WO2019101087A1

WO2019101087A1 - Slow-disk detection method, and storage array

Info

Publication number: WO2019101087A1
Application number: PCT/CN2018/116608
Authority: WO
Inventors: 连涛; 曹红强; 仇幼成
Original assignee: 华为技术有限公司
Priority date: 2017-11-22
Filing date: 2018-11-21
Publication date: 2019-05-31
Also published as: CN109815037B; CN109815037A

Abstract

A slow-disk detection method, and a storage array. The method is performed by the storage array, the storage array comprises at least one set of disks, and each set of disks comprises at least one disk. The method comprises: detecting N disk parameter values of each disk of the at least one set of disks in the storage array (210), wherein N is a positive integer greater than or equal to 2; calculating a slow-disk probability of each disk according to the N disk parameter values of the disk (220); and determining a slow disk in each set of disks according to the slow-disk probability of each disk (230). The invention detects the N disk parameter values of each disk of the at least one set of disks, so as to take multiple disk parameter values of each disk into consideration during a slow-disk detection process, thereby improving the accuracy of slow-disk detection results.

Description

Slow disk detection method and storage array

Technical field

The present application relates to the field of computer technology, and more particularly to a slow disk detection method and a memory array.

Background technique

A storage array typically includes a disk array. Affected by various factors, the disk in the disk array may have a slow response to input and output (I/O) in the latter part of the life cycle, and may even fail to respond to I/O. This type of disk is called a slow disk.

The existence of a slow disk can affect the upper layer services, for example, the processing of business data is not timely. Therefore, you need to perform slow disk detection on the slow disk in the disk array and isolate the service data carried on the slow disk.

Traditional techniques typically perform slow disk detection based on the I/O average service time. For example, if a disk has an I/O average service time in the detection period that is greater than a preset threshold, the disk is determined to be a slow disk. However, because the disk is affected by different business models, the results of slow disk detection are inaccurate.

Summary of the invention

The present application provides a slow disk detection method and a storage array, which can improve the accuracy of slow disk detection.

In a first aspect, a slow disk detection method is provided, which is performed by a storage array, the storage array including at least one disk set, each disk set including at least one disk, the method comprising: detecting the storage array N disk parameter values of each disk in at least one disk set, wherein N is a positive integer greater than or equal to 2; calculating a slow disk probability of each disk according to the N disk parameter values of each disk; The slow disk in each disk set is judged according to the slow disk probability of each disk.

The above technical solution detects N disk parameter values of each disk in at least one disk set. In the slow disk detection process, by considering the value of multiple disk parameters of each disk, the result of slow disk detection is more accurate than the traditional method of slow disk detection based only on the factor of I/O average service time. .

In a possible implementation manner, the calculating, according to the N disk parameter values of each disk, a slow disk probability of each disk, comprising: determining, by using each disk parameter value of each detected disk a parameter interval that falls within, wherein each of the disk parameters corresponds to at least one parameter interval; determining a probability corresponding to a parameter interval in which each disk parameter value of each detected disk falls, wherein each of the parameters The parameter interval in which the disk parameter values fall corresponds to a probability; calculating the slow disk probability of each disk according to the probability corresponding to the N disk parameters of each disk; determining according to the slow disk probability of each disk The slow disk in the disk set.

In a possible implementation, each disk parameter corresponding to each disk set corresponds to a weight, and the slowness of each disk is calculated according to a probability corresponding to the N disk parameters of each disk. The disk probability includes: calculating a slow disk probability of each disk according to a probability corresponding to the N disk parameters of each disk and a weight corresponding to the N disk parameters of each disk.

In the foregoing technical solution, the probability corresponding to the N disk parameters of each disk may be weighted and summed according to the weight corresponding to the N disk parameters, and the total probability that each disk is a slow disk is determined. The result of slow disk detection is more accurate by considering the N disk parameters of each disk.

In one possible implementation, each of the disk sets has the same disk characteristics.

In the above technical solution, the disk characteristics of the same disk set are the same, which is equivalent to the slow disk detection under the premise of eliminating the influence of the disk characteristics on the slow disk detection result, so that the result of the slow disk detection is more accurate.

In a possible implementation manner, at least one of the disk characteristics corresponding to different disk sets is different.

In a second aspect, a storage array is provided, the storage array includes at least one disk set, each disk set includes at least one disk, and the storage array includes: a detecting unit, configured to detect in the storage array N disk parameter values of each disk in at least one disk set, where N is a positive integer greater than or equal to 2; a computing unit configured to calculate each of the disks according to N disk parameter values of each disk The slow disk probability; the determining unit is configured to determine the slow disk in each disk set according to the slow disk probability of each disk.

In a possible implementation manner, the calculating unit is specifically configured to: determine a parameter interval in which each disk parameter value of each detected disk falls, wherein each disk parameter corresponds to at least one a parameter interval; determining a probability corresponding to a parameter interval in which each disk parameter value of each of the detected disks falls, wherein the parameter interval in which each disk parameter value falls corresponds to a probability; Calculating the probability of the slow disk of each disk according to the probability corresponding to the N disk parameters of each disk; determining the slow disk in the disk set according to the slow disk probability of each disk.

In a possible implementation, each of the disk parameters corresponding to each disk set corresponds to a weight, and the calculating unit is specifically configured to: according to the probability and location of the N disk parameters of each disk Calculate the slow disk probability of each disk by the weight corresponding to the N disk parameters of each disk.

In a third aspect, a memory array is provided, comprising a processor and a memory, the memory for storing computer instructions, the processor for executing computer instructions stored in the memory, when the computer instructions are executed, The processor is operative to perform the method of any of the first aspect or the first aspect of the first aspect.

In a fourth aspect, a computer storage medium is provided, comprising computer instructions that, when executed on a computer, cause the computer to perform the first aspect or any of the possible implementations of the first aspect Methods.

In a fifth aspect, a computer program product comprising instructions for causing said computer to perform said first aspect or any of the possible implementations of the first aspect, when said computer program product is run on a computer Methods.

DRAWINGS

1 is a diagram showing an example of the structure of a memory array to which an embodiment of the present invention is applicable.

FIG. 2 is a schematic flowchart of a slow disk detecting method according to an embodiment of the present invention.

FIG. 3 is a diagram showing an example of a grouping manner of a disk to be detected according to an embodiment of the present invention.

4 is a schematic diagram of a slow disk probability model corresponding to an I/O average service time of a disk.

FIG. 5 is a schematic diagram of a slow disk probability model corresponding to the number of I/Os in which the I/O processing time exceeds a preset threshold.

FIG. 6 is a schematic structural diagram of a memory array according to an embodiment of the present invention.

FIG. 7 is a schematic structural diagram of a memory array according to an embodiment of the present invention.

Detailed ways

The technical solutions in the present application will be described below with reference to the accompanying drawings.

1 is a diagram showing an example of the structure of a memory array to which an embodiment of the present invention is applicable. As shown in FIG. 1, the disk array 100 can be located inside the disk frame 110. The disk frame may include a plurality of disk slots, and each disk slot may have one disk 120 placed. The disk mentioned in the present application may be, for example, a solid state drive (SSD), a serial advanced technology attachment (SATA) disk, and a serial attached small computer system interface (serial attached small computer system interface, SAS) disk, near-line SAS (near line-SAS, NL-SAS) disk, etc. The storage array shown in FIG. 1 may be, for example, a redundant array of independent disks (RAID).

The embodiment of the invention provides a slow disk detection method, which can improve the accuracy of slow disk detection. The embodiments of the present invention are described in detail below with reference to FIG.

FIG. 2 is a schematic flowchart of a slow disk detecting method according to an embodiment of the present invention. The method of Figure 2 can be performed by a storage array, such as a processor in a storage array in which the disk is located. The method of FIG. 2 may include steps 210-230, which are described in detail below.

In step 210, N disk parameter values of each disk in at least one disk set in the storage array are detected, where N is a positive integer greater than or equal to 2.

It should be understood that the disks in the disk array can be divided into different disk sets according to different disk characteristics, and the disks in the disk set have the same disk characteristics. The disk characteristics can be, for example, the hardware characteristics of the disk or the type of service stored on the disk.

The embodiment of the present invention does not specifically limit the type of hardware features of the disk, and the hardware features of the disk may be various. For example, the hardware features can include at least one of the following: type of disk, speed, protocol type.

The disk type may include, for example, at least one of the following types: SSD, SATA, SAS, NL-SAS, and the like. The rotational speed of the disk may include, for example, at least one of the following rotational speeds: 5400 rpm, 7200 rpm, 10,000 rpm, 15,000 rpm. The protocol type of the disk may include, for example, at least one of the following protocols: a non-volatile memory express (NVME) protocol, an internet small computer system interface (ISCSI) protocol, and an advanced technology accessory. (advanced technology attachment, ATA) agreement.

Optionally, in some embodiments, the type of service stored on the disk may be, for example, when the disk is allocated to the upper layer application, and the data generated by the upper layer application is stored on the disk, and different upper layer applications may generate different data. Types of.

The disk characteristics corresponding to each disk set will be described in detail below with reference to FIG. Referring to Figure 3, the type of service stored on the disk can be selected as the disk domain to which the disk belongs, and the hardware characteristics of the disk are selected as the type and/or speed of the disk. As shown in FIG. 3, based on the above characteristics of the disk to be detected, the disk to be detected can be divided into four detection domains (also referred to as disk sets) as shown in FIG. The disk in the detection domain 1 meets the following conditions: the disk domain to which it belongs is disk domain 0, the disk is SAS disk, and the disk speed is 10,000 rpm. The disk in the detection domain 2 meets the following conditions: the disk domain to which it belongs is disk domain 0, the disk is SAS disk, and the disk speed is 15000 rpm. The disk in the detection domain 3 meets the following conditions: the disk domain to which it belongs is the disk domain 1 and the disk is the SDD disk. The disk in the detection domain 4 meets the following conditions: the disk domain to which it belongs is the disk domain 2, and the disk is the NVME disk. In Figure 3, at least one of the corresponding disk characteristics in different detection domains (disk sets) may have different disk characteristics.

During the actual detection process, the hard disk in the detection domain 1 - detection domain 4 may be sequentially detected. Alternatively, it is also possible to perform slow disk detection on the disks in the four detection domains in parallel.

In the embodiment of the present invention, the object detected by the slow disk is selected as a hardware feature of the disk and/or a disk set having the same service type stored on the disk. And detecting multiple parameter values for the disk in the disk set, so that the result of the slow disk detection is more accurate.

It should be understood that the N disk parameter values of each disk can be N factors for measuring whether the disk is a slow disk. The N disk parameters may include, for example, some or all of the following factors: I/O average service time, I/O processing time exceeding the preset threshold, and number of unresponsive I/Os. Wait. The average I/O service time can be, for example, the average time that the disk itself needs to process a single I/O. The number of I/Os whose I/O processing time exceeds a preset threshold can be, for example, the time that the disk processing I/O exceeds the setting. The number of I/Os of the threshold time, and the number of unresponsive I/Os may be, for example, the number of I/Os that are not responding to all I/Os of the disk.

In step 220, the slow disk probability of each disk is calculated according to the N disk parameter values of each disk.

Optionally, in some embodiments, the parameter interval in which each disk parameter value of each disk detected may be determined, wherein each disk parameter may correspond to at least one parameter interval. The slow disk probability of each disk can be calculated by the probability corresponding to the parameter interval in which each disk parameter value of each disk falls. Detailed description will be made below in conjunction with specific embodiments.

Taking the I/O average service time as a parameter value of the disk, for a single disk, generally, the longer the average I/O service time, the greater the probability that the disk is a slow disk. Figure 4 is a slow disk probability model corresponding to the I/O average service time of the disk. As shown in FIG. 4, when the I/O average service time of the disk is in the interval 0-S ₁ , the probability that the disk is a slow disk is x%. When the I/O average service time of the disk is in the interval S ₁ -S ₂ , the probability that the disk is a slow disk is y%. When the I/O average service time of the disk is in the interval S ₂ -S ₃ , the probability that the disk is a slow disk is z%. As can be seen from Figure 4, as the average I/O service time of the disk is longer, the probability that the disk is a slow disk is greater. Therefore, the disk I/O average service time can be used as a measure of whether a disk is a slow disk.

For example, if the I/O processing time exceeds the preset threshold, the number of I/Os is a parameter value of each disk. For a single disk, the I/O processing time exceeds the preset threshold. The more you have, the more likely the disk is to be a slow disk. FIG. 5 is a slow disk probability model corresponding to the number of I/Os in which the I/O processing time exceeds a preset threshold. As shown in FIG. 5, the probability that the number of I/Os in the normal disk exceeds t ₃ (the set threshold time) by the total number of I/Os is p% (also referred to as slow disk). Rate), the probability that the I/O processing time in the slow disk exceeds t ₃ (the set threshold time) by the total number of I/Os is m%. The number of I/Os in which the I/O processing time in the slow disk exceeds the preset threshold is greater than the number of I/Os in the normal disk that exceeds the preset threshold. The greater the probability that the disk is a slow disk. . Therefore, the number of I/Os that the I/O processing time exceeds the preset threshold can be used as a measure of whether the disk is a slow disk.

The importance of the K factors mentioned above may not be the same, so as shown in Table 1, different weights may be assigned to each factor according to its importance.

Table 1

Then, according to the weight of each factor, the probability that each of the N disks is a slow disk under K factors can be weighted and summed to determine the total probability that each disk is a slow disk.

As one example, as shown in Table 2, to the disk D _{_1,} x _1% probability that D ₁ is represented by slow disk, y _1% D ₁ represents the probability of the disk at a slower factor 2 at a factor. For disk D2, x ₂ % represents the probability that D ₁ is a slow disk under factor 1, y ₂ % represents the probability that D ₁ is a slow disk under factor 2, and so on.

Table 2

Then, based on the weights of the various factors, the total probability of determining each disk as a slow disk can be calculated. As shown in Table 3, the probability that the disk D ₁ is determined to be a slow disk can be calculated by the following equation: P(D ₁ ) = x ₁ % * w ₁ + y ₁ % * w ₂ + .... Where x ₁ % represents the probability that D ₁ is a slow disk under factor 1, w ₁ represents the weight of factor 1, y ₁ % represents the probability that D ₁ is a slow disk under factor 2, and w ₂ represents the weight of factor 2 value.

table 3

In step 230, the slow disk in each disk set is determined according to the slow disk probability of each disk.

In the embodiment of the present application, N disk parameter values of each disk in at least one disk set are detected. In the slow disk detection process, by considering the value of multiple disk parameters of each disk, the result of slow disk detection is more accurate than the traditional method of slow disk detection based only on the factor of I/O average service time. .

The slow disk probability of each disk mentioned in step 230 may be the probability that the disk is a slow disk in one detection period, and may also include the probability that the disk is a slow disk in each detection period in multiple detection periods. For example, the suspected slow disk corresponding to each detection period may be selected from each disk according to the probability that each disk is a slow disk in each detection cycle of a plurality of detection cycles and a preset threshold. Then, the slow disk can be determined from the suspected slow disks corresponding to the multiple detection cycles (for example, the number of suspected disks in each disk is greater than the preset number of times, or the disk having the largest number of suspected disks in each disk is determined to be slow. plate).

The suspected slow disk corresponding to each detection period described above can be understood as a disk whose total probability of slow disks is greater than a preset threshold in each detection period (hereinafter, the probability that the disk is a slow disk in a certain detection period) The condition that is greater than the preset threshold is called condition one). The suspected slow disk corresponding to each cycle may be a slow disk or multiple slow disks. If a disk is a suspected slow disk corresponding to a certain detection period, the detection cycle may also be referred to as a slow cycle of the disk, indicating that the disk has a slow I/O response time during the detection cycle.

The embodiment of the present invention does not directly determine the slow disk detection result of a certain detection period as the final slow disk detection result, but comprehensively determines whether the slow disk is included in the disk to be detected based on the slow disk detection result of the multiple detection cycles. This slow disk detection method will make the slow disk detection result more accurate.

The above suspected slow disk corresponding to each detection cycle is determined based on the condition one above. Optionally, in some embodiments, the suspected slow disk corresponding to the detection period may also be determined by considering various other conditions. For example, it may be determined whether the data of the detection period satisfies the following condition 2: the probability of the disk with the highest probability of concentration in each disk is greater than the U (U is greater than 1) times of other disks in the disk set. If the condition 2 is not satisfied, it is determined that there is no suspected slow disk in the detection period; if the condition two is satisfied, it is determined whether there is a disk that satisfies the condition 1 in the detection period; if yes, the disk is determined as the detection The suspected slow disk corresponding to the cycle. For another example, it may be determined whether the data of the detection period satisfies the following condition three: at least L (L is a positive integer not less than 1) disks in the disk set to undertake data services. If the condition 3 is not satisfied, it is determined that there is no suspected slow disk in the detection period; if the condition three is satisfied, it is determined whether there is a disk that satisfies the condition 1 in the detection period; if yes, the disk is determined as the detection The suspected slow disk corresponding to the cycle. It should be noted that the foregoing condition 1, condition 2, and condition three may be combined in any combination, and the embodiment of the present invention is not limited thereto.

The slow disk detecting method provided by the embodiment of the present invention is described in detail above with reference to FIG. 2 to FIG. 5 . The memory array provided by the embodiment of the present invention will be described in detail below with reference to FIG. 6 . The memory array of Figure 6 can be used to perform the various steps above. The memory array of FIG. 6 may include a detecting unit 610, a calculating unit 620, and a determining unit 630.

The detecting unit 610 is configured to detect N disk parameter values of each disk in the at least one disk set in the storage array, where N is a positive integer greater than or equal to 2.

The calculating unit 620 is configured to calculate a slow disk probability of each disk according to the N disk parameter values of each disk.

The determining unit 630 is configured to determine, according to the slow disk probability of each disk, the slow disk in each disk set.

Optionally, in some embodiments, the calculating unit 620 may be specifically configured to: determine a parameter interval in which each disk parameter value of each detected disk falls, wherein each of the disk parameters corresponds to at least a parameter interval; determining a probability corresponding to a parameter interval in which each of the detected disk parameter values falls, wherein the parameter interval in which each disk parameter value falls corresponds to a probability; Calculating the probability of the slow disk of each disk according to the probability corresponding to the N disk parameters of each disk; determining the slow disk in the disk set according to the slow disk probability of each disk.

Optionally, in some embodiments, each disk parameter corresponding to each disk set corresponds to a weight, and the calculating unit 620 is further configured to: respond to the N disk parameters of each disk. The probability of the disk and the weight corresponding to the N disk parameters of each disk calculate the slow disk probability of each disk.

Optionally, in some embodiments, each of the disk sets has the same disk characteristics.

Optionally, in some embodiments, at least one of the disk characteristics corresponding to different disk sets is different.

FIG. 7 is a schematic structural diagram of a memory array according to an embodiment of the present invention. The memory array 700 of Figure 7 can perform the slow disk detection method described in any of the embodiments of Figures 2 through 5. The memory array 700 of FIG. 7 can include a memory 710 and a processor 720. Memory 710 can be used to store programs. The processor 720 can be configured to execute a program stored in the memory 710. When the program stored in the memory 710 is executed, the processor 720 can be used to execute the slow disk detection method described in any of the above embodiments.

It should be understood that in the embodiment of the present invention, the term "and/or" is merely an association relationship describing an associated object, indicating that there may be three relationships. For example, A and/or B may indicate that A exists separately, and A and B exist simultaneously, and B cases exist alone. In addition, the character "/" in this article generally indicates that the contextual object is an "or" relationship.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present invention are generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer readable storage medium or transferred from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be from a website site, computer, server or data center Transmission to another website site, computer, server or data center via wired (eg coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (eg infrared, wireless, microwave, etc.). The computer readable storage medium can be any available media that can be accessed by a computer or a data storage device such as a server, data center, or the like that includes one or more available media. The usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a digital video disc (DVD)), or a semiconductor medium (such as a solid state disk (SSD)). .

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods to implement the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present application.

A person skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including The instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program code. .

The foregoing is only a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application. It should be covered by the scope of protection of this application. Therefore, the scope of protection of the present application should be determined by the scope of the claims.

Claims

A method for slow disk detection is performed by a storage array, wherein the storage array includes at least one disk set, each disk set includes at least one disk, and the method includes:

Detecting N disk parameter values of each disk in at least one disk set in the storage array, where N is a positive integer greater than or equal to 2;

Calculating a slow disk probability of each disk according to the N disk parameter values of each disk;

The slow disk in each disk set is judged according to the slow disk probability of each disk.
The method of claim 1, wherein the calculating the slow disk probability of each disk according to the N disk parameter values of each disk comprises:

Determining a parameter interval in which each disk parameter value of each of the detected disks falls, wherein each disk parameter corresponds to at least one parameter interval;

Determining a probability corresponding to a parameter interval in which each of the detected disk parameter values falls, wherein the parameter interval in which each disk parameter value falls corresponds to a probability;

Calculating a slow disk probability of each disk according to a probability corresponding to the N disk parameters of each disk;

Determining the slow disk in the disk set according to the slow disk probability of each disk.
The method according to claim 1 or 2, wherein each disk parameter corresponding to each disk set corresponds to a weight,

Calculating the slow disk probability of each disk according to the probability corresponding to the N disk parameters of each disk, including: a probability corresponding to the N disk parameters of each disk and each of the disks The weight corresponding to the N disk parameters calculates the slow disk probability of each disk.
The method of any of claims 1 to 3, wherein each of the sets of disks has the same disk characteristics.
The method according to any one of claims 1 to 4, characterized in that at least one of the disk characteristics corresponding to different disk sets is different.
A storage array, wherein the storage array includes at least one disk set, each disk set includes at least one disk, and the storage array includes:

a detecting unit, configured to detect N disk parameter values of each disk in the at least one disk set in the storage array, where N is a positive integer greater than or equal to 2;

a calculating unit, configured to calculate a slow disk probability of each disk according to the N disk parameter values of each disk;

a determining unit, configured to determine, according to the slow disk probability of each disk, the slow disk in each disk set.
The storage array of claim 6, wherein the computing unit is specifically configured to:

Determining a parameter interval in which each disk parameter value of each of the detected disks falls, wherein each disk parameter corresponds to at least one parameter interval;

Determining a probability corresponding to a parameter interval in which each of the detected disk parameter values falls, wherein the parameter interval in which each disk parameter value falls corresponds to a probability;

Calculating a slow disk probability of each disk according to a probability corresponding to the N disk parameters of each disk;

Determining the slow disk in the disk set according to the slow disk probability of each disk.
The storage array according to claim 7, wherein each disk parameter corresponding to each disk set corresponds to a weight,

The calculating unit is specifically configured to: calculate a slow disk probability of each disk according to a probability corresponding to the N disk parameters of each disk and a weight corresponding to the N disk parameters of each disk.
A storage array according to any one of claims 6 to 8 wherein each of said sets of disks has the same disk characteristics.
The storage array according to any one of claims 6 to 9, wherein at least one of the disk characteristics corresponding to different disk sets is different.