JP5939620B2 - Computer system, server device, load balancing method, and program - Google Patents

Description

  The present invention relates to a computer system including a plurality of server devices, a server device used therefor, a load distribution method for distributing loads in the computer system, and a program for realizing them.

  Conventionally, a system constructed by a plurality of application servers (hereinafter referred to as “AP servers”) and a load balancer is used in order to cope with a large amount of processing requests from a plurality of clients. Among these, the load balancer is a device that centrally manages processing requests from clients, and transmits a request to the AP server in a Push type. Processing requests transmitted from a plurality of clients are equally distributed to each AP by the load balancer. As a result, system load distribution is achieved.

  However, since the load balancer does not have a function that considers the resource status and load status of each AP server, there may be a case where a processing request exceeding the processable amount is transmitted to each AP server. In this case, there is a possibility that the processing speed decreases or the processing becomes impossible due to overflow.

  For this reason, Patent Document 1 proposes a system in which a load balancer is not installed, and the AP server itself has a function of processing processing requests in a distributed manner. Specifically, in the system disclosed in Patent Document 1, an AP server that has received a processing request from a client determines an AP server that performs processing for the processing request according to the load status of each AP server, Send a processing request to. In addition, when each AP server receives a request from another AP server, each AP server can also determine whether to perform the processing by itself or to request another AP server to perform the processing in accordance with its own load status.

  According to the system disclosed in Patent Document 1, since a situation where a processing request exceeding the processable amount is transmitted to each AP server is avoided, a decrease in processing speed and occurrence of overflow in the system are suppressed. Conceivable.

International Publication No. 2009/034994

  However, in the system disclosed in Patent Document 1, each AP server does not dynamically monitor the resource usage status of itself or another AP server. For this reason, in the system disclosed in Patent Document 1, it is difficult to perform load distribution in the entire system even if a temporary decrease in processing speed and occurrence of overflow can be suppressed.

  An example of an object of the present invention is to solve the above-described problem, and in a system composed of a plurality of server devices, a computer system and a server that can implement load distribution that dynamically considers the resource usage status of each server device An object is to provide a device, a load balancing method, and a program.

In order to achieve the above object, a server apparatus according to the present invention is a server apparatus that constitutes a system that provides a service in response to an external processing request and that executes an application program for providing the service. And
A resource information acquisition unit that acquires resource information for identifying the resource usage status of the server device;
A processing control unit that obtains a job to be processed by the application program from the outside, passes the acquired job to the application program, and executes the job;
Based on the acquired resource information, the processing control unit acquires the job to be processed so as to satisfy a first condition set to be a target of the resource usage status. Let them do, and
A resource monitoring unit that causes the application program to suspend execution of the job when the acquired resource information exceeds a second condition set as an upper limit on the usage status of the resource;
It is characterized by having.

In order to achieve the above object, a computer system according to the present invention is a system that provides a service in response to an external processing request,
A plurality of server devices for executing an application program for providing the service;
Each of the plurality of server devices includes:
A resource information acquisition unit that acquires resource information for identifying the resource usage status of the server device;
A processing control unit that obtains a job to be processed by the application program from the outside, passes the acquired job to the application program, and executes the job;
Based on the acquired resource information, the processing control unit acquires the job to be processed so as to satisfy a first condition set to be a target of the resource usage status. Let them do, and
A resource monitoring unit that causes the application program to suspend execution of the job when the acquired resource information exceeds a second condition set as an upper limit on the usage status of the resource;
It is characterized by having.

In order to achieve the above object, a load distribution method according to the present invention is a method for reducing a load in a system that provides a service in response to an external processing request,
(A) acquiring resource information for specifying a resource usage status of the server device by the server device that configures the system and executes an application program for providing the service; and
(B) Based on the resource information acquired in the step (a) by the server device, the application satisfies the first condition set to be a target of the usage status of the resource. Acquiring a job to be processed by the program from outside, step,
(C) In the application program, when the resource information acquired in the step (a) exceeds a second condition set as an upper limit on the usage status of the resource by the server device, Steps to suspend job execution, and
It is characterized by having.

Furthermore, in order to achieve the above object, a program according to the present invention configures a system that provides a service in response to an external processing request, and executes an application program for providing the service.
(A) obtaining resource information for identifying resource usage status of the computer; and
(B) Based on the resource information acquired in the step (a), the processing target of the application program is set so as to satisfy a first condition set to be a target of the usage status of the resource. Get the job from outside, step,
(C) Suspend the execution of the job in the application program when the resource information acquired in step (a) exceeds a second condition set as an upper limit on the usage status of the resource. Let the steps and
Is executed.

  As described above, according to the present invention, in a system constituted by a plurality of server devices, it is possible to implement load distribution that dynamically considers the resource usage status of each server device.

FIG. 1 is a block diagram showing an overall configuration of a computer system according to an embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of the server device according to the embodiment of the present invention. FIG. 3 is a flowchart showing operations of the computer system and the server device in the present embodiment. FIG. 4 is a block diagram illustrating an example of a computer that implements the server device according to the embodiment of the present invention.

(Embodiment)
Hereinafter, a computer system, a server device, a load distribution method, and a program according to an embodiment of the present invention will be described with reference to FIGS.

[System configuration]
First, the configuration of the computer system according to the embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing an overall configuration of a computer system according to an embodiment of the present invention.

  As shown in FIG. 1, a computer system 100 according to the present embodiment is a system for providing a service in response to a processing request from an external user. The computer system 100 includes a server device (hereinafter referred to as “AP server”) 40 that executes an application program for providing a service.

  The computer system 100 is intended to distribute the load on the AP server 40 as will be described later. In the example of FIG. 1, three server devices 40 are illustrated, but the number of server devices 40 is not particularly limited in the present embodiment.

  As shown in FIG. 1, in the present embodiment, the computer system 100 includes a database server (hereinafter referred to as “DB server”) 30 and a plurality of Web servers 20 in addition to the AP server 40. ing. Further, the Web servers 20 are not connected to each other but are connected to the DB server 20. Similarly, the AP servers 40 are not connected to each other but are connected to the DB server 20. That is, the DB server 30 is arranged so as to be interposed between the Web server 20 and the AP server 40, and the computer system 100 is constructed with a three-layer structure.

  Further, the user's client terminal 10 is connected to the Web server 20, and the user's processing request is transmitted from the client terminal 10 to the Web server 20. The Web server 20 receives the transmitted user processing request and transmits it to the DB server 30.

  As shown in FIG. 1, the DB server 30 includes a database 33 constructed in a storage area of a storage device (not shown in FIG. 1), and a data area 32 called “page” is created on the database 33. Has been. Data is written in the data area 32, and the written data is held in units called “records”. Hereinafter, the data area 32 is expressed as “page 32”. Further, data held in units of records is denoted as “record 31”.

  Writing of the record 31 to the page 32 of the DB server 30 is performed by the Web server 20 described above. When the Web server 20 receives a processing request from the client terminal 10, the Web server 20 accesses the DB server 30 and writes the received processing request as a record 31 (1 processing request = 1 record) on the database 33. Further, the database 33 is provided with a status indicating the state of the processing request as an item that needs to be written. Therefore, the Web server writes “unprocessed” in the status when writing the accepted processing request in the record 31.

  The AP server 40 acquires the record 31 whose status is not processed among the records 31 held in the page 32 of the DB server 30 as a job to be processed. At this time, the AP server 40 acquires the record 31 and simultaneously writes “processing” in the corresponding status. Further, when the processing of the acquired record 31 is completed, the AP server 40 further writes “completed” in the corresponding status.

[Device configuration]
Next, the configuration of the AP server (server device) 40 in the present embodiment will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the server device according to the embodiment of the present invention.

  As described above, each AP server 40 shown in FIG. 2 constitutes a part of the computer system 100 and executes an application program (hereinafter referred to as “business process execution application”) 42 for providing a service. . As illustrated in FIG. 2, the AP server 40 includes a resource information acquisition unit 48, a processing control unit 43, and a resource monitoring unit 47. Reference numeral 41 denotes a monitoring processing unit that collects these.

  The resource information acquisition unit 48 acquires resource information that identifies the resource usage status. In addition, the process control unit 43 acquires a job to be processed by the business process execution application 42 from the outside, and passes the acquired job to the business process execution application 42 to be executed.

  Based on the acquired resource information, the resource monitoring unit 47 sets the job to be processed so as to satisfy the first condition set to be the target of the resource usage status. Let the acquisition take place.

  In addition, when the acquired resource information exceeds the second condition set as the upper limit on the resource usage status, the resource monitoring unit 47 causes the business process execution application to suspend job execution.

  As described above, in the present embodiment, each AP server 40 independently monitors the resource usage status. In addition, each AP server performs monitoring for making the resource usage state appropriate based on the first condition, and monitoring for a sudden high load state based on the second condition. Usage is dynamically monitored.

  Here, the configuration of the AP server (server device) 40 in the present embodiment will be described more specifically. First, in the present embodiment, each AP server 40 includes, in addition to the resource information acquisition unit 48, the processing control unit 43, and the resource monitoring unit 47, a capability calculation unit 44, a job number calculation unit 45, and an input reception. Part 46.

  In this embodiment, the resource information acquisition unit 48 acquires the resource information of the AP server 40 provided with the polling interval set by the system administrator. Further, the usage status of resources to be monitored differs depending on the system environment of the AP server 40. For this reason, the usage status of the resource specified by the resource information can be appropriately selected by the system administrator according to the system environment of the AP server 40.

Specifically, the administrator can select the usage status of the resource specified by the resource information from the following (1) to (7) according to the system environment.
(1) CPU usage rate (2) Memory usage rate (3) Swap memory usage rate (4) Number of open files (5) Number of open threads (6) Hard disk usage (7) Requesters are the same and at the same time The number of processes that are running

  In the present embodiment, the resource information acquisition unit 48 can acquire resource information by executing an OS standard command, referring to a kernel parameter, or the like, for example. However, the specific method when the resource information acquisition unit 48 acquires the resource information differs depending on the operating system (OS) installed in the AP server 40 and is not particularly limited.

  The input receiving unit 46 is an interface for information input from the outside. When the system administrator inputs the first condition and the second condition using the input device or the terminal device when starting the operation of the system, the input receiving unit 46 receives these and the resource monitoring unit 47. The first condition and the second condition are appropriately set by the system administrator.

  In the present embodiment, the first condition includes an appropriate value (appropriate usage amount, appropriate usage rate) of the resource usage status, for example, an appropriate CPU usage rate, an appropriate memory usage rate, an appropriate number of open files, and the like. The first condition is an appropriate usage amount (or usage rate) of the resource that the AP server 40 seeks to use. The second condition includes an upper limit value of the resource usage status, for example, an upper limit value of the CPU usage rate, an upper limit value of the memory usage rate, an upper limit value of the number of open files, and the like. The second condition is an upper limit value of the usage amount (or usage rate) of the resource that the AP server 40 permits to use.

  The administrator can also set a polling interval when the resource information acquisition unit 48 acquires resource information. In this case, the administrator inputs the set polling interval to the input receiving unit 46 using an input device or a terminal device.

  Further, in the present embodiment, the resource monitoring unit 47 acquires the job by the processing control unit 43 so that the usage status of the resource specified by the resource information is closest to the appropriate value that is the first condition. Control. Thereby, load distribution in the computer system 100 is performed.

  Specifically, the resource monitoring unit 47 compares the resource usage specified by the resource information acquired by the resource information acquisition unit 48 with the appropriate value that is the first condition, and the resource usage is appropriate. Determine whether or not.

  If the resource usage status exceeds the appropriate value as a result of the determination, the resource monitoring unit 47 becomes a processing target for the processing control unit 43 so that the resource usage status approaches the appropriate value. Suspend job acquisition.

  Specifically, since the processing control unit 43 exchanges data with the DB server 30 by issuing an SQL statement, the resource monitoring unit 47 instructs the processing control unit 43 to select an unprocessed record. Send a message to suspend issuing. When the resource usage state falls below the appropriate value, the resource monitoring unit 47 transmits a message to the processing control unit 43 so as to resume issuing the SELECT statement that has been suspended.

  In the present embodiment, the resource monitoring unit 47 is executing the business process execution application 42 when the usage status of the resource specified by the resource information exceeds the upper limit that is the second condition. Suspend the job. Specifically, the resource monitoring unit 47 transmits a message from the processing control unit 43 to the job processing execution application 42 to suspend the job being executed. This prevents the AP server 40 from going down due to resource exhaustion.

  At this time, the job being executed is suspended in order from the job executed later. Then, the resource monitoring unit 47 executes the business process through the processing control unit 43 so that the suspended jobs are resumed in order from the previously suspended job when the resource usage state falls below the upper limit. A message is transmitted to the application 42.

  In the present embodiment, as described above, the processing control unit 43 issues a SQL statement to the DB server 30 to acquire the processing request record 31 from the client terminal 10 as a job. Then, when the process control unit 43 acquires the record 31, the process control unit 43 delivers the acquired record 31 to the business process execution application 42 of the AP server 40, unless a message for suspending the issuance of the SQL statement is received from the resource monitoring unit 47. Let it run.

  Further, the process control unit 43 acquires a record 31 whose status is “unprocessed” and which is not locked by a transaction of another AP server 40. At this time, the DBMS (Data Base Management System) lock of the DB server 30 may not be used for some reason. In this case, the process control unit 43 can explicitly declare a lock when acquiring the record 31, thereby avoiding a situation in which the same record 31 is acquired redundantly by a plurality of AP servers 40. Is done.

  As described above, when the processing control unit 43 receives a message for suspending the issuance of the SQL statement (SELECT statement) from the resource monitoring unit 47, the processing is performed until the message for resuming the issuance of the SQL statement is received. Suspend issuing.

  Further, as described above, the processing control unit 43 transmits a message when instructed by the resource monitoring unit 47 to transmit a message to suspend the job being executed. As a result, until the processing control unit 43 transmits a message indicating that the suspended job is to be resumed in response to an instruction from the resource monitoring unit 47, the business process execution application 42 executes the job in order from the job with the latest execution start time. Suspend execution.

  Furthermore, in the present embodiment, when acquiring the record 31, the process control unit 43 acquires the total number of records 31 and the total number of records 31 whose status is unprocessed from the DB server 30, and these Can be transmitted to the job number calculation unit 45 and the capability calculation unit 44.

  The capability calculation unit 44 can calculate the capability (for example, the number of simultaneous processing jobs) A of other AP servers 40 configuring the computer system 100. Specifically, the capability calculation unit 44 uses the total number R1 of records 31 received from the processing control unit 43 and the total number R3 of records 31 whose status is unprocessed, based on the following formula (1). Thus, the processing capability A of the other AP server is calculated.

A = (R1-R2) -R3 (1)
R1: Total number of records R2: Number of records processed by the local server R3: Number of records whose status is not processed

  The job number calculation unit 45 compares the capability (for example, the number of jobs that can be executed simultaneously) B of the AP server 40 provided with the job number calculation unit 45 with the capability A of the other AP server 40 calculated by the capability calculation unit 44. Depending on the comparison result, the number of jobs acquired by the control processing unit 43 can be calculated.

  Specifically, first, the job number calculation unit 45 also considers the processing capability of other AP servers 40 to determine the number of records (number of jobs) to be acquired and processed by the own AP server 40. A-B) is calculated. Then, the job number calculation unit 45 determines the number of jobs to be acquired by the control processing unit 43 as shown in the following (a) to (b) based on the calculated value of α.

(A) When α> 0 (A> B)
In this case, since the processing capability of the other AP server exceeds the processing capability of the own AP server, the job number calculation unit 45 first determines the number of records of the status of “unprocessed” R3 from the other AP server. Subtract processing capacity A. Then, the job number calculation unit 45 compares the number of records obtained by subtraction with the processing capacity (number of jobs that can be executed simultaneously) B of the own AP server, and if B is large, “the number of records obtained by subtraction”. If B is small, “B” is set as the number of jobs to be acquired next time by the control processing unit 43.
(B) When α <0 (A <B)
In this case, since the processing capability of the other AP server is lower than the processing capability of the own AP server, the job number calculation unit 45 determines the processing capability (the number of jobs that can be executed simultaneously) B of the own server next time. Set as the number of jobs to be acquired. However, at this time, the job number calculation unit 45 prevents the number of jobs to be acquired next time from exceeding the number R3 of records whose status is “unprocessed”.
(C) When α = 0 (A = B)
In this case, since the processing capability of the other AP server and the processing capability of the own AP server are the same, the job number calculation unit 45 calculates the half of the processing capability (number of jobs that can be executed simultaneously) B of the own AP server. This is set as the number of jobs to be acquired next time by the control processing unit 43. However, at this time, the job number calculation unit 45 prevents the number of jobs to be acquired next time from exceeding the number R3 of records whose status is “unprocessed”.

[System operation and device operation]
Next, operations of the computer system 100 and the AP server 40 in the present embodiment will be described with reference to FIG. FIG. 3 is a flowchart showing operations of the computer system and the server device in the present embodiment. In the following description, FIGS. 1 and 2 are referred to as appropriate. In the present embodiment, the load distribution method is implemented by operating the computer system 100. Therefore, the description of the load distribution method in the present embodiment is replaced with the following description of the operation of the computer system 100.

  First, as shown in FIG. 3, when the Web server 20 receives a processing request from the client terminal 10, the Web server 20 writes the received processing request as a record 31 in the DB server 30 (step A1). At this time, the Web server 20 writes one processing request as one record 31.

  Next, in each AP server 4, the process control unit 43 accesses the DB server 30 and records a record 31 whose status is “unprocessed” and which is not locked by a transaction of another AP server 40. Obtain (step A2).

  Note that when the record 31 is acquired in step A2, a record that is normally acquired by a server is locked by the DBMS, and a situation in which the same record is acquired by a plurality of servers does not occur. However, if the DBMS lock cannot be used for some reason, the processing control unit 43 explicitly declares the lock when acquiring the record 31.

  Next, in the AP server 40, prior to the execution of the job acquired in step A2, the resource monitoring unit 47 is based on the resource information acquired by the resource information acquiring unit 48, and the AP server 40 provided with the resource monitoring unit 47. Monitor the usage of resources. Then, the resource monitoring unit 47 determines whether or not the resource usage status (system load) of the AP server 40 is equal to or less than a preset reference value (step A3). The “reference value” is set to, for example, a value between an appropriate value that is the first condition and an upper limit value that is the second condition.

  As a result of the determination in step A3, if the resource usage status of the AP server 40 exceeds the reference value, the resource management unit 47 enters a standby state until the resource usage status falls below the reference value. On the other hand, as a result of the determination in step A3, if the resource usage status of the AP server 40 does not exceed the reference value, the resource monitoring unit 47 sends a message to the processing control unit 43 to execute the processing. .

  Upon receiving the process execution message from the resource monitoring unit 47, the process control unit 43 transmits a message to execute the process to the business process execution application 42 installed in the AP server 40 (step A4). Thereafter, the business process execution application 42 executes processing of the acquired record.

  Next, when record processing is started by the business process execution application 42, the resource monitoring unit 47 monitors the resource usage status of the AP server 40 during execution of the job (record). Then, the resource monitoring unit 47 determines whether or not the resource usage status is less than or equal to the set appropriate value (first condition) (step A5).

  If the result of determination in step A5 is not less than or equal to the appropriate value (exceeds), the resource monitoring unit 47 sends a message to the processing control unit 43 so as to suspend the next issue of the SQL statement (step A6).

  Next, when step A5 is executed, the process control unit 43 specifies the number of jobs being executed when the resource usage status of the AP server 40 exceeds the appropriate value (step A7). Then, the process control unit 43 sets the specified number of jobs as the simultaneously executable job number B, and transmits this to the capability calculation unit 44 and the job number calculation unit 45.

  Next, the capability calculation unit 44 accesses the DB server 30 through the processing control unit 43 and acquires the total number of records R1. At the same time, the capability calculation unit 44 searches for a record 31 whose status is “unprocessed” and is not locked by a transaction of another server, and also acquires the record number R3.

  Next, the capacity calculation unit 44 applies the acquired total number of records R1 and the number of records R3 whose status is “unprocessed” to the above-described formula (1) to obtain the processing capacity A of other AP servers. Calculate (step A8).

  Next, the job number calculation unit 45 calculates α (= A−B) by using the simultaneously executable job number B calculated in step A7 and the processing capability A of the other AP server calculated in step A8. calculate. Then, based on the calculated value of α, the job number calculation unit 45 determines the number of jobs to be acquired next time by the control processing unit 43 using the above-described conditions (a) to (b) ( Step A9).

  When step A9 is executed, the job number calculation unit 45 transmits the set number of jobs to be acquired to the processing control unit 43. Thereby, the process control unit 43 executes Step A2 again. Specifically, the process control unit 43 issues an SQL statement and acquires as many records 31 as the number of jobs to be acquired.

  If the result of determination in step A5 is not more than the appropriate value, the resource monitoring unit 47 sends a message to the processing control unit 43 so as to resume issuance of the SQL statement for which issuance is suspended. (Step A10). If there is no SQL statement that has been suspended, the process of step A10 is omitted.

  Next, the resource monitoring unit 47 determines whether the resource usage status of the AP server 40 is equal to or lower than the upper limit value (second condition) (step A11). As a result of the determination in step A11, when the resource usage status does not exceed the upper limit value, the resource monitoring unit 47 instructs the processing control unit 43 to execute step A2 again. Thereby, the process control unit 43 accesses the DB server 30 and acquires the record 31 whose status is “unprocessed” and which is not locked by a transaction of another AP server 40.

  On the other hand, if the resource usage status of the AP server 40 is not less than or equal to the upper limit value (if the upper limit value is exceeded) as a result of the determination in step A11, the resource monitoring unit 47 passes the business process execution application 42 through the process control unit 43. In response to this, a message is transmitted so that the job being executed is suspended in the order of execution start time (step A12).

  In addition, after the execution of step A12, the monitoring unit 47 further processes a message indicating that the suspended job is resumed in order from the first suspended one when the resource usage status of the AP server 40 falls below the upper limit value. The data is transmitted to the business process execution application 42 through the control unit 43. Thereafter, the resource monitoring unit 47 instructs the processing control unit 43 to execute Step A2 again.

[Effects of the embodiment]
As described above, in the present embodiment, each AP server 40 dynamically monitors the resource usage status without using a load balancer. In addition, each AP server performs not only monitoring based on the upper limit value of the resource usage status but also monitoring based on the appropriate value of the resource usage status, so that only a temporary decrease in processing speed and suppression of occurrence of overflow can be achieved. In addition, the load can be distributed throughout the system.

[Modification]
In the example shown in FIGS. 1 to 3 described above, the computer system 100 includes the DB server 30, but the present embodiment is not limited to this example. The computer system according to the present embodiment may include a relay server having a function of relaying a processing request received from the client terminal 10 instead of the DB server 30. In this case, each AP server 40 accesses the shared memory of the relay server instead of accessing the DB server 30 and acquires a job. Except this point, it is the same as the example shown in FIGS.

[program]
The program in the embodiment of the present invention may be a program that causes a computer to execute steps A2 to A12 shown in FIG. By installing and executing this program on a computer, the AP server in the present embodiment can be realized. In this case, the CPU (Central Processing Unit) of the computer functions as a processing control unit 43, a capacity calculation unit 44, a job number calculation unit 45, an input reception unit 46, a resource monitoring unit 47, and a resource information acquisition unit 48, To do.

  Here, a computer that realizes the AP server 40 by executing the program according to the present embodiment will be described with reference to FIG. FIG. 4 is a block diagram illustrating an example of a computer that implements the server device according to the embodiment of the present invention.

  As shown in FIG. 4, the computer 110 includes a CPU 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These units are connected to each other via a bus 121 so that data communication is possible.

  The CPU 111 performs various calculations by developing the program (code) in the present embodiment stored in the storage device 113 in the main memory 112 and executing them in a predetermined order. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Further, the program in the present embodiment is provided in a state of being stored in a computer-readable recording medium 120. Note that the program in the present embodiment may be distributed on the Internet connected via the communication interface 117.

  Specific examples of the storage device 113 include a hard disk and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and a mouse. The display controller 115 is connected to the display device 119 and controls display on the display device 119.

  The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, and reads a program from the recording medium 120 and writes a processing result in the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and another computer.

  Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital), magnetic storage media such as a flexible disk, or CD- An optical storage medium such as ROM (Compact Disk Read Only Memory) can be used.

  As described above, according to the present invention, in a system constituted by a plurality of server devices, it is possible to implement load distribution that dynamically considers the resource usage status of each server device. The present invention can be used in a system including a Web application server, and is particularly useful in the SI department of the information communication industry that constructs and maintains such a system.

10 Client Terminal 20 Web Server 30 DB Server 31 Record 32 Page 33 Database 40 AP Server (Server Device)
DESCRIPTION OF SYMBOLS 41 Monitoring processing part 42 Business process execution application 43 Processing control part 44 Capacity calculation part 45 Job number calculation part 46 Input reception part 47 Resource monitoring part 48 Resource information acquisition part 100 Computer system 110 Computer 111 CPU
112 Main Memory 113 Storage Device 114 Input Interface 115 Display Controller 116 Data Reader / Writer 117 Communication Interface 118 Input Device 119 Display Device 120 Recording Medium 121 Bus

Claims (7)

A server device that constitutes a system that provides a service in response to an external processing request, and that executes an application program for providing the service,
A resource information acquisition unit that acquires resource information for identifying the resource usage status of the server device;
From the database server that holds the record in which the processing request is written, the record whose status is not processed is acquired as a job to be processed by the application program. A processing control unit that obtains the total number of records and the total number of records whose status is unprocessed, and passes the acquired job to the application program for execution;
Let the processing control unit acquire the job to be processed so that the usage status of the resource specified by the acquired resource information is closest to the appropriate value that is the first condition. And
When the usage status of the resource specified by the acquired resource information exceeds an upper limit that is a second condition, the usage status of the resource approaches the appropriate value for the processing control unit. A resource monitoring unit that causes the application program to suspend execution of the job;
From the total number of records, the value obtained by subtracting the number of records processed by the server device and the total number of records whose status is unprocessed is obtained from other server devices constituting the system. Ability calculation part that calculates as ability,
The calculated capability of the other server device is compared with the number of records that can be executed simultaneously by the server device, which is the capability of the server device. A job number calculation unit for calculating the number of jobs when acquiring
The server apparatus characterized by the above-mentioned. The resource information is the usage status of the resource, the CPU usage rate, the memory usage rate, the swap memory usage rate, the number of open files, the number of open threads, the usage amount of the storage device provided in the server device, and Identify at least one of the number of processes that have the same request source and are running simultaneously,
The server device according to claim 1. A system that provides services in response to processing requests from outside,
A plurality of server devices for executing an application program for providing the service;
Each of the plurality of server devices includes:
A resource information acquisition unit that acquires resource information for identifying the resource usage status of the server device;
From the database server that holds the record in which the processing request is written, the record whose status is not processed is acquired as a job to be processed by the application program. A processing control unit that obtains the total number of records and the total number of records whose status is unprocessed, and passes the acquired job to the application program for execution;
Let the processing control unit acquire the job to be processed so that the usage status of the resource specified by the acquired resource information is closest to the appropriate value that is the first condition. And
When the usage status of the resource specified by the acquired resource information exceeds an upper limit that is a second condition, the usage status of the resource approaches the appropriate value for the processing control unit. A resource monitoring unit that causes the application program to suspend execution of the job;
From the total number of records, the value obtained by subtracting the number of records processed by the server device and the total number of records whose status is unprocessed is obtained from other server devices constituting the system. Ability calculation part that calculates as ability,
The calculated capability of the other server device is compared with the number of records that can be executed simultaneously by the server device, which is the capability of the server device. A job number calculation unit for calculating the number of jobs when acquiring
A computer system comprising: A method for reducing a load on a system that provides a service in response to an external processing request,
(A) acquiring resource information for specifying a resource usage status of the server device by the server device that configures the system and executes an application program for providing the service; and
(B) A processing request is written by the server device so that the usage status of the resource specified by the resource information acquired in step (a) is closest to the appropriate value that is the first condition. From the database server that holds the recorded records, the records whose status is unprocessed are acquired as jobs to be processed by the application program. At that time, from the database server, the total number of records , Get the total number of records whose status is unprocessed,
(C) When the usage status of the resource specified by the resource information acquired in the step (a) exceeds the upper limit that is the second condition, the usage status of the resource is determined by the server device. Suspending the execution of the job in the application program so as to approach the appropriate value; and
(D) A value obtained by subtracting, from the total number of records by the server device, the number of records processed by the server device and the total number of records whose status is not processed, Calculating as the capacity of other server devices constituting
(E) The server device compares the capacity of the other server device calculated in the step (d) with the number of records that can be executed simultaneously by the server device, which is the capacity of the server device; Calculating the number of jobs when acquiring the jobs in the next step (b) according to the comparison result; and
Have
The load distribution method, wherein the step (d) and the step (e) are executed after the steps (a) to (c) are executed at least once. The resource information is the usage status of the resource, the CPU usage rate, the memory usage rate, the swap memory usage rate, the number of open files, the number of open threads, the usage amount of the storage device provided in the server device, and Identify at least one of the number of processes that have the same request source and are running simultaneously,
The load distribution method according to claim 4. A computer that configures a system that provides a service in response to an external processing request and that executes an application program for providing the service.
(A) obtaining resource information for identifying resource usage status of the computer; and
(B) Holds a record in which a processing request is written so that the usage status of the resource specified by the resource information acquired in step (a) is closest to the appropriate value that is the first condition The record whose status is unprocessed is acquired from the database server as a job to be processed by the application program, and at that time, the total number of records and the status are unprocessed from the database server. The total number of records that are
(C) When the usage status of the resource specified by the resource information acquired in the step (a) exceeds an upper limit that is a second condition, the usage status of the resource approaches the appropriate value. Suspending the execution of the job in the application program,
(D) the total number of pre-Symbol record, the number of records that the computer has processed, the status is the total number of the record to untreated, a value obtained by subtracting, in addition to constituting the system Calculating as computer capacity, steps,
(E) the previous SL (d) the ability of the other computer calculated in step, is the ability of the computer compares the number of records the computer can run simultaneously, and depending on the comparison result, the following A step of calculating the number of the jobs when acquiring the jobs in the step of (b).
And execute
A program for executing the step (d) and the step (e) after the steps (a) to (c) are executed at least once. The resource information includes the usage status of the resource as the CPU usage rate, the memory usage rate, the swap memory usage rate, the number of open files, the number of open threads, the usage amount of the storage device provided in the computer , and the request. Identify at least one of the number of processes that are originally the same and are running simultaneously;
The program according to claim 6.

Images