JP2010117757A - Performance monitoring system and performance monitoring method - Google Patents

Abstract

A performance monitoring system and a performance monitoring method are provided that make it possible to grasp in real time the response time of each server or internal network on a processing path for a specific access.
The server system includes a server system and a performance monitoring server that provide services by a plurality of servers, and each server takes over the request ID between the server and a means for assigning a request ID to access. Means, and a means for outputting the individual log 121 including the request ID and time at the start and end of processing, and the performance monitoring server 200 collects the individual log 121, A log is identified from data by associating with the request ID, and based on the information on the start time and the end time of the process in each server, the server, the network between the servers, and the entire processing path And a log totaling unit 220 for calculating the response time.
[Selection] Figure 1

Description

  The present invention relates to a performance monitoring system and a performance monitoring method for monitoring the performance of a server system that provides services to a client terminal via a network through cooperation of a plurality of servers, and in particular, response time in each server and the network between the servers. The present invention relates to a technology effective when applied to a performance monitoring system and a performance monitoring method for monitoring performance based on the above.

  At present, various services including financial transactions and the like are provided to users via a network such as the Internet by a system in which, for example, a Web server is used as a front end and business servers, DB servers, external servers, and the like cooperate.

  In such a system, it is necessary to continue providing the service while maintaining a predetermined response time for a huge number of service requests from the user. For this reason, the performance of the system is generally monitored. Is. Service levels such as response time and other system performance indicators and failure rates may be negotiated between a service provider (such as a financial institution) and a system operator.

  Various methods have been proposed as means for monitoring the performance of systems such as each server and internal network. Generally, the CPU usage rate and bandwidth usage status of each server and internal network, etc. are proposed. This is resource-based monitoring. For example, a system performance monitoring method disclosed in Japanese Patent Application Laid-Open No. 2007-26303 (Patent Document 1) performs monitoring from the viewpoint of a direct service level such as response time in each server or internal network.

The system performance monitoring method disclosed in Patent Document 1 collects observation information on the operating state such as information on the CPU usage rate, individual access, service type, response time, etc. from each server or internal network switch. If the observation information necessary for the measurement is insufficient, the access generation means generates a predetermined access to generate the insufficient observation information, and when the observation information is prepared, the observation information is analyzed to measure and evaluate the performance.
JP 2007-26303 A

  In the performance monitoring method such as Patent Document 1, performance monitoring based on resources such as CPU usage rate and response based performance monitoring based on response time is performed for each service type (URL: Uniform Resource Locator) for each server. be able to.

  However, the access that is the target of the performance measurement is the actual transaction of the user in the service, and the processing contents on the server vary depending on the user's designated contents and the like. That is, even for the same service type, the processing contents in the server and the processing path in the system (the server to be accessed and the internal network that passes through) differ for each access. Accordingly, the response time information obtained in this way may vary greatly depending on the acquisition timing, and may not be suitable for use as a criterion for determining whether or not the service level is satisfied. In order to obtain an effective value, the load becomes high, such as calculating the average of many or long-time accesses, and the real-time property is also poor.

  Here, it is conceivable to perform performance measurement by acquiring a response time for the access, using the pseudo access generated by the access generation means in the performance monitoring method of Patent Document 1 as a reference access. However, in the performance monitoring method as in Patent Document 1, the response time on the client terminal basis for the access can be measured, but the access data cannot be specified from the collected observation information. The response time at each server for access cannot be acquired.

  Therefore, it is difficult to grasp in real time detailed information such as the breakdown of the response time on the client terminal basis in each server or internal network for a specific access.

  Accordingly, an object of the present invention is to provide a performance monitoring system and a performance monitoring method that make it possible to grasp in real time the response time of each server or internal network on a processing path for a specific access. The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows.

  A performance monitoring system according to a representative embodiment of the present invention includes a server system that provides services to client terminals via a network by cooperation of a plurality of servers, and a performance monitoring server that monitors the performance of the server system, Each of the servers, when the server is the server that first received an access from the client terminal, a means for giving a request ID that can be identified from the other access to the access; Means for taking over the request ID between the servers on the processing path related to the access in the server system, and outputting a log including the request ID and time as an individual log at the start and end of the processing related to the access And the performance monitoring server outputs the output by each server. From the collected log obtained by collecting the individual logs and the data of the collected log, the log output at the time of the processing related to the access at each server is identified and identified by associating with the request ID. In addition, based on the information on the start time and the end time of processing in each server in the log, the servers in the processing in the server system related to the access, the network between the servers, and the processing path And a log totaling unit for calculating the overall response time.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  According to a typical embodiment of the present invention, it becomes possible to grasp in real time the individual response time in each server or internal network on a processing path for a specific access, such as a bottleneck on the processing path It is possible to realize a response-based performance monitoring system that can evaluate the above with higher accuracy.

  Further, according to the representative embodiment of the present invention, an access through a predetermined processing path can be used as a reference access, and by grasping a response time over the processing path for the access, It is possible to realize a performance monitoring system that can evaluate the current service level without any bias and provide a highly qualified index as a standard in performance prediction and the like.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

  FIG. 1 is a diagram showing an outline of a configuration example of a performance monitoring system according to an embodiment of the present invention. The performance monitoring system is a server system that provides a service to the client terminal 400 via a network 500 such as the Internet by cooperation of a plurality of servers such as a Web server, an application server, a DB server, and a gateway server to an external system. 100 and a performance monitoring server 200 that monitors the performance of the server system 100. A plurality of server systems 100 may exist depending on the type of service to be provided.

  The server system 100 is further connected with a standard request transmission terminal 300 as a terminal for transmitting a standard request to be described later. The standard request transmission terminal 300 is connected to the server system 100 not via the network 500 that is an external network but via the internal network. However, similarly to the client terminal 400, the standard request transmission terminal 300 may be connected via the network 500. Good.

  In the example of FIG. 1, the server system 100 has a configuration in which four servers, a server A110, a server B120, a DB server C130, and a server D140, are connected via a network. Is not limited to this. Each of these servers cooperates to provide a service in response to a request from the client terminal 400 as the server system 100 by processing by middleware, an application program, or the like, or reading / writing data from / to the DB 132. The service provided by the server system 100 is not limited to one type, and there may be a plurality of types. In this case, the route of the server that performs processing may be different for each service.

  When providing a service, the server A 110 that first receives an access from the client terminal 400 assigns a unique request ID that can be distinguished from other accesses for each access. The request ID is inherited when a processing request / response for the access is transmitted / received between the servers.

  Each server outputs a log to the individual log 111 at the start and end when it receives the process request / response for the access and executes the process. The individual log 111 also outputs a request ID. This makes it possible to specify a log entry for the access in the individual log 111 written by each server. Note that a database or a file can be used for the individual log 111.

  The function of assigning the request ID and outputting the log to the individual log 111 can be implemented in each server by, for example, incorporating it in common in a module operating in a base layer that is a lower layer of the application program.

  At this time, when the server is the server that first received the access from the client terminal 400, the base layer of each server relates to the access in the server system 100 and means for assigning a request ID to the access. Means for taking over the request ID between servers on the processing path, and means for outputting a log including the request ID and time as the individual log 111 at the start and end of the processing relating to the access.

  FIG. 2 is a diagram showing an example of an outline of the server configuration and processing. Here, the server A110 is illustrated as an example. In the server A 110, for example, the OS / middleware 112 operates as the lowest layer as software, and the base layer 113 having the basic functions and the above-described means for operating as a server in the server system 100 operates as an upper layer. Further, an application program 114 that performs business processing unique to each server operates in the upper layer.

  At this time, for example, when the server A 110 receives a processing request from the client terminal 400, the base layer 113 that has received the processing from the OS / middleware 112 assigns a request ID and includes a request ID and a start time The time log is output to the individual log 111 and the process is transferred to the application program 114. The application program 114 performs predetermined processing such as business processing and DB access, and then passes the processing to the base layer 113.

  The base layer 113 outputs an end-time log including the request ID and time to the individual log 111 and transmits a processing request to another server (server B120 in the example of FIG. 2). At this time, the request ID is also transmitted to take over. As a result, the application program 114 can perform business processing or the like without knowing about the processing related to the output to the request ID and the individual log 111.

  Note that the example of FIG. 2 shows processing when a processing request is received at the server A110 that first receives access from the client terminal 400, but the same applies when a processing response is received and at other servers. is there. However, in these cases, since the request ID has already been assigned and taken over, the process of assigning the request ID is not performed.

  In the configuration example of FIG. 1, the performance monitoring server 200 includes a computer system, and includes, for example, a log totaling unit 220, an evaluation unit 230, and a user interface (I / F) unit 240. Further, the collected log 210, log record information 250, response time threshold information 260, and service route information 270 are included. The log totaling unit 220, the evaluation unit 230, and the user I / F unit 240 are implemented by one or more programs. Further, a database, a file, or the like can be used for each of the collected log 210, log record information 250, response time threshold information 260, and service route information 270.

  The collection log 210 is obtained by collecting and merging the individual logs 111 of each server in the server system 100. Various means are available for collecting logs. For example, the collection log 210 is arranged in a storage that can be commonly accessed from each server such as NAS (Network Attached Storage), and each server outputs the contents output to the individual log 111 to the collection log 210. A means such as writing asynchronously can be taken. In this case, it is possible to write data to the collection log 210 almost simultaneously with writing to the individual log 111.

  Here, as described above, request ID information is output to the individual log 111. Therefore, in the collected log 210, for each access from the client terminal 400, the log output in each individual log 111 can be associated by the request ID, and the log can be specified over the processing path for any access.

  Since each server outputs a log including the time at the start and end of the process, the response time of the process at each server can be calculated from these logs. Further, the response time in the network between the servers can be calculated from the log at the end of the previous server and the log at the start of the subsequent server between adjacent servers on the processing path.

  The log totaling unit 220 periodically identifies the log output in the process related to the predetermined access described later in each server from the data of the collected log 210 by using a request ID, Based on the information on the start time and end time of the processing at each server, the response time for each server, the network between the servers, and the entire processing path in the processing in the server system 100 related to the predetermined access is calculated. The result is calculated and the result is output to the log record information 250.

  Based on the data of the log record information 250, the contents of the service route information 270, and the response time threshold information 260, the evaluation unit 230 uses each server, the network between servers, and the process in the process in the server system 100 related to the access. It is determined whether or not the response time in the entire route exceeds a predefined response time threshold value, and it is evaluated whether or not the server system 100 satisfies a predetermined service level.

  Although details will be described later, the service route information 270 holds information on the processing route for each service provided by the server system 100 in advance. The response time threshold information 260 holds information on response time thresholds defined in advance for each server on the service processing path and the network between servers.

  The user I / F unit 240 is means for displaying a performance monitoring screen of the server system 100 on a display or the like. For example, draw processing path information for each service, display response time information for each server or server-to-server network on the processing path, and display a monitoring screen that displays a warning if the threshold is exceeded To do. In addition, information on performance monitoring based on resources such as CPU usage rate and network usage bandwidth acquired by other performance measuring means may be displayed together or switched. In addition, when there is a server or network whose response time exceeds the threshold, it may have means for automatically sending an email to a related person such as a system administrator in addition to the warning display.

  As described above, the performance monitoring server 200 according to the present embodiment acquires information on response time in each server or server-to-server network over the processing path for each access from the client terminal 400, and compares the information with a threshold value for performance monitoring. It can be performed. However, there are various patterns of access from the client terminal 400, and their processing paths are also different. Therefore, when a response time is calculated over a processing path for a certain access, for example, the processing path ends at a server on the way, which may be insufficient as a performance index for the entire server system 100.

  Therefore, in the performance monitoring server 200 of this embodiment, the standard request transmission terminal 300 periodically transmits a standard request to the server system 100. This standard request is a request for an actual service such as a stock order, for example, but it is agreed in advance with the service provider so that the actual service is not provided (such as actual trading of shares). This is a dummy transaction request.

  It is desirable that parameters and the like are set for the standard request so that the processing path has the greatest influence (the response time is large) in the server system 100. In addition, processing is always performed through the same processing path. Using this standard request as a reference access, the log output to the individual log 111 at each server for the processing related to the standard request is specified from the collection log 210 and the response time is calculated. Thereby, when performing performance evaluation or performance prediction, it is possible to acquire information on response time having a high qualification as a reference. A plurality of types of standard requests may be set.

  In order to be able to specify the log for the standard request from the collected log 210, when the server A 110 assigns a request ID for each access, the access related to the standard request is related to the standard request. Is assigned a special request ID that makes it identifiable. Whether the target access is an access related to a standard request can be determined by various methods. For example, an access from a specific user ID is an access related to a standard request. You may make it determine.

  FIG. 3 is a diagram illustrating a data configuration of the individual log 111 and an example of specific data. The individual log 111 has fields of a request ID field 301, a service name field 302, a server field 303, an application field 304, a type field 305, an I / O field 306, a time stamp field 307, and a log data field 308. Other fields may be included in addition to this. An entry is added to the individual log 111 at the start and end when each server receives a processing request / response and performs processing.

  The request ID field 301 indicates a request ID assigned for each access by the server A110. For example, in the example of FIG. 3, “A001” and “A002” are request IDs given to normal access, and “ZA01” shows a request ID given to access related to a standard request. In the present embodiment, for the access related to the standard request, the head of the request ID is set to “Z”, which is not normally used, so that it can be identified as a standard request. However, the present invention is not limited to this.

  The service name field 302 indicates the name of the service related to the target access. In some cases, a plurality of services ("domestic stock order" and "domestic stock inquiry" in the example of FIG. 3) can be processed by the same server, and this is retained for discrimination. The server field 303 and the application field 304 indicate information for identifying the server and the application program 114 that output the target individual log 111, respectively. Note that the naming rules of these pieces of information are not limited to those shown in the example of FIG.

  Whether the type field 305 is a log when a processing request for a target access is received (“Req” in the example of FIG. 3) or a log when a processing response is received (in the example of FIG. 3, “ Rsp ") indicates information for identifying the processing direction. Whether the I / O field 306 is a log at the start of processing for the target access (“IN” in the example of FIG. 3) or a log at the end (“OUT” in the example of FIG. 3). ) To identify information. With these pieces of information, it is possible to specify at what timing the target log was output.

  The time stamp field 307 indicates the time when the target log entry is output. The log data 308 indicates that the application program 114 in each server individually outputs parameter data and processing results as application log data. Therefore, the format of the value of the log data 308 may differ depending on each server.

  Since the collection log 210 corresponds to a collection and merge of the individual logs 111 of the servers, the data configuration is the same as the data configuration of the individual log 111 shown in FIG.

  FIG. 4 is a diagram showing a data configuration of the log record information 250 and an example of specific data. The log record information 250 includes fields of a request ID field 401, a service name field 402, a server field 403, an application field 404, a type field 405, and a response time field 406. Other fields may be included in addition to this. The log record information 250 holds record information of response time for each server and the network between servers for the access related to the standard request.

  The request ID field 401 and the service name field 402 are the same as the request ID field 301 and the service name field 302 shown in FIG. However, the log result information 250 includes a result obtained by the log totaling unit 220 specifying and counting only the logs related to the access related to the standard request from the collected log 210. Therefore, the request ID field 401 includes only a value indicating access related to the standard request (the head is “Z” in the present embodiment).

  The server field 403 includes a server (“A1” “B1” in the example of FIG. 4) and an inter-server network (“A1-B1” “B1-A1” in the example of FIG. 4) on the access path related to the standard request. That is, information for identifying a server or the like that is a response time calculation target is shown.

  The application field 404 indicates information for identifying the application program 114 that is a response time calculation target. When the value of the server field 403 indicates an inter-server network, since the application program 114 does not exist, the value of the application field 404 is set to “−”, for example. The type field 405 indicates information for identifying the processing direction as to whether the response time of the target entry is that when a processing request is received or when the processing response is received. In the example of FIG. 4, the value of “Req / Rsp” is the value when a processing response is received after processing request is received and processed by the application program 114 (for example, the DB server in FIG. 1). C130 and server D140).

  The response time field 406 indicates, in milliseconds, the response time calculated from the data of the collection log 210 of FIG. 3 for each type of processing direction for the target server or the network between servers.

  FIG. 5 is a diagram showing a data configuration of the response time threshold information 260 and an example of specific data. The response time threshold information 260 includes a service name field 501, a server field 502, an application field 503, a type field 504, and a threshold field 505. Other fields may be included in addition to this. The response time threshold information 260 holds a threshold of response time that is predetermined for each server on the processing path and the network between servers for access related to the standard request.

  The service name field 501, server field 502, application field 503, and type field 504 are substantially the same as the service name field 402, server field 403, application field 404, and type field 405 shown in FIG. However, in the example of FIG. 5, it is assumed that the value of “Path” in the type field 504 indicates the entire processing path of the service, not the network between servers. A threshold field 505 indicates a response time threshold in milliseconds for the target server, the network between servers, or the entire processing path.

  FIG. 6 is a diagram showing a data configuration of service route information 270 and an example of specific data. The service route information 270 includes a service name field 601, a sequence number field 602, a server field 603, an application field 604, and a server name field 605. Other fields may be included in addition to this. The service route information 270 holds information on each server and the network between servers on the processing route for access related to the standard request.

  The service name field 601, server field 603, and application field 604 are the same as the service name field 402, server field 403, and application field 404 shown in FIG. The sequence number field 602 indicates information for identifying the order in which the server of the target entry or the network between servers is processed on the processing path. By tracing the server or the server-to-server network in this order, the processing route for the target service can be specified. The server name field 605 indicates the name of the target server.

  FIG. 7 is a flowchart illustrating an example of an outline of processing in the performance monitoring server 200 according to the present embodiment. After performing the processing by periodically transmitting the standard request from the standard request transmission terminal 300, the performance monitoring server 200 collects and merges the data of the individual log 111 of each server into the collection log 210 when the processing is started. (S701). For this process, as described above, for example, the collection log 210 is placed in a storage that can be accessed in common by each server such as NAS, and each server collects the contents output to the individual log 111. Means such as asynchronously writing to the log 210 can be taken.

  Next, the log totaling unit 220 identifies and identifies the log output in the process related to the standard request in each server by the request ID for each type of standard request from the data of the collected log 210. Response time for each server, the network between servers, and the entire processing path in the processing in the server system 100 related to the standard request based on the information on the start time and the end time of the processing in each server in the log. Is calculated and stored in the log record information 250 (S702). The transmission of the standard request and the calculation of the response time can be performed at a frequency such as once per hour.

  Next, loop processing for all standard request types is started (S703). In this loop processing, loop processing is started for all servers on the processing path for access related to the target standard request and the network between servers (S704). The server on the processing path and the network between servers can be known from the service path information 270, for example.

  In the loop processing for the server and the server-to-server network, the response time held in the log record information 250 for the target server or the server-to-server network is displayed by the user I / F unit 240 at a corresponding location on the monitoring screen (S705). . Next, the evaluation unit 230 determines whether or not the response time exceeds a threshold set in advance in the response time threshold information 260 (S706). If the threshold value is exceeded, the user I / F unit 240 displays a warning at a corresponding location on the monitoring screen and notifies the system administrator or the like by e-mail or the like (S707).

  The processes in steps S705 to S707 are repeated for all servers and inter-server networks on the processing path (S708), and the processes in steps S704 to S708 are repeated for all standard request types (S709). The process ends.

  FIG. 8 is a diagram illustrating an example of a monitoring screen displayed by the user I / F unit 240. On the monitoring screen, for example, the type of each service, the processing route of each service, the servers on the processing route, and the logical arrangement of the network between the servers are displayed so that it can be easily grasped visually. The response time in the server-to-server network is displayed. Further, a warning is displayed by highlighting a portion where the response time exceeds the threshold.

  For each service, the target response time for the entire processing path and the actual result information of the response time at each measurement timing are displayed in a table format. Again, a warning is displayed by highlighting a service whose response time exceeds the threshold.

  Note that the monitoring screen of FIG. 8 is merely an example, and other display formats may be used. In addition, for example, in combination with measurement results by other resource-based performance measurement means such as CPU usage rate, when the resource usage exceeds a predetermined threshold, a warning is displayed for the corresponding server, When selected, detailed display of performance measurement results may be performed. In addition, the monitor screen is switched from the service view that displays the response time for each service to the network view that is based on the network configuration, and performance measurement values for bandwidth utilization and response time for each line are displayed and monitored based on the network configuration. It may be possible.

  As described above, according to the performance monitoring server 200 of the present embodiment, the log output from each server based on the request ID is associated with the access related to the standard request, The response time is calculated from the log at the start and end of the process. Thereby, it becomes possible to grasp in real time the response time in each server and the server-to-server network on the processing path, and the bottleneck and the like on the processing path can be more accurately evaluated. It is also possible to grasp the response time of the network between servers configured by network devices that do not have a log output function such as a hub.

  Also, by using standard requests that are always processed by a server on a predetermined processing path, the current service level of the system can be evaluated evenly by grasping the response time over the processing path for access related to the standard request. In addition, it is possible to provide a highly qualified index as a standard in performance evaluation and performance prediction.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  The present invention can be used in a performance monitoring system and a performance monitoring method for monitoring the performance of a server system that provides services to client terminals via a network through cooperation of a plurality of servers.

It is the figure which showed the outline | summary of the structural example of the performance monitoring system which is one embodiment of this invention. It is the figure which showed the example of the structure of the server in 1 embodiment of this invention, and the outline | summary of a process. It is the figure which showed the example of the data structure and specific data of the separate log in one embodiment of this invention. It is the figure which showed the data structure of the log performance information in one embodiment of this invention, and the example of concrete data. It is the figure which showed the data structure of the response time threshold value information in one embodiment of this invention, and the example of concrete data. It is the figure which showed the data structure of the service route information in one embodiment of this invention, and the example of concrete data. It is a flowchart which shows the example of the outline | summary of the process in the performance monitoring system in one embodiment of this invention. It is the figure which showed the example of the monitoring screen which the user I / F part in one embodiment of this invention displays.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Server system, 110 ... Server A, 111 ... Individual log, 112 ... OS / middleware, 113 ... Base layer, 114 ... Application program, 120 ... Server B, 121 ... Individual log, 130 ... DB server C, 131 ... Individual Log, 140 ... Server D, 141 ... Individual log, 200 ... Performance monitoring server, 210 ... Collection log, 220 ... Log tabulation unit, 230 ... Evaluation unit, 240 ... User I / F unit, 250 ... Log performance information, 260 ... Response time threshold information, 270 ... service route information, 300 ... standard request transmission terminal, 400 ... client terminal, 500 ... network,
301 ... Request ID field, 302 ... Service name field, 303 ... Server field, 304 ... Application field, 305 ... Type field, 306 ... I / O field, 307 ... Time stamp field, 308 ... Log data field,
401 ... Request ID field, 402 ... Service name field, 403 ... Server field, 404 ... Application field, 405 ... Type field, 406 ... Response time field,
501 ... Service name field, 502 ... Server field, 503 ... Application field, 504 ... Type field, 505 ... Threshold field,
601 ... Service name field, 602 ... Sequence number field, 603 ... Server field, 604 ... Application field, 605 ... Server name field.

Claims (6)

A server system that provides services to client terminals via a network by cooperation of a plurality of servers;
A performance monitoring system comprising a performance monitoring server for monitoring the performance of the server system,
Each of the servers gives a request ID such that when the server is the server that first received an access from the client terminal, the access can be identified from the other access. Means,
A second means for taking over the request ID between the servers on the processing path relating to the access in the server system;
A third means for outputting a log including the request ID and time as an individual log at the start and end of the processing relating to the access;
The performance monitoring server includes a collection log obtained by collecting the individual logs output by the servers;
From the collected log data, the log output in the process related to the access at each server is specified in association with the request ID, and the process at each server in the specified log is started. Based on the time information and the end time information, the log totaling unit that calculates the response time in each server, the network between the servers, and the entire processing path in the processing in the server system related to the access; A performance monitoring system comprising: The performance monitoring system according to claim 1,
The performance monitoring system according to claim 1, wherein the access is an access related to a standard request processed by each server on a predetermined processing path in the server system. The performance monitoring system according to claim 2,
The server assigns the request ID so that it is possible to identify that the access is related to the standard request when the request ID is assigned to the access related to the standard request. A performance monitoring system. In the performance monitoring system according to any one of claims 1 to 3,
The performance monitoring system according to claim 1, wherein the first means, the second means, and the third means in each server operate in a lower layer of an application program that operates on each server. In the performance monitoring system according to any one of claims 1 to 4,
The performance monitoring server predefines each response time in each server, the network between the servers, and the entire processing path in the processing in the server system related to the access calculated by the log aggregation unit. A performance monitoring system comprising: an evaluation unit that determines whether or not a response time threshold value is exceeded. A server system that provides services to client terminals via a network by cooperation of a plurality of servers;
A performance monitoring method in a system comprising a performance monitoring server for monitoring the performance of the server system,
Each of the servers, when the server is the server that first received an access from the client terminal, giving a request ID such that the access can be identified from other access;
Outputting the log including the request ID and time as an individual log at the start of the process related to the access, and passing the process to the application program;
Outputting the log including the request ID and time as the individual log at the end of the processing relating to the access after the processing is delivered from the application program;
Taking over the request ID to the other server on the processing path related to the access in the server system,
The performance monitoring server collecting the individual logs output by the servers in a collection log;
From the collected log data, the log output in the process related to the access at each server is specified in association with the request ID, and the process at each server in the specified log is started. And calculating the response time for each of the servers, the network between the servers, and the entire processing path in the processing in the server system related to the access based on the time information and the end time information. A performance monitoring method characterized by:

Images