【 Weak current College 】 second-tier or three-and four-layer switches difference

Layer 2 switching technology is more mature, second level switch is a data-link layer device, you can identify the MAC address in the packet information, according to the MAC address forwarding, and add those MAC addresses and corresponding port records in their own internal address tables. Specific work procedures are as follows:

(1) when you switch from one port to receive a packet, it first reads the header of the source MAC address, so that it knows the source MAC address of the machine is even on which port;

(2) to read the header of the destination MAC address and address tables to locate the appropriate ports;

(3) If a table has a MAC address with this purpose the corresponding ports, packet directly to the port;

(4) if the table is not found in the corresponding port of the packet broadcast to all ports, when the purpose of the machine on the source machine to respond to, the switch can learn end MAC address and which port corresponds, data is sent in the next time you need to broadcast to all ports.

Continuous cycle of this process, for the whole network of MAC address information to the second layer switch is to establish and maintain its own address table.

From the second layer switch works can infer the following three points:

(1) since the switch to most ports at the same time, data exchange, this requires a wide exchange of bus bandwidth, if there is a second layer switch N ports, every port bandwidth is M, switch bus bandwidth exceeding N×M, this switch can achieve wire-speed switching;

(2) learning port connection machine MAC address, write to the address table, address table size (typically two ways: first, one for BEFFERRAM MAC table item value), address table size switch access capacity;

(3) there is a second layer switch is generally contain specialized to handle packets forwarded ASIC (ApplicationspecificIntegratedCircuit) chip, so forward speed can be done very fast. As a result of various manufacturers with ASIC, directly affect product performance.

The above point is the judgment on the merits on layer switch performance technical parameters, please consider equipment selection to the attention of the comparison.

(Ii) routing technology

Router work of the OSI model of a third-tier---network layer, its mode of operation is similar with Layer 2 switching, but the router is working on the third floor, this distinction determines routing and switching in the delivery package use different control information, to implement the functionality of the way different. Works in the router also has a table, the table is marked as if going to a place, the next step should be to go there, if you can find the packet in the routing table next to where to go, the link-layer information plus forwarded out; if you do not know what the next step towards where you will drop the packet, and then returns a message to the source address.

Routing technology essence, but two functions: to decide the best routing and forwarding packets. The routing table is written to a variety of information, the routing algorithm to calculate arrival destination addresses of the best path, and then by relatively simple direct forwarding mechanism to send packets. Accept the data of the next router in accordance with the same working methods continue to forward, and so on, until the packet arrives to the router.

The routing table maintenance, there are two different ways. One is the routing information updates, some or all routing information published out of the router routing information by learning from each other, they mastered the topology of the entire network, the routing protocol is called a distance vector routing protocols, routers will own link state information is broadcast, through mutual learning and mastering the whole network routing information, and then figure out the best path forward, this type of routing protocol called link-state routing protocol.

Because the router needs to be done for path computation, General processor capability directly determine the quality of its performance. Of course this judgment or on low-end router, because high-end router often uses the distributed processing system architecture design.

(C) the third Exchange technology

In recent years in the three-tier technology advocacy, ears can afford cocoons, everywhere in the three-layer technology call, it was said that this is a very new technology, it has been said that the third Exchange, not just a router and a stack of two-layer switches, no new thing, the fact that the case? first below a simple network to look at the three-layer switches work process.

Networking is simple

Use A IP devices------------------------three-layer switches------------------------the use of IP device B
For example, A send to B, known destination IP, A subnet mask is used to obtain network address, judgment whether their purpose IP in the same network segment.

If on the same network segment, but do not know the forwarding data required by the MAC address of A send an ARP request, B returns its MAC address, A packet with the MAC encapsulation and sent to the switch, the switch using Layer 2 switching module, find MAC address table, the packet is forwarded to the appropriate port.

If the destination IP address displayed is not the same network segment, then A and b are to be achieved, in the flow of communication the cache entry has no corresponding MAC address entries, the first normal packet sent to a default gateway, the default gateway is generally in the operating system has been set, the corresponding layer 3 routing modules visible for are not the same subnet as the data was first MAC table drop is the default gateway's MAC address; then a three-layer modules receive this packet, query routing table to determine the arrival routing, B will construct a new frame header, withthe default gateway's MAC address for the source MAC address to the MAC address of host b is the destination MAC address. Recognition through a triggering mechanism to establish host A and B of the MAC address and forward the port to the corresponding gateway
Department, and log into the stream buffer entry table, later A to B's data directly to the layer 2 switching modules to complete. This is commonly referred to as a routing multiple forwarding.

These are three-layer switches work process simple overview, you can see the features of the third Exchange:
By hardware combined with high speed of data transmitted.

This is not a simple two-layer switches, and routers of superposition, three-layer routing module directly superimposed on layer 2 switching for high speed backplane bus, break the traditional router interfaces, rate limiting, rate of up to several dozens Gbit/s. Counting backplane bandwidth, these are the three layers of switch performance of two important parameters.

Simple routing software that simplifies the process routes.

Most of the data forwarded, in addition to the necessary routing by routing software processing, is also the second-tier modules express forwarding and routing software, mostly the result of processing of high-efficiency optimization software, and not simply copying the router software.

Conclusions of the

Second level switches for small local area network. This would not have many words, in a small local area network, broadcast packets has little effect on the second floor of the fast switching function of the switch, more access port and low-price for small network users with a complete solution.

The router has the advantage of the rich interface type, supported by three powerful, powerful routing capability, ideal for large networks to route, it has the advantage of choosing the best route, load sharing, link backup and and other networks for the exchange of routing information, and so on the function of the router.

Three-layer switches of the most important function is to accelerate the large LAN internal rapid forwarding of data to join the routing functionality for this purpose. If you have a large network in accordance with the Department, geography, etc. elements divided into many small local area network, which will cause a large number of Internet visits, simply use a second layer switch cannot implement Internet visits; such as a simple interface with a router, due to the limited and routing forwarding slow, will limit the speed of the network and the network size, with fast forward routing features three layer switch becomes preferred.

In General, data flow, net, require fast forwarded response networks, such as all three tiers switch to do the work, would create an excessive burden on the three-layer switches, response speed, the network of routes to the routers to finish, give full play to the advantages of different devices, is a good networking strategy, of course, the premise is that the customer's pockets are drum, or the next best thing, let three-layer switches also concurrently as Internet interconnection.

The fourth layer switching of a simple definition is: it is a function of its decision to transfer just based on MAC address (2nd tier bridge) or the source/destination IP address (Layer 3 routing), and based on the TCP/UDP (fourth floor) application port number. The fourth layer switching features like virtual IP, pointing to the physical server. It transfers of business subject to a variety of protocols, HTTP, FTP, NFS, Telnet or other protocols. These operations on the basis of the physical server, require complex load balancing algorithm. In the IP world, business types from Terminal TCP or UDP port address to decide, in the fourth level of exchange of application ranges from the side and end IP address, TCP and UDP ports.

In the fourth level of exchange for each for search using servergroup set up a virtual IP address (VIP), each group of servers support a certain application. In the domain name server (DNS) storage of each application server address is VIP, instead of the real server address.

When a user requests for application, a target server group of VIP connection request (for example, a package TCPSYN) to the server switch. Server switch in the group to select the best Server Terminal address in VIP with the actual server IP, and the connection request to the server. In this way, the same interval all packets from the server switches to map the user and the same server.

The fourth layer switching principle

The OSI model of the fourth floor is the transport layer. The transport layer is responsible for end-to-end communication, i.e. the network source and target coordinate communication between systems. In IP protocol stack this is the TCP (Transport Protocol) and UDP (User Datagram Protocol) is a protocol layer.

The fourth layer, TCP and UDP header contains the port number (portnumber), they can only distinguish each packet contains what application protocol (e.g. HTTP, FTP, etc.). Endpoint systems use this information to distinguish between the data in the package, in particular, the port number to a receiving computer system to determine the type of IP packets received and pass it to the appropriate senior software. Port number and IP address of the device combination is commonly called the "Jack (socket)". 1 and 255 port number is retained, they referred to as the "well-known" port, that is, all host TCP/IP protocol stack implementations, these port numbers are the same. In addition to the "well-known" port, standard UNIX service distribution in 256 to 1024 port range, customized application generally assign ports above 1024,. distribution of recent listing can RFc1700 "AssignedNumbers" site. TCP/UDP port number provided additional information can be used for network switch, this is the first 4 layer ExchangeFoundation. R>
"Well-known" port number, for example:
　　
Application protocol port number
FTP 20 (data)
21 (control)
TELNET　　　　23
SMTP　　　　　25
HTTP　　　　　　80
NNTP　　　　　119
NNMP　　　　　16
162（SNMPtraps）
TCP/UDP port number provided additional information can be used by network switches, fourth layer Exchange Foundation.

A fourth layer function switches to be able to play with server connected to the "virtual IP (VIP) front".

Each server and supports single or common application server group is configured with a VIP address. This VIP address is sent out and register the domain name system.

To send a service request, the fourth layer switch TCP by judgment, to identify the beginning of a session. Then it uses complex algorithms to determine the best deal with the request to the server. Once you make this decision, the switch will be a session with a specific IP address, and the real IP address of the server instead of a VIP address on the server.

Each fourth layer switches save a selected server match the source IP address and source port TCP connection table that is associated with. Then the fourth layer switch to this server forwards the connection request. All subsequent packets in the client and server between shadow and forward, until you discover the switches
Session.
　
In using a fourth layer Exchange, access to the real server connection together meet the rules set by the user, such that each server has an equal number of access or according to different server capacity to assign the transport stream.

How to choose appropriate fourth layer switching
　　
A, speed
　
In order to be effective in the enterprise network, the fourth level of Exchange must be provided with a third layer of wire-speed router comparable performance. In other words, the fourth level of Exchange must be in all ports at full speed operation of media, even in multiple Gigabit Ethernet connections.Gigabit Ethernet speeds equivalent to a per-second 488000 packets of maximum speed routing (assuming the worst case, that is, all packages for the network defined minimum size, 64 bytes long).

B, and server capacity balance algorithm
　　
According to desired capacity balance interval size, fourth floor switches will be applied to the server there are many kinds of algorithms, a simple detection loop recent connections, detect loop delay or testing server itself closed-loop feedback. All forecasts, closed-loop feedback provide servers existing business volume the most precise detection.

C, table capacity

It should be noted that for the fourth layer switch switches need to distinguish between large sending and storing capacity. table items Switch in an enterprise network core. Many second/third layer switch tend to send table size and the number of network devices. On the fourth floor switch, the amount must be multiplied by a network of different application protocol and the number of sessions. Thus send its size with the endpoint devices and the increase in the number of application types and growing rapidly. The fourth layer switch designers in the design of its products takes account of the growth of this table. Large table capacity on manufacturing support wire-speed traffic sent the fourth layer of high-performance switch is critical.
　　
D, redundancy
　
The fourth layer switch internal support redundant topologies. In the network adapter with dual-link, fault-tolerant connection may be established from one server to the network link and Server Exchange's fully redundant systems.