JP5810968B2 - Resource capacity prediction apparatus, method and program - Google Patents

Description

  The present invention relates to a resource capacity apparatus, a resource capacity prediction method, and a resource capacity prediction program for predicting resource capacity required for IT (Information Technology) service.

  One of the usage forms for providing an IT service via a network is XaaS (X as a Service). In XaaS, a contract including a service level item, which is an index for setting target values of the quality and content of an IT service, is often concluded between a service provider and a client who uses the service.

  It is desirable for the service provider to manage the resource capacity required for the IT service without excess or deficiency so as to satisfy the target value set in the service level item of the IT service. Furthermore, when there are a plurality of service level items for IT services, it is desirable for the service provider to satisfy all target values set for each.

  Sizing is a technique for estimating the required resource capacity before providing IT services. In sizing, resource capacity is estimated based on resource monitoring information obtained in a test environment before providing IT services.

  On the other hand, even if the resource capacity is estimated by sizing, it is difficult to accurately predict the monitoring status including the usage status of IT services, the operating status of various programs, and the error occurrence status before providing IT services. . That is, it is assumed that resource monitoring information is obtained in advance in a test environment, and the resource capacity is predicted based on the resource monitoring information. However, this resource capacity often does not match the resource capacity predicted based on the resource monitoring information in the production environment.

  Therefore, after providing the IT service, the service provider estimates the IT service demand peak and the change in the resource usage rate from the resource monitoring information obtained in the production environment, and the IT service demand peak changes. In addition, it is necessary to adjust the required resource capacity.

  Patent Document 1 describes a system that predicts resource capacity for each transaction from resource monitoring information. The system described in Patent Document 1 is based on the monitoring information of the resource usage rate at a plurality of times of the computer and the monitoring information of the processing count for each transaction at the plurality of times. The average value and the processing amount of each transaction per fixed time are obtained. And the system described in patent document 1 estimates the resource usage rate for every transaction by performing the multiple regression analysis using the average value of the obtained resource usage rate, and the processing amount for every transaction.

  The system described in Patent Literature 1 predicts the highest resource usage rate among all transactions as the resource capacity required for the IT service based on the estimated resource usage rate for each transaction. .

JP 2006-24017 A

  In a production environment, the usage status of IT services during busy and quiet periods, virus check and update status for maintaining IT services, operating status of resident programs such as scan desks, actual data transfer volume and communication failure frequency The network communication status such as is monitored. Since the status of these monitoring targets changes from moment to moment, information on various situations is mixed in the monitoring information obtained by monitoring.

  When the resource monitoring information in different monitoring situations coexists, the tendency of the resource monitoring information differs even if the resource monitoring information is in the same period. For example, depending on whether the usage of IT service is at peak, which resident program is running, whether there is a network communication error, etc. The average value is quite different.

  As described above, in the general monitoring method, even if the monitoring status is different, the monitoring information of the resource is acquired by regarding the same monitoring status. Then, resources required for the IT service are estimated from the resource monitoring information. Therefore, there is a problem that the resource capacity cannot be predicted accurately.

  Specifically, the monitoring information of the resources that are regarded as the same monitoring status even in different monitoring statuses have different tendencies such as the average and variance of the resource usage rate and the measurement error. For this reason, when resources required for the IT service are estimated from such monitoring information, there is a problem that variations in estimation results increase.

  Therefore, the present invention provides a resource capacity prediction device, a resource capacity prediction method, and a resource capacity prediction program capable of accurately predicting a resource capacity required for an IT service from monitoring information of resources having different monitoring conditions. Objective.

  A resource capacity prediction apparatus according to the present invention is a resource capacity prediction apparatus that predicts the resource capacity of a computer system composed of a plurality of computers, and obtains a first log relating to a resource monitoring status from the computer. Log reliability calculation means for calculating log reliability at a plurality of times based on information on the monitoring status of the computer at a plurality of times included in the first log; Obtain a third log on the measurement of log and computer service level items, information on resource utilization at multiple times of computers included in the second log, and multiple computers of computers included in the third log Information on measured values of service level items at time and log reliability Based at a plurality of times calculated by the calculating means and the reliability of the log, characterized in that a resource usage rate transition estimating means for estimating the utilization of the resources of the computer.

  A resource capacity prediction method according to the present invention is a resource capacity prediction method for predicting the resource capacity of a computer system composed of a plurality of computers, and obtains a first log relating to a resource monitoring status from the computer. The reliability of the log at a plurality of times is calculated based on the monitoring status information of the computer at the plurality of times included in the first log, and the second log regarding the use of the resource and the service level of the computer from the computer Obtain a third log related to the measurement of items, information on resource usage at multiple times of computers included in the second log, and service level items at multiple times of computers included in the third log Information on measured values and the log at multiple calculated times Based of the degrees of reliability, and estimates the utilization of computer resources.

  A resource capacity prediction program according to the present invention is a resource capacity prediction program applied to a prediction computer that predicts resource capacity of a computer system composed of a plurality of computers. A log reliability calculation process for obtaining a first log related to the monitoring status and calculating the reliability of the log at a plurality of times based on the monitoring status information of the computer included in the first log; and Obtaining from the computer a second log relating to the use of resources and a third log relating to the measurement of the service level item of the computer, and information on resource utilization at multiple times of the computer included in the second log; Computers included in the third log Resource usage transitions that estimate the resource usage of a computer based on information on measured values of service level items at multiple times and log reliability at multiple times calculated by the log reliability calculation process An estimation process is executed.

  According to the present invention, it is possible to accurately predict the resource capacity required for the IT service from the monitoring information of resources having different monitoring conditions.

It is explanatory drawing which shows the structural example of the distributed system with which a resource capacity prediction apparatus is applied. It is a block diagram which shows the example of a resource management apparatus. It is a block diagram which shows the example of a resource capacity prediction apparatus. It is explanatory drawing which shows the example of a computer system. It is explanatory drawing which shows the example of the service relevant data table stored in service DB. It is explanatory drawing which shows the example of the execution history table stored in execution state DB. It is explanatory drawing which shows the example of the resource management table stored in resource DB. It is explanatory drawing which shows the example of the monitoring condition log data table stored in monitoring log | history DB. It is explanatory drawing which shows the example of the log data table stored in monitoring log | history DB. It is explanatory drawing which shows the example of the resource transition estimation data table stored in capacity estimation DB. It is explanatory drawing which shows the example of the resource capacity prediction data table stored in capacity estimation DB. It is explanatory drawing which shows the example of log reliability. It is a flowchart which shows the example of operation | movement in which a resource management apparatus performs resource capacity prediction. It is a flowchart which shows the example of operation | movement of a resource capacity prediction apparatus. It is a flowchart which shows the example of operation | movement of a log reliability calculation part. It is a flowchart which shows the example of operation | movement of a resource utilization rate transition estimation part. It is a flowchart which shows the example of operation | movement of a resource capacity prediction part. It is explanatory drawing which shows the example of the hardware constitutions of a resource capacity prediction apparatus and a resource management apparatus. It is a block diagram which shows the outline | summary of the resource capacity prediction apparatus by this invention.

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

  In the present embodiment, a log related to the monitoring status of a computer executing IT service processing at a plurality of times, a log related to measurement of service level items at a plurality of times of the computer, and use of various resources at the plurality of times of the computer Log is used. In the present invention, attention is also paid to the reliability of logs related to the monitoring status.

  In this embodiment, the monitoring status log includes data indicating the usage status of the IT service during busy periods and quiet periods, virus checks and updates for maintaining the IT service, and resident programs (for example, a scan desk). Data indicating the operational status of the network, data indicating the network communication status (for example, actual data transfer amount, communication failure frequency), and the like. That is, it can be said that the log related to the monitoring status is a log indicating the status of each element forming the computer system that provides the IT service.

  In the present embodiment, the log related to the measurement of service level items includes throughput, average request response time, average transaction processing time, number of transactions, operation rate, non-operation rate, average recovery time, average failure interval, and the like. . That is, it can be said that the log regarding the measurement of the service level item is a measurement value by various resources used for evaluation of the service level item and an index calculated based on the measurement value.

  In the present embodiment, the resource usage log includes resource usage rates (eg, CPU usage rate, memory usage, storage capacity of storage device, network usage).

  Then, the resource capacity prediction apparatus according to the present embodiment analyzes these logs, estimates the transition of the usage state of various resources with respect to the service level item, and capacity of various resources that satisfy all target values of the service level item. Predict. Note that predicting the capacity of various resources may be simply referred to as resource capacity prediction. This resource capacity prediction enables appropriate adjustment of the resource capacity required for the IT service.

  The target of the resource capacity prediction is various resources such as a CPU usage rate, a memory usage amount, a storage capacity of a storage device, and a network usage amount. In the present embodiment, a case will be mainly described as an example where resource capacity prediction is performed for the CPU usage rate. Further, the embodiments exemplified below do not limit the invention according to the claims. In addition, not all combinations of features described in the embodiments exemplified below are necessarily essential to the solution of the invention.

  First, the configuration of a distributed system to which the resource capacity prediction apparatus of this embodiment is applied will be described. FIG. 1 is an explanatory diagram illustrating a configuration example of a distributed system to which the resource capacity prediction apparatus of the present embodiment is applied.

  The distributed system illustrated in FIG. 1 includes a computer system 10, a resource management device 20, a resource capacity prediction device 30, and a user terminal 40. The resource management device 20 is connected to the resource capacity prediction device 30 and the computer system 10, respectively.

  The computer system 10 has one or more resources that are used to execute processing of IT services. The computer system 10 executes processing of the IT service designated by the user by combining at least one resource.

  The resource management device 20 manages resources in the computer system 10 for executing IT service processing. Upon receiving an IT service execution request from the user, the resource management apparatus 20 secures at least one resource in the computer system 10 for executing the IT service processing designated by the user. Then, the resource management device 20 causes the IT service processing to be executed using the secured resource.

  Further, the resource management device 20 receives a request from the user to review whether or not the resource capacity reserved for the IT service is appropriate, and the resource capacity prediction device 30 receives the resource capacity required for the IT service. To predict.

  In addition, the resource management device 20 acquires a prediction result of the resource capacity predicted by the resource capacity prediction device 30. If there is a difference between the reserved resource capacity and the prediction result, the resource management device 20 inquires of the user whether or not to change the reserved resource capacity to the resource capacity of the prediction result. Then, the resource management device 20 adds or deletes the resource capacity in response to receiving an instruction to change the resource capacity.

  When the resource capacity prediction device 30 receives a request for reviewing whether the resource capacity reserved for the IT service is appropriate from the resource management device 20, the resource capacity prediction device 30 satisfies the target value of the service level item of the IT service. Predict the resource capacity required for IT services.

  The user terminal 40 transmits an IT service execution request, a resource capacity review request, and a resource capacity change instruction by the user to the resource management apparatus. Further, the user terminal 40 is used for receiving the execution results of the resource management device 20, the resource capacity prediction device 30, and the computer system 10.

  The communication network 99 is realized by an Internet network, an intranet network, a local area network, or the like, for example. These may be used alone or in combination. The communication network 99 connects the resource management device 20 and the user terminal 40.

  Next, the operation of the distributed system illustrated in FIG. 1 will be described. The distributed system illustrated in FIG. 1 (hereinafter simply referred to as “distributed system”) secures at least one hardware resource or software resource in a computer system for a predetermined time in response to an IT service execution request from a user. Then, the IT service processing designated by the user is executed. Note that the distributed system may reserve both hardware resources and software resources for a predetermined time.

  In response to a request to review whether or not the resource capacity reserved for the IT service is appropriate, the distributed system needs the resource capacity required for the IT service to satisfy the target value of the service level item of the IT service. Predict.

  If there is a difference between the resource capacity reserved for the IT service and the resource capacity required for the predicted IT service, the distributed system changes the reserved resource capacity to the predicted resource capacity. Ask the user whether or not to do so. Then, in response to receiving an instruction to change the resource capacity, the distributed system adds or deletes the resource capacity.

  Next, the configuration of the resource management apparatus 20 illustrated in FIG. 1 will be described. FIG. 2 is a block diagram illustrating an example of a resource management device. 2 includes a front-end processing unit 21, a service database 22 (hereinafter referred to as service DB 22), a monitoring history database 23 (hereinafter referred to as monitoring history DB 23), and capacity estimation. A database 24 (hereinafter referred to as capacity estimation DB 24), a resource database 25 (hereinafter referred to as resource DB 25), an execution management unit 26, and an execution state database 27 (hereinafter referred to as execution state DB 27) are provided. Yes.

  The front-end processing unit 21 interacts with the user terminal 40 via the communication network 99, IT service execution instruction, resource capacity prediction instruction, resource capacity estimation result confirmation, resource capacity change Instruction and execution history stored in the execution state DB 27 are added and updated.

  The front end processing unit 21 includes a request receiving unit 211, an inquiry unit 212, and a resource securing unit 213.

  The request receiving unit 211 receives from the user terminal 40 an execution request including the processing content of the IT service that the user desires to execute and the target value of the service level item of the IT service. Then, the request receiving unit 211 instructs the resource securing unit 213 to secure at least one resource in the computer system 10 that executes the IT service processing and to execute the IT service processing using the resource. To do.

  In addition, the request reception unit 211 receives a review request including an identifier of an IT service that the user desires to review the resource capacity from the user terminal 40. Then, the request reception unit 211 instructs the resource capacity prediction apparatus 30 to predict the resource capacity necessary for the IT service.

  Further, the request reception unit 211 stores an execution history table stored in the execution state DB 27 described later in accordance with the execution state of a process for a request received for each IT service (hereinafter referred to as an IT service request reception process). On the other hand, an IT service execution history record is added or updated.

  The inquiry unit 212 confirms whether the resource capacity prediction result of the IT service instructed by the request reception unit 211 to the resource capacity prediction device 30 is stored in the capacity estimation DB 24. When the prediction result is stored, the inquiry unit 212 compares the resource capacity reserved for the IT service with the prediction result.

  When there is a difference between the resource capacity reserved for the IT service and the prediction result, the inquiry unit 212 determines the resource capacity reserved for the IT service as the resource capacity of the prediction result. Ask the user whether to change to When the inquiry unit 212 receives from the user terminal 40 a change instruction including an identifier of an IT service that the user desires to change the resource capacity, the inquiry unit 212 instructs the resource securing unit 213 to add or delete the resource capacity of the IT service. .

  Further, the inquiry unit 212 responds to the execution history table stored in the execution state DB 27 described later according to the execution state of the process responding to the content inquired to the user (hereinafter referred to as the IT service inquiry process). , IT service execution history records are added or updated.

  The resource securing unit 213 accepts a resource securing request from the request accepting unit 211 and the inquiry unit 212. Then, the resource securing unit 213 adds or updates the resource management table to the resource DB 25 to be described later in order to secure resources used to execute the IT service processing designated by the request accepting unit 211 or the query unit 212. I do.

  Also, the resource securing unit 213 executes an execution history table stored in an execution state DB 27 described later according to the execution state of a process for securing resources used in the IT service (hereinafter referred to as IT service resource securing process). In addition, the IT service execution history record is added or updated.

  The service DB 22 stores information related to the IT service requested by the user as a service related data table. The IT service related information is information including a target value set for the service level item of the IT service. In addition, in order to satisfy the target value of each service level item, the resource capacity of the resource used for providing the service is determined.

  The monitoring history DB 23 stores a log relating to the monitoring status of various resources of the computer as a monitoring status log data table. In addition, the monitoring history DB 23 stores, as a log data table, logs relating to measurement results of computer service level items and use of various resources.

  The capacity estimation DB 24 stores the prediction results of the resource capacity prediction device 30 as a resource transition estimation data table and a resource capacity prediction data table.

  The resource DB 25 stores information for securing various resources of the computer system 10 as a resource management data table.

  The execution management unit 26 refers to the resource management table stored in the resource DB 25 and manages processing executed to secure various resources used in each IT service provided by the computer system 10.

  Further, the execution management unit 26 adds or updates the resource management table to the resource DB 25 according to the execution state of the IT service resource securing process. The execution management unit 26 adds or updates the IT service execution history record in the execution history table stored in the execution state DB 27 according to the execution state of the IT service resource securing process.

  The execution state DB 27 stores information indicating the execution state of each process executed by the resource management device 20 as an execution history table.

  When the resource capacity reserved for executing the IT service is excessive or insufficient by using the resource management device 20 illustrated in FIG. 2, the resource capacity predicted by the resource capacity prediction device 30 upon receiving user approval. You can change to a city. That is, it is possible to appropriately adjust the resource capacity required for the IT service.

  Next, the configuration of the resource capacity prediction apparatus of this embodiment will be described. FIG. 3 is a block diagram illustrating an example of the resource capacity prediction apparatus according to the present embodiment. The resource capacity prediction device 30 includes a prediction process management unit 31, a monitoring status log acquisition unit 32, a log reliability calculation unit 33, a log acquisition unit 34, a resource usage rate transition estimation unit 35, and a resource capacity prediction. Part 36.

  The prediction process management unit 31 receives a resource capacity prediction instruction from the front end processing unit 21 of the resource management apparatus 20. The prediction process management unit 31 also includes a monitoring status log acquisition unit 32, a log reliability calculation unit 33, a log acquisition unit 34, a resource usage rate transition estimation unit 35, and a resource capacity prediction unit 36 related to the resource capacity prediction process. The execution instruction is performed. Further, the prediction process management unit 31 manages that the resource capacity prediction process is executed by acquiring the execution state of the target for which the execution instruction has been issued.

  In addition, the prediction process management unit 31 acquires information related to the IT service from the service DB 22 and transmits it to the monitoring status log acquisition unit 32, the log acquisition unit 34, the resource usage rate transition estimation unit 35, and the resource capacity prediction unit 36. To do.

  The monitoring status log acquisition unit 32 acquires a log related to the monitoring status of the computer executing the IT service processing from the monitoring history DB 23 according to an instruction from the prediction processing management unit 31, and transmits the log to the resource usage rate transition estimation unit 35.

  The log reliability calculation unit 33, in response to an instruction from the prediction process management unit 31, analyzes a log related to the monitoring status of various resources of the computer that executes the IT service process, and trusts the log of the computer that executes the IT service process. The degree is calculated and transmitted to the resource usage rate transition estimation unit 35. The log reliability calculation method will be described later.

  The log acquisition unit 34 acquires, from the monitoring history DB 23, a log related to measurement of service level items of a computer that executes IT service processing and a log related to use of various resources in accordance with an instruction from the prediction process management unit 31. It transmits to the rate transition estimation part 35.

  The resource usage rate transition estimation unit 35, in response to an instruction from the prediction processing management unit 31, logs reliability of the computer executing the IT service processing, logs regarding measurement of service level items of the computer executing the IT service processing, From the logs related to the use of various resources, the log related to the measurement of service level items of computers with high reliability and the log related to the use of various resources are specified. Then, the resource usage rate transition estimation unit 35 estimates the resource usage rate with respect to the measured value of the service level item based on the logs. Further, the resource usage rate transition estimation unit 35 stores the estimation result in the capacity estimation DB 24. A method for estimating the resource usage rate will be described later.

  The resource capacity prediction unit 36 acquires, from the capacity estimation DB 24, an estimation result of the resource usage rate with respect to the measured value of the service level item according to an instruction from the prediction process management unit 31. The resource capacity prediction unit 36 predicts the resource capacity required for the IT service that satisfies all target values set for the service level items of the IT service. Further, the resource capacity prediction unit 36 stores the prediction result in the capacity estimation DB 24. The resource capacity prediction method will be described later.

  That is, in the resource capacity prediction apparatus 30 described above, the resource usage rate transition estimation unit 35 determines the reliability of the computer log calculated from the log related to the monitoring status of the computer executing the IT service processing, and the service of the IT service. Analyze logs related to level item measurements and logs related to the use of various computer resources. Further, the resource usage rate transition estimation unit 35 specifies a log related to measurement of service level items of a computer with high reliability and a log related to use of various resources. Then, the resource usage rate transition estimation unit 35 estimates the usage rates of various resources with respect to the measured values of the service level items based on those logs.

  Then, the resource capacity prediction unit 36 predicts a resource capacity that satisfies all target values set in the service level items based on the estimation result. With such a configuration, it is possible to accurately predict the resource capacity required for the IT service from the monitoring information of resources having different monitoring situations.

  Next, the configuration of the computer system 10 illustrated in FIG. 1 will be described. FIG. 4 is an explanatory diagram illustrating an example of a computer system. In the example illustrated in FIG. 4, a logical configuration of the computer system 10 is illustrated.

  The computer system 10 includes a plurality of computers (processing nodes 11). The computer system 10 is realized by, for example, a cloud computing system, a grid computing system, a parallel distributed computer, a super computer, a server computer, a personal computer, and a combination of any number thereof.

  The computer system 10 has one or more resources used for executing information processing. A resource 111 included in the computer system 10 is used to execute a physical or logical processor, a physical or logical memory, a physical or logical storage, a physical or logical communication interface, an information processing apparatus, a computer cluster, and other information processing. It may be a hardware resource.

  Further, the resource 111 included in the computer system 10 may be, for example, an application program, an operating system, or other software resources in which the number of users that can be used simultaneously is limited.

  Each of the one or more processing nodes 11 may be a virtual computer system that cooperates to perform processing of the IT service. This virtual computer system is configured by combining one or more resources 111.

  One or a plurality of resources included in each processing node 11 is secured by the execution management unit 26 of the resource management apparatus 20 in order to execute the IT service upon receiving user approval. Each resource 111 is configured by occupying only one resource or two or more resources (for example, a CPU, a memory, and a hard disk) or only a part of these resources.

  Each resource 111 includes an execution unit 112 that executes part or all of the processing of the IT service, a monitoring status and usage state of the resource of the execution unit 112, and a service level item, and the monitoring result is a resource management device. 20 monitoring history DBs 23 for monitoring. Further, the computer system 10 may include one or a plurality of resources 111 that are not secured for any IT service processing.

  Next, the contents of the service related data table stored in the service DB 22 will be described. FIG. 5 is an explanatory diagram showing an example of a service-related data table stored in the service DB. The service-related data table illustrated in FIG. 5 includes a “service ID” item, a “service level (hereinafter sometimes referred to as SL) type” item, an “SL request value” item, and an “infrastructure (hereinafter referred to as“ SL ”). And “infrastructure.”) “ID” item, “resource type” item, “start date and time” item, and “end date and time” item. The service-related data table may include a “safety factor” item.

  “Service ID” is identification information for uniquely identifying an IT service. “SL type” is identification information of a service level item of the IT service. The “SL requirement value” is a target value of the service level item of the IT service. The “infrastructure ID” is identification information of a computer that executes IT service processing. “Resource type” is identification information of the resource of the computer that executes the processing of the IT service. “Start date and time” is information indicating the start date and time of the IT service. “End date and time” is information indicating the end date and time of the IT service. The “safety factor” is a factor used for correcting the predicted value.

  For example, the service related data illustrated in the second row of FIG. 5 shows the following contents. Specifically, the IT service identified by the service ID “SID001” indicates that there is a target value of the service level item “Throughput 100,000 (Tx / sec)”. Further, in the processing of the IT service, the “CPU” resource of the computer identified by “WEB001” is used for a period of “2011/03/01 8: 00: 00: 00 to 2012/02/29 20:00:00”. Indicates that Further, the prediction result of the “CPU” resource of the computer identified by “WEB001” is corrected by the safety factor “1.30”. That is, this prediction result is multiplied by 1.30.

  The service related data table of the service DB 22 stores information related to the IT service included in the execution request received from the user terminal 40 by the request receiving unit 211. This execution request includes the processing content of the IT service that the user desires to execute and the target value of the IT service level item. The service related information of the IT service stored in the service related data table of the service DB 22 is acquired by the resource capacity prediction device 30 using the IT service identifier as a search key.

  Next, the contents of the execution history table stored in the execution state DB 27 will be described. FIG. 6 is an explanatory diagram illustrating an example of an execution history table stored in the execution state DB. The execution history table illustrated in FIG. 6 includes a “request ID” item, a “service ID” item, an “end flag” item, a “status” item, and an “update date / time” item.

  “Request ID” is identification information for uniquely identifying a user request. The “end flag” is information indicating whether or not all request processing has been completed. “Status” is information indicating the progress of request processing. The “update date / time” is information indicating the date / time when the “status” is updated.

  For example, the execution history record illustrated in the second row of FIG. 6 is an IT service execution history identified by the service ID “SID001”, and indicates that all the user request processes are completed execution histories. . Further, in the execution history illustrated in the second row of FIG. 6, the final status of the request is “resource allocation processing completion notified” and has been updated to “2011/12/01 8:00:00”. Indicates.

  In the execution history table of the execution state DB 27, the request reception unit 211 adds or updates the execution history record of the IT service according to the execution state of the request reception process of the IT service.

  The execution history record of the IT service is added or updated by the inquiry unit 212 in the execution history table of the execution state DB 27 according to the execution state of the inquiry processing of the IT service.

  The execution history record of the IT service is added or updated by the resource securing unit 213 in the execution history table of the execution state DB 27 according to the execution state of the resource securing process of the IT service.

  In addition, the execution history record of the IT service is added or updated by the execution management unit 26 in the execution history table of the execution state DB 27 according to the execution state of the resource securing process of the IT service.

  The execution history table of the IT service that the user desires to browse is referred to by the user terminal 40.

  Next, the contents of the resource management table stored in the resource DB 25 will be described. FIG. 7 is an explanatory diagram showing an example of a resource management table stored in the resource DB. The resource management table illustrated in FIG. 7 includes a “resource status” item, a “service ID” item, an “infrastructure ID” item, a “resource type” item, an “reservable resource amount” item, and a “reserved resource”. It includes an “amount” item, a “start date and time” item, and an “end date and time” item.

  The “resource status” is information indicating an execution state of various resource securing processes in the computer used to execute the IT service process. “Service ID” is identification information for uniquely identifying an IT service. “Infrastructure ID” is identification information for uniquely identifying a computer. The “resource type” is identification information for uniquely identifying various resources of the computer identified by the “infrastructure ID”. The “reservable resource amount” is information indicating a resource amount that is not used by the resource identified by the “resource type”. The “reserved resource amount” is information indicating the resource amount used by the resource identified by the “resource type” in the IT service identified by the “service ID”. The “start date / time” is information indicating the date / time when the use of the “reserved resource amount” can be started in the IT service identified by the “service ID”. The “end date and time” is information indicating the date and time when the use of the “reserved resource amount” ends in the IT service identified by the “service ID”.

  For example, the resource management data illustrated in the second row of FIG. 7 indicates data for managing the resource of the IT service identified by the service ID “SID001”. Further, since the “resource status” of the resource management data illustrated in the second row of FIG. 7 processes the IT service, the execution management unit 26 is executing the process of securing the secured resource amount “2 vCPU”. It shows that.

  Further, the resource management data illustrated in the second row of FIG. 7 indicates that the “CPU” resource of the computer identified by “WEB001” is “2011/03/01 8: 00: 00: 00 to 2012/2/29 20 : 0:00 ”is used. Furthermore, the resource management data illustrated in the second row of FIG. 7 includes a resource amount “2vCPU” that can be secured for the “CPU” resource of the computer identified by “WEB001”. It shows that 2vCPU can be added and used.

  In the resource DB 25, resource management is performed in order to secure resources used for executing the IT service processing designated by the request accepting unit 211 by the resource securing unit 213 that has received a request to secure resources from the request accepting unit 211. A table is added or updated.

  Further, the resource management table of the resource DB 25 is referred to by the execution management unit 26 and is used for execution management of securing processing of various resources for each IT service on the computer system 10.

  In addition, a resource management table is added or updated in the resource DB 25 by the execution management unit 26 according to the execution state of the IT service resource securing process.

  Next, the contents of the monitoring status log data table stored in the monitoring history DB 23 will be described. FIG. 8 is an explanatory diagram illustrating an example of a monitoring status log data table stored in the monitoring history DB. The monitoring status log data table illustrated in FIG. 8 includes a “log ID” item, an “infrastructure ID” item, a “log status type” item, and a “log status measurement value” item.

  The “log ID” is identification information for uniquely identifying the date and time when the logs related to the monitoring status of various resources of the computer used for executing the IT service processing are collected. “Infrastructure ID” is identification information for uniquely identifying the computer. The “log status type” is identification information for uniquely identifying the type of monitoring status. The “log status measurement value” is a value indicating a result measured for the type of monitoring status identified by the above “log status type”.

  The “log status measurement value” is converted and stored so that the value indicating the measured result indicates a higher value as the quality status is higher. Specifically, the “log status type” includes a type that indicates a higher quality status as the measured result value is larger and a type that indicates a higher quality status as the measured result value is smaller. In the case of a type that indicates a higher quality situation as the measured result value is larger, the “log situation measured value” stores a value indicating the measured result as it is.

  On the other hand, in the case of a type that indicates a higher quality situation as the measured result value is smaller, the value indicating the measured result is subtracted from the value indicating the lowest quality situation in the “log status measurement value”. Stores the value. For example, in a certain log status type, the value indicating the highest quality status is 0, and the value indicating the lowest quality status is 100. Here, when the value indicating the measured result is 0.99, 99.01 (= 100−0.99) is stored in the log status measurement value.

  For this reason, for each log status type, the larger the value indicating the measured result, the higher the quality or the lower quality information, the highest quality status value, and the lowest quality status value are referenced. It is predetermined as possible information.

  The monitoring status log data is a log relating to the monitoring status of various resources of the computer. As shown in Fig. 8, the monitoring status log data includes data indicating the usage status of IT services during busy periods and quiet periods, virus checks and updates for maintaining IT services, and operation of resident programs such as scan desks. Data indicating status, data indicating network communication status such as actual data transfer volume (effective throughput) and communication failure frequency, data indicating log collection frequency, loss frequency, delay frequency related to the use of various computer resources, Data indicating a monitoring infrastructure type category (hereinafter referred to as a monitoring infrastructure type category).

  The monitoring infrastructure type category means numerical data given when the types of operation management software that collects computer monitoring data (that is, monitors computers) are classified based on measurement errors. The accuracy of monitoring data (that is, measurement error) varies depending on the operation management software, and thus affects the reliability of the monitoring status. Therefore, the monitoring infrastructure type category can be used as a parameter that affects the reliability of the monitoring status. That is, the monitoring infrastructure type category can be said to be a value indicating the degree of precision of the monitoring data monitored by the computer.

  Specifically, operation management software for monitoring a computer is software for monitoring a server, a network, and an application. For example, NEC WebSAM MOperations (NEC, WebSAM, and MCOoperations are registered trademarks). It is done.

  For example, as the measurement error of the operation management software is smaller, a conversion table for assigning higher numerical data is determined in advance, and the precision of data monitored by each operation management software is digitized based on the conversion table. For example, “3” may be assigned as numerical data to NEC WebSAM MOperations described above, and “2” or “1” may be assigned as numerical data to other operation management software.

  For example, the monitoring status log data illustrated in the second row of FIG. 8 is log data collected at 2011/11/11 11:00:00, and a value “2011/11/11 11 indicating the collection time”. “00:00” indicates that “1320976680.000”, which is a value obtained by converting UNIX (registered trademark) time, is used as the “log ID”. The monitoring status log data illustrated in the second row of FIG. 8 indicates that the monitoring status “collection interval (min)” collected from the computer identified by the infrastructure ID “WEB001” is “10.00”. It is the log data which shows.

  A log related to the monitoring status of the computer executing the IT service processing is acquired by the monitoring status log acquisition unit 32 from the monitoring history DB 23 using the infrastructure ID which is the selector of the computer as a search key.

  The monitoring history DB 23 stores the monitoring status of the computers monitored by the execution unit 112 of the computer system 10.

  Next, the contents of the log data table stored in the monitoring history DB 23 will be described. FIG. 9 is an explanatory diagram showing an example of a log data table stored in the monitoring history DB. The log data table illustrated in FIG. 9 includes a “log ID” item, a “log type” item, an “infrastructure ID” item, and a “measurement value” item.

  The “log ID” is identification information for uniquely identifying the date and time when the log related to the measurement of the service level item of the computer executing the IT service process and the log related to the use of various resources are collected. “Log type” is identification information for uniquely identifying the type of the service level item or the type of resource used. “Infrastructure ID” is identification information for uniquely identifying the computer. “Measured value” is a value indicating the result of measurement for the type identified by “log type”.

  The log data includes a log relating to measurement of service level items of the computer and a log relating to use of various resources.

  The log relating to the measurement of the service level item of the computer is data indicating throughput, average request response time, average transaction processing time, number of transactions, operation rate, non-operation rate, average recovery time, average failure interval, and the like.

  Logs related to the use of various resources are data indicating CPU usage, number of CPU cores used, memory usage, memory usage, storage capacity, storage capacity usage, network usage, network usage, etc. It is.

  For example, the log data illustrated in the second line of FIG. 9 is log data collected at 2011/11/11 11:00:00, and a value “2011-11 / 11 11:00 indicating this collection time. “00” is converted to UNIX (registered trademark) time, “1320976680.000” is used as the “log ID”.

  The log data illustrated in the second row of FIG. 9 is log data indicating that the value of the log type “Throughput” collected from the computer identified by the infrastructure ID “LB001” is “38000”. .

  Logs related to measurement of service level items of computers that execute IT service processing and logs related to the use of various resources are acquired from the monitoring history DB 23 by the log acquisition unit 34 using the infrastructure ID that is the identification of the computer as a search key. .

  In the monitoring history DB 23, the execution unit 112 of the computer system 10 stores the measurement result of the service level item of the computer and the usage status of various resources as the monitoring result.

  Next, the contents of the resource transition estimation data table stored in the capacity estimation DB 24 will be described. FIG. 10 is an explanatory diagram illustrating an example of a resource transition estimation data table stored in the capacity estimation DB. The resource transition estimation data table includes a “service ID” item, an “SL type” item, a “resource type” item, an “infrastructure ID” item, a “relational expression” item, and a “log reliability” item. Including.

  “Service ID” is identification information for uniquely identifying an IT service. “SL type” is identification information of a service level item of the IT service. “Resource type” is identification information of the resource of the computer that executes the processing of the IT service. The “infrastructure ID” is identification information of a computer that executes IT service processing. The “relational expression” represents a change in the value of the resource identified by the “resource type” with respect to the value of the service level item identified by the “SL type”. A method for calculating the relational expression will be described later. The “log reliability” is a value indicating the reliability of the log used when estimating the “relational expression”.

  For example, the resource transition estimation data illustrated in the second line of FIG. 10 indicates that the transition of the usage rate of the IT service resource identified by the service ID “SID001” is estimated. Further, the resource transition estimation data illustrated in the second row of FIG. 10 has a relational expression indicating a fluctuation of the resource “CPU” of the computer identified by “WEB001” with respect to the value of “Throughput”. As a result of estimation using a log of “+0.20”, it is estimated that “y = 0.0004499781x + 3.545466”.

  In the resource transition estimation data table of the capacity estimation DB 24, the resource usage rate transition estimation unit 35 stores an estimation result indicating the transition of the resource usage rate with respect to the service level item of the IT service.

  In addition, the estimation result indicating the transition of the resource usage rate for each service level item of the IT service is obtained by the resource capacity prediction unit 36 using the service ID that is the identifier of the IT service as a search key as a resource transition of the capacity estimation DB 24. Obtained from the estimated data table.

  Next, the contents of the resource capacity prediction data table stored in the capacity estimation DB 24 will be described. FIG. 11 is an explanatory diagram illustrating an example of a resource capacity prediction data table stored in the capacity estimation DB. The resource capacity prediction data table includes a “service ID” item, an “SL type” item, an “SL required value” item, a “resource type” item, an “infrastructure ID” item, and an “estimated estimated value” item. Including.

  “Service ID” is identification information for uniquely identifying an IT service. “SL type” is identification information of a service level item of the IT service. The “SL requirement value” is a target value of the service level item of the IT service. “Resource type” is identification information of the resource of the computer that executes the processing of the IT service. The “infrastructure ID” is identification information of a computer that executes IT service processing. The “predicted estimated value” is a computer identified by the “infrastructure ID” necessary to satisfy the “SL requested value” that is the requested value (target value) of the service level item identified by the “SL type”. This is a value obtained by estimating the resource capacity of the resource identified by the “resource type”.

  For example, the resource capacity prediction data illustrated in the second row of FIG. 11 indicates that the resource capacity of the IT service identified by the service ID “SID001” is predicted. Further, the resource capacity prediction data illustrated in the second row of FIG. 11 is stored in the computer identified by “WEB001” in order to satisfy the target value “Throughput 100,000 (Tx / sec)” of the service level item. The prediction result that the necessary “CPU” resource is “48.54328” is shown.

  In the resource capacity prediction data table of the capacity estimation DB 24, the resource capacity prediction unit 36 stores the prediction results of the resource capacity necessary to satisfy the target value of the service level item of the IT service.

  The IT service resource capacity prediction result instructed by the request reception unit 211 to the resource capacity prediction device 30 is estimated by the inquiry unit 212 using the service ID that is the identifier of the IT service as a search key. It is acquired from the resource capacity prediction data table of DB24.

  FIG. 12 is an explanatory diagram showing an example of log reliability. As illustrated in FIG. 12, the log reliability is calculated for each log ID by the log reliability calculation unit 33.

  Next, the operation of the resource management device 20 illustrated in FIG. 1 will be described. FIG. 13 is a flowchart illustrating an example of an operation in which the resource management device 20 of the distributed system illustrated in FIG. 1 performs resource capacity prediction.

  The execution management unit 26 adds or updates the resource management table to the resource DB 25 according to the reservation status of various resources of the computer system 10, thereby matching the resource management table of the various resources of the computer system 10. Is stored in the resource DB 25.

  First, the request receiving unit 211 receives a request for reviewing whether or not the resource capacity reserved for the IT service is appropriate from the user (specifically, the user terminal 40). The request reception unit 211 adds a request reception execution history related to IT service resource capacity analysis processing to the execution state DB 27. The request reception unit 211 instructs the resource capacity prediction apparatus 30 to predict the resource capacity necessary for the IT service. The request reception unit 211 updates the status of the execution history related to the resource capacity analysis process of the IT service to “wait for capacity analysis process” (step S101).

  Next, the resource capacity prediction device 30 receives a prediction instruction for the resource capacity of the IT service from the resource management device 20. The resource capacity prediction device 30 predicts the resource capacity necessary for the IT service in order to satisfy the target value of the service level item of the IT service. Then, the resource capacity prediction device 30 stores the resource capacity prediction result in the capacity estimation DB 24 (step S102).

  Next, the inquiry unit 212 compares whether there is a difference between the resource capacity reserved for the IT service and the prediction result (step S103). If there is a difference in resource capacity (YES in step S103), the process proceeds to step S104. On the other hand, when there is no difference in resource capacity (NO in step S103), the process proceeds to step S107.

  When there is a difference in resource capacity (YES in step S103), the inquiry unit 212 inquires of the user whether or not to change the reserved resource capacity to the predicted resource capacity. Then, the inquiry unit 212 updates the status of the execution history related to the resource capacity analysis processing of the IT service to “capacity change inquiry” (step S104).

  On the other hand, when there is no difference in resource capacity (NO in step S103), the inquiry unit 212 updates the status of the execution history related to the resource capacity analysis process of the IT service to “capacity analysis process complete”. In addition, the inquiry unit 212 updates the execution history end flag to a flag indicating that all analysis processes have ended. The inquiry unit 212 notifies the user terminal 40 that the resource analysis process has been completed, and updates the status of the execution history to “capacity analysis process completion notified” (step S107).

  After the process of step S104 is performed, the inquiry unit 212 determines whether an instruction to change the resource capacity has been received (step S105). When an instruction to change the resource capacity is received (YES in step S105), the process proceeds to step S106. On the other hand, when an instruction not to change the resource capacity is received (NO in step S105), the process proceeds to step S107.

  When receiving an instruction not to change the resource capacity (NO in step S105), the inquiry unit 212 updates the status of the execution history related to the IT service resource capacity analysis processing to “capacity change inquiry completed”. . In addition, the inquiry unit 212 updates the execution history end flag to a flag indicating that all analysis processes have ended. The inquiry unit 212 notifies the user terminal 40 that the resource analysis processing has been completed, and updates the status of the execution history to “capacity change inquiry completion notified” (step S107).

  On the other hand, when receiving an instruction to change the resource capacity (YES in step S105), the inquiry unit 212 changes the secured resource capacity to the resource capacity of the prediction result so as to change the secured resource capacity 213. To instruct. Then, the inquiry unit 212 updates the status of the execution history related to the resource capacity analysis process of the IT service to “start resource securing process”.

  Furthermore, the execution management unit 26 periodically checks each status in the resource management table of the resource DB 25. When there is resource management data whose status is “request accepted”, the execution management unit 26 updates the status of the resource management data to “in process of securing”. In addition, the execution management unit 26 executes the process of securing the computer resources so that the “reserved resource amount” set in the resource management data is obtained. Then, the execution management unit 26 updates the status of the resource management data to “normal” when the resource securing process is completed. Furthermore, the execution management unit 26 updates the status of the IT service execution history to “resource reservation processing complete” (step S106).

  After the processing of step S106, the resource securing unit 213 periodically checks the execution history whose status in the execution history table of the execution state DB 27 is “resource securing processing complete”. Then, the resource securing unit 213 updates the end flag of the execution history to a flag indicating that all analysis processes have been completed. The resource securing unit 213 notifies the user terminal 40 that the resource securing process has been completed, and updates the status of the execution history to “resource secured process completion notified” (step S107).

  Next, the operation of the resource capacity prediction apparatus of this embodiment will be described. FIG. 14 is a flowchart illustrating an example of the operation of the resource capacity prediction apparatus according to the present embodiment.

  The prediction process management unit 31 receives a prediction instruction for IT service resource capacity from the front-end processing unit 21 of the resource management apparatus 20. And the prediction process management part 31 acquires the relevant information of IT service from service DB22. The prediction process management unit 31 instructs the monitoring status log acquisition unit 32 and the log acquisition unit 34 to acquire a log (step S201).

  The monitoring status log acquisition unit 32 acquires, from the monitoring history DB 23, logs related to the monitoring status at a plurality of times of the computer executing the IT service processing according to an instruction from the prediction processing management unit 31, and sends the logs to the resource usage rate transition estimation unit 35. Send. In addition, the monitoring status log acquisition unit 32 notifies the prediction processing management unit 31 that acquisition of the monitoring status log has been completed (step S202).

  When the prediction process management unit 31 receives the monitoring status log acquisition from the monitoring status log acquisition unit 32, the prediction processing management unit 31 instructs the log reliability calculation unit 33 to calculate the log reliability related to the IT service resource. .

  The log reliability calculation unit 33 analyzes a log related to the monitoring status of various resources at a plurality of times of the computer executing the IT service processing according to an instruction from the prediction processing management unit 31. Then, the log reliability calculation unit 33 calculates the log reliability at a plurality of times of the computer executing the IT service processing, and transmits the log reliability to the resource usage rate transition estimation unit 35. Further, the log reliability calculation unit 33 notifies the prediction process management unit 31 that the calculation of the log reliability has been completed (step S203).

  On the other hand, the log acquisition unit 34, in response to an instruction from the prediction processing management unit 31, logs related to measurement of service level items at a plurality of times of a computer executing IT service processing and use of various resources at the plurality of times of the computer. Is acquired from the monitoring history DB 23 and transmitted to the resource usage rate transition estimation unit 35. In addition, the log acquisition unit 34 notifies the prediction processing management unit 31 that the log acquisition is completed (step S204).

  After the processing of step S203 and step S204, the prediction process management unit 31 receives notification of completion of acquisition from the log acquisition unit 34 and calculation completion from the log reliability calculation unit 33, and executes the IT service. The resource usage rate transition estimation unit 35 is instructed to estimate the transition of usage rates of various resources.

  The resource usage rate transition estimation unit 35, in response to an instruction from the prediction processing management unit 31, logs reliability at a plurality of times of the computer executing the IT service processing and at a plurality of times of the computer executing the IT service processing. From the log relating to the measurement of the service level item and the log relating to the use of various resources at a plurality of times of the computer, the log relating to the measurement of the service level item of the computer having high reliability and the log relating to the use of various resources are specified. Then, the resource usage rate transition estimation unit 35 estimates the usage rate of the computer resource with respect to the measured value of the service level item based on the logs. Here, the safety coefficient used by the resource capacity prediction unit 36 described later may be adjusted.

  Further, the resource usage rate transition estimation unit 35 stores the estimation result in the capacity estimation DB 24. The resource usage rate transition estimation unit 35 notifies the prediction processing management unit 31 that the transition estimation of the usage rates of various resources has been completed (steps S205 and S206).

  The prediction process management unit 31 receives from the resource usage rate transition estimation unit 35 that the transition estimation of the usage rates of various resources has been completed, and instructs the resource capacity prediction unit 36 to predict the resource capacity. The resource capacity prediction unit 36 acquires, from the capacity estimation DB, an estimation result of the resource usage rate for each service level item of the IT service according to an instruction from the prediction process management unit 31. Then, the resource capacity prediction unit 36 predicts a resource capacity that satisfies all target values set in each service level item of the IT service.

  The resource capacity prediction unit 31 stores the prediction result in the capacity estimation DB 24. Further, the resource capacity prediction unit 31 notifies the prediction process management unit 31 that the resource capacity prediction is completed (step S207).

  Next, the operation of the log reliability calculation unit 33 of this embodiment will be described in detail. FIG. 15 is a flowchart illustrating an example of the operation of the log reliability calculation unit. The log reliability calculation unit 33 starts the log reliability calculation process in response to an instruction from the prediction process management unit 31. Note that it is assumed that the log reliability calculation unit 33 receives, from the monitoring status log acquisition unit 32, logs related to monitoring statuses at a plurality of times of the computer executing the IT service processing.

  The log reliability calculation unit 33 sets, for each log status type (that is, the type of monitoring status), a log status measurement value (that is, a value measured according to the type of monitoring status) for the log related to the monitoring status. An average value, a standard deviation value, and a variation coefficient are calculated (step S301).

  The log reliability calculation unit 33 assumes that a log status type with a large variation in log status measurement values is a log status type to be noted, and based on the ratio of the coefficient of variation of each log status type, Set the weight. For example, assume that the coefficient of variation of three log status types is 0.4, 0.6, and 1.0. When the sum of the weights of all log statuses is 1, the respective weights are 0.2, 0.3, and 0.5 (step S302).

The log reliability calculation unit 33 uses the average value x _arg calculated in step S301, the standard deviation value x _stdev , and the weight w calculated in step S302 to monitor the monitoring status log identified by the same log ID. The log reliability is calculated for the object based on Equation 1 and Equation 2 below. Expression 1 shown below is a log reliability calculation expression. Also, Equation 2 shown below is an equation for standardizing log status measurement values.

  The log reliability calculation unit 33 calculates the log reliability illustrated in FIG. 12 for each log ID, and transmits the log reliability to the resource usage rate transition estimation unit 35. Further, the log reliability calculation unit 33 notifies the prediction process management unit 31 that the calculation of the log reliability has been completed.

  Next, the operation of the resource usage rate transition estimation unit 35 of this embodiment will be described in detail. FIG. 16 is a flowchart illustrating an example of the operation of the resource usage rate transition estimation unit. The resource usage rate transition estimation unit 35 starts a process of estimating transitions of usage rates of various resources with respect to the service level item of the IT service according to an instruction from the prediction processing management unit 31.

  The resource usage rate transition estimation unit 35 receives IT service related information from the prediction processing management unit 31, and receives a log related to measurement of IT service level items and a log related to usage rates of various resources from the log acquisition unit 34. It is assumed that IT reliability log data is received from the log reliability calculation unit 33.

  Further, it is assumed that a threshold for determining that the log reliability is high (hereinafter, referred to as a reliability threshold) is set in the resource usage rate transition estimation unit 35. In addition, the resource usage rate transition estimation unit 35 is set with a threshold (Counter array threshold; hereinafter referred to as a frequency threshold), which will be described later, for determining whether or not there is a certain frequency or more. To do.

  Based on the log reliability threshold, the resource usage rate transition estimation unit 35 selects only the log related to the measurement of the service level item having the log reliability equal to or higher than the threshold and the log related to the use of various resources (step S401). .

  The resource usage rate transition estimation unit 35 creates a correspondence table that sets different SL types and different resource types as a set from the related information of the IT service. Then, the resource usage rate transition estimation unit 35 determines whether one set (SL type, resource type) of log data can be acquired from the log selected in Step S401 (Step S402). If it can be obtained, the process proceeds to step S403.

  When the log data of the selected set can be acquired (“when it can be acquired” in step S402), the resource usage rate transition estimation unit 35 performs a Hough transform on the two-dimensional space having the SL type as the X axis and the resource type as the Y axis. Is converted into a two-dimensional space of (angle θ, distance ρ). Then, the resource usage rate transition estimation unit 35 stores the angle θ and the distance ρ in a two-dimensional array. Specifically, the resource usage rate transition estimation unit 35 calculates (θ1, ρ1) that satisfies ρ = x1 · cos θ + y1 · sin θ with respect to (x1, y1).

  Then, the resource usage rate transition estimation unit 35 stores the calculated (θ1, ρ1) in the Counter [θ] [ρ] array. That is, the counter [θ] [ρ] array stores the frequency values of the angle θ and the distance ρ.

  The search range of the angle θ may be 0 to π, or π / 2 to π. Further, the resource usage rate transition estimation unit 35 may store the calculated (θ1, ρ1) in the Counter [θ] [ρ] array only when the distance ρ is a distance ρ1 of 0 or more. (Step S403).

  The resource usage rate transition estimation unit 35 obtains the maximum angle θ1 ′ and the distance ρ1 ′ that are equal to or higher than a certain frequency from the Counter [θ] [ρ] array based on the threshold value of the Counter array (that is, the frequency threshold value). The search range of the angle θ may be 0 to π, or π / 2 to π. Further, the search range of the distance ρ may be from 0 to the square sum d of the maximum value of the x axis and the maximum value of the y axis, or from −d to d (step S404).

  The resource usage rate transition estimation unit 35 obtains (θ1 ′, ρ1 ′) obtained in step S404 from the two-dimensional space of (angle θ, distance ρ) by the Hough inverse transformation, and sets the resource type to the SL type as the X axis. Convert to a two-dimensional space with the Y axis.

  Specifically, the resource usage rate transition estimation unit 35 calculates (a1, b1) that satisfies Equation 3 below for (θ1 ′, ρ1 ′). That is, the relational expression shown in Expression 3 can be said to be the usage rate of the computer resource with respect to the measured value of the service level item.

  y = a1 · x + b1 = −1 / tan θ1 ′ · x + ρ1 ′ / sin θ1 ′ (Expression 3)

  Then, the resource usage rate transition estimation unit 35 stores the estimation result of y = a1 · x + b1, which is a relational expression of one set (SL type X, resource type Y), in the capacity estimation DB 24. The estimation result stored here corresponds to the “relational expression” illustrated in FIG. After that, the resource usage rate transition estimation unit 35 repeats the processes after S402.

  In step S402, when the log data of the selected set cannot be acquired (“cannot be acquired” in step S402), the resource usage rate transition estimation unit 35 performs processing for estimating the usage rates of various resources for the service level item of the IT service. The prediction processing management unit 31 is notified of the completion.

  Next, the operation of the resource capacity prediction unit 36 of this embodiment will be described in detail. FIG. 17 is a flowchart illustrating an example of the operation of the resource capacity prediction unit. The resource capacity prediction unit 36 starts a process of predicting a resource capacity that satisfies all target values of each service level item of the IT service according to an instruction from the prediction process management unit 31.

  Note that the capacity estimation DB 24 stores transition estimation results (that is, relational expressions) of the usage rates of various resources for each service level item of the IT service. This transition estimation result is referred to by the resource capacity prediction unit 36. Further, it is assumed that the resource capacity prediction unit 36 includes an array that stores provisional prediction values of each resource type. The initial value of the provisional prediction value is 0.

  The resource capacity prediction unit 36 creates a service level correspondence table in which the SL type and the SL request value (that is, the target value of the service level item) are one set from the IT service related information (step S501).

  The resource capacity prediction unit 36 determines whether or not one set of SL type and SL request value can be acquired from the service level correspondence table created in step S501 (step S502). If it can be obtained, the process proceeds to step S503, and if it cannot be obtained, the process proceeds to step S505.

  When one set of SL type and SL request value can be acquired (YES in step S502), the resource capacity prediction unit 36 selects each resource type corresponding to the one set of SL type and SL request value acquired in step S502. Get the relational expression. Then, the resource capacity prediction unit 36 calculates a resource capacity prediction value from the SL request value and the relational expression (step S503).

  For example, as illustrated in FIG. 11, it is assumed that the SL request value of the SL type “Throughput” is “100000”. Further, it is assumed that the relational expression of the resource type “CPU” of the SL type “Throughput” and the infrastructure ID “WEB001” is the relational expression “y = 0.0004499781x + 3.545466” illustrated in FIG. At this time, the resource capacity prediction unit 36 calculates the resource capacity prediction value as “48.54328 (= 0.0004499781 * 100000 + 3.545466)”.

  The resource capacity prediction unit 36 compares the resource capacity prediction value of each resource type calculated in step S503 with the provisional prediction value of each resource type. If the resource capacity prediction value calculated in step S503 is larger than the provisional prediction value, the resource capacity prediction unit 36 updates the prediction value as the provisional prediction value. On the other hand, when the resource capacity prediction value calculated in step S503 is smaller than the provisional prediction value, the resource capacity prediction unit 36 does not update the provisional prediction value. After that, the resource capacity prediction unit 36 repeats the processes after step S502.

  In step S502, when one set of SL type and SL request value cannot be acquired (NO in step S502), the resource capacity prediction unit 36 determines the provisional prediction value of each resource type as a prediction value.

  Note that the safety coefficient of each resource type may be set in advance, and the resource capacity prediction unit 36 may determine the corrected predicted value obtained by correcting the provisional predicted value of each resource with the safety coefficient as the predicted value. For example, when the safety factor is “1.3” and the resource capacity provisional prediction value is “48.54328”, the resource capacity prediction value is “63.106264 (= 48.54328 * 1.3)”. Calculated (step S505).

  The source capacity estimation unit 36 stores the estimation result in the capacity estimation DB 24. Further, the resource capacity prediction unit 36 notifies the prediction process management unit 31 that the estimation of the resource capacity is completed.

  As described above, according to the present embodiment, the log reliability calculation unit 33 obtains a log related to the resource monitoring status from each computer constituting the computer system, and at a plurality of times of the computers included in the log. Based on the monitoring status information, log reliability at a plurality of times is calculated. Then, the resource usage rate transition estimation unit 35 acquires a log relating to resource use and a log relating to measurement of service level items from each of the above computers, and information on resource usage rates and measured values of service level items at a plurality of times. And the log resource reliability at the plurality of times calculated by the log reliability calculation unit 33 are used to estimate the resource usage rate of the computer. Therefore, it is possible to accurately predict the resource capacity required for the IT service from the monitoring information of resources having different monitoring conditions.

  Further, the resource capacity prediction unit 36 refers to the service related data including the target value of the service level item of the IT service, and based on the estimation result (relational expression) of the resource usage rate transition estimation unit 35, the service of the IT service A resource capacity that satisfies all target values of level items may be predicted.

  In other words, in the present embodiment, the log reliability calculation unit 33 acquires logs related to resource monitoring statuses from a computer, and based on information on monitoring statuses of various resources at a plurality of times included in these logs. The reliability of the log at the time is calculated. The monitoring status here includes data indicating the usage status of IT services during busy and quiet periods, data indicating the status of operating resident programs such as virus checks and updates for maintaining IT services, and scan desks, And data indicating the network communication status such as the actual data transfer amount and communication failure frequency.

  In the present embodiment, the resource usage rate transition estimation unit 35 acquires a log related to measurement of service level items and a log related to resource usage from the computer. Further, the resource usage rate transition estimation unit 35 includes information on measured values of service level items at a plurality of times included in these logs, information on resource usage rates at a plurality of times of the computer, and a log reliability calculation unit. Based on the log reliability at a plurality of times calculated by 33, a log related to measurement of service level items of a computer having a log reliability equal to or higher than a certain value and a log related to use of various resources are specified. And the resource usage rate transition estimation part 35 estimates the usage rate of various resources with respect to the measured value of a service level item based on those logs. Here, the measured values of the service level items include throughput, average request response time, average transaction processing time, number of transactions, operation rate, non-operation rate, average recovery time, average failure interval, and the like. The resource usage rate here includes a CPU usage rate, a memory usage amount, a storage capacity of a storage device, a network usage amount, and the like.

  Further, in the present embodiment, the resource capacity prediction unit 36 refers to the service related data including the target value of the service level item, and with respect to the measured value of the service level item of the computer estimated by the resource usage rate transition estimation unit 35 Based on the estimation result of the resource usage rate of the computer, the resource capacity that satisfies all the target values of the service level items is predicted.

  By including these elements, the resource capacity prediction apparatus 30 according to the present embodiment can accurately predict the resource capacity required for the IT service from the monitoring information of resources having different monitoring conditions.

  Next, the hardware configuration of the resource capacity prediction apparatus of this embodiment will be described. FIG. 18 is an explanatory diagram illustrating an example of the hardware configuration of the resource capacity prediction apparatus and the resource management apparatus illustrated in FIG. Hereinafter, a general configuration of an information processing apparatus using a computer as a typical example will be described. However, in the case of a dedicated machine or an embedded device, the minimum necessary configuration is selected according to the environment.

  The resource capacity prediction device and the resource management device are connected to each other by a host controller 51, a CPU 52, a bus line (not shown), a communication I / F 53, a main memory (RAM (Random Access Memory)) 54, a BIOS. (Basic Input / Output System) 55, parallel port 56, USB (Universal Serial Bus) port 57, graphic controller 58, VRAM (Video RAM) 59, audio processor 60, I / O controller 61, keyboard and mouse adapter 62 Etc. are provided. For example, storage means such as a flexible disk (FD) drive 63, a hard disk drive 64, an optical disk drive 65, and a semiconductor memory 66 can be connected to the I / O controller 61.

  A display device 71 is connected to the graphic controller 58. As an option, an amplifier circuit 67 and a speaker 68 are connected to the audio processor 60.

  The BIOS 55 stores a boot program executed by the CPU when the resource capacity prediction device and the resource management device are started up, a program depending on the hardware of the resource capacity prediction device and the resource management device, and the like. The FD drive 63 reads a program or data from the flexible disk 69 and provides it to the main memory 54 or the hard disk drive 64 via the I / O controller 61.

  As the optical disk drive 65, for example, a Blu-ray drive, a DVD (Digital Versatile Disc) -ROM (Read Only Memory) drive, a CD (Compact Disc) -ROM drive, a DVD-RAM drive, or a CD-RAM drive is used. In this case, the optical disk 70 corresponding to each drive is used. The optical disk drive 65 can also read a program or data from the optical disk and provide it to the main memory 54 or the hard disk drive 64 via the I / O controller 61.

  A computer program provided to the resource capacity prediction device and the resource management device is stored in a recording medium such as a flexible disk, an optical disk, or a memory card and provided by a user. This computer program is read from the recording medium via the I / O controller or downloaded via the communication I / F, so that the computer program is installed and executed in the resource capacity prediction device and the resource management device. The operation that the computer program causes the information processing apparatus to perform is the same as the operation in the apparatus that has already been described.

  The computer program described above may be stored in an external storage medium. As the storage medium, for example, a magneto-optical recording medium such as MD (MiniDisc) and a tape medium are used in addition to a flexible disk, an optical disk, and a memory card. In addition, a storage device such as a hard disk or an optical disk library provided in a server system connected to a dedicated communication line or the Internet is used as a recording medium, and the computer program is transferred to the resource capacity prediction device and the resource management device via the communication line. May be provided.

  The above examples have mainly described the hardware configurations of the resource capacity prediction device and the resource management device. In addition, by installing a program having the function described in the information processing apparatus in a computer and causing the computer to operate as the information processing apparatus, the same function as the information processing apparatus described above can be realized. . Therefore, the information processing apparatus described as one embodiment in the present invention can also be realized by the computer program.

  The resource capacity prediction device and the resource management device illustrated in FIG. 1 can be realized as hardware, software, or a combination of hardware and software. A typical example of a method implemented by combining hardware and software is a method implemented by a computer system having a predetermined program. In this case, when a predetermined program is loaded and executed on the computer system, the program causes the computer system to execute processing according to the present invention.

  This program is composed of a group of instructions that can be expressed in any language, code, or notation. Such instructions can be either or both of the following: (1) conversion to another language, code, or notation; (2) replication to other media; Can be executed after the

  Of course, the present invention includes not only such a program itself but also a program product including a medium on which the program is recorded. The program for executing the functions of the present invention is stored in an arbitrary computer-readable medium such as a flexible disk, MO, CD-ROM, DVD, Blu-ray disk, hard disk device, ROM, MRAM (Magnetoresistive RAM), and RAM. Can do.

  Such a program can be downloaded from another computer system connected via a communication line or copied from another medium for storage in a computer-readable medium. Further, such a program can be compressed or divided into a plurality of pieces and stored in a single or a plurality of recording media.

  For example, in the case of the resource capacity prediction apparatus of the present embodiment, the log reliability calculation unit 33, the resource usage rate transition estimation unit 35, and the resource capacity prediction unit 36 operate according to a program (resource capacity prediction program). This is realized by the CPU 52 of the computer. For example, the program is stored in the storage means described above of the resource capacity prediction apparatus, and the CPU 52 reads the program, and according to the program, the log reliability calculation unit 33, the resource usage rate transition estimation unit 35, and the resource capacity The prediction unit 36 may operate. Further, each of the log reliability calculation unit 33, the resource usage rate transition estimation unit 35, and the resource capacity prediction unit 36 may be realized by dedicated hardware.

  Next, the outline of the present invention will be described. FIG. 19 is a block diagram showing an outline of a resource capacity prediction apparatus according to the present invention. A resource capacity prediction apparatus (for example, resource capacity prediction apparatus 30) according to the present invention predicts the resource capacity of a computer system (for example, computer system 10) composed of a plurality of computers (for example, processing nodes 11). .

  The resource capacity prediction apparatus according to the present invention acquires a first log (for example, monitoring status log data illustrated in FIG. 8) regarding a resource monitoring status from a computer, and a plurality of computers included in the first log. Log reliability calculation means 81 (for example, log reliability calculation unit 33) for calculating the reliability of logs at a plurality of times based on the monitoring status information at the time, and a second log related to resource use from the computer (For example, a log related to the use of various resources included in the log data illustrated in FIG. 9) and a third log related to the measurement of the service level item of the computer (for example, the service level item included in the log data illustrated in FIG. 9) Resource logs at multiple times of the computer included in the second log Usage rate information, information on measured values of service level items at a plurality of times of the computer included in the third log, and log reliability at a plurality of times calculated by the log reliability calculation means 81. Based on this, resource utilization rate transition estimating means 82 (for example, resource utilization rate transition estimating unit 35) for estimating the resource utilization rate (for example, relational expression) of the computer is provided.

  With such a configuration, it is possible to accurately predict the resource capacity required for the IT service from the monitoring information of resources having different monitoring situations.

  Further, the resource capacity prediction apparatus refers to service-related information (for example, service-related data illustrated in FIG. 5) including the target value of the service level item of the IT service, and is estimated by the resource usage rate transition estimation unit 82 Resource capacity prediction that predicts resource capacity that satisfies all target values of IT service level items based on estimation results (for example, relational expressions) of computer resource usage rates for measured values of computer service level items Means (for example, resource capacity prediction unit 36) may be provided.

  Further, the resource usage rate transition estimation unit 82 sets a condition that the log reliability at a plurality of times calculated by the log reliability calculation unit 81 is constant (for example, a condition that the reliability is greater than a predetermined reliability). If so, the utilization rate of the computer resource may be estimated. Specifically, the resource usage rate transition estimation unit 82 may select a log having a reliability higher than or equal to a threshold value from the acquired second logs.

  The first log related to the resource monitoring status includes data indicating the usage status of the IT service, data indicating the operating status of the resident program for maintaining the IT service, data indicating the network communication status, and various resources of the computer. It may be data including any one of data indicating the collection frequency, loss frequency, and delay frequency of the log related to use, and data indicating the monitoring infrastructure type category indicating the degree of precision of the monitoring data monitored by the computer.

  In addition, the log reliability calculation unit 81 includes each measurement value (for example, log status measurement value) at a plurality of times measured according to the type of monitoring status (for example, log status type) included in the first log. Among them, the log reliability may be calculated based on a calculation formula in which the weighting factor of the log reliability calculation formula is increased as the measurement value has a larger degree of variation in the measurement value.

  As mentioned above, although this invention was demonstrated according to embodiment, this invention is not limited to embodiment mentioned above. In addition, the effects described in the embodiments of the present invention only list the most preferable effects resulting from the present invention, and the effects of the present invention are not limited to those described in the embodiments of the present invention.

  Moreover, although the resource capacity prediction apparatus of this invention was demonstrated using the said embodiment, the technical scope of the resource capacity prediction apparatus of this invention is not limited to the range described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above embodiment. It is apparent from the scope of the claims that the embodiments added with such changes or improvements can be included in the technical scope of the present invention.

  The present invention is applicable to a resource capacity apparatus that manages the capacity of a computer system.

DESCRIPTION OF SYMBOLS 10 Computer system 11 Processing node 20 Resource management apparatus 21 Front end processing part 211 Request reception part 212 Inquiry part 213 Resource securing part 22 Service database (service DB)
23 Monitoring history database (Monitoring history DB)
24 Capacity estimation database (Capacity estimation DB)
25 Resource Database (Resource DB)
26 execution management unit 27 execution state database (execution state DB)
DESCRIPTION OF SYMBOLS 30 Resource capacity prediction apparatus 31 Prediction process management part 32 Monitoring status log acquisition part 33 Log reliability calculation part 34 Log acquisition part 35 Resource utilization rate transition estimation part 36 Resource capacity prediction part 40 User terminal 99 Communication network network

Claims (10)

A resource capacity prediction device for predicting the resource capacity of a computer system composed of a plurality of computers,
A first log related to the resource monitoring status is acquired from the computer, and the reliability of the log at a plurality of times is calculated based on the monitoring status information of the computer at a plurality of times included in the first log. Log reliability calculation means for
A second log relating to the use of resources and a third log relating to the measurement of service level items of the computer are obtained from the computer, and the usage rate of the resource at a plurality of times of the computer included in the second log is obtained. Information, information on measured values of service level items at a plurality of times of the computer included in the third log, and log reliability at a plurality of times calculated by the log reliability calculation means. And a resource usage rate transition estimating means for estimating the resource usage rate of the computer. Based on the estimation result of the resource usage rate of the computer with respect to the measured value of the service level item of the computer estimated by the resource usage rate transition estimation means with reference to the service related information including the target value of the service level item of the IT service The resource capacity prediction apparatus according to claim 1, further comprising resource capacity prediction means for predicting resource capacity that satisfies all target values of service level items of the IT service. The resource usage rate transition estimation unit estimates the resource usage rate of the computer when log reliability at a plurality of times calculated by the log reliability calculation unit satisfies a certain condition. The resource capacity prediction apparatus according to 2. The first log related to the monitoring status of resources relates to data indicating the usage status of the IT service, data indicating the operating status of the resident program for maintaining the IT service, data indicating the network communication status, use of various resources of the computer The data including one or more of data indicating log collection frequency, loss frequency, and delay frequency, and data indicating a monitoring infrastructure type category indicating a degree of precision of monitoring data monitored by a computer. 4. The resource capacity prediction apparatus according to claim 1. 5. The resource capacity prediction device according to claim 1, wherein the resource usage rate transition estimation unit selects a log having a reliability level equal to or higher than a threshold value from the acquired second logs. . The log reliability calculation means is configured such that, among the measurement values at a plurality of times measured according to the type of monitoring status included in the first log, the measurement reliability of the log is larger for the measurement value having a large variation in the measurement value. The resource capacity prediction apparatus according to any one of claims 1 to 5, wherein log reliability is calculated based on a calculation formula that increases a weight coefficient of the calculation formula. A resource capacity prediction method for predicting the resource capacity of a computer system composed of a plurality of computers,
From the computer, obtain a first log regarding the monitoring status of resources,
Based on the monitoring status information of the computer at a plurality of times included in the first log, the reliability of the log at a plurality of times is calculated,
From the computer, obtain a second log relating to resource usage and a third log relating to measurement of service level items of the computer,
Information on resource usage at a plurality of times of the computer included in the second log and information on measured values of service level items at a plurality of times of the computer included in the third log are calculated. A resource capacity prediction method characterized by estimating a resource usage rate of the computer based on log reliability at a plurality of times. The service level item of the IT service is referred to based on the estimated result of the usage rate of the computer resource with respect to the measured value of the estimated service level item of the computer with reference to the service related information including the target value of the service level item of the IT service. The resource capacity prediction method according to claim 7, wherein a resource capacity that satisfies all target values is predicted. A resource capacity prediction program applied to a prediction computer for predicting the resource capacity of a computer system composed of a plurality of computers,
In the prediction computer,
A first log related to the resource monitoring status is acquired from the computer, and the reliability of the log at a plurality of times is calculated based on the monitoring status information of the computer at a plurality of times included in the first log. Log reliability calculation processing, and
A second log relating to the use of resources and a third log relating to the measurement of service level items of the computer are obtained from the computer, and the usage rate of the resource at a plurality of times of the computer included in the second log is obtained. Information, information on measured values of service level items at a plurality of times of the computer included in the third log, and log reliability at a plurality of times calculated by the log reliability calculation process A resource capacity prediction program for executing a resource usage rate transition estimation process for estimating the resource usage rate of the computer. To the prediction computer,
Based on the estimation result of the resource usage rate of the computer with respect to the measured value of the service level item of the computer estimated by the resource usage rate transition estimation process with reference to the service related information including the target value of the service level item of the IT service The resource capacity prediction program according to claim 9, wherein a resource capacity prediction process for predicting a resource capacity that satisfies all target values of service level items of the IT service is executed.

Images