CN114328122A - IO full life cycle time delay monitoring method and related device - Google Patents

Abstract

The application discloses an IO full life cycle time delay monitoring method, which comprises the following steps: respectively setting monitoring points for an initiating end and a target end; when the monitoring point is executed in the process that the initiating terminal initiates access to the target terminal, the current IO characteristic information is obtained based on the monitoring point, and a timestamp is obtained from a high-precision time service module; and performing delay data statistics on all the timestamps and all the IO characteristic information to obtain delay monitoring data. The method and the device realize interception of information of each monitoring point of the equipment at two ends in the whole process of initiating access, further carry out time delay statistics, realize time delay monitoring of multiple ends, not only pay attention to IO performance of a single system, and improve accuracy and precision of time delay monitoring. The application also discloses an IO full-life-cycle time delay monitoring method, an IO full-life-cycle time delay monitoring device, a server and a computer readable storage medium, and the method has the beneficial effects.

Description

A kind of IO full life cycle delay monitoring method and related device

technical field

The present application relates to the field of computer technology, and in particular, to a method for monitoring IO full life cycle delay, an IO full life cycle delay monitoring device, a server, and a computer-readable storage medium.

Background technique

With the continuous development of information technology, the update of network storage system is getting faster and faster. Among them, the IO performance of the network storage system in the development and testing stage or on-site deployment may not meet expectations due to factors such as system design or hardware environment. It is too inefficient to directly check the performance influencing factors from the software and hardware environment, and it is impossible to quickly locate the performance bottleneck. Test the performance of the network storage system.

In the related art, by monitoring and collecting IO (Input/Output, input/output) delay data at various stages of the whole life cycle, the performance bottleneck point can be quickly located and an optimized solution can be guided. However, the IO monitoring technology only focuses on the IO performance consumption within a single system, and cannot collect overall performance data from the full IO life cycle. The accuracy of the test data for the network storage system is reduced, and the actual delay of the storage system cannot be accurately reflected.

Therefore, how to improve the accuracy of testing the network storage system is a key issue concerned by those skilled in the art.

SUMMARY OF THE INVENTION

The purpose of the present application is to provide an IO full life cycle delay monitoring method, an IO full life cycle delay monitoring device, a server and a computer-readable storage medium, so as to improve the accuracy and precision of the delay monitoring.

In order to solve the above-mentioned technical problems, the present application provides a method for monitoring IO full life cycle delay, including:

Set up monitoring points for the initiator and target respectively;

When the monitoring point is reached in the process that the initiator initiates the access to the target, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module;

Perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data.

Optionally, set monitoring points for the initiator and target respectively, including:

Analyze the IO access process to obtain multiple stage points;

Monitoring points are set for corresponding stage points in the initiator and the target.

Optionally, before the setting of monitoring points, the method further includes:

A monitoring point collection plug-in and a high-precision timing module are respectively deployed on the initiator and the target.

Optionally, when the monitoring point is executed in the process that the initiator initiates the access to the target, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module, including:

When the monitoring point is reached in the process that the initiator initiates the access to the target, the high-precision timing module is accessed through the monitoring point collection plug-in, and the timestamp is obtained;

Obtain current IO feature information through the monitoring point collection plug-in.

Optionally, the IO feature information includes port pair information, IO unique identifier, and current processing stage information.

Optionally, perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data, including:

Based on the same IO unique identifier and the same current processing stage information, delay data statistics are performed on all the timestamps to obtain the delay monitoring data.

Optionally, also include:

Determine the stage information whose delay time is greater than the threshold based on the delay monitoring data;

The processing maintenance request is sent according to the stage information.

The application also provides an IO full life cycle delay monitoring device, including:

The monitoring point setting module is used to set monitoring points for the initiator and the target respectively;

The execution information acquisition module is used to obtain the current IO feature information based on the monitoring point and obtain the time stamp from the high-precision timing module when the monitoring point is executed in the process that the initiator initiates the access to the target terminal. ;

A delay data statistics module, configured to perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data.

The application also provides a server, including:

memory for storing computer programs;

The processor is configured to implement the steps of the above-mentioned method for monitoring the delay in the whole life cycle of IO when executing the computer program.

The present application also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned method for monitoring the delay of an IO full life cycle are implemented.

A method for monitoring IO full life cycle delay provided by the present application includes: setting monitoring points on an initiator and a target respectively; when the initiator initiates access to the target, the monitoring point is executed to the monitoring point. , obtain the current IO feature information based on the monitoring point, and obtain the time stamp from the high-precision timing module; perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data.

By setting a monitoring point on the initiator and the target, and then when the monitoring point is executed, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module, and finally the delay statistics are performed. In the whole process of initiating an access, the information of each monitoring point of the devices at both ends is intercepted, and then the delay statistics are carried out, which realizes the delay and delay monitoring of the multi-end, instead of only focusing on the IO performance of the single system, which improves the delay. Monitoring accuracy and precision.

The present application further provides a method for monitoring IO full life cycle delay, a device for monitoring IO full life cycle delay, a server and a computer-readable storage medium, which have the above beneficial effects, and are not repeated here.

Description of drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only It is an embodiment of the present application. For those of ordinary skill in the art, other drawings can also be obtained according to the provided drawings without any creative effort.

Fig. 1 is a flow chart of a method for monitoring IO full life cycle delay provided by an embodiment of the present application;

2 is a schematic diagram of the system structure of a method for monitoring IO full life cycle delay provided by an embodiment of the present application;

3 is a schematic structural diagram of a high-precision timing module provided by an embodiment of the application;

FIG. 4 is a schematic structural diagram of an IO full life cycle delay monitoring device provided by an embodiment of the present application.

Detailed ways

The core of the present application is to provide an IO full life cycle delay monitoring method, an IO full life cycle delay monitoring device, a server and a computer-readable storage medium, so as to improve the accuracy and precision of the delay monitoring.

In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

In the related art, by monitoring and collecting IO latency data at various stages of the entire life cycle, performance bottlenecks can be quickly located and an optimized solution can be guided. However, the IO monitoring technology only focuses on the IO performance consumption within a single system, and cannot collect overall performance data from the full IO life cycle. The accuracy of the test data for the network storage system is reduced, and the actual delay of the storage system cannot be accurately reflected. The IO full life cycle refers to the time from when the server initiates IO until the completion of receiving the network IO data of the storage system.

Therefore, the present application provides a method for monitoring IO full life cycle delay. By setting a monitoring point on the initiator end and the target end, and then when the monitoring point is executed, the current IO feature information is obtained based on the monitoring point, and the current IO feature information is obtained from the high The precision timing module obtains time stamps, and finally performs delay statistics, which realizes the interception of the information of each monitoring point of the devices at both ends during the whole process of initiating access, and then performs delay statistics to realize delay monitoring for multiple ends. , instead of only focusing on the IO performance of a single system, which improves the accuracy and precision of latency monitoring.

The following describes an IO full life cycle delay monitoring method provided by the present application through an embodiment.

Please refer to FIG. 1. FIG. 1 is a flowchart of a method for monitoring IO full life cycle delay provided by an embodiment of the present application.

In this embodiment, the method may include:

S101, setting monitoring points for the initiator and the target respectively;

It can be seen that this step aims to set monitoring points in the device of the initiator end and the device of the target end in the IO processing process. That is, both ends of the IO processing process set up corresponding monitoring points instead of just focusing on the IO performance inside a single system. By setting the monitoring points, the IO performance of the initiator and the target can be monitored respectively, and the performance of all terminals in the IO processing process can be monitored.

Wherein, the monitoring point may be a hook function in the set processing flow, and when the hook function of the monitoring point is activated in the processing flow, information can be acquired based on the hook function. The monitoring point may also be a set monitoring program, which detects the processing flow executed in the device, and when the stage of execution of the processing flow conforms to the set monitoring point, corresponding data can be collected.

Further, to improve the accuracy of monitoring point setting, this step may include:

Step 1, analyze the IO access process to obtain multiple stage points;

Step 2, setting monitoring points for the corresponding stage points in the initiator end and the target end.

It can be seen that this step is intended to explain how to set up monitoring points. In this optional solution, the IO access process is analyzed to obtain multiple stage points, and monitoring points are set for the corresponding stage points in the initiator end and the target end. Among them, it is to analyze which stages of the IO access process need to be executed, and the starting point of each stage is the stage point. Further, since the IO access process is executed in different devices at different stages. Therefore, according to the stage point corresponding to the initiator, which is the monitoring point, the monitoring point needs to be set for the corresponding stage point in the target terminal.

Further, before setting the monitoring point in this embodiment, it also includes:

Deploy monitoring point collection plug-ins and high-precision timing modules on the initiator and target ends respectively.

It can be seen that in this optional solution, the main purpose is to improve the accuracy of information acquisition and avoid the problem of inaccurate time acquisition. In this optional solution, the monitoring point collection plug-in and the high-precision timing module are respectively deployed on the initiator end and the target end. Among them, the monitoring point collection plug-in is mainly used to collect the information IO characteristic information of the monitoring point. The high-precision timing module is mainly used to obtain time stamps.

S102, when the initiating terminal initiates the access to the target terminal to the monitoring point, obtain current IO feature information based on the monitoring point, and obtain the time stamp from the high-precision timing module;

On the basis of S101, this step aims to obtain the current IO feature information based on the monitoring point and obtain the time stamp from the high-precision timing module when the monitoring point is reached during the process of initiating access to the target terminal by the initiator.

It can be seen that, when the access process is executed to the corresponding monitoring point, the monitoring point is triggered, and then the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module.

Further, this step may include:

Step 1, when the monitoring point is executed in the process of initiating access to the target terminal by the initiator, the high-precision timing module is accessed through the monitoring point collection plug-in, and the time stamp is obtained;

Step 2, obtain the current IO feature information through the monitoring point collection plug-in.

It can be seen that this optional solution mainly describes how to obtain information. In this optional solution, when a monitoring point is reached during the process of initiating an access from the initiator to the target, the high-precision timing module is accessed through the monitoring point collection plug-in, and the timestamp is obtained, and the current IO characteristics are obtained through the monitoring point collection plug-in. information. Among them, the high-precision timing module is based on the time signal of the satellite system and the PPS (Pulse Per Second) signal, which can provide a unified and highly synchronized time adjustment source for remote devices.

S103, perform delay data statistics on all time stamps and all IO feature information to obtain delay monitoring data.

On the basis of S102, this step aims to perform delay data statistics on all time stamps and all IO feature information to obtain delay monitoring data.

On the basis of obtaining the timestamp and all IO feature information, the time-consuming between each monitoring point can be determined through the timestamp, and then the time-consuming of each stage in the IO access process can be determined, so as to realize the complete IO Process latency monitoring.

The IO feature information includes port pair information, IO unique identifier, and current processing stage information.

Correspondingly, this step may include:

Based on the same IO unique identifier and the same current processing stage information, delay data statistics are performed on all timestamps to obtain delay monitoring data.

It can be seen that in this optional solution, the time information of the same IO access process can be sorted out from the obtained data through the unique description of the IO and the same current processing stage information, and the time to each stage can be determined based on the time information. extension information.

In addition, this embodiment may also include:

Step 1, based on the delay monitoring data, determine the stage information whose delay time is greater than the threshold;

Step 2: Send and process the maintenance request according to the stage information.

It can be seen that this optional solution mainly describes how to send the corresponding maintenance request. In this optional solution, based on the delay monitoring data, determine the stage information whose delay time is greater than the threshold, and send and process the maintenance request according to the stage information.

To sum up, in this embodiment, a monitoring point is set on the initiator end and the target end, and then when the monitoring point is executed, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module. Delay statistics, which realizes the interception of the information of each monitoring point of the devices at both ends in the whole process of initiating an access, and then performs delay statistics to realize the delay monitoring of multiple ends, instead of only focusing on the IO of a single system. performance, improving the accuracy and precision of delay monitoring.

A method for monitoring IO full life cycle delay provided by the present application will be further described below through a specific embodiment.

Please refer to FIG. 2 . FIG. 2 is a schematic diagram of a system structure of a method for monitoring a delay in an IO full life cycle provided by an embodiment of the present application.

In this embodiment, the system corresponding to the method may include: an initiator, a target, and the like.

The initiator generally refers to a server or other service host accessing the network storage system. The target end refers to the storage system that provides services such as storage IO blocks and files. Storage network refers to storage network such as FC network or Ethernet. IO monitoring point collection is an IO data collection plug-in deployed on the initiator and target sides. The plug-in is responsible for accessing the high-precision timing module to obtain accurate synchronization time. The plug-in uses real-time kernel drivers to reduce software access latency.

Among them, the characteristic data that needs to be recorded when the IO monitoring collects the total key node information of the IO path includes the PORT_PAIR (FC (Fiber Channel, Fibre Channel) network refers to the port pair composed of the remote port and the local port, which uniquely identifies an IO link in the system , Ethernet uses IP (Internet Protocol, Internet Protocol) and port to uniquely identify an IO link, Exchange ID (the unique ID used to identify an IO in the FC network, and the packet SequenceID is used in Ethernet), Time (time stamp) to read the current time of the timing module, and the Tag identifies the processing stage in the IO full life cycle.

Among them, the high-precision timing module is used to provide high-precision and synchronized timestamps for IO monitoring and collection in different systems to ensure the consistency of IO timestamps across systems. It is implemented by GPS (Global Positioning System) or Beidou satellite system High-precision clock synchronization, the accuracy can reach within 100ns.

Please refer to FIG. 3 , which is a schematic structural diagram of a high-precision timing module provided by an embodiment of the present application.

The work of the high-precision timing module is shown in Figure 3. Based on the time signal and PPS signal of the satellite system, it can provide a unified, high-synchronization time adjustment source for remote devices; in order to ensure the time acquisition accuracy, the time information before the message is processed. Before sending, it will access the internal high-speed counter (assuming that the counter count period is F1), record the counter time T1 and T2, and send the time information to the host together with the counter count values T1 and T2 after encoding; the host driver receives the interrupt from the high-speed timing module. When requesting, first read the high-speed timer C1 inside the CPU (central processing unit, central processing unit), and then decode the time information to obtain the time T3, and read the high-speed timer inside the CPU again when the time acquisition request reads the time information. (Assuming that the count period is F2) C2, using T1, T2 and C1, C2 can eliminate the time delay introduced by the software division to the greatest extent, and ensure the time accuracy. Actual time T=T3+|T2-T1|*F1+|C2-C1|*F2.

Based on the above description, the steps in this embodiment may include:

Step 1, deploy the IO monitoring point collection plug-in and the high-precision timing module on the initiator end and the target end respectively;

Step 2, analyze and determine the IO life cycle monitoring stage point, after the IO monitoring node is determined, the IO monitoring point collection and processing call can be embedded in the relevant node processing flow of the target end and the initiator end;

Step 3, the initiator initiates IO access to the target storage system;

Step 4: When the IO access process passes the monitoring stage point embedded in Step 2, the registered IO monitoring point collection will be called to collect the timestamp and IO feature information provided by the timing module, including PORT_PAIR, Exchange ID and the current stage point Tag Attributes;

Step 5: After the entire IO life cycle is over, the collected data will be aggregated together for performance data statistics;

Step 6: Obtain the delay statistics of each stage in a complete IO processing process by analyzing the collected IO characteristic data: the same Exchange id is the same IO, and the timestamp difference between adjacent tags is the IO processing delay in this stage;

Step 7, compare and analyze the links with relatively large delay data, and confirm whether the software processing and hardware configuration are reasonable.

It can be seen that in this embodiment, a monitoring point is set at the initiator end and the target end, and then when the monitoring point is executed, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module, and finally the delay is performed. Statistics, realize the interception of the information of each monitoring point of the devices at both ends in the whole process of initiating the access, and then carry out the delay statistics, realize the delay monitoring of the multi-end, instead of only focusing on the IO performance of the single system, improve the The accuracy and precision of delay monitoring are improved.

The following describes the IO full life cycle delay monitoring device provided by the embodiments of the present application. The IO full life cycle delay monitoring device described below and the IO full life cycle delay monitoring method described above may refer to each other correspondingly.

Please refer to FIG. 4 , which is a schematic structural diagram of an IO full-life-cycle delay monitoring device provided by an embodiment of the present application.

In this embodiment, the device may include:

The monitoring point setting module 100 is used for setting monitoring points for the initiator and the target respectively;

The execution information acquisition module 200 is used to acquire the current IO feature information based on the monitoring point and obtain the timestamp from the high-precision timing module when the monitoring point is executed in the process of initiating the access to the target terminal by the initiator;

The delay data statistics module 300 is configured to perform delay data statistics on all timestamps and all IO feature information to obtain delay monitoring data.

Optionally, the monitoring point setting module 100 is specifically configured to analyze the IO access process to obtain a plurality of stage points; and set monitoring points for corresponding stage points in the initiator end and the target end.

Optionally, the device may also include:

The plug-in deployment module is used to deploy the monitoring point collection plug-in and the high-precision timing module to the initiator and the target respectively.

Optionally, the execution information acquisition module 200 is specifically configured to access the high-precision timing module through the monitoring point collection plug-in when the monitoring point is executed in the process of initiating access to the target terminal by the initiator terminal, and obtain a time stamp; Click the collection plug-in to obtain the current IO feature information.

Optionally, the delay data statistics module 300 is specifically configured to perform delay data statistics on all timestamps based on the same IO unique identifier and the same current processing stage information to obtain delay monitoring data.

Optionally, the device may also include:

The maintenance request sending module is used to determine the stage information whose delay time is greater than the threshold value based on the delay monitoring data; send and process the maintenance request according to the stage information.

The embodiment of the present application also provides a server, including:

memory for storing computer programs;

The processor is configured to implement the steps of the method for monitoring the delay of the IO full life cycle as described in the above embodiments when executing the computer program.

Embodiments of the present application further provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the IO full life cycle delay as described in the above embodiments is implemented The steps of the monitoring method.

The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method.

Professionals may further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two, in order to clearly illustrate the possibilities of hardware and software. Interchangeability, the above description has generally described the components and steps of each example in terms of function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of this application.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be directly implemented in hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in random access memory (RAM), internal memory, read only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other in the technical field. in any other known form of storage medium.

The above provides a detailed introduction to a method for monitoring IO full life cycle delay, an IO full life cycle delay monitoring device, a server, and a computer-readable storage medium provided by the present application. Specific examples are used herein to illustrate the principles and implementations of the present application, and the descriptions of the above embodiments are only used to help understand the methods and core ideas of the present application. It should be pointed out that for those of ordinary skill in the art, without departing from the principles of the present application, several improvements and modifications can also be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims (10)

1. a IO full life cycle time delay monitoring method, is characterized in that, comprises: Set up monitoring points for the initiator and target respectively; When the monitoring point is reached in the process that the initiator initiates the access to the target, the current IO feature information is obtained based on the monitoring point, and the time stamp is obtained from the high-precision timing module; Perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data. 2. IO full life cycle time delay monitoring method according to claim 1, is characterized in that, monitoring point is set respectively to initiator end and target end, comprising: Analyze the IO access process to obtain multiple stage points; Monitoring points are set for corresponding stage points in the initiator and the target. 3. IO full life cycle time delay monitoring method according to claim 1, is characterized in that, before described setting monitoring point, also comprises: A monitoring point collection plug-in and a high-precision timing module are respectively deployed on the initiator and the target. 4. IO full life cycle time delay monitoring method according to claim 3, is characterized in that, when described initiating terminal initiates visit to described target terminal, when executing to described monitoring point, obtain based on this monitoring point Current IO feature information and time stamps obtained from the high-precision timing module, including: When the monitoring point is reached in the process that the initiator initiates the access to the target, the high-precision timing module is accessed through the monitoring point collection plug-in, and the timestamp is obtained; Obtain current IO feature information through the monitoring point collection plug-in. 5. The method for monitoring IO full life cycle delay according to claim 1, wherein the IO feature information comprises port pair information, IO unique identifier, and current processing stage information. 6. IO full life cycle delay monitoring method according to claim 5, is characterized in that, carry out delay data statistics to all described time stamps and all described IO feature information, obtain delay monitoring data, comprising: Based on the same IO unique identifier and the same current processing stage information, delay data statistics are performed on all the timestamps to obtain the delay monitoring data. 7. IO full life cycle time delay monitoring method according to claim 1, is characterized in that, also comprises: Determine the stage information whose delay time is greater than the threshold based on the delay monitoring data; The processing maintenance request is sent according to the stage information. 8. an IO full life cycle time delay monitoring device, is characterized in that, comprises: The monitoring point setting module is used to set monitoring points for the initiator and the target respectively; The execution information acquisition module is used to obtain the current IO feature information based on the monitoring point and obtain the time stamp from the high-precision timing module when the monitoring point is executed in the process that the initiator initiates the access to the target terminal. ; A delay data statistics module, configured to perform delay data statistics on all the time stamps and all the IO feature information to obtain delay monitoring data. 9. A server, characterized in that, comprising: memory for storing computer programs; The processor is configured to implement the steps of the method for monitoring the delay in the whole life cycle of IO according to any one of claims 1 to 7 when executing the computer program. 10. A computer-readable storage medium, characterized in that, a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the IO according to any one of claims 1 to 7 is implemented The steps of the whole life cycle delay monitoring method.

Images