Application of load balancing technology in the cluster of mail system

Mail system is a set of separate systems, to have their own servers, in the mail system can be set up multiple domains, each domain can be set up multiple users, such as I bought a anymacro or express mail system, you can do a number of settings. Its data is placed on its own server.

In the concept, TCP/IP e-mail system is divided into user interface and message transmission, but the transmission of messages is not independent, forming a similar to the motis in the concept of MTA. The reason for this is that TCP/IP insists on end-to-end thinking throughout, and its e-mail system uses End-to-end transmission without any storage and forwarding problems, so there is no need to introduce an organization that is responsible for storing and forwarding like the MTA.

In the End-to-end way, although the initial host to participate in the whole process of message transmission, but because of the simplicity of the TCP/IP downlevel protocol, its efficiency is higher than the storage and forwarding. As with Motis, the primary problem for TCP/IP e-mail systems is the question of how to send e-mail when consignee hosts (called a remote machine) are temporarily inaccessible (such as no boot, failure, inability to establish a network connection, etc.). In the terminology of TCP/IP, this is called deferred transfer (de1ayed IBuySpy).

The advantage of an e-mail system is that even if a remote machine is inaccessible, the sender can send messages. For this purpose, TCP/IP adopts spooling buffering technology, which distinguishes the transmission of the user from the actual messages, which is actually the prototype of UA and MTA division. When a user sends a message, he first uses the user interface to generate the message and then passes it to the spooling area, which corresponds to a message from UA to Ms. Then the entire send process users do not care, waiting for the report on the delivery of the results can be. The client who is responsible for sending the message (equivalent to the sending portion of the MTA) is a background process that discovers that a send buffer has a message pending, takes it out immediately, maps the host name to an IP address, and then requests a TCP connection with the other's server, which is equivalent to the receiver portion of the MTA If the connection succeeds, the message is sent, and the other server stores the received message in the receiving mail mailbox (the equivalent of MS), and the customer deletes the corresponding message from the send buffer when it is sent. If the connection is unsuccessful, the customer notes the sending time. The client process periodically checks the send buffer, and whenever it finds an unsent message or a new message comes from the user, the customer immediately proceeds to send it. When a message is found for a long time (if some systems set the time to 3 days), the customer returns it to the sender.

How the e-mail system works

e-mail systems operate in a fundamentally different way from other Web applications. In most other network applications, network protocols are directly responsible for sending data to destinations. In an e-mail system, the sender does not wait to send a job to complete, but rather simply sends the content to be sent out.

For example: File Transfer Protocol (FTP) is like a phone call, real-time access to both sides of the conversation, if one side is temporarily not answered, then the call will fail. The e-mail system is different, the sender will send the content through their own electronic post office to send the letter to the receiver of the electronic post Office. If the receiver's electronic post office is temporarily busy, the sender's electronic post Office will hold the letter until it can be sent. When the receiver is not connected to the Internet, the receiver's electronic post office registers the letter until the receiving party takes it. It can be said that the e-mail system is like the traditional post office functions on the Internet, and is more quickly and easily implemented.

Now as an ordinary user wants to connect to the Internet, you must find an Internet service provider (ISP) to provide connection services. China Telecom is the biggest ISP. While the ISP provides the connection, it also provides an electronic post office and assigns an e-mail to the user. And there are many free e-mail services available on the Internet. It is worth mentioning that emails always have a narrative of space size, what does that mean? This is the space that the electronic post Office gives each user the temporary storage letter. Of course, the bigger the better.

At present, whether in the Enterprise network, Park network or in the WAN such as the Internet, the development of business is beyond the past most optimistic estimates, the internet boom surging, new applications, even in accordance with the optimal configuration of the network at that time, will soon feel unbearable. In particular, the core of each network, its data flow and computational strength, making a single device impossible to afford, and how to achieve a reasonable distribution of traffic between multiple network devices that perform the same function, so that it does not appear that a device is too busy, and other devices are not fully functioning, becomes a problem, Therefore, the load balancing mechanism arises.

Load balancing is built on existing network structures, providing a cheap and efficient way to extend server bandwidth and increase throughput, enhance network data processing capabilities, and improve network flexibility and availability. It mainly completes the following tasks: Solve the network congestion problem, provide the nearest service, realize the geographical independence, provide the users with better access quality, improve the server response speed, improve the utilization efficiency of the server and other resources, and avoid the single point failure in the network key parts.

Custom

In fact, load balancing is not a traditional "equilibrium", in general, it is only the possibility of congestion in a place of the load to many places to share. If you renamed it "Load sharing", perhaps better understand some. In layman's terms, the role of load balancing in the network is like taking turns on the duty system, giving the task to everyone to complete so as not to make a person sweat. However, the balance in this sense is generally static, the predetermined "rotation" strategy.

Different from the rotating duty system, dynamic load balancing analyzes the data packet in real time through some tools, grasps the data traffic condition in the network, and assigns the task rationally. Structure is divided into local load balancing and regional load balancing (global load Balancing), the former refers to the local server cluster load balancing, the latter is to be placed in different geographical location, in different networks and server clusters for load balancing.

Each service node in a server cluster runs a separate copy of the required server program, such as Web, FTP, Telnet, or e-mail server programs. For some services, such as those running on a Web server, a copy of the program runs on all the hosts in the cluster, while Network Load balancing allocates the workload among those hosts. For other services, such as e-mail, only one host handles the workload, and for these services, Network Load Balancing allows network traffic to flow to a host and move traffic to other hosts when the host fails.