CN106933718B - Method for monitoring performance and device - Google Patents

Abstract

The invention discloses a performance monitoring method and device, belonging to the technical field of computers. The method includes: receiving the application-side performance information of the operation object, and receiving the storage-side performance information of the operation object; receiving the corresponding relationship between the first identifier and the second identifier of the operation object; from the application-side performance information of each operation object Read the first identifier of the operation object; read the second identifier of the operation object from the storage-side performance information of each operation object; determine the application-side performance information and storage-side performance information corresponding to each operation object according to the corresponding relationship; Analysis of application-side performance information and storage-side performance information corresponding to each operation object; determine whether there is a problem with the application server or the storage server according to the analysis results. The invention solves the problem that it is impossible to determine whether there is a problem in the application server or the storage server, which leads to monitoring errors, and achieves the effect of improving the accuracy of monitoring.

Description

Performance monitoring method and device

technical field

The invention relates to the technical field of computers, in particular to a performance monitoring method and device.

Background technique

Users will generate data in the process of using the application. In order to store the data, the developer can configure a storage server for the application server running the application. Wherein, the application may be software such as a database.

In related technologies, a developer connects an application server to an external storage server through a network, and the storage server may be a collection of storage devices of various specifications. While users are using applications, developers can monitor the performance of application servers and storage servers. Usually, the performance of the application server can be monitored on the side of the application server, and the performance of the storage server can be monitored on the side of the storage server.

When the application server makes an error when operating an operation object, because the application server cannot monitor the performance of the storage server when operating the operation object, it cannot determine whether there is a problem with the application server or the storage server, resulting in a monitoring error.

Contents of the invention

In order to solve the problem that when an application server makes an error in operating an operation object, it is impossible to determine whether the application server or the storage server has a problem, resulting in a monitoring error. Embodiments of the present invention provide a performance monitoring method and device. Described technical scheme is as follows:

In a first aspect, a performance monitoring method is provided, the method is used in a management server, the management server is respectively connected to an application server and a storage server, and the storage server provides data storage for the application server, and the method includes:

The management server receives the application-side performance information of the operation object sent by the application server, and receives the storage-side performance information of the operation object sent by the storage server; the management server then receives the correspondence between the first identifier and the second identifier of the operation object sent by the application server ; Then, the management server reads the first identifier of the operation object from the application-side performance information of each operation object, and reads the second identifier of the operation object from the storage-side performance information of each operation object; the management server according to the corresponding Determine the application-side performance information and storage-side performance information corresponding to each operation object; finally, the management server analyzes the application-side performance information and storage-side performance information corresponding to each operation object, and determines whether it is an application server or a storage server according to the analysis results There is a problem.

Determine the application-side performance information and storage-side performance information corresponding to each operation object through the management server, and then analyze the application-side performance information and storage-side performance information corresponding to each operation object, and determine whether it is the application server or the storage side based on the analysis results. There is a problem in the server, which solves the problem that it is impossible to determine whether there is a problem in the application server or the storage server, which leads to monitoring errors, and achieves the effect of improving the accuracy of monitoring.

In a first possible implementation manner of the first aspect, the first identifier includes at least one of a name and a logical address of the operation object, and the second identifier includes a physical address of the operation object;

Correspondence The application server obtains the first correspondence between the name of each operation object and the logical address, and obtains the second correspondence between the name of each operation object and the physical address, which will include the first correspondence and the second correspondence of the same name Corresponding relations are merged to get.

The application server collects the first corresponding relationship and the second corresponding relationship respectively, and merges the first corresponding relationship and the second corresponding relationship into a corresponding relationship, so that the application-side performance information and the storage-side performance information of the operation object can be processed according to the corresponding relationship. The correspondence solves the problem that the performance information of the application side and the performance information of the storage side cannot be matched, and achieves the effect of improving the accuracy of the analysis.

In the second possible implementation of the first aspect, when the application-side performance information includes information on the application-side latency of data operations, and the storage-side performance information includes information on the storage-side latency of data operations, each operation Analysis of application-side performance information and storage-side performance information corresponding to objects, including:

Subtract the delay on the application side from the delay on the storage side to obtain a delay difference; detect whether the delay difference is greater than the first threshold; when the detection result is that the delay difference is greater than the first threshold, generate A configuration problem analysis result.

By comparing the delay difference with the first threshold, the management server can automatically determine whether there is a problem with the application server, which solves the need for developers to compare the delay difference with the first threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

In combination with the second possible implementation of the first aspect, in the third possible implementation of the first aspect, the analysis of application-side performance information and storage-side performance information corresponding to each operation object further includes:

Detecting whether the storage-side delay is greater than the second threshold; when the detection result is that the storage-side delay is greater than the second threshold, generating an analysis result indicating that the storage server has a second configuration problem.

By comparing the storage-side delay with the second threshold, the management server can automatically determine whether there is a problem with the storage server, which solves the need for developers to compare the storage-side delay with the second threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

In a second aspect, a performance monitoring device is provided, which is used in a management server. The management server is connected to an application server and a storage server respectively, and the storage server provides data storage for the application server. The device includes:

The receiving unit is configured to receive the application-side performance information of the operation object sent by the application server, and receive the storage-side performance information of the operation object sent by the storage server;

The receiving unit is further configured to receive the correspondence between the first identifier and the second identifier of the operation object sent by the application server;

A reading unit, configured to read the first identifier of the operation object from the application-side performance information of each operation object received by the receiving unit;

The reading unit is further configured to read the second identifier of the operation object from the storage-side performance information of each operation object received by the receiving unit;

A determining unit, configured to determine the application-side performance information and storage-side performance information corresponding to each operation object read by the reading unit according to the correspondence received by the receiving unit;

An analysis unit, configured to analyze application-side performance information and storage-side performance information corresponding to each operation object determined by the determination unit;

The determination unit is further configured to determine whether there is a problem with the application server or the storage server according to the analysis result obtained by the analysis unit.

Determine the application-side performance information and storage-side performance information corresponding to each operation object through the management server, and then analyze the application-side performance information and storage-side performance information corresponding to each operation object, and determine whether it is the application server or the storage side based on the analysis results. There is a problem in the server, which solves the problem that it is impossible to determine whether there is a problem in the application server or the storage server, which leads to monitoring errors, and achieves the effect of improving the accuracy of monitoring.

In a first possible implementation manner of the second aspect, the first identifier includes at least one of a name and a logical address of the operation object, and the second identifier includes a physical address of the operation object;

Correspondence The application server obtains the first correspondence between the name of each operation object and the logical address, and obtains the second correspondence between the name of each operation object and the physical address, which will include the first correspondence and the second correspondence of the same name Corresponding relations are merged to get.

The application server collects the first corresponding relationship and the second corresponding relationship respectively, and merges the first corresponding relationship and the second corresponding relationship into a corresponding relationship, so that the application-side performance information and the storage-side performance information of the operation object can be processed according to the corresponding relationship. The correspondence solves the problem that the performance information of the application side and the performance information of the storage side cannot be matched, and achieves the effect of improving the accuracy of the analysis.

In a second possible implementation of the second aspect, when the application-side performance information includes application-side delay information, and the storage-side performance information includes storage-side delay information, the analyzing unit is specifically configured to:

Subtract the delay on the application side from the delay on the storage side to obtain the delay difference; check whether the delay difference is greater than the first threshold; when the detection result is that the delay difference is greater than the first threshold, generate an analysis of the problems in the application server result.

By comparing the delay difference with the first threshold, the management server can automatically determine whether there is a problem with the application server, which solves the need for developers to compare the delay difference with the first threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

In combination with the second possible implementation of the second aspect, in the third possible implementation of the second aspect, the analysis unit is further configured to: detect whether the storage side delay is greater than the second threshold; when the detection result is the storage When the side delay is greater than the second threshold, an analysis result indicating that there is a problem with the storage server is generated.

By comparing the storage-side delay with the second threshold, the management server can automatically determine whether there is a problem with the storage server, which solves the need for developers to compare the storage-side delay with the second threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained based on these drawings without creative effort.

FIG. 1 is a first structural schematic diagram of a performance monitoring system provided by various embodiments of the present invention;

Fig. 2 is a second structural schematic diagram of the performance monitoring system provided by various embodiments of the present invention;

FIG. 3 is a third structural schematic diagram of a performance monitoring system provided by various embodiments of the present invention;

FIG. 4 is a schematic diagram of a fourth structure of a performance monitoring system provided by various embodiments of the present invention;

Fig. 5 is a fifth structural schematic diagram of the performance monitoring system provided by various embodiments of the present invention;

Fig. 6 is a sixth structural schematic diagram of a performance monitoring system provided by various embodiments of the present invention;

7 is a schematic diagram of a seventh structure of a performance monitoring system provided by various embodiments of the present invention;

Fig. 8 is a schematic structural diagram of a management server provided by an embodiment of the present invention;

Fig. 9 is a method flowchart of a performance monitoring method provided by an embodiment of the present invention;

FIG. 10A is a method flowchart of a performance monitoring method provided by another embodiment of the present invention;

Fig. 10B is a schematic diagram of an interface for displaying information provided by another embodiment of the present invention;

Fig. 11 is a structural block diagram of a performance monitoring device provided by an embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the implementation manner of the present invention will be further described in detail below in conjunction with the accompanying drawings.

Please refer to FIG. 1 , which shows a first structural diagram of a performance monitoring system provided by various exemplary embodiments of the present invention. The performance monitoring system includes: an application server 110, a storage server 120 and a management server 130, and the management server 130 is connected to the application server 110 and the storage server 120 respectively. Wherein, the storage server 120 is connected to the application server 110 and provides data storage for the application server 110 .

Please refer to the second schematic structural diagram of the performance monitoring system shown in FIG. The side performance information is sent to the management server 130. A storage data collection program 121 runs on the storage server 120 , and the storage data collection program 121 is used to collect storage-side performance information of the storage server 130 and send the storage-side performance information to the management server 130 . The management server 130 is configured to analyze the performance information of the application side and the performance information of the storage side, and determine whether there is a problem in the application server 110 or the storage server 120 according to the analysis result.

The applications running on the application server 110 may be local applications, such as local databases, virtual desktops, virtual machines, etc., or networked applications, such as cloud databases, Internet of Things applications, Internet of Vehicles applications, etc. It should be added that the present invention is only illustrated by using the application server 110, and in actual implementation, the application server 110 may also be replaced by a terminal, and the application at this time may be an application installed in the terminal.

Please refer to the third structural diagram of the performance monitoring system shown in FIG. 3 , the performance monitoring system may further include a result outputter 140 , and the management server 130 sends the analysis results to the result outputter 140 for output. Wherein, the result outputter 140 may be a display, a player, etc., which is not limited in the present invention.

Please refer to the fourth schematic structural diagram of the performance monitoring system shown in FIG. The application-side performance information is sent to the application data collection program 111 .

Please refer to the fifth structural diagram of the performance monitoring system shown in Figure 5, the application data collection program 111 may include an agent (Agent) data collection program 111a and a data transmission program 111b, and the agent data collection program 111a is connected to the application performance statistics program 112 , the data transfer program 111b is connected to the management server 130 . The proxy data collection program 111a is used to collect the application-side performance information sent by the application performance statistics program 112, process the application-side performance information to obtain the final application-side performance information, and store it in the storage organization. For example, the proxy data collection program 111a organizes the final application-side performance information into a file in xml format, and stores the file in the database.

The data transmission program 111b is configured to analyze the application-side performance information collected by the proxy data collection program 111a, and send the analyzed application-side performance information to the management server 130 according to a predetermined protocol. Specifically, when the performance information on the application side is organized into a file in xml format, the data transfer program 111b parses the file into performance information on the application side, and transfers the information through the secure file transfer protocol (English: secure File Transfer Protocol; sFTP for short) or secure The shell protocol (English: Secure Shell; SSH for short) protocol sends the application-side performance information to the management server 130 .

Please refer to the sixth schematic structural diagram of the performance monitoring system shown in FIG. 6 , the performance monitoring system may include at least one application server 110 and at least one storage server 120 . In FIG. 6 , the performance monitoring system includes two application servers 110 and two storage servers 120 as an example for illustration.

Please refer to the seventh schematic structural diagram of the performance monitoring system shown in FIG. 7 , wherein the application server 110 is specifically a cloud application service 110 , and the storage server 120 is specifically a cloud storage service 120 . The performance monitoring system may include at least one cloud application service 110 and at least one cloud storage service 120 . In FIG. 7 , the performance monitoring system includes one cloud application service 110 and one cloud storage service 120 as an example for illustration.

For the convenience of description, the implementation of sending application-side performance information from the application server 110 to the management server 130 will be described below. The application data collection program 111 directly collects the application side performance information of the application server 110 and sends it to the management server 130. It may also include the application performance statistics program 112 collecting the application side performance information and sending it to the application data collection program 111. The application data collection program 111 An implementation manner of sending application-side performance information to the management server 130 . In the same way, the following describes the implementation of the storage server 120 sending storage-side performance information to the management server 130. This implementation may include the implementation of the storage server 120 directly collecting storage-side performance information and sending it to the management server 130, or may include storing An implementation manner in which the data collection program 111 directly collects storage-side performance information of the storage server and sends it to the management server 130 .

Please refer to FIG. 8 , which shows a schematic structural diagram of a management server provided by an exemplary embodiment of the present invention. The management server may be the management server shown in FIGS. 1 to 7 , and the management server may include: a receiver 810 , a processor 820 and a transmitter 830 . Those skilled in the art can understand that the structure of the management server shown in FIG. 8 is not limited to the management server, and may include more or less components than those shown in the figure, or combine some components, or arrange different components. For example, the management server further includes a storage 840 and the like. in:

The processor 820 is the control center of the management server, and uses various interfaces and lines to connect various parts of the entire management server, by running or executing software programs and/or modules stored in the memory 840, and calling applications stored in the memory 840 Performance information on the storage side and performance information on the storage side, and perform data analysis functions. Optionally, the processor 820 may include one or more processing cores; optionally, the processor 820 may integrate an application processor and a modem processor, wherein the application processor mainly processes operating systems and application programs, etc., modulation The demodulation processor mainly handles wireless communication. It can be understood that the above-mentioned modem processor may not be integrated into the processor 820, and the above-mentioned modem processor may be separately implemented as a baseband chip.

Memory 840 can be used for software programs as well as modules. The processor 820 executes data analysis functions by running software programs and modules stored in the memory 840 . The memory 840 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system 841, a receiving module 842, a reading module 843, a determination module 844, an analysis module 845 and at least one application program required by other functions 846, etc.; the storage data area can store data created according to the use of the management server (such as application-side performance information, storage-side performance information, etc.) and the like. In addition, the memory 840 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (English: Static Random Access Memory, referred to as: SRAM), electrically erasable and programmable Read-only memory (English: Electrically Erasable Programmable Read-Only Memory, referred to as: EEPROM), erasable programmable read-only memory (English: Erasable Programmable Read Only Memory, referred to as: EPROM), programmable read-only memory (English: Programmable Read-Only Memory, referred to as: PROM), read-only memory (English: ReadOnly Memory, referred to as: ROM), magnetic memory, flash memory, magnetic disk or optical disk.

Transmitter 830 may include a radio frequency transmitting component, such as an antenna. The transmitter 830 is configured to carry the analysis result in a wireless signal for transmission. The wireless signal may be a time-frequency resource in the mobile communication system.

Receiver 810 may include radio frequency receiving components, such as an antenna. The receiver 810 is configured to receive application-side performance information and storage-side performance information carried in wireless signals. The wireless signal may be a time-frequency resource in the mobile communication system.

Although not shown, the management server may also optionally include a power supply, a bluetooth module, etc., which will not be repeated here.

Please refer to FIG. 9 , which shows a flow chart of a performance monitoring method provided by an embodiment of the present invention. The performance monitoring method can be applied to the management server shown in FIGS. 1 to 7 . The performance monitoring method includes:

Step 901: Receive the application-side performance information of the operation object sent by the application server, and receive the storage-side performance information of the operation object sent by the storage server.

Step 902, receiving the correspondence between the first identifier and the second identifier of the operation object sent by the application server.

Step 903, read the first identifier of the operation object from the application-side performance information of each operation object.

Step 904, read the second identifier of the operation object from the storage-side performance information of each operation object.

Step 905: Determine application-side performance information and storage-side performance information corresponding to each operation object according to the corresponding relationship.

Step 906, analyzing the application-side performance information and storage-side performance information corresponding to each operation object.

Step 907, determine whether there is a problem with the application server or the storage server according to the analysis result.

To sum up, in the performance monitoring method provided by the embodiment of the present invention, the management server determines the application-side performance information and storage-side performance information corresponding to each operation object, and then the application-side performance information and storage-side performance information corresponding to each operation object Analysis of performance information can determine whether there is a problem with the application server or the storage server based on the analysis results, which solves the problem of being unable to determine whether the problem is with the application server or the storage server, which leads to monitoring errors, and achieves the effect of improving the accuracy of monitoring.

Since the performance monitoring method provided in this embodiment can be applied to any application, and for different applications, the application-side performance information collected by the application server is different, and correspondingly, the storage-side performance information collected by the storage server is also different. For ease of understanding and description, this embodiment takes the application being a database as an example to illustrate the performance monitoring method of the database. The performance information on the application side is used to describe the performance of the application server, and the performance information on the storage side is used to describe the performance of the storage server.

Please refer to FIG. 10A , which shows a flow chart of a performance monitoring method provided by another embodiment of the present invention. The performance monitoring method can be applied to the management server shown in FIGS. 1 to 7 . The performance monitoring method includes:

Step 1001, receiving the application-side performance information of the operation object sent by the application server, and receiving the storage-side performance information of the operation object sent by the storage server.

When the application is a database, the application-side performance information may include but not limited to: application-side delay information and application-side average throughput information; correspondingly, storage-side performance information may include but not limited to: storage-side delay information and storage Side average throughput information.

The application-side delay information refers to the operation delay of the database in the application server on an operation object, and the storage-side delay information refers to the operation delay of the storage server on an operation object under the control of the application server. The operation object may be a data file, or one or more storage device LUNs, which is not limited in this embodiment. The operation performed on the operation object may be one of a read operation, a write operation, and a recovery (Redo) operation.

The average throughput information on the application side is a parameter to measure the average data throughput of the application server during the operation of an operation object, which may include but not limited to: the number of read/write/recovery operations per second (English: Input/Output Operations Per Second; Abbreviation: IOPS), bandwidth, etc. Among them, bandwidth refers to the amount of data that is read/written/restored per second.

The process of collecting performance information on the application side by the application server and collecting performance information on the storage side by the storage server will be introduced respectively below.

First, the performance information of the application side is the information of the delay of the application side, and the performance information of the storage side is the information of the delay of the storage side.

1) Every first time period, the application server obtains the application-side delay of the operation on each operation object in the first time period, and the start time of the application-side delay is the moment when the application server receives the operation input by the operation device , the end time is the time when the operation result returned by the storage server is output to the operation device; the first identification of the operation object is acquired; the information of the application side delay carrying the application side delay and the first identification is generated, and the application side delay The information is sent to the management server. Wherein, the first identifier includes at least one of the name of the operation object and the logical address of the operation object.

When the application server receives an operation on an operation object input by an external operation device, it needs to send the operation to the storage server. The storage server processes the operation and returns the operation result to the application server. Output the result of the operation. For example, when the application server receives a read operation on a data file input by the operating device through the application-side I/O interface, it reads the first identifier of the data file carried in the read operation, and according to the data file in the first identifier The name of the data file determines the physical address of the data file, and sends a read operation carrying the physical address to the storage server through the storage-side I/O interface, and the storage server reads the content stored in the physical address, and sends the The content is returned to the application server as an operation result, and the application server outputs the content to the operation device through the application-side I/O interface.

According to the above response process to the operation, it can be seen that the application-side delay from receiving the operation input by the operating device to outputting the operation result to the operating device can represent the performance of the application server. Therefore, the application server can use the application-side delay information as the application The side performance information is sent to the management server.

Specifically, the application server counts the application-side delay of an operation on an operation object, obtains the first identifier carried in the operation, and generates information including the application-side delay and the application-side delay of the first identifier. Optionally, the application server may also obtain the application server ID, and add the application server ID to the delay information on the application side.

Since the application server can perform operations on multiple operation objects, it will consume more resources to collect the application-side delay information of each operation object. The application-side latency of the operation performed by the operation object is sampled, and only the application-side latency information is generated for the operation during the sampling period to save resources. Wherein, the first time period can be set and modified by itself.

2) Every first time period, the storage server obtains the storage-side delay of the operation on each operation object in the first time period, and the start time of the storage-side delay is the moment when the storage server receives the operation sent by the application server, The end time is the moment when the operation result is returned to the application server; obtain the second identifier of the operation object; generate storage-side delay information carrying the storage-side delay and the second identifier, and send the storage-side delay information to the management server. Wherein, the second identifier includes the physical address of the operation object, and the second identifier is obtained after the application server converts the first identifier in the operation object.

According to the above response process to the operation, it can be seen that the storage-side delay from receiving the operation sent by the application server to outputting the operation result to the application server can represent the performance of the storage server. Therefore, the storage server can use the storage-side delay information as a storage The side performance information is sent to the management server.

Specifically, the storage server counts the storage-side delay of operations on an operation object, obtains the second identifier carried in the operation, and generates information including the storage-side delay and the storage-side delay of the second identifier. Optionally, the storage server may also acquire the storage server identifier, and add the storage server identifier to the storage-side delay information.

Since the storage server can perform operations on multiple operation objects, it will consume more resources to collect the storage-side delay information of each operation object. Therefore, preferably, the storage server can also perform operations on a certain The storage-side latency of the operation performed by the operation object is sampled, and only the storage-side latency information is generated for the operation during the sampling period to save resources.

It should be noted that the sampling time of the application server and the storage server is the same, so that both application-side delay information and storage-side delay information of the same operation object can be sampled.

This embodiment does not limit the order in which the application server obtains the application-side delay information and the storage server obtains the storage-side delay information.

Second, the application side performance information is the application side average throughput information, and the storage side performance information is the storage side average throughput information.

1) Every second time period, the application server collects the average throughput data of each application side that is counted in the second time period; acquires the first identifier of the operation object; generates an application that carries the average throughput data of the application side and the first identifier The average throughput information on the application side is sent to the management server.

According to the above response process to the operation, the application-side IOPS and application-side bandwidth can represent the performance of the application server. Therefore, the application server can send the average application-side throughput information as the application-side performance information to the management server.

Specifically, when the average throughput information on the application side is IOPS on the application side, the application server counts the number of operations per second on the I/O interface on the application side when performing an operation on an operation object, obtains the IOPS on the application side, and obtains the IOPS carried in the operation. The first identifier of the application-side average throughput information carrying the application-side IOPS and the first identifier is generated. When the average throughput information of the application side is the bandwidth of the application side, the application server counts the amount of data per second on the application side when performing an operation on an operation object, obtains the bandwidth of the application side, and obtains the first identifier carried in the operation to generate There is the application-side bandwidth and the first-identified application-side average throughput information. Optionally, the application server may also obtain the application server ID, and add the application server ID to the average throughput information on the application side.

Wherein, the second time period may be set and modified by itself, and the duration of the second time period may be the same as or different from that of the first time period, which is not limited in this embodiment.

2) Every second time period, the storage server collects each piece of storage-side throughput data counted in the second time period; obtains the second identifier of the operation object; generates storage-side average throughput data and the second identifier that carry the storage-side Average throughput information, which sends the average throughput information on the storage side to the management server.

According to the above response process to the operation, the storage IOPS and the storage bandwidth can represent the performance of the storage server. Therefore, the storage server can send the average storage throughput information to the management server as the storage performance information.

Specifically, when the storage-side average throughput information is the storage-side IOPS, the storage server counts the number of operations per second on the storage-side I/O interface when performing an operation on an operation object, obtains the storage-side IOPS, and obtains the IOPS carried in the operation. The second identifier of the storage-side average throughput information carrying the storage-side IOPS and the second identifier is generated. When the average throughput information on the storage side is the bandwidth on the storage side, the storage server counts the amount of data per second on the storage side when performing an operation on an operation object to obtain the bandwidth on the storage side, and obtains the second identifier carried in the operation to generate the There is the storage-side bandwidth and the storage-side average throughput information of the second identifier. Optionally, the storage data server may also obtain the storage server ID, and add the storage server ID to the average throughput information on the storage side.

In this embodiment, the order in which the application server obtains the average throughput information on the application side and the storage server obtains the average throughput information on the storage side is not limited.

It should be noted that the application server can actively send application-side performance information to the management server, and can also send application-side performance information when receiving a request from the management server; similarly, the storage server can actively send storage-side performance information to the management server , the storage side performance information may also be sent when receiving the request sent by the management server, which is not limited in this embodiment.

Step 1002, receiving the correspondence between the first identifier and the second identifier of the operation object sent by the application server, wherein the first identifier includes at least one of the name and logical address of the operation object, and the second identifier includes the physical address of the operation object ; and the corresponding relationship is to obtain the first corresponding relationship between the name of each operation object and the logical address by the application server, and obtain the second corresponding relationship between the name of each operation object and the physical address, which will include the first corresponding relationship of the same name and The second corresponding relationship is obtained by merging.

Wherein, the first correspondence may include: the name and logical address of the operation object, and the second correspondence may include: the name and physical address of the operation object. After the application server combines the first correspondence and the second correspondence, the obtained The corresponding relationship may be: the name of the operation object, the logical address, and the physical address.

Since the first identifier includes at least one of the name and logical address of the operation object, and the second identifier includes the physical address of the operation object, the above correspondence may indicate a correspondence between the first identifier and the second identifier.

After receiving the corresponding relationship, the management server may store the corresponding relationship.

It should be noted that due to disk damage, backup, and other reasons, the physical address of the object in the storage server may change. Therefore, the application server needs to update the corresponding relationship every predetermined time period. In a possible implementation manner, the application server may update the corresponding relationship every first time period, and send the updated corresponding relationship and application-side performance information to the management server; in another possible implementation manner , because the frequency of change of the geographical address of the object may be low, therefore, the application server can set a third time period, update the corresponding relationship every third time period, and send the updated corresponding relationship to the management server, wherein, the first The duration of the three time periods is longer than the duration of the first time period. For example, the duration of the first time period is 5 minutes, the duration of the third time period is 1 day, etc., which are not limited in this embodiment.

Step 1003, read the first identifier of the operation object from the application-side performance information of each operation object.

The management server receives at least one piece of application-side performance information sent by the application server, and receives at least one piece of storage-side performance information sent by the storage server. At this time, the management server needs to determine the application-side performance information and storage-side performance information corresponding to the same operation object.

Specifically, since the application-side performance information carries the first identifier, the management server can directly read the first identifier from the application-side performance information.

Step 1004, read the second identifier of the operation object from the storage-side performance information of each operation object.

Specifically, since the storage-side performance information carries the second identifier, the management server can directly read the second identifier from the storage-side performance information.

This embodiment does not limit the execution sequence of step 1003 and step 1004.

Step 1005: Determine the application-side performance information and storage-side performance information corresponding to each operation object according to the corresponding relationship.

After obtaining each first identifier and each second identifier, the management server determines the first identifier and the second identifier of each operation object according to the corresponding relationship. For the first identifier and the second identifier of each operation object, determine the application-side performance information including the first identifier and the storage-side performance information including the second identifier as the application-side performance information and storage-side performance information corresponding to the operation object performance information.

It should be noted that step 1002 only needs to be executed before step 1005, and this embodiment does not limit the execution sequence of step 1002 and other steps.

After determining the application-side performance information and storage-side performance information corresponding to each operation object, the management server may send the application-side performance information and storage-side performance information corresponding to each operation object to the result outputter. The result exporter can output the application-side performance information and storage-side performance information corresponding to each operation object, and mark the application-side performance information according to the application server ID, and mark the storage-side performance information according to the storage server ID.

When the result output device is a display, it can directly display the performance information of the application side and the performance information of the storage side in the interface, so that developers can intuitively understand the relationship between the performance information of the application side and the performance information of the storage side, and improve information acquisition. efficiency. The following takes the result output device as the display as an example to introduce the display method.

In the first display mode, the monitor displays the above content in chronological order. In the second display manner, the display can display the above content in various dimensions. For example, space utilization dimension, I/O heat dimension, I/O size distribution dimension, delay dimension, etc.

Please refer to the interface schematic diagram of display information shown in FIG. 10B , where (1) in FIG. 10B shows IOPS, (2) shows delay, and (3) shows bandwidth.

Developers can observe the content output by the result outputter. When the developer determines that there is a problem with the application server according to the content, he can input the first configuration information to the application server, and the application server fixes the problem according to the first configuration information. For example, when the developer determines that there is a problem with the task priority of the application server according to the content, he may input task priority configuration information to the application server, and the application server reconfigures the task priority according to the configuration information.

When the developer determines that there is a problem with the storage server according to the content, he may input second configuration information to the storage server, and the storage server fixes the problem according to the second configuration information. For example, when the developer determines that the storage space of the storage server is insufficient according to the content, he may input space configuration information to the storage server, and the storage server re-allocates storage space according to the space configuration information.

Since the efficiency and accuracy of the developers to determine whether there is a problem with the application server or the storage server based on the content are low, therefore, in a possible implementation, the management server can continue to analyze the above content to determine whether it is the application server If there is still a problem with the storage server, step 1006 is performed at this time.

Step 1006, when the application-side performance information includes application-side delay information, and the storage-side performance information includes storage-side delay information, subtract the storage-side delay from the application-side delay to obtain a delay difference; detect the delay Whether the difference is greater than the first threshold; when the detection result is that the delay difference is greater than the first threshold, an analysis result indicating that there is a problem with the application server is generated.

Wherein, the delay difference is the time consumed by the application server itself to respond to the operation. When the time delay difference is large, it means that the application server cannot respond to the operation in time, and an analysis result of the problem with the application server is generated; when the time delay difference is small, it means that the application server can respond to the operation in time, and an analysis result of the problem with the application server is generated result.

For example, if the delay on the application side is 3s, the delay on the storage side is 2s, and the calculated delay difference is 1s, if the first threshold is 0.5s, the analysis result of the problem with the application server is generated; if the first threshold is 2s, Generate an analysis result that there is no problem with the application server.

Step 1007 , detecting whether the storage-side delay is greater than a second threshold; when the detection result is that the storage-side delay is greater than the second threshold, an analysis result indicating that there is a problem with the storage server is generated.

Among them, the storage side delay is the time consumed by the storage server itself to respond to the operation. When the delay on the storage side is large, it means that the storage server cannot respond to the operation in time, and an analysis result indicating that there is a problem with the storage server is generated; when the delay on the storage side is small, it means that the storage server can respond to the operation in time, and there is no problem with the storage server analysis results.

For example, if the delay on the storage side is 2s, if the second threshold is 1s, an analysis result indicating that there is a problem with the storage server is generated; if the second threshold is 3s, an analysis result indicating that there is no problem with the storage server is generated.

Step 1008, determine whether there is a problem with the application server or the storage server according to the analysis result.

It should be noted that the management server may also store the analysis result in the storage server, and subsequently, other devices may obtain the analysis result from the storage server.

Optionally, the application server can also obtain and collect other performance information of the application server. The following takes other performance information as waiting event information as an example for illustration. The performance monitoring method provided in this embodiment also includes: receiving the operation sent by the application server The waiting event information of the object; determine whether there is a problem with the application server according to the waiting event information.

Since the above-mentioned waiting event information is counted on the application server, the application server can directly send the waiting event information to the management server. Wherein, the waiting event information refers to the time consumed by the application server under some kind of waiting. For example, the waiting event information may include but not limited to: central processing unit (English: Central Processing Unit; abbreviation: CPU) waiting time, CPU waiting task sorting, I/O waiting time, non-I/O waiting time, idle time, sampling time consuming.

In a possible implementation manner, the management server does not analyze the waiting event information, but directly sends the waiting event information to the result output device, and the result output device displays the waiting event information for developers to analyze and determine whether it is Whether there is a problem with the application server or the storage server. In another possible implementation, the management server sets a waiting threshold for each waiting event, compares the waiting event information with the corresponding waiting threshold, and determines that there is a problem with the application server when the waiting event information is greater than the corresponding waiting threshold; When the waiting event information is less than the corresponding waiting threshold, it is determined that there is no problem with the application server.

Optionally, the application server may also acquire the application server ID, and add the application server ID to the waiting event information.

To sum up, in the performance monitoring method provided by the embodiment of the present invention, the management server determines the application-side performance information and storage-side performance information corresponding to each operation object, and then the application-side performance information and storage-side performance information corresponding to each operation object Analysis of performance information can determine whether there is a problem with the application server or the storage server based on the analysis results, which solves the problem of being unable to determine whether the problem is with the application server or the storage server, which leads to monitoring errors, and achieves the effect of improving the accuracy of monitoring.

The application server collects the first corresponding relationship and the second corresponding relationship respectively, and merges the first corresponding relationship and the second corresponding relationship into a corresponding relationship, so that the application-side performance information and the storage-side performance information of the operation object can be processed according to the corresponding relationship. The correspondence solves the problem that the performance information of the application side and the performance information of the storage side cannot be matched, and achieves the effect of improving the accuracy of the analysis.

By comparing the delay difference with the first threshold, the management server can automatically determine whether there is a problem with the application server, which solves the need for developers to compare the delay difference with the first threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

By comparing the storage-side delay with the second threshold, the management server can automatically determine whether there is a problem with the storage server, which solves the need for developers to compare the storage-side delay with the second threshold to determine the low efficiency and accuracy of the problem problem, and achieve the effect of improving the efficiency and accuracy of determining the problem.

Please refer to FIG. 11 , which shows a structural block diagram of a performance monitoring device provided by an embodiment of the present invention. The performance monitoring device can be implemented as all or a part of a management server through software, hardware or a combination of the two. The performance monitoring device includes: a receiving unit 1110 , a reading unit 1120 , a determining unit 1130 and an analyzing unit 1140 .

A receiving unit 1110, configured to implement the functions of the above step 901 and step 902;

The reading unit 1120 is configured to realize the functions of the above step 903 and step 904;

A determining unit 1130, configured to implement the functions of the above steps 905 and 907;

The analysis unit 1140 is configured to realize the function of the above step 906.

In another optional embodiment, the above-mentioned receiving unit 1110 is used to realize the functions of step 1001 and step 1002; the above-mentioned reading unit 1120 is used to realize the functions of step 1003 and step 1004; the above-mentioned determining unit 1130 is used to Realize the functions of step 1005 and step 1008; the analysis unit 1140 is used to realize the functions of step 1006 and step 1007.

Relevant details may be combined with reference to the foregoing method embodiments.

It should be noted that the above-mentioned receiving unit 1110 can be realized by the processor of the management server executing the receiving module in the memory; the above-mentioned reading unit 1120 can be realized by the processor of the management server executing the reading module in the memory; the above-mentioned The determination unit 1130 can be realized by the processor of the management server executing the determination module in the memory; the analysis unit 1140 can be realized by the processor of the management server executing the analysis module in the memory.

Those of ordinary skill in the art can understand that all or part of the steps for implementing the above embodiments can be completed by hardware, and can also be completed by instructing related hardware through a program. The program can be stored in a computer-readable storage medium. The above-mentioned The storage medium mentioned may be a read-only memory, a magnetic disk or an optical disk, and the like.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within range.

Claims (10)

1. A performance monitoring method, characterized in that, it is used in a management server, and the management server is connected to an application server and a storage server respectively, and the storage server provides data storage for the application server, and the method includes: receiving the application-side performance information of the operation object sent by the application server, and receiving the storage-side performance information of the operation object sent by the storage server; receiving the correspondence between the first identifier and the second identifier of the operation object sent by the application server; Reading the first identifier of the operation object from the application-side performance information of each operation object; Read the second identifier of the operation object from the storage-side performance information of each operation object; Determining application-side performance information and storage-side performance information corresponding to each operation object according to the correspondence relationship; Analyze the application-side performance information and storage-side performance information corresponding to each operation object; According to the analysis result, it is determined whether there is a problem with the application server or the storage server. 2. The method according to claim 1, wherein the first identifier includes the name and logical address of the operation object, and the second identifier includes the physical address of the operation object; The correspondence is obtained by merging the first correspondence including the same name and the second correspondence, the first correspondence is the correspondence between the name of each operation object and the logical address, and the second correspondence It is the corresponding relationship between the name of each operation object and the physical address. 3. The method according to claim 1, characterized in that, the method further comprises: receiving the waiting event information of the operation object sent by the application server; Determine whether there is a problem with the application server according to the waiting event information. 4. The method according to any one of claims 1 to 3, wherein when the application-side performance information includes application-side delay information, and the storage-side performance information includes storage-side delay information, the Describes the analysis of application-side performance information and storage-side performance information corresponding to each operation object, including: Subtracting the storage-side delay from the application-side delay to obtain a delay difference; Detecting whether the delay difference is greater than a first threshold; When the detection result is that the delay difference is greater than the first threshold, an analysis result indicating that there is a problem with the application server is generated. 5. The method according to claim 4, wherein the analysis of the application-side performance information and storage-side performance information corresponding to each operation object further comprises: Detecting whether the storage-side delay is greater than a second threshold; When the detection result is that the storage side delay is greater than the second threshold, an analysis result indicating that there is a problem with the storage server is generated. 6. A performance monitoring device, characterized in that it is used in a management server, the management server is respectively connected to an application server and a storage server, and the storage server provides data storage for the application server, and the device includes: a receiving unit, configured to receive the application-side performance information of the operation object sent by the application server, and receive the storage-side performance information of the operation object sent by the storage server; The receiving unit is further configured to receive the correspondence between the first identifier and the second identifier of the operation object sent by the application server; A reading unit, configured to read the first identifier of the operation object from the application-side performance information of each operation object received by the receiving unit; The reading unit is further configured to read the second identifier of the operation object from the storage-side performance information of each operation object received by the receiving unit; A determining unit, configured to determine application-side performance information and storage-side performance information corresponding to each operation object read by the reading unit according to the correspondence received by the receiving unit; An analyzing unit, configured to analyze application-side performance information and storage-side performance information corresponding to each operation object determined by the determining unit; The determination unit is further configured to determine whether there is a problem with the application server or the storage server according to the analysis result obtained by the analysis unit. 7. The device according to claim 6, wherein the first identifier includes the name and logical address of the operation object, and the second identifier includes the physical address of the operation object; The correspondence is obtained by merging the first correspondence including the same name and the second correspondence, the first correspondence is the correspondence between the name of each operation object and the logical address, and the second correspondence It is the corresponding relationship between the name of each operation object and the physical address. 8. The device of claim 6, wherein: The receiving unit is further configured to receive the waiting event information of the operation object sent by the application server; The determining unit is further configured to determine whether there is a problem with the application server according to the waiting event information received by the receiving unit. 9. The device according to any one of claims 6 to 8, wherein when the application-side performance information includes application-side delay information, and the storage-side performance information includes storage-side delay information, the The analysis unit described above is specifically used for: Subtracting the storage-side delay from the application-side delay to obtain a delay difference; Detecting whether the delay difference is greater than a first threshold; When the detection result is that the delay difference is greater than the first threshold, an analysis result indicating that there is a problem with the application server is generated. 10. The device according to claim 9, wherein the analysis unit is also used for: Detecting whether the storage-side delay is greater than a second threshold; When the detection result is that the storage side delay is greater than the second threshold, an analysis result indicating that there is a problem with the storage server is generated.