CN107493333B - SDN-based web system resource management method and system in cloud computing environment - Google Patents

Abstract

The invention discloses a web system resource management method and system in an SDN-based cloud computing environment. The method includes: S1, constructing a web application system in the SDN-based cloud computing environment; S2, obtaining each The number of instantaneous session connections and instantaneous link traffic of the server; S3. If the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold of the number of instantaneous session connections, perform a resource balancing operation, otherwise go to step S4; S4, obtain Average number of session connections and average link traffic; S5. Compare the average number of session connections of each server with the maximum threshold of the average number of session connections and the aggregate threshold of the average number of session connections; S6. Perform resource balancing operations or resources according to the comparison results Aggregate operation. The beneficial effects of the invention are that the server resources are saved, and the utilization efficiency of the existing resources is effectively improved.

Description

A method and system for web system resource management in a cloud computing environment based on SDN

technical field

The invention relates to the technical field of SDN, in particular to a method and system for managing web system resources in an SDN-based cloud computing environment.

Background technique

Cloud computing is a product of the development and integration of traditional computer and network technologies such as distributed computing, parallel computing, utility computing, network storage, virtualization, load balancing, and hot backup redundancy. This model provides usable and convenient network access, using on-demand computing resources, including networks, servers, storage, application software, services, etc. Due to the advantages of high reliability, easy scalability, low maintenance and use costs, and virtualization, cloud computing technology has developed very rapidly in recent years, and is widely regarded as another technological revolution in the information field.

In the cloud computing environment, enterprises build business systems mainly through virtual web application clusters. A virtual web application cluster uses virtualization technology to create a group of servers running the same web application on the cloud platform at the same time, but it looks like a server to the outside world. However, the virtual web application server cluster currently constructed by the cloud computing platform only completes the cluster construction, and does not realize the automatic dynamic management of cluster resources. Under normal circumstances, users need to independently monitor the operation of the server cluster, obtain the operation parameters of the server cluster, and judge by themselves whether to expand or aggregate the server and network resources. Due to the problem of concurrent access by multiple users in the web application system, it is necessary to dynamically adjust the scale of server clusters and network link traffic in this case. Under the current traditional cloud computing network conditions, the data forwarding plane and the control plane are closely integrated, which cannot Automatic session and traffic load balancing operations, such as monitoring of too many sessions caused by multiple users concurrently, link traffic congestion, and low link utilization, cannot automatically adjust resources according to policies. At present, there are two main ways to adjust the web server cluster used by the cloud platform. For virtual server resources, when the load is too heavy, a resource request is sent to the cloud platform, and the platform starts the virtual machine and adds it to the server cluster; for network link resources, it needs to The platform resets resources such as routers and switches. The above adjustment methods for resources cannot automatically manage resources in real time, which seriously affects the utilization efficiency of the cloud computing platform and also causes a very poor user experience.

SUMMARY OF THE INVENTION

The present invention provides a web system resource management method and system in an SDN-based cloud computing environment, and solves the technical problems of the prior art.

The technical scheme that the present invention solves the above-mentioned technical problems is as follows:

A web system resource management method in an SDN-based cloud computing environment, comprising:

S1. Construct a web application system under an SDN-based cloud computing environment, where the web application system includes: multiple servers and SDN controllers deployed on a cloud computing platform;

S2, obtaining the instantaneous session connection number and instantaneous link traffic of each of the servers at preset time intervals;

S3. Compare the number of instantaneous session connections obtained by each of the servers for two consecutive times with the maximum threshold of the number of instantaneous session connections, respectively, and compare the instantaneous link traffic obtained by each of the servers for two consecutive times with the instantaneous link traffic respectively. The maximum threshold is compared, and if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold for the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the maximum instantaneous link traffic threshold, start the resource balancing module to perform the resource balancing operation through the SDN controller, otherwise go to step S4;

S4, obtain the average session connection number and the average link flow of each of the servers according to the instantaneous session connection number and the instantaneous link flow obtained by each of the servers for a continuous preset number of times;

S5. Compare the average session connection number of each server with the maximum average session connection number threshold and the average session connection number aggregation threshold, respectively, and compare the link average traffic of each server with the link average traffic maximum Compare the threshold with the link average traffic aggregation threshold to obtain the comparison result;

S6. According to the comparison result, start the resource balancing module to perform a resource balancing operation through the SDN controller or start the resource aggregation module to perform a resource aggregation operation through the SDN controller.

The beneficial effects of the present invention are: the technical solution automatically adjusts the scale of the Web application cluster according to the two strategies of the number of session connections and the link traffic. Due to artificial reasons, the server cluster cannot be adjusted when the load increases. On the other hand, compared with the traditional cloud computing platform, the load balance can only be achieved by adding servers, which saves server resources and effectively improves the utilization of existing resources. efficiency.

On the basis of the above technical solutions, the present invention can also be improved as follows.

Preferably, step S6 includes:

When the comparison result is that the average session connection number of any server is greater than the maximum threshold of the average session connection number or the average link traffic of any server is greater than the maximum link average traffic threshold, the resource balancing module is started to pass the SDN controller performs a resource balancing operation;

When the comparison result is that the average session connection number of at least one server is less than the average session connection number aggregation threshold and the link average traffic of the at least one server is less than the link average traffic aggregation threshold, start the resource aggregation The module performs resource aggregation operations through the SDN controller.

Preferably, the method for performing resource balancing operation by the SDN controller includes:

The SDN controller searches the global resource information base to obtain the first server with the smallest comprehensive usage of session and link resources;

When the number of resources to be balanced for the number of session connections is less than the difference between the maximum number of balanced session connections and the average number of session connections of the first server and the number of resources to be balanced for link traffic is less than the maximum link traffic of the first server When the difference between the number of balances and the average traffic of the link is determined, the excess load is balanced to the first server through the flow table method.

The number of resources to be balanced for the number of session connections is the difference between the maximum threshold of the average number of session connections and the average number of session connections of the two or more servers, and the number of resources to be balanced for the link traffic is the The difference between the maximum threshold of the average traffic on the road and the average traffic on the links of the two or more servers.

Preferably, the method for performing resource aggregation operation by the SDN controller includes:

The two servers with the smallest comprehensive usage of session and link resources are obtained by searching from the global resource information base through the SDN controller;

When the aggregated server session connection number is between the average session connection number aggregation threshold and the average session connection maximum threshold and the aggregated server link traffic is between the link average traffic aggregation threshold and the link average traffic When between the maximum thresholds, the two links corresponding to the two servers are aggregated, and the resources occupied by any one of the two links and the server corresponding to the link are released,

The aggregated server session connection number is the sum of the average session connection numbers of the two servers, and the aggregated server link traffic is the sum of the average link traffic of the two servers.

Preferably, the multiple servers include: an application server, a database server and a load balancing server.

An SDN-based web system resource management system in a cloud computing environment, comprising:

a building module for building a web application system in an SDN-based cloud computing environment, where the web application system includes: each server and an SDN controller deployed on the cloud computing platform;

a first obtaining module, configured to obtain the instantaneous session connection number and instantaneous link traffic of each of the servers of the web application system at preset time intervals;

The first comparison module is configured to compare the number of instantaneous session connections obtained by each of the servers for two consecutive times with the maximum threshold of the number of instantaneous session connections, and compare the instantaneous link traffic obtained by each of the servers for two consecutive times with the instantaneous number of connections. Compare with the maximum threshold of link traffic, if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold of the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the instantaneous link traffic. If the maximum threshold of road traffic is reached, the resource balancing module is started to perform resource balancing operation through the SDN controller, otherwise the second comparison module is called;

The second obtaining module is configured to obtain the average session connection number and link average flow of each of the servers according to the instantaneous session connection number and instantaneous link flow obtained by each of the servers for a continuous preset number of times;

The second comparison module is configured to compare the average number of session connections of each of the servers with the maximum threshold of the average number of session connections, and compare the average traffic of the links of each of the servers with the aggregate threshold of the average number of session connections, get the comparison result;

An operation module, configured to start the resource balancing module to perform a resource balancing operation through the SDN controller or start the resource aggregation module to perform a resource aggregation operation through the SDN controller according to the comparison result.

Preferably, the operation module is specifically used for:

When the comparison result is that the average session connection number of any server is greater than the maximum threshold of the average session connection number or the average link traffic of any server is greater than the maximum link average traffic threshold, the resource balancing module is started to pass the SDN controller performs a resource balancing operation;

When the comparison result is that the average session connections of at least one server are all less than the average session connection number aggregation threshold and the link average traffic of the at least one server is all less than the link average traffic aggregation threshold, start the The resource aggregation module performs resource aggregation operations through the SDN controller.

Preferably, the resource balancing module is specifically used for:

The SDN controller searches the global resource information base to obtain the first server with the smallest comprehensive usage of session and link resources;

When the number of resources to be balanced for the number of session connections is less than the difference between the maximum number of balanced session connections and the average number of session connections of the first server and the number of resources to be balanced for link traffic is less than the maximum link traffic of the first server When the difference between the number of balances and the average traffic of the link is determined, the excess load is balanced to the first server through the flow table method.

The number of resources to be balanced for the number of session connections is the difference between the maximum threshold of the average number of session connections and the average number of session connections of the two or more servers, and the number of resources to be balanced for the link traffic is the The difference between the maximum threshold of the average traffic on the road and the average traffic on the links of the two or more servers.

Preferably, the resource aggregation module is specifically used for:

The two servers with the smallest comprehensive usage of session and link resources are obtained by searching from the global resource information base through the SDN controller;

When the aggregated server session connection number is between the average session connection number aggregation threshold and the average session connection maximum threshold and the aggregated server link traffic is between the link average traffic aggregation threshold and the link average traffic When between the maximum thresholds, the two links corresponding to the two servers are aggregated, and the resources occupied by any one of the two links and the server corresponding to the link are released,

The aggregated server session connection number is the sum of the average session connection numbers of the two servers, and the aggregated server link traffic is the sum of the average link traffic of the two servers.

Preferably, the plurality of the servers include: an application server, a database server and a load balancing server.

Description of drawings

1 is a schematic flowchart of a method for managing web system resources in an SDN-based cloud computing environment provided by an embodiment of the present invention;

2 is a schematic flowchart of a method for managing web system resources in an SDN-based cloud computing environment provided by another embodiment of the present invention;

3 is a schematic structural diagram of a web application system provided by another embodiment of the present invention;

FIG. 4 is a schematic structural diagram of a web system resource management system in an SDN-based cloud computing environment provided by another embodiment of the present invention.

Detailed ways

The principles and features of the present invention will be described below with reference to the accompanying drawings. The examples are only used to explain the present invention, but not to limit the scope of the present invention.

As shown in Figure 1, a web system resource management method in an SDN-based cloud computing environment includes:

S101. Build a web application system in an SDN-based cloud computing environment, where the web application system includes: multiple servers and SDN controllers deployed on the cloud computing platform;

S102, obtaining the instantaneous session connection number and instantaneous link traffic of each server at each preset time interval;

S103: Compare the number of instantaneous session connections obtained by each server for two consecutive times with the maximum threshold of the number of instantaneous session connections, respectively, and compare the instantaneous link traffic obtained by each server for two consecutive times with the maximum threshold of instantaneous link traffic respectively , if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold for the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the maximum instantaneous link traffic threshold, the resource balancing module is activated to pass The SDN controller performs a resource balancing operation, otherwise, step S104 is performed;

S104, obtaining the average number of session connections and the average link traffic of each server according to the instantaneous session connection number and instantaneous link flow obtained by each server continuously preset times;

S105: Compare the average number of session connections of each server with the maximum threshold of the average number of session connections and the aggregated threshold of the average number of session connections, respectively, and compare the average traffic of each server with the maximum threshold of average traffic of the link and the maximum threshold of average traffic of the link respectively. The average traffic aggregation threshold is compared to obtain the comparison result;

S106. According to the comparison result, start the resource balancing module to perform a resource balancing operation through the SDN controller or start the resource aggregation module to perform a resource aggregation operation through the SDN controller.

Deploy application servers, database servers, load balancing servers, SDN controllers (or SDN controller clusters) on the cloud computing platform, and SDN controllers (or SDN controller clusters) are deployed in each of the servers, and each of the The servers are all virtual machines provided by the cloud platform, and each of the servers is connected through a virtual switch or a physical switch, and the virtual switch and the physical switch are dedicated SDN switches or SDN hybrid switches.

After the system is constructed, the instantaneous session connection number S _current and instantaneous link traffic B _current of each server and its corresponding link of the web application system are continuously obtained with time t as the sampling interval.

Set the maximum threshold value B _max of instantaneous link traffic and the maximum threshold value S _max of the number of instantaneous session connections. Compare the two consecutive instantaneous link traffic and the number of instantaneous session connections with the corresponding maximum thresholds. If two consecutive instantaneous link traffic B _current >B _max or the number of instantaneous session connections S _current > S _max for two consecutive times, the resource balancing operation is performed, otherwise, the following steps are performed.

Assuming that the N measurement results obtained within the time interval T are S _current1 , _{S current2 ···}

Assuming that N times of traffic measurement results are obtained within the time interval T as B _current1 , B _current2 . . . B _currentn , the average link traffic in the time T is calculated by formula (2) as:

The SDN controller centrally collects and calculates the instantaneous session connection and link traffic load of the application server in the system, and the average session and link traffic load of the server within a certain time interval. Balance and aggregation realize the dynamic management method of web application system resources based on the dual strategy of session and link traffic, which effectively improves the utilization efficiency of cloud platform resources and improves the user application experience.

As shown in Figure 2, a method for managing web system resources in an SDN-based cloud computing environment includes:

S201, constructing a web application system in an SDN-based cloud computing environment, where the web application system includes: multiple servers and SDN controllers deployed on the cloud computing platform;

S202, obtaining the instantaneous session connection number and instantaneous link traffic of each server at each preset time interval;

S203. Compare the number of instantaneous session connections obtained twice in a row by each server with the maximum threshold of the number of instantaneous session connections, respectively, and compare the instantaneous link traffic obtained twice in a row by each server with the maximum threshold of instantaneous link traffic respectively , if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold for the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the maximum instantaneous link traffic threshold, the resource balancing module is activated to pass The SDN controller performs a resource balancing operation, otherwise, step S204 is performed;

S204, obtaining the average number of session connections and the average link flow of each server according to the instantaneous session connection number and instantaneous link flow obtained by each server for a continuous preset number of times;

S205. Compare the average number of session connections of each server with the maximum threshold of the average number of session connections and the aggregated threshold of the average number of session connections, respectively, and compare the average traffic of each server with the maximum threshold of average traffic of the link and the maximum threshold of average traffic of the link, respectively. The average traffic aggregation threshold is compared to obtain the comparison result;

S206. When the comparison result is that the average session connection number of any server is greater than the maximum threshold of the average session connection number or the average link traffic of any server is greater than the maximum link average traffic threshold, start the resource balancing module to perform resource balancing through the SDN controller operate;

S207. When the comparison result is that the average session connection number of at least one server is less than the average session connection number aggregation threshold and the link average traffic of at least one server is less than the link average traffic aggregation threshold, start the resource aggregation module to perform resource aggregation through the SDN controller operate.

Deploy application servers, database servers, load balancing servers, SDN controllers (or SDN controller clusters) on the cloud computing platform, and SDN controllers (or SDN controller clusters) are deployed in each of the servers, and each of the The servers are all virtual machines provided by the cloud platform, and each of the servers is connected through a virtual switch or a physical switch, and the virtual switch and the physical switch are dedicated SDN switches or SDN hybrid switches.

After the system is constructed, the instantaneous session connection number S _current and instantaneous link traffic B _current of each server and its corresponding link of the web application system are continuously obtained with time t as the sampling interval.

Set the maximum threshold value B _max of instantaneous link traffic and the maximum threshold value S _max of the number of instantaneous session connections. Compare the two consecutive instantaneous link traffic and the number of instantaneous session connections with the corresponding maximum thresholds. If two consecutive instantaneous link traffic B _current >B _max or the number of instantaneous session connections S _current > S _max for two consecutive times, the resource balancing operation is performed, otherwise, the following steps are performed.

Assuming that the N measurement results obtained within the time interval T are S _current1 , _{S current2 ···}

Assuming that N times of traffic measurement results are obtained within the time interval T as B _current1 , B _current2 . . . B _currentn , the average link traffic in the time T is calculated by formula (2) as:

Let the maximum threshold of the average number of session connections be S _{avg_max} , the aggregate threshold of the average number of session connections to be S _{avg_min} , the maximum threshold of average link traffic to be B _{avg_max} , and the aggregate threshold of average link traffic to be B _{avg_min} , if S _avg >S _{avg_max} or B _avg >B _{avg_max} , perform resource balancing operation. If S _avg < S _{avg_min} and B _avg < B _{avg_min} , query the system session load and link traffic status information through the SDN controller. If there are more than two groups that meet the above aggregation conditions, then Perform resource aggregation operations.

The resource balancing operation process includes: the SDN controller searches for the server with the smallest comprehensive usage of session and link resources through the global resource information base, and sets the maximum balanced number of session connections of the server as S _{balance_max} , and the maximum balanced number of link traffic as B _{balance_max} , The average number of session connections is S _avg1 , and the average link traffic is B _avg1 . Calculate the number of resources to be balanced: S _balance =S _{avg_max} -S _avg , B _balance =B _{avg_max} -B _avg , if S _balance < S _{balance_max} -S _avg1 and B _balance <B _{balance_max} -B _avg1 , the SDN controller balances the excess load to the server through the flow table method, otherwise calculates the number of remaining balanced loads S _{rest_balance} =S _balance -(S _{balance_max} -S _avg1 ), B _{rest_balance} =B _balance -(B _{balance_max} -B _avg1 ), continue the resource balancing operation until all traffic and sessions are balanced to other servers, if there is no available balance server, start a new application server through the cloud platform and establish it through the controller Related links to achieve resource balancing operation.

The resource aggregation operation process includes: the SDN controller searches for the two servers with the smallest comprehensive usage of session and link resources through the global resource information base. Let the average session connection numbers of the two servers be S _avg1 and S _avg2 , and the average link traffic is B _avg1 and B _avg2 , calculate the number of server sessions after aggregation S _total =S _avg1 +S _avg2 , link traffic after aggregation B _total =B _avg1 +B _avg2 , if S _total >S _{avg_max} or B _total >B _{avg_max} , no operation is performed ; If S _{avg_min} <S _total <S _{avg_max} and B _{avg_min} <B _total <B _{avg_max} , the SDN controller aggregates the two links and releases the resources occupied by one of the links and the server, if S _total < S _{avg_min} And B _total <S _{avg_min} , continue to perform resource aggregation operation until the resource of this type cannot be found.

In the example, an Openstack cloud computing platform consisting of 5 physical servers is configured, and a single physical server is configured with an Intel Xeon E5620 2.4G 4-core 8-thread processor, 64GB RAM, SAS 2.0T RAID5 hard disks, and the virtual machine is created with uniform specifications and allocated. Intel Xeon E5620*1 CPU, 5GB memory, 30GB hard disk, running Ubuntu 14.04LTS operating system. The initial web application system consists of 6 virtual servers, including 1 load balancing server, installed with Haproxy software; 4 web application servers, installed with tomcat middleware and Sun PetStore application, 1 SDN controller server, installed with Opendaylight open source controller . The initial deployment state of the entire Web application system is shown in Figure 3.

In the actual test, set the average session information statistics time period=15min, obtain the instantaneous session connection number and link instantaneous traffic every 1min (ie t=1min) within the interval, and calculate the number of 15 session connections within the time interval after 15min The average value is taken as the average session connection number S _avg , and the average value of 15 link traffic is taken as the average link traffic B _avg . In addition, set the maximum number of instantaneous session connections to 50, the maximum instantaneous link traffic to 70M, the minimum aggregation threshold for average session connections to 8, the minimum aggregation threshold for average link traffic to 25M, the maximum average session connection threshold to 20, and the link The maximum threshold of average traffic is 80M.

First, use the openstack platform to deploy a web application system composed of 6 virtual servers. Monitor the number of session connections and link traffic of each server in the web application system. Here, it is necessary to collect and calculate the number of instantaneous session connections, instantaneous link traffic, average session connections, and average link traffic. The number of connections and the instantaneous traffic of the link are judged to see if the maximum threshold is exceeded twice in a row. If it exceeds, the resource balancing operation will be performed. If it does not exceed, the average load of the server and link within the specified time period will be judged. Calculate the average number of session connections S _avg and the average link traffic B _avg within the specified time interval, and compare S _avg , B _avg with the maximum threshold S _max of the average number of session connections, the maximum threshold of average link traffic B _max , and the average session connection Compare the minimum aggregation threshold S _min and the minimum aggregation threshold B _min of average link traffic. If S _avg > S _max or B _avg > B _max , the resource balancing operation is performed. If there are redundant servers and links that meet the requirements, then Balance the excess sessions and link traffic to other servers. If they do not exist, start a new virtual server through the cloud platform and add it to the load balancing to achieve load balancing; if all servers satisfy S _min < S _avg < S _max and S _min < B _avg < S _max , keep the scale of the existing web application system unchanged, and do not need to perform resource management operations; if there are more than two servers that satisfy S _avg < S _min and B _avg < B _min , perform the resource aggregation operation, change the access path through the SDN controller, aggregate the access sessions and traffic of the two servers to access one server, and remove the application system from the other server.

In the above specific implementation process, the number of access session connections and link traffic is automatically generated. In the initial state, the entire system includes 1 load balancing server, 4 Web application servers, and 1 SDN control server (for deploying SDN controllers or Cluster), an OpenvSwitch virtual switch that supports SDN, in which, the load balancing server and the web application server are connected to the OpenvSwitch switch, and the OpenvSwitch switch is connected to the SDN controller through a dedicated channel. In the experiment, the method of gradually increasing the load is used to generate session connections and link traffic. The minimum aggregated connection session threshold is 8, the minimum aggregated link traffic threshold is 25M, the maximum average session connection threshold is 20, and the maximum link average traffic threshold is 80M. Initially, the number of session connections per server per minute is about 5, and the traffic per link is about 20M. At this time, the session load and link traffic are both less than the minimum aggregation threshold. The aggregation operation is triggered, and the flow table is issued through the SDN controller. Change the external access path, realize resource aggregation, aggregate two servers and their links into one, and reduce the entire system by 2 web application servers; then gradually increase the pressure, so that the average number of session connections per server reaches 10, and each link When the traffic reaches about 45M, the number of session connections and link traffic are between the minimum aggregation threshold and the maximum threshold to meet resource management requirements, so no related operations are required; after that, increase the session connection and link traffic load of individual servers. The average number of sessions on two of the servers reaches 25, the link traffic reaches about 90M, the number of session connections between the two servers remains 10, and the link traffic remains about 45M. At this time, a balancing operation is triggered to balance the two server sessions whose load exceeds the maximum threshold. Then, continue to increase the number of server session connections and link traffic load, so that the average number of sessions per server reaches 25, and the link traffic reaches At about 90M, all servers exceed the maximum threshold and cannot be balanced. At this time, a new server is automatically added through the cloud platform and connected to the OpenvSwitch switch to balance the excess traffic to the newly added server.

Through the above experiments, it is proved that this method can automatically adjust the scale of the Web application cluster according to the two strategies of the number of session connections and the link traffic. The resulting problem is that the server cluster cannot be adjusted when the load increases. On the other hand, compared with the traditional cloud computing platform, the load balance can only be achieved by adding servers, which saves server resources and effectively improves the utilization efficiency of existing resources.

As shown in Figure 4, an SDN-based web system resource management system in a cloud computing environment includes:

The building module 401 is used to build a web application system in an SDN-based cloud computing environment, where the web application system includes: each server and an SDN controller deployed on the cloud computing platform;

The first obtaining module 402 is configured to obtain the instantaneous session connection number and instantaneous link traffic of each server of the web application system at preset time intervals;

The first comparison module 403 is configured to compare the number of instantaneous session connections obtained twice in a row by each server with the maximum threshold of the number of instantaneous session connections, and compare the instantaneous link traffic obtained twice by each server with the instantaneous link traffic The maximum threshold is compared. If the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the maximum instantaneous link traffic threshold, the system will be activated. The resource balancing module 407 performs the resource balancing operation through the SDN controller, otherwise the second comparison module is called;

The second obtaining module 404 is configured to obtain the average session connection number and link average flow of each server according to the instantaneous session connection number and instantaneous link flow obtained by each server continuously preset times;

The second comparison module 405 is configured to compare the average number of session connections of each server with the maximum threshold of the average number of session connections, and compare the average traffic of each server with the aggregate threshold of the average number of session connections to obtain a comparison result ;

The operation module 406 is configured to, according to the comparison result, start the resource balancing module 407 to perform a resource balancing operation through the SDN controller or start the resource aggregation module 408 to perform a resource aggregation operation through the SDN controller.

Specifically, the operation module 406 is specifically used for:

When the comparison result is that the average number of session connections of any server is greater than the maximum threshold of the average number of session connections or the average link traffic of any server is greater than the maximum threshold of average link traffic, the resource balancing module 407 is started to perform a resource balancing operation through the SDN controller ;

When the comparison result is that the average session connection number of at least one server is less than the average session connection number aggregation threshold and the average link traffic of at least one server is less than the link average traffic aggregation threshold, the resource aggregation module 408 is started to perform resource aggregation through the SDN controller. Aggregate operation.

Specifically, the resource balancing module 407 is specifically used for:

Obtain the first server with the smallest comprehensive usage of session and link resources from the global resource information base through the SDN controller;

When the number of resources to be balanced for the number of session connections is less than the difference between the maximum number of balanced session connections of the first server and the average number of session connections, and the number of resources to be balanced for link traffic is less than the maximum number of balanced link traffic of the first server and the number of links When the difference between the average traffic of the road, the excess load is balanced to the first server through the flow table method.

Among them, the number of resources to be balanced for the number of session connections is the difference between the maximum threshold of the average number of session connections and the average number of session connections of two or more servers, and the number of resources to be balanced for link traffic is the maximum threshold of the average link traffic and the number of more than two servers. The difference between the average traffic of the link.

Specifically, the resource aggregation module 408 is specifically used for:

The two servers with the smallest comprehensive usage of session and link resources are obtained from the global resource information base through the SDN controller;

When the number of server session connections after aggregation is between the aggregation threshold of average session connections and the maximum threshold of average session connections and the server link traffic after aggregation is between the aggregation threshold of average link traffic and the maximum threshold of average link traffic, the two The two links corresponding to each server are aggregated, and the resources occupied by any one of the two links and the server corresponding to the link are released.

The number of session connections of the server after aggregation is the sum of the average number of session connections of the two servers, and the link traffic of the server after aggregation is the sum of the average link traffic of the two servers.

Specifically, the multiple servers include: an application server, a database server, and a load balancing server.

The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (6)

1. a web system resource management method under an SDN-based cloud computing environment, is characterized in that, comprising: S1. Construct a web application system under an SDN-based cloud computing environment, where the web application system includes: multiple servers and SDN controllers deployed on a cloud computing platform; S2, obtaining the instantaneous session connection number and instantaneous link traffic of each of the servers at preset time intervals; S3. Compare the number of instantaneous session connections obtained by each of the servers for two consecutive times with the maximum threshold of the number of instantaneous session connections, respectively, and compare the instantaneous link traffic obtained by each of the servers for two consecutive times with the instantaneous link traffic respectively. The maximum threshold is compared, and if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold for the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the maximum instantaneous link traffic threshold, then start the resource balancing module to perform the resource balancing operation through the SDN controller, otherwise go to step S4; S4, obtain the average session connection number and the average link flow of each of the servers according to the instantaneous session connection number and the instantaneous link flow obtained by each of the servers for a continuous preset number of times; S5. Compare the average session connection number of each server with the maximum average session connection number threshold and the average session connection number aggregation threshold, respectively, and compare the link average traffic of each server with the link average traffic maximum Compare the threshold with the link average traffic aggregation threshold to obtain the comparison result; S6. According to the comparison result, start the resource balancing module to perform a resource balancing operation through the SDN controller or start the resource aggregation module to perform a resource aggregation operation through the SDN controller; Step S6 includes: When the comparison result is that the average session connection number of any server is greater than the maximum threshold of the average session connection number or the average link traffic of any server is greater than the maximum link average traffic threshold, the resource balancing module is started to pass the SDN controller performs a resource balancing operation; When the comparison result is that the average session connection number of at least one server is less than the average session connection number aggregation threshold and the link average traffic of the at least one server is less than the link average traffic aggregation threshold, start the resource aggregation The module performs a resource aggregation operation through the SDN controller; The method for performing resource balancing operation by the SDN controller includes: The SDN controller searches the global resource information base to obtain the first server with the smallest comprehensive usage of session and link resources; When the number of resources to be balanced for the number of session connections is less than the difference between the maximum number of balanced session connections and the average number of session connections of the first server and the number of resources to be balanced for link traffic is less than the maximum link traffic of the first server When the difference between the number of balances and the average traffic of the link is determined, the excess load is balanced to the first server through the flow table method. The number of resources to be balanced for the number of session connections is the difference between the maximum threshold of the average number of session connections and the average number of session connections for two or more of the servers, and the number of resources to be balanced for the link traffic is the number of resources for the link traffic to be balanced. The difference between the maximum threshold of the average traffic on the road and the average traffic on the links of two or more servers. 2. The method for managing web system resources in an SDN-based cloud computing environment according to claim 1, wherein the method for performing resource aggregation operations by the SDN controller comprises: The two servers with the smallest comprehensive usage of session and link resources are obtained by searching from the global resource information base through the SDN controller; When the aggregated server session connection number is between the average session connection number aggregation threshold and the average session connection maximum threshold and the aggregated server link traffic is between the link average traffic aggregation threshold and the link average traffic When between the maximum thresholds, the two links corresponding to the two servers are aggregated, and the resources occupied by any one of the two links and the server corresponding to the link are released, The aggregated server session connection number is the sum of the average session connection numbers of the two servers, and the aggregated server link traffic is the sum of the average link traffic of the two servers. 3. The method for managing web system resources in an SDN-based cloud computing environment according to claim 1 or 2, wherein the plurality of servers comprise: an application server, a database server, and a load balancing server. 4. a web system resource management system under an SDN-based cloud computing environment, is characterized in that, comprising: a building module for building a web application system in an SDN-based cloud computing environment, where the web application system includes: each server and an SDN controller deployed on the cloud computing platform; a first obtaining module, configured to obtain the instantaneous session connection number and instantaneous link traffic of each of the servers of the web application system at preset time intervals; The first comparison module is configured to compare the number of instantaneous session connections obtained by each of the servers for two consecutive times with the maximum threshold of the number of instantaneous session connections, and compare the instantaneous link traffic obtained by each of the servers for two consecutive times with the instantaneous number of connections. Compare with the maximum threshold of link traffic, if the number of instantaneous session connections obtained by any server for two consecutive times is greater than the maximum threshold of the number of instantaneous session connections or the instantaneous link traffic obtained by any server for two consecutive times is greater than the instantaneous link traffic. If the maximum threshold of road traffic is reached, the resource balancing module is started to perform resource balancing operation through the SDN controller, otherwise the second comparison module is called; The second obtaining module is configured to obtain the average session connection number and link average flow of each of the servers according to the instantaneous session connection number and instantaneous link flow obtained by each of the servers for a continuous preset number of times; The second comparison module is configured to compare the average number of session connections of each of the servers with the maximum threshold of the average number of session connections, and compare the average traffic of the links of each of the servers with the aggregate threshold of the average number of session connections, get the comparison result; an operation module, configured to, according to the comparison result, start the resource balancing module to perform a resource balancing operation through the SDN controller or start the resource aggregation module to perform a resource aggregation operation through the SDN controller; The operation module is specifically used for: When the comparison result is that the average session connection number of any server is greater than the maximum threshold of the average session connection number or the average link traffic of any server is greater than the maximum link average traffic threshold, the resource balancing module is started to pass the SDN controller performs a resource balancing operation; When the comparison result is that the average session connections of at least one server are all less than the average session connection number aggregation threshold and the link average traffic of the at least one server is all less than the link average traffic aggregation threshold, start the The resource aggregation module performs a resource aggregation operation through the SDN controller; The resource balancing module is specifically used for: The SDN controller searches the global resource information base to obtain the first server with the smallest comprehensive usage of session and link resources; When the number of resources to be balanced for the number of session connections is less than the difference between the maximum number of balanced session connections and the average number of session connections of the first server and the number of resources to be balanced for link traffic is less than the maximum link traffic of the first server When the difference between the number of balances and the average traffic of the link is determined, the excess load is balanced to the first server through the flow table method, The number of resources to be balanced for the number of session connections is the difference between the maximum threshold of the average number of session connections and the average number of session connections for two or more of the servers, and the number of resources to be balanced for the link traffic is the number of resources for the link traffic to be balanced. The difference between the maximum threshold of the average traffic on the road and the average traffic on the links of two or more servers. 5. the web system resource management system under a kind of SDN-based cloud computing environment according to claim 4, is characterized in that, described resource aggregation module is specifically used for: The two servers with the smallest comprehensive usage of session and link resources are obtained by searching from the global resource information base through the SDN controller; When the aggregated server session connection number is between the average session connection number aggregation threshold and the average session connection maximum threshold and the aggregated server link traffic is between the link average traffic aggregation threshold and the link average traffic When between the maximum thresholds, the two links corresponding to the two servers are aggregated, and the resources occupied by any one of the two links and the server corresponding to the link are released, The aggregated server session connection number is the sum of the average session connection numbers of the two servers, and the aggregated server link traffic is the sum of the average link traffic of the two servers. 6 . The web system resource management system in an SDN-based cloud computing environment according to claim 4 or 5 , wherein the multiple servers include: an application server, a database server, and a load balancing server. 7 .

Images