CN103188163A - Load balancing method and device - Google Patents

Abstract

The invention provides a load balancing method and device in order to solve the problem that in the prior art, client side applications can not be well guaranteed in a load balancing process. The method includes that the load balancing device receives a server access request sent by a client side, chooses from a plurality of network exits according to a preset load balancing algorithm, and transmits the server access request according to the chosen exit, wherein the plurality of network exits are respectively connected with a plurality of networks through a plurality of physical links, home networks of a server are included in the plurality of networks, other non-home networks and the home networks are linked in a communication mode, and due to the fact that the load balancing device chooses from the plurality of network exits according to the load balancing algorithm and transmits the server access request sent by the client side, the client side applications are well guaranteed in the load balancing process.

Description

A load balancing method and device

technical field

The invention relates to the field of data services, in particular to a load balancing method and device.

Background technique

Due to the rapid growth of business volume, access volume and data flow of each core part of the existing network, its processing capacity and calculation intensity also increase correspondingly, making it impossible for a single server device to bear it. In this case, if you throw away the existing equipment and do a lot of hardware upgrades, it will cause a waste of existing resources, and if you face the next increase in business volume, this will lead to another high cost of hardware upgrades. Cost input, even the equipment with excellent performance cannot meet the needs of the current business growth.

To this end, a new technology-load balancing (also known as load balancing, the English name is LoadBalance) has emerged, which means to balance and distribute the load (work tasks) to multiple operating units for execution, such as Web Servers, FTP servers, enterprise key application servers and other mission-critical servers, etc., so as to complete work tasks together. The existing load balancing technology applications are all forward applications, such as realizing server load balancing according to request content and load conditions, or realizing automatic route selection on the server side according to network conditions. The core is to ensure the normal business of the server, but there is no guarantee The scheme applied by the client.

It can be seen that the following problems exist in the prior art: the client application cannot be well guaranteed during the load balancing process.

Contents of the invention

The purpose of the present invention is to provide a load balancing method and device for the problem existing in the prior art that the client application cannot be well guaranteed during the load balancing process, the method comprising:

The load balancing device receives the server access request sent by the client;

The load balancing device selects from its multiple network outlets according to the preset load balancing algorithm, and forwards server access requests through the selected outlets. The multiple network outlets are respectively connected to multiple networks through multiple physical links, and multiple The network includes the home network of the server, and other non-home networks have communication connections with the home network.

Further, the load balancing device selects from multiple network outlets according to the preset load balancing algorithm, and forwards the server access request through the selected outlet, specifically:

When the load balancing device judges that the client belongs to the specified client set, it selects from multiple network outlets according to the preset load balancing algorithm, and forwards the server access request through the selected outlet.

Further, it also includes: when the load balancing device judges that the client does not belong to the specified client set, directly forwarding the server access request through the preset network egress.

Further, the load balancing device selects from multiple network outlets according to the preset load balancing algorithm, and forwards the server access request through the selected outlet, specifically:

The access address in this server access request is the first access, and the load balancing device selects from multiple network outlets according to the routing delay and hop count corresponding to each network outlet, as well as the preset load balancing algorithm, and passes the selected The egress forwards server access requests.

Further, the load balancing device selects from multiple network outlets, and forwards the server access request through the selected outlet, specifically:

The access address in this server access request is not the first access, and the load balancing device forwards the server access request according to the previously selected network exit.

The embodiment of the present invention also provides a load balancing device, including:

The receiving module is used to receive the server access request sent by the client;

The forwarding module is used to select from multiple network outlets according to the preset load balancing algorithm, and forward the server access request through the selected outlet, wherein the multiple network outlets are respectively connected to multiple networks through multiple physical links, and multiple One network includes the home network of the server, and other non-home networks have communication connections with the home network.

Further, the forwarding module is also used to select from multiple network exits according to a preset load balancing algorithm when judging that the client belongs to the specified client set, and forward the server access request through the selected exit.

Further, the forwarding module is also used to directly forward the server access request through the preset network exit when judging that the client does not belong to the specified client set.

Further, the forwarding module is also used to select the access address in the server access request as the first access, according to the routing delay and hop count corresponding to each network exit, and the preset load balancing algorithm, from multiple network exits , and forward the server access request through the selected exit.

Further, the forwarding module is also used to forward the server access request according to the previously selected network exit when the access address in the server access request is not the first access.

Since the load balancing device selects from multiple network outlets according to the load balancing algorithm, and forwards the server access request sent by the client, the client application is well guaranteed during the load balancing process.

Description of drawings

Fig. 1 represents the network topology diagram of the load balancing system provided by the present invention;

Fig. 2 represents the method flowchart provided by the present invention;

Fig. 3 shows the structure diagram of the device provided by the present invention.

Detailed ways

The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings. The embodiments of the present invention provide a method and device for load balancing to solve the existing problems in the prior art that cannot be performed well during the load balancing process. The problem of safeguarding client applications.

The load balancing system of the present embodiment comprises, as shown in Figure 1, the client 11, the load balancing device 12, the server 3, there is a communication link between the client 11 and the load balancing device 12, and the load balancing device 12 has a plurality of networks and connect multiple networks 14 via multiple physical links 13 through multiple network outlets. The multiple networks include the home network of the server, and there are communication connections between other non-home networks and the home network. The method includes that the load balancing device receives the server access request sent by the client, and the load balancing device selects from a plurality of network outlets according to a preset load balancing algorithm, and forwards the server access request through the selected outlet, wherein the plurality of The network egresses are respectively connected to multiple networks via multiple physical links, and the multiple networks include the home network of the server, and communication connections between other non-home networks and the home network.

The flow chart of the method provided by the embodiment of the present invention is shown in Figure 2 below, including the following steps:

Step 101, the client sends a server access request to the load balancing device.

Step 102, the load balancing device checks the source address to determine whether it is a VIP client according to the server access request, if so, execute step 106, otherwise execute step 103.

In this step, the source IP address of the client is extracted from the server access request. If the source IP address is in the specified IP address range for VIP customers, the client can be considered as a VIP customer, that is, the load balancing device can judge the client The end belongs to the specified client set, so that in the subsequent steps, according to the preset load balancing algorithm, it can select from multiple network exits, and forward the server access request through the selected exit. For example, the load balancing device has 3 network exits, which are network exit A, network exit B, and network exit C, and the client is a VIP customer, then in the subsequent steps, network exit B can be selected according to the preset load balancing algorithm (of course, It may be network exit A or network exit C), and forward the server access request through network exit B. The specific implementation will be described in detail in the subsequent steps.

Step 103, the load balancing device forwards the server access request to the general egress.

In this step, if the source IP address is not in the specified IP address segment for VIP customers, it can be considered that the client is not a VIP customer, that is, when the load balancing device judges that the client does not belong to the specified The network egress forwards server access requests. For example, when the client is a non-VIP client, the load balancing device does not perform balancing processing, but directly forwards the server access request to the network egress A. Network egress A is pre-configured as the network egress of non-VIP clients.

Step 104 , the load balancing device judges whether the server can be accessed normally, and if so, executes step 110 , otherwise executes step 105 .

Step 105, the load balancing device returns an error message to the client, and the process ends.

Step 106, the load balancing device invokes the load balancing group with multiple egresses, and checks the IP address of the target server.

When the load balancing device judges that the client belongs to the specified client set, it calls the load balancing group with multiple egresses to prepare for balancing processing. The load balancing group is preset with a load balancing algorithm. Through the preset load balancing algorithm, the following steps , according to the preset load balancing algorithm, select from multiple network exits, select the optimal path, and forward the server access request through the network exit corresponding to the selected optimal path. In this step, the IP address of the target server is also extracted from the server access request.

Step 107, the load balancing device judges whether the address of the target server has an access record, if yes, execute step 109, otherwise execute step 108.

The load balancing device will record the address of the target server that has been visited, and compare the IP address of the target server with the record to check whether the address of the target server has been visited this time. If it has been accessed, it indicates that the access address in this server access request is not the first access,

Step 108, the load balancing device selects the optimal path according to the routing delay and hop count corresponding to each network egress, and the preset load balancing algorithm, and then executes step 110.

This step shows that the access address in this server access request is the first access, and the load balancing group will use the routing delay and hop count corresponding to the network egress as parameters, select from multiple network egresses according to the load balancing algorithm, and select the most For example, the route delay corresponding to network exit A is 50ms and the number of hops is 5; the route delay corresponding to network exit B is 30ms and the number of hops is 3; the route delay corresponding to network exit C is 60ms and the number of hops is 6. The optimal path selected by the load balancing algorithm is the path corresponding to the network egress B, and the load balancing device forwards the server access request through the network egress B.

Step 109 , the load balancing device checks whether the previously selected optimal path is available, if so, execute step 110 , otherwise execute step 108 .

This step indicates that the access address in this server access request is not the first access, and the load balancing device forwards the server access request according to the previously selected network exit B. At the same time, the export health check is performed to check whether the path corresponding to the network exit B is available, so as to prevent the server access request from being unable to be sent to the corresponding server due to the unavailable path.

Step 110, the server responds to the request and returns the request data flow to end.

The solution of this embodiment is a reverse application of load balancing, that is, a load balancing device connected to multiple network outlets is set up on the client side, so that the client, especially the VIP client, can automatically select a route when accessing the network. It makes up for the defect that it is difficult to guarantee the important Internet services of VIP customers, and expands the application scope of load balancing technology on the client side. This method not only saves the equipment cost of VIP customers and the maintenance cost of multiple links, but also ensures normal service even when a certain egress is abnormal.

The embodiment of the present invention also provides a load balancing device, as shown in FIG. 3 , including:

A receiving module 201, configured to receive a server access request sent by a client;

The forwarding module 202 is configured to select from multiple network outlets according to a preset load balancing algorithm, and forward the server access request through the selected outlet, wherein the multiple network outlets are respectively connected to multiple networks via multiple physical links, The multiple networks include the home network of the server, and other non-home networks have communication connections with the home network.

Further, the forwarding module 202 is also configured to select from multiple network egresses according to a preset load balancing algorithm when judging that the client belongs to the specified client set, and forward the server access request through the selected egress.

Further, the forwarding module 202 is further configured to directly forward the server access request through a preset network egress when judging that the client does not belong to the specified client set.

Further, the forwarding module 202 is also used for the access address in this server access request to be the first access, according to the routing delay and hop count corresponding to each network exit, and the preset load balancing algorithm, from multiple network exits Select, and forward server access requests through the selected exit.

Further, the forwarding module 202 is also used for forwarding the server access request according to the previously selected network exit if the access address in the server access request is not the first visit.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: it still Modifications or equivalent replacements can be made to the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims (10)

1. A load balancing method, characterized in that, comprising: The load balancing device receives the server access request sent by the client; The load balancing device selects from its multiple network outlets according to the preset load balancing algorithm, and forwards server access requests through the selected outlets. The multiple network outlets are respectively connected to multiple networks through multiple physical links, and multiple The network includes the home network of the server, and other non-home networks have communication connections with the home network. 2. The method according to claim 1, wherein the load balancing device selects from a plurality of network outlets according to a preset load balancing algorithm, and forwards the server access request through the selected outlet as follows: When the load balancing device judges that the client belongs to the specified client set, it selects from multiple network outlets according to the preset load balancing algorithm, and forwards the server access request through the selected outlet. 3. The method according to claim 2, further comprising: when the load balancing device judges that the client does not belong to the specified client set, directly forwarding the server access request through a preset network egress. 4. The method according to claim 1, wherein the load balancing device selects from a plurality of network outlets according to a preset load balancing algorithm, and forwards the server access request through the selected outlet as follows: The access address in this server access request is the first access, and the load balancing device selects from multiple network outlets according to the routing delay and hop count corresponding to each network outlet, as well as the preset load balancing algorithm, and passes the selected The egress forwards server access requests. 5. The method according to claim 4, wherein the load balancing device selects from a plurality of network outlets, and forwards the server access request through the selected outlet as follows: The access address in this server access request is not the first access, and the load balancing device forwards the server access request according to the previously selected network exit. 6. A load balancing device, characterized in that, comprising: The receiving module is used to receive the server access request sent by the client; The forwarding module is used to select from multiple network outlets according to the preset load balancing algorithm, and forward the server access request through the selected outlet, wherein the multiple network outlets are respectively connected to multiple networks through multiple physical links, and multiple One network includes the home network of the server, and other non-home networks have communication connections with the home network. 7. The device according to claim 6, wherein the forwarding module is further used for judging that the client belongs to the specified client set, according to the preset load balancing algorithm, select from multiple network outlets, and pass The selected egress forwards server access requests. 8. The device according to claim 7, wherein the forwarding module is further configured to directly forward the server access request through a preset network exit when judging that the client does not belong to the specified client set. 9. The device according to claim 6, characterized in that the forwarding module is also used for the access address in this server access request to be the first access, according to the routing delay and hop count corresponding to each network exit, and the preset The load balancing algorithm selects from multiple network outlets and forwards server access requests through the selected outlets. 10. The device according to claim 9, wherein the forwarding module is further used to forward the server access request according to the previously selected network exit when the access address in the server access request is not the first access.

Images