JP6551042B2 - Monitoring server, monitoring system, monitoring method, and program - Google Patents

Description

  The present invention relates to technology for monitoring a system composed of a plurality of servers.

  When any service is running on a system consisting of servers or server groups, when the input load from the outside to the system, its own processing load, or response performance is monitored and the load exceeds a certain level There is known a technology for performing measures for load reduction.

  Patent Document 1 discloses a technique for measuring the communication processing time of each communication program operating on a target system, and monitoring the load condition of the communication server based on the communication processing time. According to the technology disclosed in Patent Document 1, there is provided an alternative server operation request unit that automatically requests operation of the alternative server when the communication processing time here exceeds the upper limit value set in advance. Implement load reduction processing for monitoring and load increase.

  The technology disclosed in Patent Document 2 receives a telegram from each terminal, performs processing corresponding to the received telegram, and executes a plurality of servers for which reassignment of resources is executed according to a load change accompanying the reception of the telegram. Prepare. The technology disclosed in Patent Document 2 classifies a message received from each terminal based on an input classification table, and outputs the classified item as input time-series information for each classified item. The technology disclosed in Patent Document 2 is a load prediction that records input time-series information, load time-series information representing load change for each resource, and a predicted value of a load fluctuation amount applied to a server after a predetermined time corresponding to a message. Based on the rules, forecast the minimum usage for each resource. According to the technology disclosed in Patent Document 2, thereby, the technology described in Search Document 1 effectively allocates resources even if load fluctuation occurs, and maintains a service level such as response time at a predetermined value. it can.

Patent No. 3692571 JP 2005-128866 A

  When a significant reduction in response performance occurs in the server immediately, the techniques disclosed in Patent Literature 1 and Patent Literature 2 implement a load reduction measure after occurrence of degradation of response performance has started. However, since the server performance has already deteriorated, the techniques disclosed in Patent Document 1 and Patent Document 2 cannot prevent the server from adversely affecting the user and other systems in the above situation.

  The technique disclosed in Patent Document 1 measures the response performance of the service on the target system, and when it is attempted to reduce the load based on it, it is difficult to follow the rapid increase in the system load. .

  The technique disclosed in Patent Document 2 allocates resources such as a front-end server and an application server based on input classification information and load prediction rules. For example, when the input load of an input message increases rapidly, the front-end server Resource allocation can not keep up with the load. Moreover, until the number of times when the input load of the input message increases rapidly is not recorded as the load prediction rule, the response performance is degraded.

  In other words, the techniques of Patent Document 1 and Patent Document 2 have a problem that it is not possible to prevent a decrease in response performance due to a rapid input load or an improvement in processing inability.

  The objective of this invention is providing the monitoring server which solves the said subject and prevents the response performance fall accompanying the rapid input load and the improvement of processing impossibility.

  A monitoring server according to an aspect of the present invention monitors a system including a first server that performs processing based on an input, and a second server that receives a processing result of the first server and performs processing based on the processing result. A receiving means for receiving the input amount of the input; first characteristic information indicating characteristics of the past output quantity with respect to the past input quantity, which is a characteristic of the first server; and the first Storage means for storing second characteristic information indicating the characteristics of the delay time of the past output with respect to the past input quantity, the input quantity, the first characteristic information, and the second characteristic. The configuration of the second server is changed based on the output amount corresponding to the input amount and the output time based on the information, and the output amount and the output time estimated by the estimation means. Yo And a request means for requesting to said second server.

  A monitoring system according to an aspect of the present invention includes a first server that performs processing based on an input, a second server that receives a processing result of the first server, and performs processing based on the processing result, and a monitoring server. The monitoring server includes receiving means for receiving the input amount of the input, first characteristic information indicating the characteristic of the past output amount with respect to the past input amount, which is a characteristic of the first server, and the first Storage means for storing second characteristic information indicating the characteristics of the delay time of the past output with respect to the past input quantity, the input quantity, the first characteristic information, and the second characteristic. The configuration of the second server is changed based on the output amount corresponding to the input amount and the output time based on the information, and the output amount and the output time estimated by the estimation means. like Having a request means for requesting the serial second server.

  In the monitoring method according to one aspect of the present invention, a computer monitors a system including a first server that performs processing based on an input and a second server that receives a processing result of the first server and performs processing based on the processing result. The computer is characterized by the first characteristic information indicating the characteristic of the first server and the characteristic of the past output quantity with respect to the past input quantity, and the characteristic of the first server. Second characteristic information indicating a characteristic of a delay time of a past output with respect to a past input quantity, and the computer, based on the input quantity, the first characteristic information, and the second characteristic information, Estimating the output amount and the output time corresponding to the input amount, and changing the configuration of the second server based on the estimated output amount and the output time. Requests to said second server.

  A program according to an aspect of the present invention is a computer communicably connected to a first server that performs processing based on an input, and a second server that receives processing results of the first server and performs processing based on the processing results First characteristic information indicating the characteristic of the first server and the characteristic of the past output quantity with respect to the past input quantity, and the past output corresponding to the characteristic of the first server and the past input quantity. Processing of acquiring second characteristic information indicating characteristics of delay time, processing of receiving an input amount of the input, and the input amount, the first characteristic information, and the second characteristic information. Requests the second server to change the configuration of the second server based on the process of estimating the output amount and the output time corresponding to the input amount, and the estimated output amount and the output time. place And, to the execution.

  The object of the present invention is also achieved by a program stored in the above-mentioned computer readable recording medium.

  The present invention prevents the response performance deterioration associated with the rapid improvement of input load and impossibility of processing.

It is a block diagram showing composition of surveillance system 10 in a 1st embodiment of the present invention. It is a flowchart which shows the operation example of the monitoring system 10 in the 1st Embodiment of this invention. It is a block diagram which shows the structure of the monitoring system 10 in the 2nd Embodiment of this invention. It is a flowchart which shows the operation example of the monitoring system 10 in the 2nd Embodiment of this invention. It is a flowchart which shows the operation example of the monitoring system 10 in the 2nd Embodiment of this invention. It is a figure which shows the example of the 1st characteristic information which the 1st memory | storage means 320A in the 2nd Embodiment of this invention memorize | stores. It is a figure which shows the example of the 2nd characteristic information which the 1st memory | storage means 320A in the 2nd Embodiment of this invention memorize | stores. It is an example of data showing the result of having added together the estimation result which the estimation means 330 in the 2nd Embodiment of this invention estimated. It is a figure which shows an example of the structure change process characteristic information which the 2nd memory | storage means 350 in the 2nd Embodiment of this invention memorize | stores. It is a schematic block diagram showing an example of composition of computer 1000 which mounts surveillance system 10 in each embodiment of the present invention.

  Hereinafter, embodiments of a monitoring server and a system will be described with reference to the drawings. In addition, since the component which attached | subjected the same code | symbol in each embodiment performs the same operation | movement, re-explanation may be abbreviate | omitted.

First Embodiment
Hereinafter, the first embodiment will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration example of a monitoring system 10 according to a first embodiment of the present invention. Referring to FIG. 1, the monitoring system 10 in the present embodiment includes a first server 200 that performs processing based on an input, a second server 100 that receives a processing result of the first server and performs processing based on the processing result, The monitoring server 300 monitors the system 500 including the server 200 and the second server 100.

  The monitoring server 300 includes a receiving unit 310, an estimation unit 320, a storage unit 330, and a request unit 340.

  The receiving unit 310 receives an input amount of input. The storage unit 320 includes first characteristic information indicating the characteristics of the first server and the characteristics of the past output quantity with respect to the past input quantity, and the past characteristics with respect to the past input quantity and the characteristics of the first server. Storing second characteristic information indicating characteristics of delay time of the output of

  The estimation unit 330 estimates an output amount and an output time corresponding to the input amount based on the input amount received by the reception unit 310 and the first characteristic information and the second characteristic information read from the storage unit 320. The request unit 340 requests the second server 100 to change the configuration of the second server 100 based on the output amount and the output time estimated by the estimation unit 330.

  The operation of the present embodiment will be described using FIG. FIG. 2 is a flowchart showing an operation example of the monitoring server 300 in the first embodiment of the present invention.

  The receiving unit 310 receives the input amount of the external input (step S101).

  The estimation unit 330 estimates an output amount and an output time corresponding to the input amount based on the input amount received by the reception unit 310 and the first characteristic information and the second characteristic information read from the storage unit 320 (step S102).

  The request unit 340 requests the second server 100 to change the configuration of the second server 100 based on the output amount and output time estimated by the estimation unit 330 (step S103).

  As described above, according to the present embodiment, the second server 100 is requested to change the configuration of the second server 100 based on the estimated output amount and output time. Therefore, the second server 100 can be reconfigured along the estimated output amount and output time. That is, according to the present embodiment, the response performance deterioration associated with the rapid input load and the impossibility of processing can be prevented.

Second Embodiment
Hereinafter, the second embodiment will be described in detail with reference to the drawings.

  FIG. 3 is a block diagram showing a configuration example of the monitoring system 10 in the second embodiment of the present invention. Referring to FIG. 3, the monitoring system 10 according to the second embodiment of the present invention includes a target server 100A, a pre-stage server 200A, a monitoring server 300, and an external server group 400. The target server 100A, the upstream server 200A, the monitoring server 300, and the external server group 400 are connected via a wired or wireless network, respectively, to perform communication. The second server 100 in the first embodiment has the same configuration and function as the target server 200 in the present embodiment. The monitoring server 300 in the first embodiment has the same configuration and function as the monitoring server 300 in the present embodiment.

  The target server 100A includes a second processing unit 110 and a configuration changing unit 120. The first processing unit 110 receives an output from the upstream server 200A, performs a process according to the output content, and outputs a processing result. The first processing unit 110 is, for example, an application that creates a table, draws a graphic, and the like. The configuration change unit 120 performs the configuration change process of the target server 100A with the content requested by the request unit 340.

  The target server 100A receives the output from the upstream server 200A, performs processing according to the input content, and outputs the processing result. Here, the target server 100A may transmit the processing result to the external server group 400, or may transmit it to the former server 200A again as an input to the former server 200A. This depends on the implementation contents of the application 210 on the former server 200A.

  The upstream server 200A includes upstream servers 200A1 to 200An (n is a natural number). Each of the upstream servers 200A1 to 200An includes a first processing unit 210 and a monitoring unit 220. The first processing unit 210 receives an input from the external server group 400, performs a process according to the input content, and outputs a processing result to the target server 100A. The first processing unit 210 is, for example, an application. Each of the upstream servers 200A1 to An receives an external input, performs processing according to the input content, and outputs a processing result to the target server 100A.

  The monitoring unit 220 monitors an input amount indicating an amount of input from the external server group 400 to the former server having the monitoring unit 220, for example, the former server 200A1, at fixed time intervals. The monitoring unit 220 transmits the monitored input amount to the receiving unit 310.

  The monitoring server 300 includes a reception unit 310, a first storage unit 320A, an estimation unit 330, a request unit 340, and a management unit 370. The request unit 340 includes a second storage unit 350 and a selection unit 360.

  The receiving unit 310 receives the input amount of the external input monitored by the monitoring unit 220. The first storage unit 320A stores the first characteristic information and the second characteristic information. The first storage unit 320A is, for example, a database. The first storage unit 320A stores first characteristic information and second characteristic information, which are characteristic information related to input and output with respect to the target server 100A of the pre-stage servers 200A1 to An prepared in advance.

  The first characteristic information is information indicating a characteristic between an input amount from the external server group 400 to the upstream server 200A and an output amount from the upstream server 200A to the target server 100A with respect to the input amount. The second characteristic information is information indicating the amount of input from the external server group 400 to the upstream server 200A and the characteristic of output time indicating the delay time of the output from the upstream server 200A to the target server 100A with respect to the input amount.

  FIG. 6 is a diagram illustrating an example of the first characteristic information and the second characteristic information stored in the first storage unit 320A illustrated in FIG. FIG. 6A is a diagram showing an example of the first characteristic information stored in the first storage unit 320A. FIG. 6B is a diagram showing an example of the second characteristic information stored in the first storage unit 320A. As shown in FIG. 6A, the first characteristic information is how much output amount with respect to how much input amount when the amount of OS (Operating System) resource amount is allocated. It includes the characteristic information of the former server 200A. As illustrated in FIG. 6A, in the first characteristic information, the characteristic of the output amount per time according to the input amount per hour to the upstream server 200A is described for each resource allocated to the upstream server 200A.

  As shown in FIG. 6A, the first characteristic information includes a large amount of input from the external server group 400 when, for example, the allocated resources of the upstream server 200A1 are small (hereinafter referred to as “small allocated resources”) ( (Hereinafter described as “large input amount”) indicates that the output amount to the target server 100A is not large (hereinafter referred to as “low output amount”). Further, as shown in FIG. 6A, the first characteristic information indicates, for example, a characteristic that the output amount is large when the upstream server 200A1 has a large allocation resource, and the input amount is large.

  As shown in FIG. 6B, the second characteristic information is a pre-stage server, such as how long a delay time from the observation of the input by the pre-stage server 200A until the output to the target server 100A arrives It contains the characteristic information of 200A. As shown in FIG. 6B, the second characteristic information is actually input from the former server 200A according to the input amount per hour to the former server 200A, and after being output, the second characteristic information is input to the target server 100A. The delay time until it becomes is noted for each resource allocated to the former server 200A.

  As shown in FIG. 6B, for example, when the first stage server 200A1 has a small allocation resource, the second characteristic information has a large delay time to the target server 100A when the input amount is large (hereinafter, "large output time"). Description). Further, as shown in FIG. 6B, the second characteristic information indicates that, for example, when the upstream server 200A1 has a large allocation resource, and the input amount is large, the output time is short.

  That is, as shown in FIG. 6A, for example, the characteristic of the output amount per time according to the input amount per hour to the former server 200A1 is described for each resource allocated to the former server 200A1. Further, as shown in FIG. 6B, for example, according to the input amount per hour to the former server 200A1, the input is actually processed from the former server 200A1, and after it is output, it becomes an input to the target server 100A. The delay time is noted for each resource allocated to the former server 200A1.

  From the above, the first characteristic information represents, for example, a characteristic that the output amount as a result of processing changes according to the input amount. The first characteristic information represents a characteristic that the operation that the output amount changes according to the input amount changes depending on the resource allocated to the server. For example, in a general server, the delay time until output is increased according to the input amount, such as the load on the server increases as the input amount increases, and the processing time increases. Represents the characteristic that it depends on the resource allocated to the server. Moreover, although a general server may output to a plurality of servers as a result of the input, the first characteristic information describes the amount of output to only the target server 100A. Here, the first characteristic information and the second characteristic information may be prepared in advance. In that case, the first characteristic information and the second characteristic information are, for example, measured in advance or created based on the design specifications.

  Returning to the description with reference to FIG. 3, the estimation unit 330 acquires the input amount received by the reception unit 310 and the first characteristic information and the second characteristic information stored in the first storage unit 320 </ b> A. The output amount and output time to be output to the target server 100A are estimated based on The estimation means 330 estimates the output amount and the output time for all the pre-stage servers 200A, adds up all the estimation results, and transmits all the estimation results obtained by the addition to the request means 340. Here, a result obtained by adding together the estimation results estimated by the estimation means 330 is shown in FIG. FIG. 7 is an example of data representing the result of adding the estimation results. The detailed description of FIG. 7 will be described later.

  From the above, the estimation unit 330 outputs the output time indicating the time when the received amount of input to the pre-stage servers 200A1 to An is processed in the pre-stage servers 200A1 to An and output to the target server 100A. Estimate the ability. The request unit 340 requests the target server 100A to change the configuration of the second server 100 based on the output amount and output time estimated by the estimation unit 330.

  The second storage unit 350 stores configuration change processing characteristic information which is information indicating the characteristics of the configuration change processing. The configuration change process characteristic information includes, for example, a configuration change process type, a required time, a corresponding input amount, parameters necessary for executing the configuration change process, and the like regarding a plurality of configuration change processes that the target server 100A can take. The configuration change process type (hereinafter referred to as “process type”) is, for example, addition of a CPU core, addition of a memory or storage, and priority change of OS scheduling for the second processing unit 110 operating on the target server 100A. Configuration change processing such as changing the operation target server of the second processing unit 110 may be mentioned. The configuration change process is not limited to the above process, and any process that has a performance improvement effect on the target server 100A application 110 may be used.

  The required time is a required time required to execute the configuration change process described in the configuration change process type. Although processing such as adding a CPU core takes only ms order time, processing such as changing the server on which the application 110 is operating takes time in the order of seconds to minutes, etc. The required time depends on the type of configuration change processing Is different. The available input amount (hereinafter referred to as “available amount”) is an input amount that can be handled by the target server 100A without significant deterioration in response performance after the configuration change. The available amount may be a difference amount before and after the configuration change, or may be an amount that can be accepted after the configuration change.

  Parameters necessary for executing configuration change processing (hereinafter referred to as “parameters”) are parameters and the like necessary for performing configuration change processing for each configuration change type, and for example, for configuration change processing such as CPU core addition On the other hand, the parameters required to execute the configuration change process include the number of cores to be actually added.

  FIG. 8 is a diagram showing an example of the configuration change processing characteristic information stored in the second storage unit 350. As shown in FIG. As shown in FIG. 8, the configuration change process characteristic information includes, for example, a process, a parameter, a required time, and an available amount. As shown in FIG. 8, the processing is “add CPU” and “add memory”. The parameters are “1 core”, “2 cores” and “4 cores” in “CPU addition” and “100 MB”, “500 MB” and “1 GB” in “memory addition”.

  The required time is "50 ms" for "1 core" in "CPU addition", "100 ms" for "2 cores" in "CPU addition", and "4 cores" in "CPU addition" “200 ms”. The supportable amount is "5 MByte / sec" for "1 core" in "CPU addition", and "10 MByte / sec" for "2 core" in "CPU addition", and "4 in" CPU addition " If it is a core, it will be "20 MByte / sec".

  Here, the allowable amount shown in FIG. 8 is assumed to be an acceptable amount after the configuration change. Further, the configuration change processing characteristic information is not limited to the items and values shown in FIG. Further, for example, when there are a plurality of target servers 100A, the second storage unit 350 stores a plurality of configuration change processing characteristic information corresponding to the respective target servers 100A.

  The selection unit 360 determines the necessity of the configuration change based on the output amount and output time estimated by the estimation unit 330, and stores the estimated output amount and output time, and the second storage unit 350, if necessary. Configuration change processing is selected based on the configuration change processing characteristic information. The selection unit 360 requests the target server 100A to change the configuration in the selected configuration change process. The external server group 400 inputs to the former server 200A through a network connection.

  The management unit 370 manages the amount of OS resources currently allocated to each server (hereinafter referred to as “resource amount”). Here, the resource amount is, for example, a CPU speed and the number of assigned CPU cores, a memory capacity, a hard disk capacity, or an available network communication band.

  The operation of the present embodiment will be described with reference to FIGS. 4 and 5. FIG.4 and FIG.5 is a flowchart which shows the operation example of the monitoring system 10 in the 2nd Embodiment of this invention.

  The monitoring unit 220 monitors an input amount from the external server group 400 to the upstream server 200A (step S101). The monitoring unit 220 monitors, for example, an input (external input) through the network connection from the external server group 400 with respect to the upstream servers 200A1 to An at regular intervals, and acquires an input amount. The input amount monitored by the monitoring unit 220 is, for example, the value per hour of the communication amount transmitted to each of the upstream servers 200A1 to An, the number of processing requests per hour, or the connection from the external server 400 per hour. The number of requests.

  The receiving unit 310 receives the input amount to the upstream server 200A acquired by the monitoring unit 210 (step S102). The receiving unit 310 receives, for example, the number of processing requests per hour to the former server 200A1. The receiving unit 310 receives, for example, the input amount per time to the upstream server 200A acquired and transmitted by the monitoring unit 220.

  The management unit 370 acquires the amount of resources currently allocated to each server of the upstream server 200A, that is, the upstream servers A1 to An (step S104).

  The estimation unit 330 acquires the first characteristic information and the second characteristic information stored in the first storage unit 320A (step S105). Here, the process of step S105 may be performed, for example, after step S102.

  The estimation unit 330 receives the input amount to the front servers A1 to An received by the reception unit 310, the acquired first characteristic information and second characteristic information, and the resource amount of the front server 200A acquired by the management unit 370. From this, the output amount and output time to the target server 100A are estimated (step S106). The estimating unit 330 determines in advance the input amounts to the pre-stage servers 200A1 to An acquired in step S102, the OS resource amounts currently allocated to the pre-stage servers 200A1 to An acquired in step S104, and the pre-acquisition unit acquired in step S105. Based on the prepared first characteristic information and second characteristic information, the input load to the target server 100A and the occurrence time (delay time) of the input load are estimated.

  Hereinafter, the estimation unit 330 estimates the output amount and output time for the pre-stage server 200A1 when the resource amount of the pre-stage server 200A1 is "small allocation resource" and the input amount to the pre-stage server 200A1 is 30 MBytes / sec. Describe. In this case, the estimation unit 330 estimates that the output amount is 10 MBytes / sec and the output time is 300 ms by using the first characteristic information and the second characteristic information shown in FIGS. 6A and 6B.

  The estimation unit 330 repeats steps S104 to S106 for all the upstream servers 200A (step S103). In the present embodiment, the estimation unit 330 estimates the input amount and output time in all of the upstream servers 200A1 to 200An. For example, the estimation unit 330 may perform a plurality of estimation processes for estimating the output amount and the output time, or may perform each process one by one.

  The estimation means 330 estimates the output amount and output time of all the pre-stage servers 200A, and then adds all the estimation results (step S107). As described above, FIG. 7 is an example of data representing the result of adding the estimation results. As shown in FIG. 7, the input amount from the upstream servers 200A1 to An added to the target server 100A and the estimated value of the arrival time thereof are, for example, change amounts for each time. This is because the arrival time of the input load differs depending on each of the upstream servers 200A1 to An.

  As shown in FIG. 7, for example, the estimation unit 330 calculates an estimated value of an output amount of 10 MByte / sec after 100 ms, that is, an output time of 100 ms for the upstream server 200A2. For example, the estimation unit 330 estimates that the output time is 200 ms and the output amount is 30 MByte / sec for the former server 200A3, and that the output time is 100 ms and 5 MBytes / sec for the former server 200A4. Calculate the value. As shown in FIG. 7, the curve graph is obtained by superimposing a plurality of predicted values of the upstream servers 200A1 to An, for example.

  As FIG. 7 shows, the horizontal axis of this graph is the output time from the present time, ie, arrival time, and the vertical axis is the output quantity from the current, ie, the output estimated quantity. When the process of step S107 is finished, the process of the monitoring system 10 proceeds to step S201. The estimation unit 330 transmits a plurality of estimation values and an estimation result obtained by adding the estimation values to the selection unit 360 (step S201).

  The selection unit 360 determines whether it is necessary to cope with the configuration change of the target server 100A based on the transmitted plurality of estimated values and the added estimation result (step S202). For example, the selection unit 360 estimates the output amount such as the maximum value of the estimated value of the current output amount, the integral value of the estimated value of the output amount within a certain time, or the increment of the estimated value of the output amount within the certain time. A determination is made based on whether any value calculated based on the value exceeds a predetermined threshold. In addition, for example, even if the selecting unit 360 acquires the available capacity of the current target server 100A, and compares the current applicable capacity with the output volume from the previous stage server 200A, the presence or absence of the configuration change is determined. good.

  When the calculated value exceeds the threshold value, the selection unit 360 determines to perform the response by the configuration change (“Yes” in step S202). This is because if the calculated value exceeds the threshold, the target server 100A may cause a significant deterioration in response performance.

  The selection unit 360 acquires configuration change processing characteristic information in which the characteristics of the configuration change processing that can be handled by the target server 100A are described from the second storage unit 350 (step S203). The selection unit 360 is configured to be implemented this time based on the obtained output amounts to the target server and the estimated values of the output time, and the plurality of configuration change processings that can be taken by the target server 100A described in the configuration change processing characteristic information. A change process is selected (step S204). For example, the selection unit 360 calculates the time when the output amount exceeding the threshold reaches based on the output amount to the target server 100A and the estimated value of the output time. The selection means 360 selects, based on the configuration change processing characteristic information, the configuration change processing provided with the available capacity that can correspond to the maximum value of the output estimation amount in the required time completed by that time.

  Here, it is assumed that the current available capacity of the target server 100A is 10 MBytes / sec, the output volume from the upstream server 200A3 is 30 MBytes / sec, and the output time is an estimated value of 200 ms. In this case, the selection unit 360 compares the current available amount of the target server 100A with the output amount, and performs a determination to implement the response by the configuration change. The selection unit 360 selects the configuration change process based on the estimated value and the configuration change process characteristic information shown in FIG. The selection unit 360 selects a configuration change process in which the correspondence amount is 30 MBytes / sec or more and the required time is 200 ms or less from the configuration change process characteristic information. As shown in FIG. 8, when “add CPU” is “4 cores”, the selection unit 360 selects the configuration process because the above condition is satisfied.

  The selection unit 360 requests the change unit 120 to change the configuration in the selected configuration change process (step S205). The changing unit 120 performs the configuration change based on the configuration change process selected by the selection unit 360 (step S206). Thus, the configuration change of the target server 100A is performed. For example, parameter information included in configuration change processing characteristic information is used as parameters necessary for performing the configuration change processing.

  If the calculated value does not exceed the threshold value, the selection unit 360 determines that no action is required due to the configuration change (“No” in step S202). If it is determined that the selection means 360 is unnecessary, the process ends. Next time, after a certain time, if the estimation result is estimated by the processing of steps S101 to S107 of FIG. 4 and the estimation result of the input amount to the target server 100A and the arrival time by the input / output load estimation means 320 is transmitted again Then, the process is performed again from step S201 of FIG.

  As described above, the monitoring system 10 in the present embodiment estimates the output amount and output time of the upstream server 200A, and changes the configuration from the estimated values of the output amount and output time and the configuration change processing characteristic information stored in advance. Select and implement the process. As a result, it is possible to prevent the response performance deterioration due to the rapid input load and the impossibility of processing.

  In addition, the monitoring system 10 can detect a rapid increase in input load that causes a decrease in the response performance of the target server 100A at an early stage. This is based on the monitoring result of the input amount by the monitoring unit 220, the resource amount of the upstream server 200A by the management unit 370, and the first characteristic information and the second characteristic information by the first storage unit 320A. This is because 320 estimates the amount and time of output to the target server 100A.

  Here, when there is no configuration change process that satisfies the condition in the configuration change process characteristic information, the selection unit 360 is, for example, from the viewpoint of the configuration change process that is completed by the time when the input amount exceeding the threshold value is reached. Select and request configuration change processing. Even when the change unit 120 performs the processing, when the changeable input amount is insufficient, the selection unit 360 continuously selects and requests other configuration change units. As a result, although the increase in the input load can not be completely coped with in time, the first configuration change processing enables some degree of processing enhancement by the load arrival time. After that, the second configuration change is continuously completed, so that the processing enhancement that can finally cope with the input prediction amount is realized. Therefore, it is possible to achieve the purpose of preventing significant deterioration of the response performance of the target server 100A.

  In addition, although not shown in FIG. 5, if there is a change in the resource amount of the target server 100A due to the implementation of the configuration change, the corresponding information on the management means 370 is changed. Further, when the threshold used for the determination of step S202 in FIG. 5 is affected by the configuration change, the threshold is updated. In general, the ability to respond to input amounts is improved by changing the configuration, so the threshold value must be changed to a larger value. Once the configuration change has been carried out to enhance the ability to cope with the input load, for example, when an input amount significantly below the threshold value is continuously observed at the time of determination in step S202 of FIG. You may return to the initial state.

  FIG. 9 is a schematic block diagram illustrating a configuration example of a computer 1000 that implements the monitoring system 10 according to each embodiment of the present invention. The computer 1000 includes a CPU 1001, a main storage device 1002, an auxiliary storage device 1003, an interface 1004, an input device 1005, and a display device 1006.

  The monitoring server 300 and the like of each embodiment are mounted on the computer 1000. The operation of the monitoring server 300 or the like is stored in the auxiliary storage device 1003 in the form of a program. The CPU 1001 reads out the program from the auxiliary storage device 1003, develops it in the main storage device 1002, and executes the above processing according to the program.

  The auxiliary storage device 1003 is an example of a non-temporary tangible medium. Other examples of non-transitory tangible media include magnetic disks connected via interface 1004, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memories, and the like. When this program is distributed to the computer 1000 through a communication line, the computer 1000 that has received the distribution may deploy the program in the main storage device 1002 and execute the above-described processing.

  The interface 1004 is connected to the CPU 1001 and connected to a network or an external storage medium. External data may be taken into the CPU 1001 via the interface 1004. The input device 1005 is, for example, a keyboard, a mouse, or a touch panel. The display 1006 displays a screen corresponding to drawing data processed by the CPU 1001, a graphics processing unit (GPU) (not shown) or the like, such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display. Device. Note that the hardware configuration illustrated in FIG. 11 is merely an example, and each unit illustrated in FIG. 1 may be configured with independent logic circuits.

  Also, the program may be for realizing a part of the above-mentioned processing. Furthermore, the program may be a differential program that realizes the above-described processing in combination with another program already stored in the auxiliary storage device 1003.

  The present invention has been described above using the embodiment, but the present invention is not necessarily limited to the above embodiment. The configurations and details of the present invention can be variously modified and implemented (within the scope of the technical idea) which can be understood by those skilled in the art within the scope of the present invention.

10 Monitoring system 100 Second server 100A Target server 110 Second processing means 120 Changing means 200 First server 200A Pre-stage server 210 First processing means 220 Monitoring means 300 Monitoring server 310 Receiving means 320 Storage means 320A First storage means 330 Estimating means 340 request means 350 second storage means 360 selection means 370 management means 400 external server group 500 system 1000 computer 1001 CPU
1002 main storage unit 1003 auxiliary storage unit 1004 interface 1005 input device 1006 display unit

Claims (10)

A monitoring server for monitoring a system comprising: a first server processing based on an input; and a second server receiving processing results of the first server and processing based on the processing results,
Receiving means for receiving the input amount of the input;
The first characteristic information indicating a characteristic of past output amount to the previous input quantity a characteristic of the first server, and, in the past output for past input quantities a characteristic of the first servers Storage means for storing second characteristic information indicating characteristics of delay time;
And estimating means for estimating the basis of the received input quantity of the input and said first characteristic information and the second characteristic information, power output and force out that corresponds to the input amount of input the received time,
Request means for requesting the second server to change the configuration of the second server based on the output amount and the output time estimated by the estimation means;
Monitoring server with Said storage means, said first characteristic information and the second characteristic information, a plurality of stored for each amount allocated resources to the first servers,
The estimating means on the basis of the resource amount currently allocated to the first servers, from among the plurality of first characteristic information and the plurality of second characteristic information, the specific first The output amount and the output time corresponding to the input amount are selected based on the specific first characteristic information and the specific second characteristic information by selecting the characteristic information and the specific second characteristic information. presume,
The monitoring server according to claim 1. The system comprises a plurality of the first servers,
The estimation unit estimates the output amount and the output time corresponding to the input amount for each of the plurality of first servers, and information obtained by adding up the estimated output amounts for each of the output times Generate
The monitoring server according to claim 1, wherein the request unit requests the second server to change a configuration of the second server based on information generated by the estimation unit. For each of the second plurality of configuration change processing performed on servers candidates, the required time it takes to complete the configuration change process, a second that associates and stores the expected performance enhancement degree by the configuration change And storage means,
The request means preferentially selects a configuration change process capable of improving the performance as much as possible corresponding to the output amount estimated by the estimation means by the output time estimated by the estimation means, and instructs the configuration change The monitoring server according to any one of claims 1 to 3, wherein: A first server that processes based on the input;
A second server that receives the processing result of the first server and performs processing based on the processing result;
And a monitoring server,
The monitoring server is
Receiving means for receiving the input amount of the input;
The first characteristic information indicating a characteristic of past output amount to the previous input quantity a characteristic of the first server, and, in the past output for past input quantities a characteristic of the first servers Storage means for storing second characteristic information indicating characteristics of delay time;
And estimating means for estimating the basis of the received input quantity of the input and said first characteristic information and the second characteristic information, power output and force out that corresponds to the input amount of input the received time,
Request means for requesting the second server to change the configuration of the second server based on the output amount and the output time estimated by the estimation means;
Monitoring system with. Said storage means, said first characteristic information and the second characteristic information, a plurality of stored for each amount allocated resources to the first servers,
The estimating means on the basis of the resource amount currently allocated to the first servers, from among the plurality of first characteristic information and the plurality of second characteristic information, the specific first The output amount and the output time corresponding to the input amount are selected based on the specific first characteristic information and the specific second characteristic information by selecting the characteristic information and the specific second characteristic information. presume,
The monitoring system according to claim 5. The monitoring system comprises a plurality of the first servers,
The estimation unit estimates the output amount and the output time corresponding to the input amount for each of the plurality of first servers, and information obtained by adding up the estimated output amounts for each of the output times Generate
The monitoring system according to claim 5 or 6, wherein the request unit requests the second server to change the configuration of the second server based on the information generated by the estimation unit. For each of the second plurality of configuration change processing performed on servers candidates, the required time it takes to complete the configuration change process, a second that associates and stores the expected performance enhancement degree by the configuration change And storage means,
The request means preferentially selects a configuration change process capable of improving the performance as much as possible corresponding to the output amount estimated by the estimation means by the output time estimated by the estimation means, and instructs the configuration change The monitoring system according to any one of claims 5 to 7, wherein: A monitoring method in which a computer monitors a system comprising: a first server that performs processing based on an input; and a second server that receives the processing result of the first server and performs processing based on the processing result,
The computer receives an input amount of the input,
The computer, the first characteristic information indicating a characteristic of past output amount to the previous input quantity a characteristic of the first server, and, for the input of the past a characteristic of the first servers Obtain second characteristic information indicating characteristics of delay time of the past output,
The computer, on the basis of said the input quantity of the input thus received first characteristic information and the second characteristic information, to estimate to that output force and output time corresponding to the input amount of the input thus received,
The monitoring method, wherein the computer requests the second server to change the configuration of the second server based on the estimated output amount and the output time. A computer that is communicably connected to a first server that performs processing based on an input and a second server that receives processing results of the first server and processes based on the processing results;
The first characteristic information indicating a characteristic of past output amount to the previous input quantity a characteristic of the first server, and, in the past output for past input quantities a characteristic of the first servers A process of acquiring second property information indicating a property of delay time;
A process of receiving an input amount of the input;
A process of estimating the basis of the received input quantity of the input and said first characteristic information and the second characteristic information, power output and force out that corresponds to the input amount of input the received time,
A process of requesting the second server to change the configuration of the second server based on the estimated output amount and the output time;
A program that runs

Images