Simply put, the third-layer switching technology is: second-layer switching technology + third-layer forwarding technology. It solves the situation where subnets in network segments must rely on routers for management after the division of network segments in LANs, and solves the network bottleneck problem caused by the low speed and complexity of traditional routers.
What is a three-layer exchange
Three-layer switching (also known as multi-layer switching technology, or IP switching technology) is proposed relative to the traditional switching concept. As we all know, traditional switching technology operates in the second layer of the OSI network standard model, the data link layer, while the third layer switching technology implements high-speed forwarding of data packets in the third layer of the network model. Simply put, the third-layer switching technology is: second-layer switching technology + third-layer forwarding technology.
The emergence of layer three switching technology has solved the situation where subnets in network segments must rely on routers for management after the division of network segments in LANs, and solved the network bottleneck problem caused by the low speed and complexity of traditional routers.
Three-layer exchange principle
A device with layer three switching function is a layer two switch with layer three routing function, but it is an organic combination of the two, not simply superimposing the hardware and software of the router device on the LAN switch.
The principle is: suppose that two sites A and B using IP protocol communicate through a third-layer switch. When sending site A starts sending, compares its IP address with the IP address of Station B to determine whether Station B is in the same subnet as it. If the destination station B and the sending station A are in the same subnet, the second layer forwarding will be performed. If the two sites are not in the same subnet, if the sending station A wants to communicate with the destination station B, the sending station A must send an ARP (address resolution) packet to the "default gateway", and the IP address of the "default gateway" is actually the layer three switching module of the layer three switch. When the sending station A broadcasts an ARP request to the IP address of the "default gateway", if the layer three switching module already knows the MAC address of Station B during the previous communication process, it replys to the sending station A of the MAC address of B. Otherwise, the third-layer switching module broadcasts an ARP request to the B station according to the routing information. After the B station obtains this ARP request, the B station replys to the third-layer switching module. The third-layer switching module saves this address and replys to the sending station A, and at the same time sends the MAC address of the B station to the MAC address table of the second-layer switching engine. From then on, all the data packets sent by A to B are handed over to Layer 2 for exchange processing, and the information can be exchanged at high speed. Since Layer 3 processing is only required during the routing process, most of the data is forwarded through Layer 2 switching, the Layer 3 switch is very fast, close to that of Layer 2 switch, and is much lower than the price of the same router.
Types of Layer 3 switches
Layer three switches can be divided into two categories: pure hardware and pure software according to their different data processing.
(1) The three-layer technology of pure hardware is relatively complex and has high cost, but it has fast speed, good performance and strong load capacity. The principle is to use ASIC chips and use hardware to search and refresh routing tables.
When the data is received by the port interface chip, first look up the corresponding destination MAC address in the layer 2 switching chip. If found, layer 2 forwarding will be performed, otherwise the data will be sent to the layer 3 engine. In the third-layer engine, the ASIC chip searches for the corresponding routing table information, compares it with the destination IP address of the data, and then sends the ARP packet to the destination host, obtains the MAC address of the host, sends the MAC address to the second-layer chip, and forwards the packet by the second-layer chip.
(2) The software-based three-layer switch technology is simpler, but the speed is slower, and it is not suitable as the backbone. The principle is to use the CPU to search for routing tables in software.
Principle of software layer three switch
When the data is received by the port interface chip, first look up the corresponding destination MAC address in the layer 2 switching chip. If found, layer 2 forwarding will be performed otherwise the data will be sent to the CPU. The CPU searches for the corresponding routing table information, compares it with the destination IP address of the data, and then sends the ARP packet to the destination host to obtain the MAC address of the host, sends the MAC address to the second-layer chip, and forwards the packet by the second-layer chip. Because low-cost CPUs are slower to process, this kind of layer three switch is slower to process.
Market product selection
In recent years, broadband IP network construction has become a hot topic. Here, take the third-layer switch products suitable for positioning at the access layer or small and medium-sized aggregation layer as an example, and introduce some specific technologies for layer three switches. The mainstream Layer 3 switches on the market mainly include Cisco's Catalyst 2948G-L3, Extreme's Summit24, and AlliedTesyn's Rapier24, etc. These three-layer switch products have their own characteristics and cover most of the application characteristics of layer 3 switches. Of course, when choosing a third-layer switch, users can judge and select the above products or other manufacturers according to their needs, such as Nortel Networks' Passport/Acceler series, the original Cabletron's SSR series (after Cabletron scored 4, most SSR layer three switches have been incorporated into Riverstone), Avaya's Cajun M series, 3Com's Superstack3 4005 series, etc. In addition, domestic network manufacturers Shenzhou Digital Network, TCL Network, Shanghai Radio and Television Yingxin, Unigroup Network, and Shouxin have launched three-layer switch products. The following is an introduction to three of the products, so that you can have a more comprehensive understanding of the three-layer switch and choose the right model for your situation.
Cisco Catalyst 2948G-L3 switches provide a complete solution in combination with industry standard IOS. They fully support IOS access control list ACL on version 12.0 (10) or above. In conjunction with the core Catalyst 6000, they can complete the construction of end-to-end comprehensive broadband metropolitan area networks (Catalyst 6000 uses MSFC module to complete its multi-layer switching services, and has stopped using RSM routing switching modules, and IOS version 6.1 or above fully support ACLs).
Extreme's three-layer switching product solution can provide unique Ethernet bandwidth allocation capabilities, with cutting units of 500kbps or 200kbps. Service providers can charge based on bandwidth usage, enabling fixed delay transmission of audio and video.
The PPPoE features provided by AlliedTesyn's Rapier24 Layer 3 switch enrich and improve user authentication and billing methods. They can be suitable for a variety of access networks, flexible applications, and easy to achieve business choices. At the same time, they also protect the existing investments of users. They can also cooperate with NAT (network address conversion) and DHCP Server and other functions, which are optimistic about many service providers.
In short, it only took a few years to develop the concept of the three-layer switch from the introduction of the concept to the popularization of today's application, but its extended functions have been continuously enriched in combination with practical applications. With the development of ASIC hardware chip technology and the promotion of practical applications, the technology and products of three-layer exchange will also be further developed.
LAN switch technology and application analysis
At present, many companies are competing to establish corporate intranets, connect to the Internet, and implement e-commerce. To establish such a virtual business platform, we need to comprehensively use today's most advanced computer technology, network technology and communication technology. With the development of this trend, switches (Switch), as the "cornerstone" for building network platforms, have also gained new "definition" and vitality.
Classification and functions of switches
In a broad sense, there are two types of switches: wide area network switch and LAN switch. WAN switches are mainly used in the telecommunications field and provide a basic platform for communications. Local area network switches are used in local area networks and are used to connect terminal devices, such as PCs and network printers. From the transmission medium and transmission speed, it can be divided into Ethernet switches, fast Ethernet switches, Gigabit Ethernet switches, FDDI switches, ATM switches and token ring switches, etc. In terms of scale applications, it can be divided into enterprise-level switches, department-level switches and workgroup switches. The scales divided by each manufacturer are not completely consistent. Generally speaking, enterprise-level switches are rack-type, department-level switches can be rack-type (small slots) or fixed configuration, while workgroup-level switches are fixed configuration (simple functions). On the other hand, from the perspective of application scale, when as a backbone switch, switches that support large enterprises with more than 500 information points are enterprise-level switches, switches that support medium-sized enterprises with less than 300 information points are department-level switches, and switches that support less than 100 information points are working group-level switches. If not specified below, the switches mentioned refer to local area network switches.
As we all know, the switch works on the second layer of the OSI reference model - the data link layer. The main functions include physical addressing, network topology, error verification, frame sequence and flow control. Physical addressing (corresponding to network addressing) defines the addressing method of the device at the data link layer; the network topology includes descriptions of the data link layer, which defines the physical connection method of the device, such as star topology or bus topology, etc.; error verification alerts to the upper layer protocol where transmission errors occur; data frame sequences are reorganized and frames other than sequences are transmitted; flow control can delay the data transmission ability so that the receiving device will not collapse because it receives information flows exceeding its processing capacity at a certain moment. Currently, the switch also has some new functions, such as support for VLANs, support for link aggregation, and some even have the function of a firewall. This is the function of the third layer switch. The so-called third-layer switch adds routing function when partitioning VLANs based on protocols.
Analysis of the current status and trend of switch technology
Layer 3 switch is key to adopting intranet, which combines the advantages of both Layer 2 switches and Layer 3 routers into a flexible solution that provides line-speed performance at all levels. This integrated structure also introduces policy management attributes, which not only relate to the second and third layers, but also provides traffic priority processing, security, and a variety of other flexible functions, such as link aggregation, dynamic deployment of VLANs and intranets. The third layer switch is divided into three parts: interface layer, switching layer and routing layer.
The interface layer contains all important LAN interfaces: 10/100M Ethernet, Gigabit Ethernet, FDDI and ATM. The switching layer integrates multiple LAN interfaces and is supplemented by policy management, and also provides link aggregation, VLAN and Tagging mechanisms. The routing layer provides the main LAN routing protocols: IP, IPX and AppleTalk, and through policy management, it provides traditional routing or direct-to-layer third-layer forwarding technology. Policy management and administrative management enable network administrators to adjust networks according to the specific needs of the enterprise.
Compared with the third layer, the degree to which the second layer is adopted determines the so-called network control classification. A pure second layer solution is the cheapest solution, but it also provides the least control in terms of molecular networking and broadcast restrictions. The third layer switch can provide dynamic integration support for all levels in the classification. Traditional universal routers can also achieve this goal when used with external switches, but compared to this solution, Layer 3 switches require less configuration, less space, less cabling, cheaper, and provide higher and more reliable performance.
The third-layer switch basically has all the functions of a traditional switch, and the third-layer switch shall prevail. The specific switch technology implementation includes:
1. Programmable ASIC
ASIC is a purpose-specific integrated circuit dedicated to optimizing layer 2 processing and is at the heart of today's networking solutions. It integrates multiple functions on one chip, with the advantages of simplicity of design, high reliability, low power consumption, higher performance and lower cost.
2. Distributed pipeline
With distributed pipelines, multiple distributed forwarding engines can quickly and independently transmit data packets.
In a single pipeline, multiple ASIC chips process multiple frames simultaneously. This concurrency and pipeline can bring forwarding performance to a new level: implementing line-speed performance for on-demand (Unicast), Broadcast and Multicast on all ports.
3. Dynamically scalable memory
For advanced LAN switching products, real performance is based on intelligent memory systems. The third layer switch directly associates a portion of the memory with the forwarding engine. More interface modules are added, including their respective forwarding engines, and the memory is expanded accordingly. And through pipelined ASIC processing, the cache is dynamically constructed, increasing the memory usage rate, and the system can also handle large burst data streams without losing packets.
4. Advanced queue mechanism
Even if the network device has outstanding performance, it will be damaged by the congestion on the network segments it connects to. Traditionally, traffic through a port must be saved in a cache with only one output queue, regardless of its priority, and must be processed in a first-in-first-out manner. When the queue is full, any excess will be discarded. In addition, when the queue becomes longer, the delay also increases. This feature makes it very difficult to run real-time transaction processing and multimedia applications on traditional Ethernet. For this reason, many network device manufacturers have developed new technologies that can provide different levels of service on one Ethernet segment while providing control over delay and jitter. This introduces a mechanism for each port to have a different level of queue.
This queue can better distinguish different traffic levels so that the network can be matched closer to high-performance applications. Packets like multimedia and real-time data streams are placed into high priority queues. Using the weighted fair queuing algorithm, high priority queues can be processed more frequently without ignoring low priority queues. Users of traditional applications do not perceive changes in response time and throughput, while those using emergency applications can receive timely responses.
[1][2] Next page
Article entry: csh Editor in charge: csh
5. Automatic traffic classification
Some data streams are more important than others. Using automatic traffic classification, the Layer 3 switch can instruct the packet pipeline to distinguish user-specified data streams, thereby enabling low-latency, high-priority transmission and avoiding congestion.
6. Intelligent License Control
Layer 3 switches provide multiple security mechanisms and use traffic classifiers, where administrators can limit any identified data flow, including restricting access to servers and excluding useless protocol broadcasts. This is a breakthrough in the field of network technology, namely providing a line-speed firewall.
7. Dynamic traffic supervision
Traffic classification, priority processing and resource retention enable enterprise network and intranet administrators to focus their energy on more important things, namely traditional and next-generation applications. But there is one thing that needs to be done, that is traffic supervision. Traffic supervision is not a policy mechanism because it is actually a protection mechanism. It monitors traffic and network congestion and responds dynamically to these situations to ensure that all network elements (end users and the network itself) are under control and operate optimally.
To prioritize on congested LANs, many Layer 3 switches use the IEEE 802.1p service level. To avoid congestion, high-performance layer 3 switches even use more advanced technologies to dynamically monitor the size of the output queue to find out whether a port will become crowded. By controlling the size and congestion of the queue, the network can maintain the limits required for delay-sensitive data flows.
8. Extensible RMON implementation
Support for RMON has become an indispensable part of proactive and extensive network management. The MIB defined in RFC 1757 contains statistical data for the physical layer and MAC layer. RMON 2 defined in RFC 2021 extends the acquisition of statistical data above the network layer.
9. Vector processing technology
Vector processing technology is used to speed up the processing speed of data frames. The architecture of the third layer switch not only adds the control capability of the third layer on the second layer, but also adds multiple vector controls in multiple directions, thus greatly enhancing the vector processing function. The vector processing of layer 3 switches has many advantages: fast frame processing speed. Thanks to ASIC-based packet classification, forwarding and interpretation technology, the work of frame decoding by software is minimized, and this method can achieve much higher performance compared to pure software design. Highly adaptable functional control. Vector processing works in conjunction with programmable ASICs, enabling support for new future standards with minimal overhead. For example, support for IPv6 is already part of vector logic. Enhanced management features. Multi-directional vector processing also includes built-in network management agents and RMON.
10. Multi-RISC processor
In high reliability switches, a dedicated high-performance RISC processor is absolutely required. In fact, the performance provided by the combination of frame processors (FPs) and vector logic is unparalleled.
A standalone application processor (AP) assists FP. Like FP, AP is also a high-performance RISC processor. AP controls all operations except frame forwarding: high-level bridging and routing, such as spanning tree and OSPF protocols, as well as SNMP and HTTP operations, etc. The benefits of using AP and FP are obvious because management and computing work does not affect data forwarding, resulting in high throughput and low latency.
Through the above technical analysis, it is not difficult to see that high performance, security, ease of use, manageability, stackability, service quality and fault tolerance are the technical characteristics of current switches. With the emergence of a new generation of applications that show the arrival of the multimedia era, such as video conferencing, real-time multicast, Internet telephone, programmable switching and automatic call forwarding, where should exchange technology develop? One thing that is certain is that high bandwidth, security, service quality and intelligence should be the technical directions that the new generation of switches should pursue. However, it is worth mentioning that some manufacturers are now moving towards the direction of distributed network computing for switches.
Web Switch
The development of the Internet is changing rapidly. In order to cope with increasing loads and new application needs, Web switches have emerged to provide management, routing and load balancing transmission for data center devices (including Internet servers, firewalls, cache servers and gateways, etc.). Unlike traditional network devices, traditional network devices focus on completing the exchange of individual frames and data packets at high speed, while web exchange focuses on tracking and processing of web sessions. In addition to the connection and packet routing provided by traditional layer 2/3 switches, web switches can also provide complete strategies lacking in traditional LAN switches and routers, combining local and global server load balancing, access control, quality of service assurance (QoS), and bandwidth management capabilities. Currently, web switches have evolved from pure transport layer (layer 4) devices to intelligence with content-based (layer 7) switching. Redirecting web requests using content or user categories is a feature of the web server. However, the development of Internet transmission and commerce far exceeds the improvement of computer processing capabilities. Offloading content classification to a web switch balances the infrastructure of the entire website
。
Previous page [1][2]
Article entry: csh Editor in charge: csh