bandwidth is consumed in conflict. 802.3 is not actually 10 Mb/s, while 802.4 (almost) is actually 10 Mb/s.
A subtle and important issue related to the bridge bottleneck problem is the timer value of the previous layers. Assume that the network layer on the LAN 802.4 wants to send a long packet (frame sequence ). After the last frame is sent, it starts a timer and waits for confirmation. If the packet must be forwarded to the slow 802.5 network through the bridge, the timer may be reached before the last frame is forwarded to the low-speed LAN. The network layer may resend the entire message for Frame loss. After several failed transfers, the network layer will discard the transfer and tell the destination site of the transport layer that has been shut down.
Third, among all the problems, the most serious problem is that the three 802lans have different maximum frame lengths. For 802.3, the maximum frame length depends on the configuration parameters, but for the standard 10 m/BS system, the maximum payload is 1500 bytes. The maximum frame length of 802.4 is fixed to 8191 bytes. 802.5 There is no upper limit, as long as the transfer time of the site does not exceed the token holding time. The maximum frame length is 5000 bytes if the default token time is 10 ms. An obvious problem arises: what will happen when a long frame must be forwarded to a LAN that cannot receive long frames? In this layer, frames are not divided into small segments. All Protocols assume that the frame either arrives or does not arrive, and there is no provision to restructure smaller units into frames. This does not mean that such a Protocol cannot be designed. You can design and have such a protocol, but 802 does not provide this function. This problem cannot be solved. frames that cannot be forwarded for too long must be discarded. This is also the degree of transparency.
4. Two Bridges
1. Transparent Bridge
The first type of bridge 802 is transparent bridge (transparentbridge) or spanning tree bridge (spanningtreebridge ). The primary concern for people who support this design is that it is completely transparent. According to their point of view, after purchasing the IEEE Standard bridge, multiple LAN units only need to insert the connection plug into the bridge. You do not need to change the hardware and software, do not need to set the address switch, do not need to load route tables or parameters. In short, do nothing, just need to insert the cable to complete, the operation of the existing LAN is not affected by any bridge. This is incredible. They finally succeeded.
A transparent bridge works in a hybrid manner and receives each frame transmitted over all the LAN connected to it. When a frame arrives, the bridge must decide whether to discard or forward it. If you want to forward data, you must determine the LAN to which the data is sent. You need to query the destination address in a large hash in the bridge. This table lists each possible destination and which output line (LAN) it belongs ). When the bridge is inserted, all the scattered lists are empty. Because the bridge does not know the location of any destination, floodingalgorithm is used ): output each incoming frame with an unknown destination to all the LANs connected to this bridge (except the LAN that sends the frame ). Over time, the bridge will understand the location of each destination. Once you know the destination location, the frames sent to the destination are placed on the appropriate LAN instead of distributed.
The transparent Bridge uses the backwardlearning algorithm ). The bridge works in a hybrid manner, so it can see the frames transmitted on any connected LAN. View the source address to know which machine can be accessed on which LAN, so add the previous item to the hash list.
When the computer and the bridge are powered on, powered off, or migrated, the network topology changes accordingly. To handle the dynamic topology problem, the frame arrival time is specified in each hash entry. This item is updated at the current time whenever the destination has arrived in the table frame. In this way, the arrival time of the last frame of the machine can be known from the time of each item in the table. A process in the bridge regularly scans the hash list to clear all table items that are several minutes earlier than the current time. Therefore, if a computer is removed from the LAN and re-connected to the LAN elsewhere, then it can resume normal operation within a few minutes without manual intervention. This algorithm also means that if the machine has no action within a few minutes, the frames sent to it will have to be distributed until it sends a frame by itself.
The process of selecting the route to the frame depends on the sent LAN (source LAN) and the destination LAN (destination LAN), as shown below:
1. If the source LAN is the same as the target LAN, the frame is discarded.
2. If the source LAN is different from the target LAN, the frame is forwarded.
3. If the target LAN is unknown, it will spread.
To improve reliability, two or more parallel bridges are set between the lan. However, this configuration causes other problems because a loop is generated in the topology, infinite loops may occur. The solution is the following spanning tree algorithm.
Build a tree bridge
The solution to the above-mentioned infinite loop problem is to allow bridges to communicate with each other and overwrite the actual topology with a spanning tree to reach each LAN. Use the Spanning Tree to ensure that there is only one path between any two LANs. Once the bridge has agreed on the Spanning Tree, all the transfers between the LAN will follow this tree. Since there is only a unique path from each source to each destination, there is no loop.
To build a spanning tree, you must first select a bridge as the root of the Spanning Tree. The implementation method is that each bridge broadcasts its serial number (which is set by the manufacturer and globally unique), and selects the bridge with the smallest serial number as the root. Next, construct the Spanning Tree Based on the shortest path from the root to each bridge. If a bridge or LAN fails, the calculation is recalculated.
The bridge uses the BPDU (bridgeprotocoldataunit) to communicate with each other. Before the bridge makes its own configuration decision, each bridge and each port need the following configuration data:
Bridge: bridge ID (unique identifier)
Port: port ID (unique identifier)
Port relative priority
Cost of each port (high bandwidth = low cost)
After each bridge is configured, the bridge automatically determines the Spanning Tree Based on the configuration parameters. This process has three phases:
1. Select the root bridge
A bridge with the minimum bridge ID is selected as the root bridge. The bridge ID should be unique, but if the two bridges have the same minimum ID, the bridge with a small MAC address will be selected as the root.
2. Select the root port on all other bridges.
Each bridge except the root bridge needs to select a root port, which should be the most suitable port for communication with the root bridge. Calculate the cost of each port to the root bridge, and use the minimum port as the root port.
3. Select "specified (designated) Bridge" and "specified port" for each LAN"
If there is only one bridge connected to a LAN, it must be the designated Bridge of the LAN. If there is more than one bridge, the minimum cost to the root bridge will be selected as the designated Bridge of the LAN. Specify the port connection to specify the bridge and the corresponding LAN (if there is more than one such port, the low-priority is selected ).
A port must be one of the following:
1. Root Port
2. The specified port of a LAN
3. Blocked ports
When a bridge powers up, it assumes that it is the root bridge, sends a cbpdu (configurationbridgeprotocoldataunit) to inform it of the root bridge ID. A bridge receives a cbpdu with a root bridge ID less than the id it knows. It updates its own table. If the frame arrives from the root port (upload), it will send it to all specified ports (download) distribution. When a bridge receives a cbpdu with a root bridge ID greater than its known ID, the information is discarded. If the frame arrives from the specified port, a frame is sent back to inform the real root bridge of its lower ID.
When the network is reconfigured intentionally or due to a line failure, the above process will repeat to generate a new spanning tree.
2. Select a bridge as the source route
The advantage of the transparent bridge is that it is easy to install. You only need to plug in the cable to complete the installation. However, on the other hand, such bridges do not use the best bandwidth because they only use a subset of the topology (Spanning Tree ). The relative importance of these two (or other) factors led to the split within the 802 committee. People who support CSMA/CD and the Token Bus have chosen transparent bridges, while those who prefer a bridge called sourcerouting ).
The core idea of source route selection is to assume that the sender of each frame knows whether the receiver is on the same LAN. When sending a frame to another LAN, the source machine sets the destination address to 1 as a flag. In addition, it also adds the actual path of the frame to the frame header.
The Source Route Selection Bridge only cares about the frames whose destination address is 1. When such frames are seen, It scans the routes in the frame header to find the LAN number sent to this frame. If the LAN number sent to this frame is followed by the net bridge number, the frame is forwarded to the LAN next to the route table. If the LAN number is not followed by the net bridge, the frame is not forwarded. This algorithm has three possible implementations: software, hardware, and hybrid. The prices and performance of these three implementations are different. First, there is no interface hardware overhead, but the CPU needs to process all the incoming frames quickly. The last implementation requires a special large-scale chip, which shares much of the work of the Bridge. Therefore, the bridge can adopt a slow CPU or connect more LAN.
The premise of source route selection is that each machine on the internet knows the optimal path of all other machines. How to obtain these routes is an important part of the source route selection algorithm. The basic idea of obtaining a routing algorithm is: if you do not know the location of the destination address, the source machine will publish a broadcast frame and ask where it is. Each bridge forwards the lookup frame (discoveryframe) so that the frame can reach every LAN on the Internet. When the reply is returned, the Network Bridge records their own identifiers in the Reply frame. Therefore, the sender of the broadcast frame can obtain the exact route and select the best route from it.
Although this algorithm can find the best route (it finds all the routes), it also faces the frame explosion problem. A similar situation may occur for transparent bridges, but not so serious. The spread is based on the spanning tree, so the total number of frames transmitted is a linear function of the network size, rather than an exponential function for source route selection. Once the host finds a route to a specific destination, it stores it in the high-speed buffer without any further search. Although this method can greatly curb frame explosion, it adds a transactional burden to all hosts, and the entire algorithm is certainly not transparent.
3. Comparison of the two bridges
A transparent bridge is generally used to connect to an Ethernet segment, while a source route is used to select a bridge to connect to a ring CIDR block.
5. Remote Bridge
A bridge is sometimes used to connect two or more remote LANs. For example, if a company is distributed in multiple cities and the company has a local LAN in each city, the ideal situation is that all the LANs are connected, the entire system is like a large LAN.
This goal can be achieved by setting a bridge in each LAN and connecting them two locations using a point-to-point connection (such as a telephone line rented by a telephone company. Different protocols can be used for point-to-point connections. One way is to use a standard point-to-point data link protocol to add the complete MAC frame to the payload. If all the LANs are the same, this method works best. The only problem is that frames must be sent to the correct LAN. Another method is to remove the MAC header and tail from the source bridge, add the remaining parts to the Point-to-Point Protocol payload, and generate a new header and tail in the target bridge. Its disadvantage is that the checksum to reach the target host is not the checksum calculated by the source host. Therefore, errors caused by a damage to the bridge memory may not be detected.
Basic working principle of Bridges
The device connected to the data link layer is a bridge. In the network connection, it acts as a data receiving, address filtering, and data forwarding device. It is used to exchange data between multiple network systems.
Basic Features of Bridges
1. The bridge implements LAN interconnection on the data link layer;
2. The bridge can interconnect two networks with different data link layer protocols, different transmission media, and different transmission rates;
3. The bridge implements communication between interconnected networks by receiving, storing, filtering and forwarding addresses;
4. The network to be connected by the bridge should adopt the same protocol above the data link layer;
5. The bridge can separate the broadcast traffic between two networks to improve the performance and security of the interconnected network.
Working principle of the two bridges
Author: Anonymous, source: IT expert network forum, responsible editor: Bai Zhifei,
A bridge is like a smart repeater. A repeater receives signals from a network cable, scales them up, and sends them to the next cable. They do not care about the content of the messages they forward. In comparison, bridges are more sensitive to Information uploaded from checkpoints.
1. Transparent Bridge
The first type of bridge 802 is transparent bridge (transparentbridge) or spanning tree bridge (spanningtreebridge ). The primary concern for people who support this design is that it is completely transparent. According to their point of view, after purchasing the IEEE Standard bridge, multiple LAN units only need to insert the connection plug into the bridge. You do not need to change the hardware and software, do not need to set the address switch, do not need to load route tables or parameters. In short, do nothing, just need to insert the cable to complete, the operation of the existing LAN is not affected by any bridge. This is incredible. They finally succeeded.
A transparent bridge works in a hybrid manner and receives each frame transmitted over all the LAN connected to it. When a frame arrives, the bridge must decide whether to discard or forward it. If you want to forward data, you must determine the LAN to which the data is sent. You need to query the destination address in a large hash in the bridge. This table lists each possible destination and which output line (LAN) it belongs ). When the bridge is inserted, all the scattered lists are empty. Because the bridge does not know the location of any destination, floodingalgorithm is used ): output each incoming frame with an unknown destination to all the LANs connected to this bridge (except the LAN that sends the frame ). Over time, the bridge will understand the location of each destination. Once you know the destination location, the frames sent to the destination are placed on the appropriate LAN instead of distributed.
The transparent Bridge uses the backwardlearning algorithm ). The bridge works in a hybrid manner, so it can see the frames transmitted on any connected LAN. View the source address to know which machine can be accessed on which LAN, so add the previous item to the hash list.
When the computer and the bridge are powered on, powered off, or migrated, the network topology changes accordingly. To handle the dynamic topology problem, the frame arrival time is specified in each hash entry. This item is updated at the current time whenever the destination has arrived in the table frame. In this way, the arrival time of the last frame of the machine can be known from the time of each item in the table. A process in the bridge regularly scans the hash list to clear all table items that are several minutes earlier than the current time. Therefore, if a computer is removed from the LAN and re-connected to the LAN elsewhere, then it can resume normal operation within a few minutes without manual intervention. This algorithm also means that if the machine has no action within a few minutes, the frames sent to it will have to be distributed until it sends a frame by itself.
The process of selecting the route to the frame depends on the sent LAN (source LAN) and the destination LAN (destination LAN), as shown below:
1. If the source LAN is the same as the target LAN, the frame is discarded.
2. If the source LAN is different from the target LAN, the frame is forwarded.
3. If the target LAN is unknown, it will spread.
To improve reliability, two or more parallel bridges are set between the lan. However, this configuration causes other problems because a loop is generated in the topology, infinite loops may occur. The solution is the following spanning tree algorithm.
The solution to the above-mentioned infinite loop problem is to allow bridges to communicate with each other and overwrite the actual topology with a spanning tree to reach each LAN. Use the Spanning Tree to ensure that there is only one path between any two LANs. Once the bridge has agreed on the Spanning Tree, all the transfers between the LAN will follow this tree. Since there is only a unique path from each source to each destination, there is no loop.
To build a spanning tree, you must first select a bridge as the root of the Spanning Tree. The implementation method is that each bridge broadcasts its serial number (which is set by the manufacturer and globally unique), and selects the bridge with the smallest serial number as the root. Next, construct the Spanning Tree Based on the shortest path from the root to each bridge. If a bridge or LAN fails, the calculation is recalculated.
The bridge uses the BPDU (bridgeprotocoldataunit) to communicate with each other. Before the bridge makes its own configuration decision, each bridge and each port need the following configuration data:
Bridge: bridge ID (unique identifier)
Port: port ID (unique identifier)
Port cost relative to priority (high bandwidth = low cost)
After each bridge is configured, the bridge automatically determines the Spanning Tree Based on the configuration parameters. This process has three phases:
1) Select the root bridge
A bridge with the minimum bridge ID is selected as the root bridge. The bridge ID should be unique, but if the two bridges have the same minimum ID, the bridge with a small MAC address will be selected as the root.
2) Select the root port on all other bridges.
Each bridge except the root bridge needs to select a root port, which should be the most suitable port for communication with the root bridge. Calculate the cost of each port to the root bridge, and use the minimum port as the root port.
3) Select "specified (designated) Bridge" and "specified port" for each LAN"
If there is only one bridge connected to a LAN, it must be the designated Bridge of the LAN. If there is more than one bridge, the minimum cost to the root bridge will be selected as the designated Bridge of the LAN. Specify the port connection to specify the bridge and the corresponding LAN (if there is more than one such port, the low-priority is selected ).
A port must be one of the following:
1) Root Port
2) the specified port of a LAN
3) blocked Port
When a bridge powers up, it assumes that it is the root bridge, sends a cbpdu (configurationbridgeprotocoldataunit) to inform it of the root bridge ID. A bridge receives a cbpdu with a root bridge ID less than the id it knows. It updates its own table. If the frame arrives from the root port (upload), it will send it to all specified ports (download) distribution. When a bridge receives a cbpdu with a root bridge ID greater than its known ID, the information is discarded. If the frame arrives from the specified port, a frame is sent back to inform the real root bridge of its lower ID.
When the network is reconfigured intentionally or due to a line failure, the above process will repeat to generate a new spanning tree.
2. Select a bridge as the source route
The advantage of the transparent bridge is that it is easy to install. You only need to plug in the cable to complete the installation. However, on the other hand, such bridges do not use the best bandwidth because they only use a subset of the topology (Spanning Tree ). The relative importance of these two (or other) factors led to the split within the 802 committee. People who support CSMA/CD and the Token Bus have chosen transparent bridges, while those who prefer a bridge called sourcerouting ).
The core idea of source route selection is to assume that the sender of each frame knows whether the receiver is on the same LAN. When sending a frame to another LAN, the source machine sets the destination address to 1 as a flag. In addition, it also adds the actual path of the frame to the frame header.
The Source Route Selection Bridge only cares about the frames whose destination address is 1. When such frames are seen, It scans the routes in the frame header to find the LAN number sent to this frame. If the LAN number sent to this frame is followed by the net bridge number, the frame is forwarded to the LAN next to the route table. If the LAN number is not followed by the net bridge, the frame is not forwarded. This algorithm has three possible implementations: software, hardware, and hybrid. The prices and performance of these three implementations are different. First, there is no interface hardware overhead, but the CPU needs to process all the incoming frames quickly. The last implementation requires a special large-scale chip, which shares much of the work of the Bridge. Therefore, the bridge can adopt a slow CPU or connect more LAN.
The premise of source route selection is that each machine on the internet knows the optimal path of all other machines. How to obtain these routes is an important part of the source route selection algorithm. The basic idea of obtaining a routing algorithm is: if you do not know the location of the destination address, the source machine will publish a broadcast frame and ask where it is. Each bridge forwards the lookup frame (discoveryframe) so that the frame can reach every LAN on the Internet. When the reply is returned, the Network Bridge records their own identifiers in the Reply frame. Therefore, the sender of the broadcast frame can obtain the exact route and select the best route from it.
Although this algorithm can find the best route (it finds all the routes), it also faces the frame explosion problem. A similar situation may occur for transparent bridges, but not so serious. The spread is based on the spanning tree, so the total number of frames transmitted is a linear function of the network size, rather than an exponential function for source route selection. Once the host finds a route to a specific destination, it stores it in the high-speed buffer without any further search. Although this method can greatly curb frame explosion, it adds a transactional burden to all hosts, and the entire algorithm is certainly not transparent.
3. Comparison of the two bridges
A transparent bridge is generally used to connect to an Ethernet segment, while a source route is used to select a bridge to connect to a ring CIDR block.