WO2013047339A1 - Load balancing device, method and program - Google Patents

Abstract

This load balancing device is provided with means for: receiving data which is to any of one or more servers from a terminal; assessing whether to transfer data to any of the servers or to store the same on the basis of server load information and/or priority of the data; transmitting the data to any of the servers and transmitting an ACK to the terminal if the data have been assessed to be transferred; storing the data and responding with an ACK to the terminal if the data have been assessed to be stored; determining the timing to transmit the stored data to any of the servers on the basis of the server load information and the communication status; and transmitting the stored data to any of the servers on the basis of the determined timing and transmitting an ACK to the terminal.

Description

Load balancing apparatus, method and program

The present invention relates to a load distribution apparatus, method, and program.

In recent years, in addition to personal computers and servers, various devices such as home appliances and sensors are connected to the Internet, and various services are being developed. Among services using information from devices such as sensors and home appliances, those using sensors include services such as observation of weather information (temperature and humidity, etc.) and security. For devices intended for home appliances, services such as remote maintenance can be cited. In such various services, data collection is the main. The number of devices such as sensors and home appliances is enormous, and it is important to ensure the scalability of servers for collecting information. Server redundancy has been implemented to ensure server scalability. A load balancer is deployed in front of the server, and access from terminals is distributed to multiple servers to improve scalability. If access becomes enormous and server resources are insufficient, it is also scaled out.

For example, Patent Document 1 discloses a load distribution control device that stores a received request in a buffer and transfers the request to a server whose response waiting request number does not exceed a threshold value.

JP 2008-040718 A

Scalability can be improved by increasing server redundancy and server resources. However, from the viewpoint of cost, if the number of servers is increased in redundant server operation, the cost increases.

Also, the load distribution device of Patent Document 1 is inefficient because all received requests are temporarily stored in a buffer and sequentially processed even when the server load is low. In addition, since data is processed in the order stored in the buffer, data with low priority is processed first, and processing of data with high priority may be delayed.

The present invention has been made in view of the above problems, and an object thereof is to provide a load distribution apparatus, method, and program capable of performing more efficient load distribution processing based on server load and data priority. And

The present invention relates to a means for receiving data from one of one or more servers from a terminal, and to transfer the data to one of the servers based on at least one of server load information and the priority of the data. If it is determined that the data is transferred with a determination unit that determines whether to store the data, the data is transmitted to any server, and a confirmation response is transmitted to the terminal. Storage means for storing the data and returning a confirmation response to the terminal; timing determination means for determining a timing for transmitting the stored data to any one of the servers based on the server load information and the communication status; , Means for transmitting the accumulated data to any server based on the determined timing, and transmitting an acknowledgment to the terminal A load balancer, characterized in that it comprises.

The present invention receives data from one of one or more servers from a terminal and transfers or stores the data to any server based on at least one of server load information and the priority of the data. Determining whether to transfer the data, transmitting the data to any server, transmitting an acknowledgment to the terminal, and determining to store the data, storing the data, An acknowledgment is sent back to the terminal, the timing for transmitting the accumulated data to any server is determined based on the server load information and the communication status, and the accumulated is determined based on the determined timing. The load distribution method is characterized in that data is transmitted to any one of the servers, and an acknowledgment is transmitted to the terminal.

According to the present invention, the computer transfers the data to any server based on at least one of the process of receiving data from the terminal to one of the servers, the server load information, and the priority of the data. Determination process for determining whether to store or to store, if it is determined to transfer the data, the process of transmitting the data to any server and transmitting an acknowledgment to the terminal, the case of determining to store the data A storage process for storing the data and returning a confirmation response to the terminal; a timing determination process for determining a timing for transmitting the stored data to one of the servers based on the server load information and the communication status; The accumulated data is transmitted to any server based on the determined timing, and an acknowledgment is transmitted to the terminal Is a program for causing to execute that acknowledgment transmission process.

According to the present invention, more efficient load distribution processing can be performed based on server load and data priority.

FIG. 1 is a schematic diagram of a data processing system including a load distribution apparatus according to an embodiment of the present invention. FIG. 2 is a functional block diagram of the terminal, the load balancer, and the server. FIG. 3 is a diagram in which the correspondence between the number of communication sessions and the CPU usage rate is plotted. FIG. 4 is a plot of the correspondence between the number of communication sessions and the memory usage rate. FIG. 5 is a graph plotting the correspondence between the number of communication sessions and the number of disk I / Os. FIG. 6 is a diagram showing an operation sequence when receiving a data transmission request from a terminal and transferring data to a server. FIG. 7 is a flowchart for determining whether data is transferred to the server or temporarily stored in the load balancer. FIG. 8 is a diagram illustrating an operation sequence when a data transmission request is received from a terminal and data is temporarily stored in the load balancer. FIG. 9 is a diagram showing an operation sequence in a case where data temporarily stored in the load balancer is transmitted to the server.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic diagram of a data processing system including a load distribution apparatus according to an embodiment of the present invention. This data processing system includes one or more terminals 1 ₁ to 1 _m , a load balancer 10, and one or more servers 100 ₁ to 100n.

Terminal 1 ₁ to terminal 1 _m transmit data to server 100 ₁ to server 100n. The terminal 1 ₁ to the terminal 1 _m transmit priority information such as the importance of data and the delay constraint information related to the data arrival delay constraint to the server to the load distribution device 10 at the time of data transmission request. The load distribution apparatus 10 that has received data from the terminals 1 ₁ to 1 _m selects and selects the server 100 that is the data transfer destination based on the priority of the data indicated by the priority information and the load state of the server. Data is transferred to the server 100. Further, if the load distribution apparatus 10 receives and accumulates data from the terminal 1 according to the load on the server and confirms that the load on the server 100 ₁ to the server 100n is below a certain level, the server server 100 ₁ to Data is transmitted to the server 100n.

FIG. 2 is a functional block diagram of the terminal 1, the load balancer 10, and the server 100.

The terminal 1 includes a data transmission unit 2 and a data storage unit 3. The data transmission unit 2 acquires data from the data storage unit 3 and transmits a data transmission request message to the load balancer 10 in order to transmit data to any of the servers 100. The data transmission unit 2 includes a data amount confirmation unit 21 and a priority determination unit 22. The data amount confirmation unit 21 includes the data amount to be transmitted in the data transmission request message. The priority determination unit 22 includes priority information indicating the priority of data, such as the importance of data and delay constraints (for example, allowable delay time), in the data transmission request message. The data storage unit 3 stores data to be transmitted.

The load distribution apparatus 10 includes a communication session reception unit 11, a server load monitoring unit 12, a transfer destination server determination unit 13, a data transfer unit 14, a session state monitoring unit 15, a data reception unit 16, a data storage unit 17, and a data transmission unit 18. Is provided.

The communication session acceptance unit 11 includes a priority confirmation unit 111 and a data amount confirmation unit 112. The priority confirmation unit 111 confirms data priority information included in the received request message. The data amount confirmation unit 112 confirms information on the amount of data included in the received request message. The communication session reception unit 11 receives a data transmission request from the terminal 1. The communication session accepting unit 11 acquires the latest session number and session number threshold value of the servers 100 ₁ to 100n acquired by the server load monitoring unit 12, and obtains the data priority and the latest session number of the server 100 ₁ to server 100n. Based on the session number threshold, whether to transfer data to one of the servers 100 or to store the data is determined.

In this determination, when the priority of data is high, such as when the latest number of sessions is smaller than the session number threshold or the data arrival delay time to the server is severe, the communication session accepting unit 11 transfers the data to the server 100. A data transfer instruction including the identification information of the terminal 1 that has decided to perform the data transmission request is sent to the data transfer unit 14. When performing data transfer, the server load monitoring unit 12 is also notified of data transfer.

In addition, when the latest number of sessions exceeds the session number threshold, the delay constraint is not strict, and the priority of data is low, the communication session accepting unit 11 determines to accumulate data, and receives and accumulates data. The data receiving unit 16 is instructed to do so.

Further, when the communication session accepting unit 11 confirms the data amount information included in the received data transmission request, the communication session accepting unit 11 notifies the session state monitoring unit 15 of the terminal identification information and the data amount.

The server load monitoring unit 12 acquires the number of communication sessions established in the server 100 and server load information, obtains the correspondence between the number of communication sessions and the server load, and based on this, the server 100 monitors the data It is determined whether or not data transfer should be performed in consideration of the priority. The server load information acquired from the server 100 ₁ to the server 100n is a CPU usage rate, a memory usage amount, and a disk I / O number indicating the number of times data is read from and written to the disk. Further, the number of sessions with which each server 100 is communicating is acquired from the session state monitoring unit 15. The method for acquiring the server load information and the number of communication sessions can be arbitrarily set. For example, it may be acquired when a data transmission request from the terminal 1 is received, may be acquired at regular intervals, or may be acquired according to the number of communication sessions. The acquired data may be accumulated and stored as measurement data for a certain period of time. The server load information is related to the number of communication sessions when the server load occurs. You may make it relate server load information and the number of communication sessions based on the acquisition time etc. of each information.

As shown in FIGS. 3, 4, and 5, the server load monitoring unit 12 determines the number of communication sessions, the CPU usage rate, and the memory for each server based on the acquired server load information and the corresponding number of communication sessions. The correspondence between the usage rate and the number of disk I / Os is plotted. The CPU usage rate, memory usage rate, and disk I / O count threshold values are set such that high priority data can be transferred to any of the servers 100 ₁ to 100 n. This value may be calculated by the load balancer or entered by the user. The server load monitoring unit 12 obtains a threshold value of the number of communication sessions corresponding to each determined threshold value.

As shown in FIG. 3, FIG. 4, and FIG. 5, the threshold for the number of communication sessions is obtained by plotting the relationship between the number of communication sessions and the CPU usage rate, and by plotting the relationship between the number of communication sessions and the memory usage rate. , Plotting the relationship between the number of communication sessions and the number of disk I / Os, an approximate curve is added to each of them, and the number of sessions determined by the intersection of the CPU usage rate, the memory usage rate, and the disk I / O threshold is calculated. It is defined as a threshold value. Also, a method of taking a minimum value or an average value within the range of the number of sessions that can be taken by the CPU usage rate, the memory usage rate, and the disk I / O thresholds may be used.

The server load monitoring unit 12 notifies the communication session reception unit 11 of the threshold value of the number of sessions and the number of communication sessions when the data transmission request from the terminal 1 is received. When the communication session reception unit 11 determines to transfer data to any of the servers 100 ₁ to 100n, the communication session reception unit 11 receives a notification of data transfer. Upon receiving this notification, the server load monitoring unit 12 notifies the transfer destination server determining unit 13 of the relationship between the number of communication sessions and the server load information. In addition, the server load monitoring unit 12 notifies the session state monitoring unit 15 of the relationship between the number of communication sessions and server load information.

The transfer destination server determination unit 13 acquires the number of communication sessions and server load information notified from the server load monitoring unit 12 and determines the transfer destination server. The method for determining the transfer destination server is as follows. The latest number of communication sessions of each server, server load information (CPU usage rate, memory usage rate, disk I / O number) acquired from the server load monitoring unit 12 and shown in FIG. 3, FIG. 4, and FIG. From the data indicating the relationship between the number of communication sessions and server load information (CPU usage rate, memory usage rate, disk I / O count), the latest communication session count does not exceed the session count threshold, and the CPU usage rate and memory usage A server in which the rate and the number of disk I / Os do not exceed the CPU threshold, the memory usage threshold, and the disk I / O threshold is selected. When there are a plurality of corresponding servers, the server with the lowest load may be selected as the transfer destination. When the transfer destination server is determined, the transfer destination server determination unit 13 notifies the data transfer unit 14 of the transfer server identification information.

The data transfer unit 14 receives a data transfer instruction from the communication session reception unit 11, acquires the identification information of the transmission source terminal, and acquires the identification information of the transfer destination server from the transfer destination server determination unit 13. The data transfer unit 14 relays communication between the server and the terminal using both pieces of identification information. Further, the data transfer start time and data transfer amount of each communication are held.

The session state monitoring unit 15 calculates the end time of each communication session performing data transfer from the data amount, the data transfer amount, the data transfer start time, and the current time, and acquires the number of communication sessions acquired from the server load monitoring unit 12 Based on the data indicating the relationship of the server load information, a timing at which the server load (CPU usage rate, memory usage rate, disk I / O) becomes equal to or less than a threshold value is calculated. An instruction is given to transmit the data stored in the data storage unit 17 at the calculated timing. For example, the above timing may be calculated for each server, and the server may be instructed to transmit data to the server with the earliest calculated timing.

When the data receiving unit 16 receives an instruction to receive and store data from the communication session receiving unit 11, the data receiving unit 16 receives the data and stores it in the data storage unit 17 in association with the identification information of the terminal. When the data transmission from the terminal is completed, the load distribution apparatus 10 transmits a confirmation response indicating that the temporary data is accumulated to the terminal.

The data accumulating unit 17 accumulates data received from the data receiving unit 16.

The data transmission unit 18 transmits data according to the data transmission timing instructed from the session state monitoring unit 15. After the data transmission is completed, a confirmation response indicating that the data transmission to the server is completed is transmitted to the terminal.

The server 100 includes a data receiving unit 101, a data storage unit 102, and a server load monitoring unit 103. The data reception unit 101 receives data transmitted from the load distribution apparatus 10 and accumulates it in the data accumulation unit 102. The data accumulation unit 102 accumulates data received from the data reception unit 101. The server load monitoring unit 103 notifies the CPU usage rate, the memory usage rate, and the number of disk I / Os in response to a request from the load distribution apparatus 10.

An outline of the operation of the data processing system according to the embodiment of the present invention will be described. First, a data transmission request is transmitted from the terminal 1 ₁ to the terminal 1 _m to the load distribution apparatus 10. When receiving the data transmission request, the load balancer 10 determines whether to transfer the data to any of the servers 100 ₁ to 100 n or to store the data in the load balancer 10 based on the priority of the data and the server load information. . When determining that the data is to be transferred, the load distribution apparatus 10 determines a transfer destination server and transfers the data. When it is determined that the data is stored in the load distribution device 10, the load distribution device 10 stores temporary data. When the data is accumulated, the load distribution apparatus 10 transmits a confirmation response indicating that the temporary data is accumulated to the terminal. Then, the load balancer 10 calculates the timing at which the server load (CPU usage rate, memory usage rate, disk I / O) falls below a certain value, and stores the data stored in the load balancing device 10 at the calculated timing. It is transmitted to any _{one of the} servers 100 ₁ to 100n. When the data transmission is completed, an acknowledgment indicating that the data transmission is completed is transmitted.

Next, the operation when the load distribution apparatus 10 receives a data transmission request from a terminal and transfers data to the server will be described with reference to the operation sequence of FIG. 6 and the flowchart of FIG. The flowchart of FIG. 7 shows a process in which the load balancer 10 receives a data transmission request from a terminal and determines whether to transfer data to a server or temporarily store it in the load balancer 10.

There is data transmission request from the terminal 1 _1, illustrating a case where the data is sent to either server 100 as an example. The operation is the same when there is a data transmission request from a plurality of terminals 1 ₁ to 1 _m .

First, when the communication session accepting unit 11 of the load distribution apparatus 10 receives a data transmission request from the terminal 1 (step S1 in FIG. 6 and step A in FIG. 7), the server load monitoring unit 12 of the load distribution apparatus 10 receives the data transmission request for each server. The latest session number and the session number threshold value are acquired and compared (step S2 in FIG. 6 and step B in FIG. 7). When it is confirmed that the latest number of communication sessions is smaller than the communication session number threshold for any of the servers, it is determined to transfer data to the corresponding server (step S4 in FIG. 6, step C in FIG. 7).

In addition, if the latest number of communication sessions exceeds the threshold value of the number of communication sessions for any server, and if the priority of the data is equal to or higher than a certain level (step S3 in FIG. 6, step D in FIG. 7), to the server Data transfer is determined (step 4 in FIG. 6 and step F in FIG. 7). The communication session accepting unit 11 notifies the server load monitoring unit 12 that data is transferred.

Upon receiving this notification, the server load monitoring unit 12 sends the latest communication session count, server load information (CPU usage rate, memory usage rate, disk I / O) to the transfer destination server determination unit 13, FIG. 3, FIG. 4, FIG. Data indicating the relationship between the number of communication sessions and server load information as shown in FIG.

The transfer destination server determination unit 13 sets a server as a data transfer destination based on the notified server load information. In the example shown in FIG. 6, it is selected the server 100 ₁ as the transfer destination (step S5 in FIG. 6). After server selection, load balancer 10 transfers the data from the terminal 1 to the server 100 ₁ (step S6 in FIG. 6).

Next, with reference to the operation sequence of FIG. 8 and the flowchart of FIG. 7, the operation of the load balancer 10 when receiving a data transmission request from a terminal and temporarily storing data in the load balancer 10 will be described. explain.

There is data transmission request from the terminal 1 _1, a case for temporarily storing the data at the load balancer 10 as an example. The operation is the same when there is a data transmission request from a plurality of terminals 1 ₁ to 1 _m .

First, when the communication session accepting unit 11 of the load distribution apparatus 10 receives a data transmission request from the terminal 1 (step S1 in FIG. 8 and step A in FIG. 7), the server load monitoring unit 12 of the load distribution apparatus 10 receives a request for each server. The latest session number and the session number threshold value are acquired and compared (step S2 in FIG. 8, step B in FIG. 7). In the server 100 ₁ to the server 100n, when it is confirmed that the latest number of communication sessions exceeds the communication session number threshold, the priority of data is confirmed (step S3 in FIG. 4 and step D in FIG. 7). When it is confirmed that the priority of the data is equal to or less than a certain value, it is determined that the data is stored in the load balancer 10 (step S4 in FIG. 8, step E in FIG. 7). When data from the terminal 1 is received and data transmission is completed, a confirmation response indicating that temporary data is stored in the load balancer is returned from the data receiver 16 of the load balancer 10 (step S5 in FIG. 8).

Next, the operation in the case where the load distribution apparatus 10 transmits data temporarily stored in the load distribution apparatus 10 to the server will be described with reference to the operation sequence of FIG.

After the data receiving unit 16 of the load distribution apparatus 10 accumulates data in the data accumulating unit 17, the session state monitoring unit 15 acquires the amount of data being transferred from the communication session accepting unit 11, and monitors the server load. The communication session number threshold is acquired from the unit 12, and the data transfer start time and data transfer amount of each communication are acquired from the data transfer unit 14 (step S1 in FIG. 9). After that, calculate the time required for the transfer from the data transfer amount of each communication, the current time, and the data transfer start time, calculate the time to transfer the remaining data of each communication based on that time, and calculate the server load (number of communication sessions) Etc.) below the communication session threshold (step S2 in FIG. 9).

The timing is calculated by calculating the remaining data transfer time based on the data transfer amount of each communication and the time required for the transfer. For example, if the data amount is 10 MB, the current data transfer amount is 4 MB, and the transfer time is 10 seconds, the remaining 6 MB data transfer time is 15 seconds (= 10 seconds × 6 ÷ 4). This calculation is performed in each communication, the time below the communication session number threshold of the server is obtained, and the data transmission timing is determined. In the example of FIG. 9, the server 100 _1, it determines the data transmission timing. For transmission timing determination, a server is selected in consideration of delay constraints of accumulated data. Data transmission unit 18 at the timing of the determined data transmission, transmits the data to the server 100 ₁ acquires data from the data storage unit 17 (step S3 in FIG. 9). When the data transmission is completed, a confirmation response indicating that the data transmission to the server is completed is transmitted to the terminal 1 that is the transmission source of the accumulated data (step S4 in FIG. 9).

As described above, according to the present embodiment, when the load on the server increases, the data is temporarily accumulated in the load balancer according to the priority of the data without increasing the server resources, and the server By transmitting data to the server when the load decreases, it is possible to effectively use server resources. Also, when temporarily accumulating in the load balancer, the terminal side even if data is lost in the load balancer by sending an acknowledgment to the terminal indicating that it is temporarily accumulated in the load balancer Thus, the reliability of data communication is improved.

The communication session reception unit 11, the server load monitoring unit 12, the transfer destination server determination unit 13, the data transfer unit 14, the session state monitoring unit 15, the data reception unit 16, the data of the load distribution apparatus 10 according to the embodiment of the present invention described above. The storage unit 17 and the data transmission unit 18 may be realized by a CPU (Central （Processing の Unit) of the load distribution apparatus 10 reading and executing an operation program or the like stored in the storage unit, or configured by hardware. May be. Only some functions of the above-described embodiments can be realized by a computer program.

Some or all of the above embodiments can be described as in the following supplementary notes, but are not limited thereto.

(Appendix 1)
Means for receiving data from the terminal to any of the one or more servers;
Based on at least one of server load information and priority of the data, determination means for determining whether to transfer or store the data to any server, and when determining that the data is to be transferred, Means for transmitting to the server and transmitting an acknowledgment to the terminal;
If it is determined to store the data, storage means for storing the data and returning a confirmation response to the terminal;
Timing determining means for determining a timing for transmitting the accumulated data to any server based on the server load information and communication status;
A load distribution apparatus comprising: an acknowledgment transmission means for transmitting the accumulated data to any server based on the determined timing and transmitting an acknowledgment to the terminal.

(Appendix 2)
The load distribution apparatus according to appendix 1, wherein the determination unit uses at least one of a CPU usage rate, a memory usage rate, and a disk I / O number of each server as the server load information.

(Appendix 3)
For each server, a threshold is set for at least one of the CPU usage rate, the memory usage rate, and the number of disk I / Os, and the communication session of the server when the set threshold value is acquired is acquired. Means further comprising a threshold,
The determination means compares the threshold value of the communication session with the number of communication sessions acquired for each server, and determines whether to transfer or store the data to any server. The load balancer described in 1.

(Appendix 4)
The load balancing apparatus according to any one of appendices 1 to 3, wherein the determination unit determines to transfer the data to any server when the priority of the data exceeds a predetermined value.

(Appendix 5)
The determination unit determines that the data is to be accumulated when the priority of the data does not exceed a predetermined value and the server load based on the server load information exceeds a predetermined value. 5. The load distribution apparatus according to any one of items 1 to 4.

(Appendix 6)
The confirmation response sent to the terminal by the storage means indicates that data has been stored;
6. The load balancer according to any one of appendices 1 to 5, wherein

(Appendix 7)
The timing determination unit calculates a data transfer end time based on the data amount of the data, the transferred data amount, and the time required for the transfer, and the server load is smaller than a predetermined value based on the calculated data transfer end time. The load distribution apparatus according to any one of appendices 1 to 6, wherein the estimated time is determined as a timing at which the accumulated data is transmitted to a server.

(Appendix 8)
Any one of Supplementary notes 1 to 7, wherein the confirmation response transmission means transmits the confirmation response indicating that data transmission to any one of the one or more servers is completed to the terminal. The load balancer according to claim 1.

(Appendix 9)
Receiving data from the terminal to any of the one or more servers;
A determination step of determining whether to transfer or store the data to any server based on at least one of server load information and priority of the data;
If it is determined to transfer the data, transmitting the data to any server and transmitting an acknowledgment to the terminal;
If it is determined to store the data, the storage step of storing the data and returning a confirmation response to the terminal;
A timing determining step for determining a timing for transmitting the accumulated data to any of the servers based on the server load information and the communication status;
A load distribution method comprising: an acknowledgment transmission step of transmitting the accumulated data to any of the servers based on the determined timing and transmitting an acknowledgment to the terminal.

(Appendix 10)
The load balancing method according to claim 9, wherein the determination step uses at least one of a CPU usage rate, a memory usage rate, and a disk I / O number of each server as the server load information.

(Appendix 11)
For each server, a threshold is set for at least one of the CPU usage rate, the memory usage rate, and the number of disk I / Os, and the communication session of the server when the set threshold value is acquired is acquired. And further comprising a step of setting a threshold,
The determination step compares the threshold value of the communication session with the number of communication sessions acquired for each server, and determines whether to transfer or store the data to any server. Load balancing method described in 1.

(Appendix 12)
The load balancing method according to any one of appendices 9 to 11, wherein the determination step determines that the data is transferred to any server when the priority of the data exceeds a predetermined value.

(Appendix 13)
The determining step determines that the data is to be accumulated when the priority of the data does not exceed a predetermined value and the server load based on the server load information exceeds a predetermined value. 13. The load distribution method according to any one of items 1 to 12.

(Appendix 14)
The confirmation response transmitted to the terminal by the accumulation step indicates that data has been accumulated.
14. The load distribution method according to any one of appendices 9 to 13, characterized in that:

(Appendix 15)
The timing determination step calculates a data transfer end time based on the data amount of the data, the transferred data amount, and the time required for the transfer, and the server load is smaller than a predetermined value based on the calculated data transfer end time. The load balancing method according to any one of appendices 9 to 14, wherein the estimated time is estimated and the estimated time is determined as a timing for transmitting the accumulated data to a server.

(Appendix 16)
Any one of Supplementary notes 9 to 15, wherein the confirmation response transmitting step transmits the confirmation response indicating that data transmission to any one of the one or more servers is completed to the terminal. The load distribution method according to claim 1.

(Appendix 17)
On the computer,
Processing to receive data from a terminal to any of one or more servers,
A determination process for determining whether to transfer or store the data to any server based on at least one of server load information and the priority of the data;
When it is determined that the data is to be transferred, a process of transmitting the data to any server and transmitting an acknowledgment to the terminal;
If it is determined that the data is to be stored, the data is stored, and a storage process for returning a confirmation response to the terminal;
A timing determination process for determining a timing for transmitting the accumulated data to any server based on the server load information and the communication status;
A program that transmits the accumulated data to any server based on the determined timing, and executes an acknowledgment transmission process for transmitting an acknowledgment to the terminal.

(Appendix 18)
The program according to claim 17, wherein the determination process uses at least one of a CPU usage rate, a memory usage rate, and a disk I / O number of each server as the server load information.

(Appendix 19)
For each server, a threshold is set for at least one of the CPU usage rate, the memory usage rate, and the number of disk I / Os, and the communication session of the server when the set threshold value is acquired is acquired. Further execute the process of setting the threshold,
The determination process compares the threshold value of the communication session with the number of communication sessions acquired for each server, and determines whether the data is transferred to any server or stored. The program described in.

(Appendix 20)
The program according to any one of appendices 17 to 19, wherein the determination processing determines that the data is transferred to any server when the priority of the data exceeds a predetermined value.

(Appendix 21)
The determination process is performed when the priority of the data does not exceed a predetermined value, and when the server load based on the server load information exceeds a predetermined value, it is determined to store the data. 21. The program according to any one of 1 to 20.

(Appendix 22)
The confirmation response transmitted to the terminal by the accumulation process indicates that data has been accumulated.
The program according to any one of appendices 17 to 21, characterized in that:

(Appendix 23)
The timing determination process calculates a data transfer end time based on the data amount of the data, the transferred data amount, and the time required for the transfer, and the server load is smaller than a predetermined value based on the calculated data transfer end time. 23. The program according to any one of appendices 17 to 22, wherein the program is estimated as a timing for transmitting the accumulated data to a server.

(Appendix 24)
Any one of appendices 17 to 23, wherein the confirmation response transmission process transmits the confirmation response indicating completion of data transmission to any one of the one or more servers to the terminal. The program according to item 1.

Although the present invention has been described with reference to the preferred embodiments, the present invention is not necessarily limited to the above-described embodiments, and various modifications can be made within the scope of the technical idea.
This application claims the priority on the basis of Japanese application Japanese Patent Application No. 2011-208510 for which it applied on September 26, 2011, and takes in those the indications of all here.

DESCRIPTION OF SYMBOLS 1 Terminal 2 Data transmission part 3 Data storage part 10 Load distribution apparatus 11 Communication session reception part 12 Server load monitoring part 13 Transfer destination server determination part 14 Data transfer part 15 Session state monitoring part 16 Data reception part 17 Data storage part 18 Data transmission Unit 101 Data receiving unit 102 Data storage unit 103 Server load monitoring unit 111 Priority confirmation unit 112 Data amount confirmation unit

Claims (10)

Means for receiving data from the terminal to any of the one or more servers;
Based on at least one of server load information and priority of the data, determination means for determining whether to transfer or store the data to any server, and when determining that the data is to be transferred, Means for transmitting to the server and transmitting an acknowledgment to the terminal;
If it is determined to store the data, storage means for storing the data and returning a confirmation response to the terminal;
Timing determining means for determining a timing for transmitting the accumulated data to any server based on the server load information and communication status;
A load distribution apparatus comprising: an acknowledgment transmission means for transmitting the accumulated data to any server based on the determined timing and transmitting an acknowledgment to the terminal. 2. The load distribution apparatus according to claim 1, wherein the determination unit uses at least one of a CPU usage rate, a memory usage rate, and a disk I / O number of each server as the server load information. For each server, a threshold is set for at least one of the CPU usage rate, the memory usage rate, and the number of disk I / Os, and the communication session of the server when the set threshold value is acquired is acquired. Means further comprising a threshold,
The determination means compares the threshold value of the communication session with the number of communication sessions acquired for each server, and determines whether to transfer or store the data to any server. 2. The load balancer according to 2. 4. The load distribution apparatus according to claim 1, wherein the determination unit determines to transfer the data to any server when the priority of the data exceeds a predetermined value. 5. . The determination means determines that the data is to be accumulated when the priority of the data does not exceed a predetermined value and the server load based on the server load information exceeds a predetermined value. The load distribution apparatus according to any one of 1 to 4. The confirmation response sent to the terminal by the storage means indicates that data has been stored;
The load distribution apparatus according to any one of claims 1 to 5, wherein The timing determination unit calculates a data transfer end time based on the data amount of the data, the transferred data amount, and the time required for the transfer, and the server load is smaller than a predetermined value based on the calculated data transfer end time. The load balancing apparatus according to claim 1, wherein the estimated time is determined, and the estimated time is determined as a timing for transmitting the accumulated data to a server. 8. The confirmation response transmitting unit transmits the confirmation response indicating that data transmission to any one of the one or more servers is completed to the terminal. The load distribution apparatus according to any one of the above. Receive data from the terminal to any of one or more servers,
Based on at least one of server load information and the priority of the data, determine whether to transfer or store the data to any server,
If it is determined that the data is to be transferred, the data is sent to any server, an acknowledgment is sent to the terminal,
If it is determined that the data is to be stored, the data is stored and a confirmation response is returned to the terminal;
Based on the server load information and communication status, determine the timing to transmit the accumulated data to any server,
A load distribution method comprising: transmitting the accumulated data to any server based on the determined timing, and transmitting an acknowledgment to the terminal. On the computer,
Processing to receive data from a terminal to any of one or more servers,
A determination process for determining whether to transfer or store the data to any server based on at least one of server load information and the priority of the data;
When it is determined that the data is to be transferred, a process of transmitting the data to any server and transmitting an acknowledgment to the terminal;
If it is determined that the data is to be stored, the data is stored, and a storage process for returning a confirmation response to the terminal;
A timing determination process for determining a timing for transmitting the accumulated data to any server based on the server load information and the communication status;
A program that transmits the accumulated data to any server based on the determined timing, and executes an acknowledgment transmission process for transmitting an acknowledgment to the terminal.

Images