KR100856664B1 - System for offering application service provider service in grid-base and method for assigning server pool thereof - Google Patents

Abstract

The present invention relates to an ASP service device based on Grid technology and a method for allocating a server pool thereof, and more particularly, to manage a plurality of server resources in a single pool according to the type of application service, and to allocate an application. It is a technology that efficiently uses the server pool while satisfying the quality of service requirements of customers when providing services and unallocated application services. To this end, the present invention provides a server pool for performing the processing of the allocated application service and the non-assigned application service, the application service unit that receives the application service processing request from the customer and provides the application service processing result, and the application service processing According to the request, the server is dynamically allocated according to the utilization rate and processing load of the server providing the allotted application service among the provisioned application services among the application services, and the non-allocated type when the application services are provided. It is characterized by including an application platform for dynamically allocating server resources by determining whether processing within the delivery date can be completed using the progress of the server providing the application service.

Description

ASP service device based on grid technology and server pool allocation method {SYSTEM FOR OFFERING APPLICATION SERVICE PROVIDER SERVICE IN GRID-BASE AND METHOD FOR ASSIGNING SERVER POOL THEREOF}

1 is a block diagram of a grid-based ASP service apparatus according to an embodiment of the present invention.

FIG. 2 is a detailed configuration diagram of the server pool of FIG. 1. FIG.

3 is a flowchart illustrating a service providing method of the ASP service apparatus of FIG. 1.

FIG. 4 is a flowchart of an allocated application service processing process of the ASP service apparatus of FIG. 1. FIG.

5 is a flow chart of an unassigned application service processing of the ASP service device of FIG.

*** Explanation of symbols for the main parts of the drawing ***

100: Applied Service Department 101: Acquisition Department

102: identification unit 103: schedule reservation unit

104: result provider 105: service management unit

200: application platform unit 210: application processing unit

211: schedule setting unit 212: job distribution unit

213: provisioning unit 214: exception handling unit

215: execution control unit 216: resource management unit

220: service adapter unit 230: monitoring unit

300: server pool 301: allocation server pool

302: Unassigned Server Pool 303: Free Server Pool

400: storage

The present invention relates to an ASP service device based on Grid technology and a method for allocating a server pool thereof, and more particularly, to manage a plurality of server resources in a single pool according to the type of application service, and to allocate an application. It is a technology that efficiently uses the server pool while satisfying the quality of service requirements of customers when providing services and unallocated application services.

In general, grid technology is a technology that virtualizes distributed computing resources such as high-performance processors, mass storage devices, applications, and data into supercomputing resources that can be shared with each other over a network. It is a technology that integrates computing resources and utilizes them for various purposes.

Recently, in some areas such as finance, securities, and pharmaceuticals other than science and technology, attempts are being made to share and utilize the computational resources inside the enterprise as grid technology, and efforts are being made to integrate grid technology into business areas.

On the other hand, ASP (Application Service Provider) generally refers to a service provider that distributes and manages software-based services and applications to customers through a communication network.

In the conventional ASP business method, additional computing resources must be secured so that application processing can be performed in a timely manner in response to an unscheduled request of the customer. Therefore, the utilization rate of the computing resources is maintained at a very low level. If a lot of calculations are needed, it is difficult to effectively provide a service.

Conventionally, management of computing resources including application application installation is relied on by a server administrator's manual work, and application requests are handled through dedicated computing resources.

Therefore, it is not practical to manage various application applications and construct an application service processing environment of server resources immediately when needed, and thus it is not practical to provide application services based on grid technology.

In addition, the conventional server has a problem that the utilization rate of the server is low by allocating dedicated server resources for each application service.

That is, conventionally, since the server resources are sufficiently configured to provide the service even at the time of the request in preparation for the unscheduled request of the customer, when the service is provided that does not require a lot of server resources, the server is configured with unnecessary server resources. On the other hand, it causes a waste of resources, when a high performance operation is required a lot of server resources, there is a problem that can not provide enough of the service only the server resources configured.

An object of the present invention, which is created to meet the above-mentioned demands, is to manage server resources in the form of a virtualized shared server pool when providing ASP services based on grid technology, and vary according to the size of the workload. By allocating server resources, we can satisfy the service quality of our customers and increase server utilization.

In addition, another object of the present invention is to allocate a server variably in response to a variable load when requesting an allocated application service, and to meet a delivery date promised to a customer when requesting an unassigned application service.

Grid-based ASP service apparatus according to the present invention for achieving the object as described above, the server pool consisting of a plurality of servers for performing the processing of the application service, the application service processing request from the customer to obtain the application service The server dynamically allocates a server according to the utilization rate and processing load of an application service unit providing a processing result and a server providing the allocated application service when the application service is allocated among the application services according to the application service processing request. And an application platform unit for dynamically allocating server resources by determining whether the processing can be completed within the delivery period by using a task progress rate of the server providing the unassigned application service when the unassigned application service is provided among the application services. It is characterized by the configuration.

In addition, the server pool allocation method in the grid-based ASP service apparatus according to the present invention, the first step of determining the addition, removal, maintenance of the server by monitoring the server utilization rate and processing load providing the allocated application service, and In case of adding a server, it is determined whether an available server resource exists in a server pool for an allocated application service that is providing the allocated application service, and when the available server resource exists, the allocated application using the available server resource. And a second process of continuously providing a service, and a third process of stopping the provision of the allocated allotted application service and marking the server as an available server resource when it is necessary to remove the server. do.

In addition, the server pool allocation method in the grid-based ASP service apparatus according to the present invention, the first process of monitoring the progress of the server providing the unassigned application service, the process using the processing progress and server resource information A second process of calculating an estimated time required for processing of an unassigned application service and determining whether the unassigned application service can be completed on time using the result of the calculation; And a third process of processing the unallocated application service in a deadline by using available server resources in a server pool for an unallocated application service that provides a type application service.

An embodiment using the server allocation method of the grid-based ASP service apparatus of the present invention having the above-described configuration is as follows.

1 is a block diagram of a grid-based ASP service apparatus according to an embodiment of the present invention.

The grid-based ASP service apparatus of the present invention includes an application service unit 100, an application platform unit 200, a server pool 300, and a storage 400.

The application service unit 100 obtains a service provision policy for an application service from a grid ASP service provider, sets a service rule, obtains an application application, stores it in the storage 400, and sets a search directory. After performing the basic setting for, if the application service processing request is generated from the customer, obtain and identify the application service processing request, schedule the application service processing schedule, and the application platform processing result through the application platform unit 200 Provide to customers.

In this case, the service provision policy includes an application service type, a service processing class, a minimum server specification, a minimum memory capacity, and an operating system (OS) type. Here, the application service unit 100 is preferably provided such as a portal site so that customers can easily request and receive the results.

To this end, the application service unit 100 includes an acquisition unit 101, an identification unit 102, a schedule reservation unit 103, a result providing unit 104, and a service management unit 105.

The acquisition unit 101 obtains an application service processing request from a customer and transmits it to the identification unit 102.

Identification unit 102 is the type of application services received from the acquisition unit 101 (assignment type, unassigned type), service processing class (delivery date (designated delivery time desired), delivery date not specified), server requirements, memory capacity , Storage requirements, and estimated processing time are identified to create a profile for processing the application request. Here, among the types of application services, an allocated application service means a service that requires a dedicated server allocation to provide a continuous service, and an unassigned application service uses a server only for processing time of an application request. It means unnecessary services.

The schedule reservation unit 103 pre-sets a schedule for processing an application request in cooperation with the schedule setting unit 211 of the application service processing unit 210.

The result providing unit 104 provides the customer with the application service processing result processed based on the grid technology through the application platform unit 200.

The service manager 105 registers an initial application application and a service provision policy in the storage 400 to provide an application service and manages the application service.

The application platform unit 200 divides or merges the processing source data unit according to the type of application service, sets a processing schedule in consideration of available server resources of the server pool, and is responsible for configuring server resources, distributing work, and monitoring according to the schedule. . To this end, the application platform 200 includes an application processor 210, a service adapter 220, and a monitoring unit 230.

If there is an application service processing request from the application service unit 100, the application processing unit 210 sets a schedule, distributes work, and provisions for performing application service processing using the server pool 300 according to the type of the requested service. (provisioning) function, etc. To this end, the application processing unit 210 includes a schedule setting unit 211, a job distribution unit 212, a provisioning unit 213, an exception processing unit 214, an execution control unit 215, and a resource management unit 216. .

The schedule setting unit 211 establishes a processing schedule in conjunction with the schedule reservation unit 103, the job distribution unit 212 allocates server resources and distributes jobs according to a schedule, and the provisioning unit 213 is a server pool. Dynamically configure and manage the execution environment of the server resources of the exception, the exception processing unit 214 handles exceptions that occur during processing, the execution control unit 215 controls the operation of each component (211 ~ 216) The detailed job processing order check and parallel processing are performed. The resource manager 216 stores a schedule table, server resource allocation information, server resource work status, and the like.

In particular, the execution controller 215 uses the monitoring result of the monitoring unit 230 to maintain the server if the service processing load is an appropriate level, and to remove some servers when the service processing load is low and the server utilization is low, If the processing load is above a certain level and the server utilization rate is high, additional servers should be added.

For example, the Allocate Application Service includes services that the user's request volume varies over time, such as CDN (CONTENTS DELIVERY NETWORK) service, content download, media streaming, etc., and more than a certain number of server resources for the Allocate Application Service. Assign to a dedicated server in advance. Such an allocated application service has a fixed server resource, and the server may be insufficient due to a heavy load during a crowded time. Accordingly, the execution control unit 215 of the present invention controls to dynamically allocate server resources in response to variable traffic when providing an allocated application service.

On the other hand, unassigned application services include graphics rendering, media file encoding services, and the like, and include application services that require a lot of parallel computation. Such unassigned application services can be completed within a delivery date specified by a customer. It is important. Accordingly, the execution control unit 215 of the present invention calculates the estimated time required for processing the unallocated application service using the progress of the server resources monitored by the monitoring unit 230 and adds server resources according to the result. This controls the application service to be processed on time.

The service adapter unit 220 divides an application service processing job into detailed job units or merges the parallel job processing results when it is necessary to perform parallel processing using a plurality of servers.

The monitoring unit 230 monitors the work progress status of the server resources, the server resource status, the billing information, and the like, and transmits the processing result to the storage 400 and the result providing unit 104.

The server pool 300 manages server resources as virtual shared resources and executes applications according to types of applications. To this end, the server pool 300 includes an allocated server pool 301, an unassigned server pool 302, and a spare server pool 303.

The configuration of the server pool 300 will be described in detail with reference to FIG. 2.

First, the assignable server pool 301 includes server resources pre-allocated for the allocated application service. For example, a certain number of server resources in allocated server pool 1 or allocated server pool 1 are allocated for the processing of the CDN service, and allocated server pool 2 or allocated server pool 2 in the processing of the media streaming service. A certain number of server resources are allocated. In addition, when the server pool 301 receives a variable server request from a customer, the server server normally provides a service using a minimum amount of server resources, and dynamically allocates server resources in response to a load during a busy time. It is desirable to be able to efficiently handle the variable load.

Unassigned server pool 302 includes server resources for unassigned application services and then allocates server resources in unassigned server pool 302 according to the amount of work data requested by the customer or server usage as a result of service provision. Then assign it and use it. In addition, the unassigned server pool 302 may dynamically allocate and process server resources in accordance with the desired processing completion time for an unallocated application service requiring processing completion desired time from a customer.

The spare server pool 303 reserves server resources separate from the assigned server pool 301 and the unassigned server pool 302, thereby assigning the assigned server pool 301 and the unassigned server pool 302. When the server resources of the server is insufficient, the server resources in the free server pool 303 are provided.

The storage 400 stores application application and service provision policies, stores data of application service processing, and stores original data and intermediate or final results for application requests.

Hereinafter, a method of providing a service of a grid technology based ASP service apparatus according to an embodiment of the present invention will be described with reference to FIGS. 3 to 5.

3 is a flowchart illustrating a service providing method of the ASP service apparatus of FIG. 1.

When the application service processing request from the customer is obtained through the acquisition unit 101, the acquisition unit 101 transmits the application service processing request to the identification unit 102, the identification unit 102 is the content of the service requested application service processing Identify the type of application service (allocation type, non-allocation type), service processing class (delivery date designation, delivery date not specified), server requirements, memory capacity, storage requirements, and estimated processing time required (S100).

Thereafter, the schedule reservation unit 103 sets the priority of the application service processing according to the identified service type for each service type, parallel processing level, server specification storage usage, etc. and schedules the schedule in conjunction with the schedule setting unit 211. (S200).

Subsequently, the execution controller 215 determines the type of the requested service according to the reserved schedule (S300), and performs application service processing according to the service type (S400, S500). That is, as a result of the determination in step S300, if the requested application service belongs to the server-allocated application service, the allocated-type processing is performed (S400), and if the requested application service belongs to the server-unassigned application service, the unassigned processing is performed. Perform (S500). That is, the application processing unit 210 allocates and distributes application service processing tasks to server resources according to application service types to process grid technology based ASP services. The method of allocating the server during the allocation process of the process S400 and the non-allocation process of S500 will be described in detail later with reference to FIGS. 4 and 5.

Thereafter, when the allocated or unassigned processing for the requested application service is completed, the customer is notified of the application service processing result (S600).

Hereinafter, a method of allocating server resources when providing an allocated application service will be described in detail with reference to FIG. 4. In particular, FIG. 4 illustrates a method of guaranteeing a quality of service at the time of job processing by variably allocating a server when providing an allocated application service.

Initially, allocated application services are provided using the allocated server resources in the allocated server pool 301. During the provision of such an allocated application service, the monitoring unit 230 monitors the server utilization rate and processing load during service provision (S401), and the execution control unit 215 determines whether to add, remove, or maintain the server (S402). ). That is, the execution control unit 215 uses the monitoring result of the monitoring unit 230, and if the service processing load and the server utilization rate are within the range of the first threshold value and the second threshold value, the number of servers that provide the allocated application service. If the service processing load and processing load is greater than or equal to the second threshold, additional servers are added to increase the number of servers that provide the allocated application service, and the service processing load and processing load are If the threshold is less than 1, some servers in operation may be removed to reduce the number of servers providing the allocated application service.

When it is determined in step S402 that the server needs to be removed, the execution control unit 215 stops driving server resources in the server pool 301 providing the application service (S403) and schedules the corresponding server resources as available server resources. In the table (S404).

As a result of the determination in step S402, when it is necessary to add a server, the execution control unit 215 determines whether there is an available server resource in the server pool 301 for providing the application service (S405).

As a result of the determination in step S405, when no available server resource exists in the service server pool 301, the execution controller 215 determines whether there is available server resource in the free server pool 303 (S406). If the available server resources do not exist even in the free server pool 303, the result provider 104 notifies the operator of the necessity of expanding the server (S407).

On the other hand, when the available server resources exist in the service server pool 301 as a result of the determination in step S405 and the available server resources exist in the spare server pool 303 as a result of the determination in step S406, the execution control unit 215 ) Determines whether server provisioning is necessary for the available server resource (S408).

That is, the execution control unit 215 determines whether an application for providing the allocated application service is installed in the available server resource, and determines that provisioning is required if the application is not installed, and provisioning is performed if the application is installed. I do not think it is necessary. At this time, the execution control unit 215 preferably reflects the information on the available server resources in the schedule table.

As a result of the determination in step S408, if server provisioning is not required, the server allocates an application service using the corresponding server resource (S410), and if server provisioning is required, the server for the server resource through the provisioning unit 213. After configuring a work environment and installing an application (S409), the application service is processed (S410).

As described above, the present invention provides a variable load by providing a variable allocation of servers according to the traffic amount in consideration of the variable load even when a customer initially specifies the minimum fixed dedicated server number when providing an allocated application service. Can be efficiently processed.

Hereinafter, a method of allocating server resources when providing an unallocated application service will be described in detail with reference to FIG. 5.

First, when providing an unassigned application service, the monitoring unit 230 calculates a processing time required for a corresponding service by monitoring a work processing progress rate of a server providing a service (S501).

As a result of the calculation in step S501, it is determined whether the corresponding application service can be completed within the delivery date (process completion desired time) (S502), and if it is possible within the delivery date, proceed as it is and if the server for the service for additional allocation is impossible It is determined whether there is an available server resource in the pool 302 (S503).

As a result of the determination in step S503, when no available server resource exists in the service server pool 302, the execution controller 215 determines whether there is available server resource in the free server pool 303 (S504). If the available server resources do not exist even in the free server pool 303, the result provider 104 notifies the operator of the necessity of server expansion (S505).

On the other hand, when the available server resources exist in the service server pool 302 as a result of the determination in step S503, and the available server resources exist in the spare server pool 303 as a result of the determination in step S504, the execution control unit 215 ) Determines whether server provisioning is necessary for the available server resource (S506).

That is, the execution control unit 215 determines whether an application for providing the allocated application service is installed in the available server resource, and determines that provisioning is required if the application is not installed, and provisioning is performed if the application is installed. I do not think it is necessary. At this time, the execution control unit 215 is preferably to reflect the information on the available server resources in the schedule table.

As a result of the determination in step S506, when server provisioning is not necessary, an allocated application service is processed using the corresponding server resource (S508), and when server provisioning is required, the server is provided with respect to the corresponding server resource through the provisioning unit 213. After configuring a work environment and installing an application (S507), the application service is processed (S508).

As shown in FIG. 5, the present invention allows a customer to specify a service processing grade (delivery date (delivery completion time), non-delivery date) at the time of providing an unassigned application service, and to complete the work within the processing completion time required by the customer. do. To this end, the execution control unit 215 predicts the expected time required for the requested work based on the available resource status (job progress rate) of the server resources according to the job processing schedule table and whether the customer can provide the desired delivery date in the service system. If it is determined that the service processing cannot be completed within the expected time, the server should be expanded.

As described above, when the grid technology is applied to the ASP service field for processing an application that requires a large amount of computing resources, the present invention dynamically allocates server resources according to the type of application service and accurately predicts a predetermined time for processing. Allows you to complete application service provision.

As described above, the present invention manages server resources in the form of a virtualized shared server pool when providing an ASP service based on grid technology, and efficiently allocates server resources according to the size of a workload. It satisfies the quality and improves the economics by increasing the server utilization rate.

The present invention is not limited to the above-described embodiments, but can be modified and modified within the scope not departing from the gist of the present invention, and such modifications and changes should be regarded as belonging to the following claims.

Claims (19)

A server pool consisting of a plurality of servers for processing application services; An application service unit which obtains the application service processing request from a customer and provides the application service processing result; And According to the application service processing request, the server dynamically allocates the server according to the utilization rate and processing load of the server providing the allocated application service when providing the allocated application service among the application services, and the unassigned application among the application services. It is configured to include an application platform for dynamically allocating server resources by determining whether it is possible to complete the processing within the delivery period using the work progress of the server providing the unassigned application service at the time of service provision, The server pool, An assignable server pool including dedicated server resources for the assigned application service; An unallocated server pool including a plurality of server resources for the unallocated application service; And Grid technology-based ASP service apparatus comprising a free server pool that provides a free server resources when the available resources of the server resources of the allocated server pool or the unallocated server pool. The method of claim 1, Grid for ASP based on grid technology, characterized in that further comprising a storage for storing the application for the application service processing, the data of the application service processing, the original data for the application service processing, the service provision policy by the application service provider Service device. The method of claim 2, wherein the application service unit, Upon receiving a service registration request from the application service provider, an ASP service device based on grid technology, wherein the service providing policy and the application for processing the application service are delivered to the storage. The method of claim 2, wherein the application service unit, An acquisition unit for obtaining an application service processing request from a customer; An identification unit for identifying a content of an application service processing request received from the acquisition unit; A schedule reservation unit that checks available resources of server resources for processing the application service and reserves the schedule in advance; A result providing unit providing the application service processing result to the customer; And Grid service technology-based ASP service device characterized in that it comprises a service management unit for registering the application and service provision policy for the application service processing in the storage and performs the management of application services. The method of claim 4, wherein the identification unit A profile for processing the application service request by identifying the type of the service, the service processing class, the server requirements, the memory capacity of the server resources, the storage requirements, and the expected processing time of the application service An ASP service apparatus based on grid technology, which is generated. The method of claim 1, The assigned application service corresponds to a service requiring a dedicated server allocation for continuous service provision, and the unassigned application service uses a server only during processing time of an application request, and thus corresponds to a service requiring no server assignment. An ASP service apparatus based on grid technology. delete The method of claim 1, wherein the application platform unit, An application processing unit which performs schedule setting, job distribution, exception processing, and provisioning functions in association with the application service unit; A service adapter unit for dividing the data to be worked on for parallel processing and collecting and merging the parallel processing results from the server resources when the requested application service is an unassigned application service; And Grid technology-based ASP service characterized in that it comprises a monitoring unit for monitoring the application service processing result, the utilization rate of the server resources, the processing load of the server resources, the work status of the server resources, and the work progress of the server resources Device. The method of claim 8, wherein the application processing unit, A schedule setting unit configured to set a schedule previously reserved according to available resources of server resources for application service processing; A job distribution unit for allocating server resources and distributing jobs according to the schedule; A provisioning unit that dynamically installs and manages an execution environment of server resources of the server pool by installing an application for processing the application service; An execution control unit controlling driving of the schedule setting unit, the job distribution unit, and the provisioning unit, performing a detailed job processing order and determining whether to perform parallel processing, and controlling dynamic allocation of the server according to the application service type; A resource management unit for storing a schedule table including schedule information set by the schedule setting unit, the server resource allocation information, and a server resource work state; An ASP service apparatus based on grid technology. The method of claim 9, wherein the execution control unit, Determining the addition, maintenance, and removal of the server according to the utilization rate and processing load of the server providing the allocated application service, and if the server needs to be added, it is determined whether there are available server resources in the corresponding service pool. If there is no server resource, it is determined whether an available server resource exists in the free server pool, and if the available server resource exists, it is determined whether server provisioning is required for the available server resource. ASP technology device based on grid technology, characterized in that the control to process the application service after installation. The method of claim 9, wherein the execution control unit, Calculate the estimated time required for the service using the task progress rate of the server that is providing the unallocated application service to determine whether the in-delivery process can be completed, and if the in-delivery process cannot be completed, add the server to the non-allocated application. An ASP service apparatus based on grid technology, which controls a service to be completed in a delivery date. The method of claim 11, wherein the execution control unit, At the time of adding the server, it is determined whether there is an available server resource in the corresponding service pool, and if it is not present, it is determined whether there is an available server resource in the free server pool, and if the available server resource exists, the available Grid technology-based ASP service device characterized in that it determines whether server provisioning is required for server resources, and if the provisioning is required to control the application service after installing the application. delete delete delete delete A first step of monitoring a processing progress of a server that is providing an unassigned application service; A second step of calculating an estimated time required for processing of the unassigned application service using the processing progress rate and server resource information and determining whether the unassigned application service can be completed within the delivery date using the calculation result; And If the in-delivery cannot be completed, a third process of processing the unassigned application service in-time using available server resources in the unassigned application service server pool for providing the unassigned application service; Make up, The third process is Step 3-1 of determining whether an available server resource exists in the server pool for unallocated application services; A third step of determining whether an available server resource exists in a free server pool when the available server resource does not exist as a result of the determination of step 3-1; Determining whether a server provisioning is necessary for the available server resource when an available server resource exists in the free server pool or an available server resource exists in the server pool for unassigned application service; And When the server provisioning is required, grid technology-based ASP service apparatus comprising the step 3-4 of installing an application for the server resources and configuring a server working environment and providing the unallocated application service Server pool allocation method in. delete The method of claim 17, As a result of the determination of step 3-2, if the available server resources do not exist in the free server pool, grid technology based ASP, characterized in that further comprising the step of notifying the operator needs to increase the server How server pools are allocated in service units.

Images