How the switch works Switching technology is a switching product with the characteristics of simplicity, low price, high performance and high port density. It embodies the complex switching technology of the bridge technology and operates at the second layer of the OSI reference model. Unlike the bridge, the forwarding delay of the switch is very small, and the operation is close to the performance of a single local area network, far exceeding the forwarding performance between ordinary bridged Internet. Cell exchange ATM technology represents a good medicine for many problems in network and communication. ATM uses a fixed length of 53 bytes for cell exchange. Because the length is fixed, it is easy to implement with hardware. ATM uses a dedicated non-differential connection , Running in parallel, you can establish multiple nodes at the same time through a switch, but it will not affect the communication ability between each node. ATM also allows the communication ability between the source node and the target node. ATM uses a statistical time division circuit for complex It can greatly improve the utilization rate of the channel. The bandwidth of the ATM can reach 25M, 155M, 622M or even several GB of transfer capacity. Types and selection of LAN switches LAN switches can be divided into: If divided according to the application field of the switch, it can be divided into: The local area network computer is the core equipment of the network system. For users, the most important indicators of LAN switches are factors such as port configuration, data, data exchange capability, and packet exchange speed. Therefore, you should pay attention to the following items when choosing a switch " Stackable switches have the advantages of rapid deployment, good value, scalability, and ease of management. They are currently widely used, especially in e-commerce applications. However, evaluating a stackable switch based solely on the price of each user port does not reflect all circumstances. In fact, if the manufacturer calculates the price per port or makes other assertions based on configurations that do not meet the actual conditions, judging a product based only on price factors may mislead users. Internal scalability: To what extent do interconnections between stacked devices limit scalability? Bandwidth expansion: How many ports can be upgraded from 10Mbps Ethernet to 100Mbps Ethernet ports before the switch is overloaded? External scalability: To what extent does the uplink of the switch limit the flow of data to other parts of the network? Take a switch with 24 user ports as an example. Here we intentionally use the term "user port", because in the actual configuration, a certain number of ports have been used for the server or internal and external connections. If these 24 ports all have a throughput of 10Mbps, and if the stackable device is connected to a Gigabit Ethernet uplink, then upgrading the 8 ports to 100Mbps Ethernet will saturate the uplink. The calculation process is as follows: 8 100Mbps ports equal 800Mbps throughput, and the remaining 16 10Mbps ports equal 160Mbps throughput. The total throughput of 960Mbps means that no further Fast Ethernet upgrades can be implemented without blocking the 1Gbps uplink. In addition, many layer 3 stackable switches do not support differentiated services, service types, or multi-protocol label switching, so these stackable switches require external routers or additional layer 3 switches to obtain the required management features. Manageability Many studies have proved that the costs related to operation and management consume more money during the life of the product than the initial purchase cost of the product. Therefore, manageability has become another key factor in assessing overall value. The inherent advantage of stackable switches is that it is easier to manage a single logical entity than to manage multiple devices that must be independently configured and monitored. However, there are still some other factors that need to be studied here, including the quality of service (QoS) used to prioritize data flows, the ability to enforce policies, the ability to manage VLAN transport flows, and ease of management and operability. QoS features focus on saving the required bandwidth and forwarding transport streams to support the needs of different service levels. Generally stackable switches support the IEEE 802.1p and 802.1Q standards that describe priority and VLAN. But when it comes to support for the resource reservation protocol (the general mechanism used to reserve the specified bandwidth for specific data streams), the results are different, and resource reservation is not necessary to ensure that enough bandwidth is available for use when establishing a connection. Short. Support for strategy Policies refer to the rules that control the behavior of the switch. Network administrators use policies to allocate bandwidth, priority, and control network access to application streams. The focus is on meeting the bandwidth management policies required by service level agreements and the way to publish policies to the switch. Because configuring one switch at a time requires a lot of manpower and input errors may occur, the policy must be published to the switch group. When a new strategy is needed, a large number of switches should be able to be quickly modified. Therefore, it is necessary to check whether the stackable switches support directory-enabled network and light directory access protocols, as well as general open policy services (a more feature-rich protocol expected to be implemented ). Users can use VLANs to manage transport streams across multiple switches or stacked devices. Current products all support VLANs defined by 802.1Q tags, and many products also support other VLANs based on switch ports, media access control addresses, Layer 3 protocols or policies. Switching is the direction of current network technology development. Routing technology is an important part of the switching network. The correct selection of routing technology in the switching network will directly affect the overall performance of the network. Therefore, the routing technology is increasingly valued by manufacturers and network designers. 4. Conclusion In summary, each of the three router technologies has its own characteristics and advantages, and users can choose according to their actual needs. It should be emphasized that routing technology is still a very important part of the switching network in the present, and even in the foreseeable future. The correct selection of routing technology will directly affect the overall performance of the network and must be given sufficient attention. LED wall lights attach to
the outside facade of a building, home and wall. They complement any area
bringing light to spaces that need light the most. Their designs are varied so
at to seamlessly blend in to your style and taste. These lights are LED
compatible giving you the ability to have maximum illumination for minimal electric
usage. Led Wall Light,Outside Wall Lights,Outdoor Led Wall Lights,Modern Outdoor Lighting ZHONGSHAN G-LIGHTS LIGHTING CO., LTD. , https://www.glightsled.com
Switching technology allows shared and highly dedicated LAN segments to adjust bandwidth to alleviate bottlenecks in the flow of information between LANs. There are already Ethernet, Fast Ethernet, FDDI and ATM technology exchange products.
Three kinds of switching technologies Port switching The port switching technology first appeared in slot-type hubs. The backplane of this type of hub is usually divided into multiple Ethernet segments, which are not connected by bridges or routers. The networks are not interconnected. After the Ethernet main module is inserted, it is usually allocated to a network segment of a backplane. Port switching is used to distribute and balance the ports of the Ethernet module among multiple network segments on the back bench. According to the degree of support, the port can also be subdivided into:
* Module switching: network segment migration of the entire module * Port group switching: usually the ports on the module are divided into several groups, and each group of ports allows network segment migration.
* Port-level switching: each port can be migrated between different network segments. This switching technology is based on the first layer of OSI, which has the advantages of flexibility and load balancing. If properly configured, it can also be fault-tolerant to a certain extent, but it does not change the characteristics of the shared transmission medium, so it cannot be called It is a real exchange.
Frame exchange Frame exchange is currently the most widely used LAN switching technology. It provides a parallel transmission mechanism by micro-segmenting traditional transmission media to reduce collision domains and obtain high bandwidth. Generally speaking, the products of each company are The implementation techniques are all different in travel, but there are several ways to process network frames:
* True exchange: provide wire-speed processing capability, the switch only reads the first 14 bytes of the network frame, and then transfers the network frame to the corresponding broken port.
* Store and forward: Error detection and control by reading network frames.
The switching speed of the former method is very fast, but it lacks more advanced control of network frames, lacks intelligence and security, and cannot support the switching of ports with different rates. Therefore, various manufacturers regard the latter technology as Focus.
* Ethernet switch * Token ring switch * FDDI switch * ATM switch * Fast Ethernet switch
* Desktop switch * workgroup switch * backbone switch * enterprise switch * segment switch * port switch * network switch
1. The number of exchange ports 2. The type of exchange ports 3. The expansion capability of the system 4. The connection method of the backbone 5. The total exchange capacity of the switch 6. Whether the routing capability is required 7. Whether the hot swap capability is required 8. Whether fault tolerance is required Ability 9. Can it be compatible with existing equipment, smoothly connected 10. Network management capabilities
A more comprehensive way to evaluate stackable switches is to increase the assessment of scalability and manageable performance. When it comes to scalability, many manufacturers list the number of user ports to support their guarantee of high reliability. This is not enough. It is also necessary to consider whether this configuration has an uplink. It is also necessary to determine whether the switch is blocked or Saturated location. There are three points to consider here:
1. Three routing technologies There are currently three routing technologies in the switching network, the first of which is the most conservative method, that is, the method of combining the third layer router and the second layer switch. The second layer switch is strictly limited to the bridge structure and is used for data exchange between different nodes in the same virtual network. At the second layer of the OSI reference model, that is, the data link layer, the virtual LAN function is implemented. The function is left to the router to realize, the router completes the data transmission between the virtual networks and establishes the connection between the LAN and the enterprise backbone network.
The second method uses distributed routing technology. Its characteristic is that it uses a multi-layer switch, combining the bridge of the second layer with the routing of the third layer, and some documents also call the multi-layer switch as the third layer switch. Its own routing function supports virtual LAN, and supports most nodes within the same virtual network or communication between different virtual networks, reducing the number of routers used between work groups and departments. But it still cannot completely get rid of the use of traditional routers, because multi-layer switches can only provide a subset of the protocol, security, traffic management, and WAN connection functions that high-end routers can provide. For example, CISCO's 7000 series routers can handle 12 protocols and support point-to-point, circuit-switched, and cell-switched WAN communications, while Alantec's Powerhub multilayer switches can only handle three protocols: DECnet, IP, and IPX, and there is no WAN interface. Therefore, routers are required as gateways for WANs in multi-layer switch networks and complete more complex routing functions.
The third routing technology in the switching network uses a completely new structure: a combination of routing servers and border switches. We know that traditional routers accomplish both packet forwarding and routing. The routing server-based network is completed by two independent devices to perform the above two functions: the border switch completes the forwarding of information packets, and the routing information is determined by the more expensive routing server. The border switch only accesses the routing server when it cannot find the address of the target node in its address table. At this time, the router responds with a correct address to it, and the switch caches the information for backup. It should be pointed out that at present, the communication protocol between the routing server and the switch is not uniform, and the products of different manufacturers are not compatible.
Second, compare and evaluate the above three routing technologies have their own characteristics, network designers can choose according to the actual situation. In order to make people better understand them, we compare them in the following four aspects.
1. Network scale The size of the network is the decisive factor in choosing which routing technology to use for networking. The method of combining the second layer switch with the traditional router is suitable for a small-scale network, and its characteristic is economical and practical. However, when the backbone network is expanded to a relatively large network, the overhead of the second layer virtual LAN will increase significantly.
With the expansion of the backbone network, the intelligent advantages of the multi-layer switch are brought into full play. Because it only forwards broadcasts to those network segments that belong to a specific subnet, the number of broadcast traffic on the backbone network is reduced. Because the virtual network composed of multi-layer switches has a filtering function and can save the bandwidth of the backbone network and the clock of the end site, the security of the virtual network is better. In addition, compared with the first method, since the switch can be responsible for both switching and routing within the working group and department, it saves the number of traditional routers. The distributed router method and routing server are also more suitable for large-scale distributed networks.
2. Delay The increase in network delay will cause a decrease in network performance. The size of the network delay is generally proportional to the size of the job that the device must process before forwarding traffic. For Layer 2 Ethernet switches, because Layer 2 virtual networks essentially use bridges instead of routers, they are relatively fast. When performing a simple MAC address lookup, a packet (64 bytes) of The delay is less than 100 microseconds. The use of Layer 3 routers increases the search for headers and the execution of certain algorithms, thus greatly increasing the delay of the packet, with a delay time of up to 5 milliseconds.
It can be seen that for a network consisting of a combination of a layer 2 switch and a layer 3 router, the traffic has fairly good performance when passing through the switch; the performance is poor when the traffic flows from one switch to another through the router.
Almost all Layer 2 switches and software can be used together to form nodes into a virtual network (broadcast domain), and thus improve network performance. The traffic between nodes in the same virtual network is exchanged at the MAC layer, and the delay is small. When the nodes between different virtual networks exchange information, the packet transmission must pass through the router, and the network delay is large at this time.
Distributed routing technology allows switches to pass information between Layer 3 protocol subnet ID virtual networks, which can overcome the bottleneck formed by the above routers.
The routing server method uses border switches to make path selections. Occasionally, when the border switch does not know the destination address, it sends an inquiry packet to the routing server, and then there is a delay in pathfinding. Under normal circumstances, the switch can directly look up the address in the cache address table, and then forward the packet directly. In this case, the delay generated is basically the same as the delay of the MAC layer switch.
3. Management Routing information is stored in each router in the network, and each protocol has a corresponding list. The network administrator must configure the routers one by one, including the setting of filters, adding and modifying routing tables, etc. In addition, the man-machine interface is a text-based interface, so when the enterprise network is expanded to a larger scale, the management and configuration of the router is quite time-consuming.
The disadvantage of distributed routing technology is that its management overhead and the number of routing and switching tables increase exponentially. In order to overcome this shortcoming, manufacturers intend to adopt the following measures: formulate traffic control strategies in the central console and automatically propagate through the network, thereby avoiding configuring each device one by one and adding a graphical human-machine interface.
The characteristics of the routing server are easy to manage. Only the configuration of a routing server can provide high-quality services and virtual network management. For example, Cabletron's Securefast management program can allow network management personnel to use the screen to assign access rights to different groups of users, and notify all switches by executing the software.
Another advantage of the routing server method is that it allows network administrators to transparently formulate traffic management strategies without having to care about the types of end-station users. For example, the network administrator can assign the node on the Ethernet switch and the server on the ATM to the same virtual local area network without entering the MAC or IP address of the Ethernet node or the VPI / VCI of the ATM node. 4. Price Price is another major factor that people consider when networking. The following authors give a comparison of the average price of each port of the three routing products of 50, 250, and 500 ports produced by several companies (see Tables 1 and 2). The price of each port here is the result of dividing the total price of network equipment by the number of ports. The network equipment includes Ethernet switches, ATM switches, routing servers, and Layer 3 routers.
The routing server networking method is only given by Newbridge, and the average price of each port of its 50, 250, and 500 port devices is 1920, 1520, and 1435 US dollars, respectively.
From the data given above, it can be seen that the solution based on the second-tier switching and router networking is the cheapest, the distributed routing technology networking has the highest price, and the routing server networking price is moderate. We can also find that, when using layer 2 switches and layer 3 routers for networking, with the expansion of the network scale, the average price of each port is getting smaller and smaller, and the routing server networking is similar. However, the average router price in the distributed router networking mode is not greatly affected by the network size.
Third, the connection with the ATM backbone As the routing server and the distributed routing of the switching LAN are connected to the ATM backbone, there is currently no unified standard, so the connection methods provided by the companies are also different.
The common method is to send all the virtual network traffic of the Ethernet or Token Ring LAN switch to the router equipped with the ATM interface card, but the disadvantage of this method is that the router will become the bottleneck of the entire network and affect the overall performance of the network .
The better way is that Ethernet LAN switches are equipped with their own ATM interfaces, allowing LAN switches and ATM switches to establish a connection directly without having to go through a router. This is an obvious improvement. However, the data transmission between different virtual networks still needs to pass through the router, and the bottleneck still exists.
At present, there are two standards for traditional traffic transmission on ATM: one is the LAN emulation developed by the ATM Forum, and the other is the traditional IP standard on ATM (IPOverATM) developed by the International Computer Interconnection Network Engineering Task Force IETF. LAN emulation runs on the MAC layer of media access control. Its biggest advantage is that it can ensure that the traffic of Ethernet and Token Ring runs normally on the ATM network without any changes to the application program and the human-machine interface. The IPOverATM standard has the same purpose as LAN emulation. Unlike LAN emulation, it only allows ATM traffic to run on IP networks. However, they have not completely solved the transmission of traffic between different virtual networks, and it is still necessary to set up routers between different virtual networks: the router assembles the cells into packets, completes the routing, and then sends the packets before sending Restored into cells, this is significantly less efficient. In order to eliminate the bottleneck formed by routers, the ATM Forum has developed a multi-protocol transmission standard (MPOA) on ATM, which aims to solve the transmission of multiple protocols on ATM, including IP, IPX / SPX, and Appletalk. The routing traffic between different virtual networks of MPOA is based on the traffic information of the network layer (such as IP subnet address) in order to avoid the use of external routers.