Information determination method and device, computer readable storage medium and computer program product

Abstract

The invention provides an information determination method. The method comprises the following steps: acquiring first time delay information of a plurality of storage devices corresponding to a second server; for the first storage device whose first time delay information satisfies a target condition, monitoring multiple pieces of second time delay information of the first storage device in a target time period; determining whether the first storage device is abnormal or not based on the multiple pieces of second time delay information; and for the second storage device of which the first time delay information does not meet the target condition, whether the second storage device is abnormal is determined based on the first time delay information of the second storage device, so that the problem that whether the hard disk is abnormal cannot be accurately determined in related technologies is solved. The embodiment of the invention further provides a first server, a computer readable storage medium and a computer program product.

Description

Technical Field

[0001] This application relates to information determination technology in the field of computer technology, and more particularly to an information determination method, device, computer-readable storage medium, and computer program product. Background Technology

[0002] In the field of distributed storage, after a client sends a read / write request to a metadata server, the metadata server typically identifies the target hard drive from among multiple hard drives on the data server to read or write data to that target hard drive. However, after prolonged data read / write operations, hard drives can experience aging and fragmentation, leading to a significant decrease in read / write speeds (at this point, the hard drive can be termed a "slow drive"). Therefore, accurately identifying "slow drives" is crucial. Currently, related technologies typically compare hard drive performance data with threshold parameters to determine if the hard drive is abnormal, i.e., whether it is a "slow drive." However, this method cannot accurately determine whether a hard drive is abnormal or a "slow drive." Summary of the Invention

[0003] This application provides an information determination method, device, computer-readable storage medium, and computer program product, which can solve the problem in related technologies that it is impossible to accurately determine whether a hard disk is abnormal.

[0004] The technical solution of this application embodiment is implemented as follows: This application provides an information determination method, the method comprising: Obtain the first latency information of multiple storage devices corresponding to the second server; For a first storage device whose first latency information meets the target condition, monitor multiple second latency information of the first storage device within the target time period; Based on the multiple second delay information, it is determined whether the first storage device is abnormal; For a second storage device whose first latency information does not meet the target condition, determine whether the second storage device is abnormal based on the first latency information of the second storage device.

[0005] The method in the above scheme further includes: Based on multiple first delay information, a first value is determined for each first delay information; wherein, the first value characterizes the outlier degree of each first delay information; The first time delay information, which determines that the first value is greater than or equal to the first target threshold, satisfies the target condition; The first time delay information, which is determined to be less than the first target threshold, does not meet the target condition.

[0006] In the above scheme, determining whether the first storage device is abnormal based on the plurality of second latency information includes: Based on the plurality of second time delay information, a second value is determined; wherein, the second value characterizes the degree of dispersion among the plurality of second time delay information; If the second value is equal to the second target threshold, it is determined that the first storage device is not abnormal; If the second value is not equal to the second target threshold, the first operating characteristic information of the first storage device is determined; wherein, the first operating characteristic information characterizes the operating status of the first storage device; Based on the first operational characteristic information, it is determined whether the first storage device is abnormal.

[0007] In the above scheme, determining whether the second storage device is abnormal based on the first latency information of the second storage device includes: Based on the first latency information of the second storage device, a target score is determined for the first latency information; If the target score is greater than or equal to the third target threshold, it is determined that the second storage device is not abnormal; If the target score is less than the third target threshold, the second operating characteristic information of the second storage device is determined; wherein, the second operating characteristic information characterizes the operating status of the second storage device; Based on the second operational characteristic information, it is determined whether the second storage device is abnormal.

[0008] In the above scheme, determining the first operational characteristic information of the first storage device includes: Obtain first operating information and first health status information of the first storage device; wherein, the first health status information represents the wear and tear status of the first storage device; The first operational information and the first health status information are fused together to obtain the first operational feature information.

[0009] In the above scheme, determining whether the first storage device is abnormal based on the first operational characteristic information includes: The third operational characteristic information of the historical abnormal storage device is determined; wherein, the third operational characteristic information characterizes the operational status of the historical abnormal storage device; If the first operational feature information matches the third operational feature information, it is determined that the first storage device is abnormal. If the first operational characteristic information does not match the third operational characteristic information, it is determined that the first storage device is not abnormal.

[0010] In the above scheme, determining the third operational characteristic information of the historical abnormal storage device includes: The second operating information and the second health status information of the historical abnormal storage device are obtained; wherein, the second health status information characterizes the wear and tear status of the historical abnormal storage device. The second operational information and the second health status information are fused together to obtain the third operational feature information.

[0011] The method in the above scheme further includes: Receive a data write request sent by the client for the data to be processed; wherein the data write request carries an identifier of the data to be processed; The storage device with the abnormality is removed from the plurality of storage devices to obtain the target storage device; The target storage location is determined from the plurality of storage locations based on the identifier and the number of storage locations of the target storage device.

[0012] A first server, the device comprising: a processor, a memory, and a communication bus; The communication bus is used to realize the communication connection between the processor and the memory; The processor is used to execute the information determination program in the memory to implement the steps of the information determination method described above.

[0013] A computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the steps of the information determination method described above.

[0014] A computer program product comprising a computer program that, when executed by a processor, implements the aforementioned information determination method.

[0015] The information determination method, device, computer-readable storage medium, and computer program product provided in this application embodiment can obtain first latency information of multiple storage devices corresponding to a second server; for a first storage device whose first latency information meets the target condition, monitor multiple second latency information of the first storage device within a target time period; determine whether the first storage device is abnormal based on the multiple second latency information; and for a second storage device whose first latency information does not meet the target condition, determine whether the second storage device is abnormal based on the first latency information of the second storage device. Thus, the first latency information of multiple storage devices corresponding to the second server can be obtained first, and different methods can be used to determine whether the first storage device whose first latency information meets the target condition and the second storage device whose first latency information does not meet the target condition are abnormal. This allows for the determination of abnormal conditions of different types of storage devices among multiple storage devices using different methods, rather than simply comparing the hard drive performance data with threshold parameters as in related technologies to determine whether the hard drive is abnormal. This solves the problem in related technologies of not being able to accurately determine whether the hard drive is abnormal, thereby ensuring that data can be written to a normal hard drive and improving data processing efficiency. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating an information determination method provided in an embodiment of this application; Figure 2 This is a flowchart illustrating another information determination method provided in an embodiment of this application; Figure 3 This is a system architecture diagram corresponding to an information determination method provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an information determination device provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of a first server provided in an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0019] [In the following description, the terms "first," "second," and "third" are used merely to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that "first," "second," and "third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0020] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0021] This application provides an information determination method, which can be applied to a first server, as shown in the following embodiments. Figure 1 As shown, the method may include the following steps: Step 101: Obtain the first latency information of multiple storage devices corresponding to the second server.

[0022] In this embodiment of the application, the first server may refer to a metadata server (MDS) in a distributed storage system, the second server may refer to a data server (DS) in a distributed storage system, and the storage device may refer to a hard disk installed on the second server.

[0023] It should be noted that there can be multiple second servers, i.e., DS; the hard drives can include mechanical hard drives and solid-state drives.

[0024] In the embodiments of this application, the first latency information may refer to the input / output (I / O) latency generated by multiple storage devices when performing read / write operations. It should be noted that each storage device corresponds to one first latency information.

[0025] Step 102: For the first storage device whose first delay information meets the target conditions, monitor multiple second delay information of the first storage device within the target time period.

[0026] In this embodiment, the target condition may refer to a high outlier value for the first delay information, meaning that for a given first delay information, the difference between it and other first delay information is significant; the first storage device refers to a device among multiple storage devices where all first delay information satisfies the target condition. There may be multiple first storage devices.

[0027] It should be noted that the higher the outlier value of the first delay information, the higher the probability that the corresponding storage device is abnormal. In this case, it is necessary to combine other parameters for further evaluation. However, even if the outlier value of the first delay information is low, it cannot be directly determined that there is no abnormality, and further judgment is still required.

[0028] In this application embodiment, if the first latency information of a certain storage device meets the target condition, it indicates that the storage device is likely to be a slow disk. At this time, multiple second latency information of the device (i.e., the first storage device) can be continuously monitored in the future (i.e., the target time period).

[0029] It should be noted that the target time period can be preset according to actual needs; multiple second delay information can be generated after the first storage device performs multiple read / write operations within the target time period. Each read / write operation will generate one second delay information.

[0030] Step 103: Based on multiple second delay information, determine whether there is an anomaly in the first storage device.

[0031] In one feasible embodiment, after monitoring and obtaining multiple second delay information of the first storage device within a target time period, the multiple second delay information can be compared to determine the degree of difference between each second delay information and other second delay information, and the degree of difference can be used to determine whether the first storage device is abnormal.

[0032] In another feasible embodiment, the token bucket algorithm can be directly used to process multiple second latency information and the first target threshold, and the results of the algorithm output can be used to determine whether the first storage device is abnormal, that is, whether the first storage device is a slow disk.

[0033] Step 104: For the second storage device whose first latency information does not meet the target conditions, determine whether the second storage device is abnormal based on the first latency information of the second storage device.

[0034] In this embodiment of the application, the second storage device refers to the device among multiple storage devices where all the first latency information does not meet the target condition. There may be multiple second storage devices.

[0035] In this embodiment of the application, if the first latency information of a certain storage device does not meet the target condition, it means that the storage device is less likely to be a slow disk, that is, the probability of the storage device being a slow disk is relatively small. At this time, it is not necessary to continuously monitor multiple second latency information of the device (i.e. the second storage device) in the future period of time. Instead, the target vector machine algorithm can be directly used to process the first latency information of the second storage device, and the results of the algorithm output can be used to determine whether the second storage device is abnormal.

[0036] In this embodiment, the first server can actively determine whether the hard drive installed on the second server is abnormal by actively sending read and write operation requests to the second server, instead of passively waiting for read and write operation requests from the client as in related technologies. This would prevent the active and real-time identification of slow disks with abnormalities, thereby improving the identification efficiency of slow disks and ensuring that data can be written to normal hard drives.

[0037] The information determination method provided in the embodiments of this application can first obtain the first latency information of multiple storage devices corresponding to the second server, and use different methods to determine whether the first storage device whose first latency information meets the target condition and the second storage device whose first latency information does not meet the target condition are abnormal. In this way, different methods can be used to determine the abnormality of different types of storage devices among multiple storage devices, instead of simply comparing the hard drive performance data with the threshold parameter to determine whether the hard drive is abnormal, as in related technologies. This solves the problem in related technologies that it is impossible to accurately determine whether the hard drive is abnormal, thereby ensuring that data can be written to normal hard drives and improving data processing efficiency.

[0038] Based on the foregoing embodiments, embodiments of this application provide an information determination method, which can be applied to a first server, referring to... Figure 2 As shown, the method may include the following steps: Step 201: The first server obtains the first latency information of multiple storage devices corresponding to the second server.

[0039] In the embodiments of this application, such as Figure 3 As shown, the first server can periodically send data processing requests to the second server through the probe component. After receiving the data processing request, the second server can perform corresponding operations on the data according to the data processing request and return the latency information of the operation to the first server.

[0040] It should be noted that after obtaining the first latency information, the detection component of the first server can send it to the computing component of the first server for subsequent data processing.

[0041] Step 202: The first server determines the first value corresponding to each first delay information based on multiple first delay information.

[0042] The first value represents the degree of outlier status of each first time delay information.

[0043] In the embodiments of this application, the first value may refer to the outlier factor. Specifically, the computing component can directly use the Local Outlier Factor (LOF) algorithm to process multiple first time delay information and output multiple outlier factors, i.e., multiple first values.

[0044] In the embodiments of this application, there is a one-to-one correspondence between the multiple first delay information and the multiple first values, that is, one first delay information corresponds to one first value.

[0045] It should be noted that the first value can characterize the degree of outlier of the first delay information. The larger the first value, the higher the degree of outlier of the corresponding first delay information. Conversely, the smaller the first value, the lower the degree of outlier of the corresponding first delay information.

[0046] Step 203: The first server determines that the first delay information with a first value greater than or equal to the first target threshold satisfies the target condition.

[0047] In this embodiment of the application, if the first value corresponding to a certain first delay information is greater than or equal to the first target threshold, then it can be determined that the first delay information meets the target condition, that is, the outlier degree of the first delay information is high.

[0048] It should be noted that the first target threshold can be preset based on historical data or actual needs. In one feasible approach, the first target threshold can be 1.

[0049] Step 204: The first server determines that the first delay information with a first value less than the first target threshold does not meet the target condition.

[0050] In this embodiment of the application, if the first value corresponding to a certain first delay information is less than the first target threshold, then it can be determined that the first delay information does not meet the target condition, that is, the outlier degree of the first delay information is low.

[0051] Step 205: For the first storage device whose first latency information meets the target conditions, the first server monitors multiple second latency information of the first storage device within the target time period.

[0052] In the embodiments of this application, the target time period can refer to a future period of time starting from the current time. For example, if the current time is 16:00, then the target time period can refer to 16:00-18:00.

[0053] In this application embodiment, the first delay information satisfying the target condition may refer to the outlier degree of the first delay information being greater than or equal to the first target threshold.

[0054] It should be noted that there can be multiple first storage devices, and each first storage device corresponds to multiple second latency information. Specifically, for each first storage device, the first server can send data processing requests to the second server corresponding to each first storage device at the same time interval within the target time period, thereby obtaining multiple second latency information for each first storage device within the target time period.

[0055] In other embodiments of this application, the first server may also send data processing requests at different time intervals within the target time period to obtain multiple second latency information for each first storage device.

[0056] It should be noted that the number of second latency information items to be acquired by the first server is pre-set, and the number of second latency information items corresponding to each first server is the same. Furthermore, for each first storage device, there is a corresponding relationship between the multiple second latency information items.

[0057] For example, for the first storage device A1, multiple second delay information pieces acquired in chronological order may include second delay information B1, second delay information C1, and second delay information D1; for the first storage device A2, multiple second delay information pieces acquired in chronological order may include second delay information B2, second delay information C2, and second delay information D2. Second delay information B1 and second delay information B2 have a corresponding relationship, meaning that the data processing request corresponding to second delay information B1 and the data processing request corresponding to second delay information B2 are sent at the same time. In other words, the first server will send data processing requests to the second server a corresponding to the first storage device A1 and the second server b corresponding to the second storage device A2 at the same time. Afterwards, after performing data processing operations, the first storage device A1 and the second storage device A2 will return second delay information B1 and second delay information B2 to the first server. Correspondingly, second delay information C1 and second delay information C2 also have a corresponding relationship, as do second delay information D1 and second delay information D2.

[0058] Step 206: The first server determines the second value based on multiple second delay information.

[0059] The second value represents the degree of dispersion among multiple second time delay information.

[0060] In this embodiment, the LOF algorithm can be used first to process the second latency information of all first storage devices to obtain the outlier degree of each second latency information corresponding to each first storage device. Then, for each first storage device, the token bucket algorithm can be used to process the outlier degree of multiple second latency information and a first target threshold to obtain a second value. It should be noted that the dispersion among multiple second latency information is actually the degree of difference among multiple second latency information.

[0061] In this embodiment, the second value may refer to the number of remaining tokens in the token bucket. It should be noted that the initial number of tokens in the token bucket is the same as the number of second delay information items; that is, the number of tokens in the token bucket corresponds to the number of second delay information items.

[0062] In this embodiment, the outlier degree of each second delay information and the first target threshold can be input into the token bucket. For each second delay information whose outlier degree value is less than the first target threshold, one token is consumed. That is, the number of tokens consumed is the number of second delay information whose outlier degree values ​​are less than the first target threshold. If the outlier degree values ​​of all second delay information are less than the first target threshold, all tokens in the token bucket will be consumed. Conversely, if the outlier degree value is greater than or equal to the first target threshold, no token is consumed, thus obtaining the number of tokens remaining in the token bucket (i.e., the second value).

[0063] It should be noted that if the outlier values ​​of multiple second latency information corresponding to a certain first storage device are all less than the first target threshold, all tokens in the token bucket will be consumed. At this time, the number of remaining tokens in the token bucket (i.e., the second value) is 0. Correspondingly, if the outlier value of some of the multiple second latency information corresponding to a certain first storage device is greater than or equal to the first target threshold, some tokens in the token bucket will not be consumed. At this time, the number of remaining tokens in the token bucket (i.e., the second value) is not 0.

[0064] In other embodiments of this application, step 207 or steps 208-209 may be performed after step 206.

[0065] Step 207: If the second value is equal to the second target threshold, the first server determines that the first storage device is not abnormal.

[0066] In this embodiment of the application, the second target threshold is 0. The second value equal to the second target threshold means that the number of tokens remaining in the token bucket is 0, that is, the number of tokens in the token bucket has been completely consumed. At this time, it can be determined that the outlier values ​​of multiple second latency information of the first storage device in the target time period are all small, that is, it means that the read and write performance of the first storage device in the target time period is good and stable. Then it can be determined that the first storage device is not abnormal, that is, the first storage device is not a slow disk.

[0067] Step 208: If the second value is not equal to the second target threshold, the first server determines the first operating characteristic information of the first storage device.

[0068] Among them, the first operational characteristic information represents the operational status of the first storage device.

[0069] In this embodiment of the application, if the second value is not equal to the second target threshold (i.e., not equal to 0), it means that there are still some unused tokens in the token bucket, which means that among the multiple second latency information corresponding to the first storage device, some second latency information has a high outlier value. In other words, the read and write performance of the first storage device is unstable during the target time period, which means that the first storage device is very likely to be abnormal. That is, the probability of the first storage device being abnormal is high. At this time, it is necessary to determine the first operating characteristic information of the first storage device and further determine whether the first storage device is abnormal based on the first operating characteristic information.

[0070] It should be noted that the first running feature information can be represented by the first running feature vector.

[0071] In other embodiments of this application, the "first server determines the first operating characteristic information of the first storage device" in step 208 can be implemented through steps 208a-208b.

[0072] Step 208a: The first server obtains the first operating information and the first health status information of the first storage device.

[0073] Among them, the first health status information represents the wear and tear of the first storage device.

[0074] In this embodiment, the first operating information may include temperature information and operating logs of the first storage device (i.e., a hard disk). In one possible implementation, if the first storage device is a mechanical hard disk, the first operating information may further include vibration information of the mechanical hard disk.

[0075] In this embodiment of the application, the first health status information may refer to Self-Monitoring Analysis and Reporting Technology (SMART) data.

[0076] Step 208b: The first server performs fusion processing on the first running information and the first health status information to obtain the first running characteristic information.

[0077] In this embodiment, a target multimodal fusion algorithm can be used to first extract features from the first operating information and the first health status information of the first storage device to obtain a first feature vector and a second feature vector. Then, the first feature vector and the second feature vector are fused to obtain the first operating feature vector, i.e., the first operating feature information. The target multimodal fusion algorithm is a multimodal fusion algorithm.

[0078] It should be noted that each first storage device corresponds to a first operational characteristic information.

[0079] Step 209: The first server determines whether there is an anomaly in the first storage device based on the first operational characteristic information.

[0080] In one feasible approach, the first operational feature information can be input into a gradient boosting decision tree model, and the presence of anomalies in the first storage device can be determined based on the model's output.

[0081] In another possible approach, the operational characteristics of a historically abnormal storage device that has been identified as having anomalies can be determined, and the presence of anomalies in the first storage device can be jointly determined based on the first operational characteristics and the operational characteristics of the historically abnormal storage device.

[0082] In other embodiments of this application, step 209 can be implemented by steps 209a-209c.

[0083] Step 209a: The first server determines the third operational characteristic information of the historical abnormal storage device.

[0084] Among them, the third operational characteristic information represents the operational status of historical abnormal storage devices.

[0085] In this application embodiment, the historical abnormal storage device can refer to a hard disk that has been identified as having an abnormality, that is, a hard disk that has been identified as a slow disk; the third running characteristic information can also be represented in the form of a vector.

[0086] In one feasible approach, the third operational characteristic information of the historical anomaly storage device may be pre-stored in the first server.

[0087] In another possible implementation, step 209a can also be implemented via steps 209a1-209a2.

[0088] Step 209a1: The first server obtains the second operating information and the second health status information of the historical abnormal storage device.

[0089] Among them, the second health status information represents the wear and tear of historically abnormal storage devices.

[0090] In this embodiment of the application, the second operational information may include temperature information and operational logs of the historical abnormal storage device.

[0091] It should be noted that the type of the historical abnormal storage device must be the same as the type of the first storage device. That is, if the first storage device is a mechanical hard drive, then the historical abnormal storage device must also be a mechanical hard drive. Similarly, if the first storage device is a solid-state drive, then the historical abnormal storage device must also be a solid-state drive. If the historical abnormal storage device is a mechanical hard drive, then the second operating information may also include the vibration information of the abnormal mechanical hard drive.

[0092] In this embodiment of the application, the second health status information may refer to the SMART data of the historical abnormal storage device.

[0093] Step 209a2: The first server performs fusion processing on the second running information and the second health status information to obtain the third running characteristic information.

[0094] In this embodiment of the application, a target multimodal fusion algorithm can be used to first extract features from the second operating information and the second health status information of the historical abnormal storage device to obtain a third feature vector and a fourth feature vector. Then, the third feature vector and the fourth feature vector are fused to obtain a third operating feature vector, that is, the third operating feature information.

[0095] In the embodiments of this application, step 209b or step 209c can be performed after step 209a.

[0096] Step 209b: If the first operating characteristic information matches the third operating characteristic information, the first server determines that the first storage device is abnormal.

[0097] In this embodiment of the application, after obtaining the first operating feature information (i.e., the first operating feature vector) and the third operating feature information (i.e., the third operating feature vector), the Euclidean distance between the first operating feature vector and the third operating feature vector can be calculated. If the Euclidean distance is greater than or equal to a preset threshold, it indicates that the first operating feature vector (i.e., the first operating feature information) matches the third operating feature vector (i.e., the third operating feature information), that is, the operating status of the first storage device is highly similar to the operating status of historical abnormal storage devices. At this time, it can be directly determined that the first storage device is abnormal, that is, the first storage device is a slow disk.

[0098] Step 209c: If the first running characteristic information does not match the third running characteristic information, the first server determines that the first storage device does not have any abnormalities.

[0099] In this embodiment of the application, if the first operating characteristic information does not match the third operating characteristic information, it indicates that the operating status of the first storage device is not very similar to the operating status of historical abnormal storage devices. In this case, it can be determined that the first storage device is not abnormal, that is, the first storage device is not a slow disk.

[0100] Step 210: For the second storage device whose first latency information does not meet the target conditions, the first server determines the target score for the first latency information based on the first latency information of the second storage device.

[0101] In this embodiment of the application, the fact that the first delay information does not meet the target condition may mean that the outlier degree of the first delay information is less than the first target threshold.

[0102] It should be noted that there can be multiple second storage devices.

[0103] In this embodiment of the application, for the second storage device, a One-Class Support Vector Machine (One-Class SVM) algorithm can be used to process the first latency information of the second storage device to obtain a score (i.e., a target score) for the first latency information. This score can characterize the degree of anomalousness of the first latency information.

[0104] In the embodiments of this application, step 211 or steps 212-213 can be performed after step 210.

[0105] Step 211: If the target score is greater than or equal to the third target threshold, the first server determines that there is no abnormality in the second storage device.

[0106] In this embodiment, the third target threshold can be preset according to actual needs. If the target score (i.e., the score of the first latency information) is greater than or equal to the third target threshold, it indicates that the score is within the normal range. In this case, it can be determined that the corresponding second storage device is not abnormal.

[0107] Step 212: If the target score is less than the third target threshold, the first server determines the second operating characteristic information of the second storage device.

[0108] Among them, the second operational characteristic information characterizes the operational status of the second storage device.

[0109] In this embodiment, if the target score (i.e., the score of the first latency information) is less than the third target threshold, it indicates that the score is abnormal. In this case, the second operational characteristic information of the second storage device can be determined, and the presence of any abnormality in the second storage device can be further determined based on the second operational characteristic information. The second operational characteristic information can also be represented by a vector.

[0110] It should be noted that the specific process for determining the second operating characteristic information of the second storage device can be referred to steps 208a-208b.

[0111] Step 213: The first server determines whether there is an anomaly in the second storage device based on the second operational characteristic information.

[0112] In one feasible approach, the second operational feature information can be input into a gradient boosting decision tree model, and the presence of anomalies in the second storage device can be determined based on the model's output.

[0113] In another possible approach, the operational characteristics of a historically abnormal storage device that has been identified as having anomalies can be determined, and the existence of anomalies in the second storage device can be jointly determined based on the second operational characteristics and the operational characteristics of the historically abnormal storage device.

[0114] It should be noted that the determination process of step 213 can refer to steps 209a-209c.

[0115] In this embodiment, by combining algorithms such as LOF algorithm, single-class support vector machine algorithm and multimodal data fusion, not only can slow disks among multiple hard drives be accurately identified, but data processing efficiency can also be improved. At the same time, the slow disk identification method provided in this embodiment can be applied to hard drives of various brands and models, rather than the slow disk identification method in related technologies that is only applicable to some models of hard drives, which makes it impossible to accurately identify the hard drives with abnormalities, thereby affecting data processing efficiency.

[0116] In the embodiments of this application, such as Figure 3As shown, after the computing component determines whether there is an anomaly in each storage device corresponding to the second server, it can report the abnormal status of each storage device to the status management component. Then, the status management component will update the dynamic mapping table between "node x: disk m-status" in the database according to the abnormal status. In this way, by continuously updating the dynamic mapping table in the data table and using a high-speed cache to store the latest disk status information, the latest status of the disk can be quickly found, thereby avoiding data being written to disks in abnormal status.

[0117] Here, node x can refer to the xth second server, hard disk m can refer to the mth hard disk on the xth second server, and status can refer to the abnormal status of the mth hard disk. The status can include two cases: 0 and 1. "0" means that the hard disk is not abnormal (i.e., it is not a slow disk), and "1" means that the hard disk is abnormal (i.e., it is a slow disk).

[0118] In other embodiments of this application, steps 214-216 may be performed after step 213.

[0119] Step 214: The first server receives a data write request for the data to be processed sent by the client.

[0120] The data write request carries an identifier of the data to be processed.

[0121] In this embodiment of the application, the identifier of the data to be processed may refer to the name of the data to be processed.

[0122] Step 215: The first server removes the storage devices with abnormalities from multiple storage devices to obtain the target storage device.

[0123] In this embodiment of the application, the target server can refer to a storage device (i.e., a hard drive) that is not an anomaly among multiple storage devices. Specifically, after receiving a data write request, the first server, as follows: Figure 3 As shown, a dynamic mapping table of "node: hard disk-status" can be obtained through the storage gateway, and storage devices with abnormalities can be removed from the multiple storage devices corresponding to the second server based on the dynamic mapping table, thereby obtaining the target storage device with a normal status, that is, without any abnormalities.

[0124] Step 216: The first server determines the target storage location from multiple storage locations based on the identifier and the number of multiple storage locations of the target storage device.

[0125] In this embodiment, the first server can obtain the correspondence between the object storage device (OSD), i.e., the identifier of the storage device and multiple storage locations of the storage device (also known as PlacementGroup Identifier, PGID). Then, based on this correspondence, it can determine the multiple storage locations corresponding to the target storage device, i.e., the multiple storage locations corresponding to the hard disks without abnormalities. Then, the storage gateway can perform a hash operation on the name (i.e., identifier) ​​of the data to be processed and the number of multiple storage locations of the target storage device to determine the target storage location where the data to be processed is to be written from the multiple storage locations. After that, the distributed gateway node will write the data to be processed into the target storage location.

[0126] In this embodiment of the application, when there are slow disks among multiple hard drives, the data writing location can be calculated based on the updated hard drive status (normal, slow disk) label and the mapping table between the hard drives, so as to improve the performance of the object storage cluster and ensure that the data can be written to the normal hard drives.

[0127] In this embodiment, machine learning algorithms can be combined to more accurately determine whether a hard drive is slow. By tagging the hard drive, dynamically maintaining the status mapping table, and calculating the data write location based on the status mapping table, the data is written to the hard drive in normal status. This can proactively eliminate the impact of slow drives on performance and improve the performance of the object storage cluster through a targeted approach.

[0128] The information determination method provided in the embodiments of this application can first obtain the first latency information of multiple storage devices corresponding to the second server, and use different methods to determine whether the first storage device whose first latency information meets the target condition and the second storage device whose first latency information does not meet the target condition are abnormal. In this way, different methods can be used to determine the abnormality of different types of storage devices among multiple storage devices, instead of simply comparing the hard drive performance data with the threshold parameter to determine whether the hard drive is abnormal, as in related technologies. This solves the problem in related technologies that it is impossible to accurately determine whether the hard drive is abnormal, thereby ensuring that data can be written to normal hard drives and improving data processing efficiency.

[0129] Based on the foregoing embodiments, this application provides an information determining device, which can be applied to... Figure 1 and 2 In the information determination method provided in the corresponding embodiment, refer to Figure 4 As shown, the information determining device 3 may include: an acquisition unit 31, a monitoring unit 32, a first determining unit 33, and a second determining unit 34, wherein: Acquisition unit 31 is used to acquire first latency information of multiple storage devices corresponding to the second server; Monitoring unit 32 is used to monitor multiple second delay information of the first storage device within a target time period for the first storage device whose first delay information meets the target conditions; The first determining unit 33 is used to determine whether there is an anomaly in the first storage device based on multiple second delay information. The second determining unit 34 is used to determine whether there is an anomaly in the second storage device based on the first delay information of the second storage device, for the second storage device whose first delay information does not meet the target conditions.

[0130] In other embodiments of this application, the first determining unit 33 is further configured to perform the following steps: Based on multiple first delay information, a first value is determined for each first delay information; wherein, the first value characterizes the outlier degree of each first delay information. The first time delay information, whose first value is greater than or equal to the first target threshold, satisfies the target condition. The first time delay information whose first value is less than the first target threshold does not meet the target condition.

[0131] In other embodiments of this application, the first determining unit 33 is further configured to perform the following steps: A second value is determined based on multiple second time delay information; wherein, the second value characterizes the degree of dispersion among the multiple second time delay information. If the second value equals the second target threshold, it is determined that the first storage device is not abnormal. If the second value is not equal to the second target threshold, the first operating characteristic information of the first storage device is determined; wherein, the first operating characteristic information characterizes the operating status of the first storage device; Based on the first operational characteristic information, determine whether the first storage device is abnormal.

[0132] In other embodiments of this application, the second determining unit 34 is further configured to perform the following steps: Based on the first latency information of the second storage device, a target score is determined for the first latency information; If the target score is greater than or equal to the third target threshold, it is determined that the second storage device is not abnormal; If the target score is less than the third target threshold, the second operational characteristic information of the second storage device is determined; wherein, the second operational characteristic information characterizes the operational status of the second storage device; Based on the second operational characteristic information, determine whether the second storage device is abnormal.

[0133] In other embodiments of this application, the first determining unit 33 is further configured to perform the following steps: Obtain first operating information and first health status information of the first storage device; wherein, the first health status information represents the wear and tear status of the first storage device; The first operational information and the first health status information are fused together to obtain the first operational characteristic information.

[0134] In other embodiments of this application, the first determining unit 33 is further configured to perform the following steps: The third operational characteristic information of the historical abnormal storage device is determined; wherein, the third operational characteristic information characterizes the operational status of the historical abnormal storage device; If the first operational characteristic information matches the third operational characteristic information, it is determined that the first storage device is abnormal; If the first operational characteristic information does not match the third operational characteristic information, it is determined that the first storage device does not have any abnormalities.

[0135] In other embodiments of this application, the first determining unit 33 is further configured to perform the following steps: Acquire the second operating information and the second health status information of the historical abnormal storage device; wherein, the second health status information characterizes the wear and tear status of the historical abnormal storage device; The second operational information and the second health status information are fused together to obtain the third operational characteristic information.

[0136] In other embodiments of this application, the second determining unit 34 is further configured to perform the following steps: Receive a data write request sent by the client for the data to be processed; wherein the data write request carries the identifier of the data to be processed; The target storage device is obtained by removing the faulty storage device from multiple storage devices. The target storage location is determined from multiple storage locations based on the identifier and the number of storage locations of the target storage device.

[0137] It should be noted that a detailed explanation of the steps performed by each unit can be found in [reference needed]. Figure 1 and 2 The description of the information determination method provided in the corresponding embodiments will not be repeated here.

[0138] The information determination device provided in the embodiments of this application can first obtain the first latency information of multiple storage devices corresponding to the second server, and use different methods to determine whether the first storage device whose first latency information meets the target condition and the second storage device whose first latency information does not meet the target condition are abnormal. In this way, different methods can be used to determine the abnormality of different types of storage devices among multiple storage devices, instead of simply comparing the performance data of the hard disk with the threshold parameter to determine whether the hard disk is abnormal, as in related technologies. This solves the problem in related technologies that it is impossible to accurately determine whether the hard disk is abnormal, thereby ensuring that data can be written to normal hard disks and improving data processing efficiency.

[0139] Based on the foregoing embodiments, embodiments of this application provide a first server, which can be applied to... Figure 1 and 2 In the information determination method provided in the corresponding embodiment, refer to Figure 5 As shown, the first server 4 may include: a processor 41, a memory 42, and a communication bus 43, wherein: Communication bus 43 is used to realize the communication connection between processor 41 and memory 42; The processor 41 is used to execute the information determination program in the memory 42 to perform the following steps: Obtain the first latency information of multiple storage devices corresponding to the second server; For a first storage device whose first latency information meets the target condition, monitor multiple second latency information of the first storage device within the target time period; Based on multiple second delay information, determine whether the first storage device is abnormal; For a second storage device whose first latency information does not meet the target conditions, determine whether the second storage device is abnormal based on the first latency information of the second storage device.

[0140] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: Based on multiple first delay information, a first value is determined for each first delay information; wherein, the first value characterizes the outlier degree of each first delay information. The first time delay information, whose first value is greater than or equal to the first target threshold, satisfies the target condition. The first time delay information whose first value is less than the first target threshold does not meet the target condition.

[0141] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: A second value is determined based on multiple second time delay information; wherein, the second value characterizes the degree of dispersion among the multiple second time delay information. If the second value equals the second target threshold, it is determined that the first storage device is not abnormal. If the second value is not equal to the second target threshold, the first operating characteristic information of the first storage device is determined; wherein, the first operating characteristic information characterizes the operating status of the first storage device; Based on the first operational characteristic information, determine whether the first storage device is abnormal.

[0142] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: Based on the first latency information of the second storage device, a target score is determined for the first latency information; If the target score is greater than or equal to the third target threshold, it is determined that the second storage device is not abnormal; If the target score is less than the third target threshold, the second operational characteristic information of the second storage device is determined; wherein, the second operational characteristic information characterizes the operational status of the second storage device; Based on the second operational characteristic information, determine whether the second storage device is abnormal.

[0143] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: Obtain first operating information and first health status information of the first storage device; wherein, the first health status information represents the wear and tear status of the first storage device; The first operational information and the first health status information are fused together to obtain the first operational characteristic information.

[0144] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: The third operational characteristic information of the historical abnormal storage device is determined; wherein, the third operational characteristic information characterizes the operational status of the historical abnormal storage device; If the first operational characteristic information matches the third operational characteristic information, it is determined that the first storage device is abnormal; If the first operational characteristic information does not match the third operational characteristic information, it is determined that the first storage device does not have any abnormalities.

[0145] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: Acquire the second operating information and the second health status information of the historical abnormal storage device; wherein, the second health status information characterizes the wear and tear status of the historical abnormal storage device; The second operational information and the second health status information are fused together to obtain the third operational characteristic information.

[0146] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps: Receive a data write request sent by the client for the data to be processed; wherein the data write request carries the identifier of the data to be processed; The target storage device is obtained by removing the faulty storage device from multiple storage devices. The target storage location is determined from multiple storage locations based on the identifier and the number of storage locations of the target storage device.

[0147] It should be noted that a detailed description of the steps performed by the processor can be found in [reference needed]. Figure 1 and 2 The information determination method provided in the corresponding embodiments will not be described again here.

[0148] The first server provided in the embodiments of this application can first obtain the first latency information of multiple storage devices corresponding to the second server, and use different methods to determine whether the first storage device whose first latency information meets the target condition and the second storage device whose first latency information does not meet the target condition are abnormal. In this way, different methods can be used to determine the abnormal situation of different types of storage devices among multiple storage devices, instead of simply comparing the performance data of the hard disk with the threshold parameter to determine whether the hard disk is abnormal, as in related technologies. This solves the problem in related technologies that it is impossible to accurately determine whether the hard disk is abnormal, thereby ensuring that data can be written to normal hard disks and improving data processing efficiency.

[0149] Based on the foregoing embodiments, embodiments of this application provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement... Figure 1 and 2 The corresponding embodiments provide the steps of the information determination method.

[0150] Based on the foregoing embodiments, embodiments of this application provide a computer program product, which includes a computer program that is implemented when executed by processor 41. Figure 1 and 2 The corresponding embodiments provide the steps of the information determination method.

[0151] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0152] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0153] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0154] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0155] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A method for determining information, characterized in that, The method is applied to a first server, and the method includes: Obtain the first latency information of multiple storage devices corresponding to the second server; For a first storage device whose first latency information meets the target condition, monitor multiple second latency information of the first storage device within the target time period; Based on the multiple second delay information, it is determined whether the first storage device is abnormal; For a second storage device whose first latency information does not meet the target condition, determine whether the second storage device is abnormal based on the first latency information of the second storage device.

2. The method according to claim 1, characterized in that, The method further includes: Based on multiple first delay information, a first value is determined for each first delay information; wherein, the first value characterizes the outlier degree of each first delay information; The first time delay information, which determines that the first value is greater than or equal to the first target threshold, satisfies the target condition; The first time delay information, which is determined to be less than the first target threshold, does not meet the target condition.

3. The method according to claim 1, characterized in that, The step of determining whether the first storage device is abnormal based on the plurality of second delay information includes: Based on the plurality of second time delay information, a second value is determined; wherein, the second value characterizes the degree of dispersion among the plurality of second time delay information; If the second value is equal to the second target threshold, it is determined that the first storage device is not abnormal; If the second value is not equal to the second target threshold, the first operating characteristic information of the first storage device is determined; wherein, the first operating characteristic information characterizes the operating status of the first storage device; Based on the first operational characteristic information, it is determined whether the first storage device is abnormal.

4. The method according to claim 1, characterized in that, The step of determining whether the second storage device is abnormal based on the first latency information of the second storage device includes: Based on the first latency information of the second storage device, a target score is determined for the first latency information; If the target score is greater than or equal to the third target threshold, it is determined that the second storage device is not abnormal; If the target score is less than the third target threshold, the second operating characteristic information of the second storage device is determined; wherein, the second operating characteristic information characterizes the operating status of the second storage device; Based on the second operational characteristic information, it is determined whether the second storage device is abnormal.

5. The method according to claim 3, characterized in that, The determination of the first operational characteristic information of the first storage device includes: Obtain first operating information and first health status information of the first storage device; wherein, the first health status information characterizes the wear and tear status of the first storage device; The first operational information and the first health status information are fused together to obtain the first operational feature information.

6. The method according to claim 3, characterized in that, The step of determining whether the first storage device is abnormal based on the first operational characteristic information includes: The third operational characteristic information of the historical abnormal storage device is determined; wherein, the third operational characteristic information characterizes the operational status of the historical abnormal storage device; If the first operational feature information matches the third operational feature information, it is determined that the first storage device is abnormal. If the first operational feature information does not match the third operational feature information, it is determined that the first storage device is not abnormal. Accordingly, the third operational characteristic information for determining the historical abnormal storage device includes: The second operating information and the second health status information of the historical abnormal storage device are obtained; wherein, the second health status information characterizes the wear and tear status of the historical abnormal storage device. The second operational information and the second health status information are fused together to obtain the third operational feature information.

7. The method according to claim 1, characterized in that, The method further includes: Receive a data write request sent by the client for the data to be processed; wherein the data write request carries an identifier of the data to be processed; The storage device with the abnormality is removed from the plurality of storage devices to obtain the target storage device; The target storage location is determined from the plurality of storage locations based on the identifier and the number of storage locations of the target storage device.

8. A first server, characterized in that, The device includes: a processor, a memory, and a communication bus; The communication bus is used to realize the communication connection between the processor and the memory; The processor is used to execute an information determination program in memory to implement the steps of the information determination method as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the information determination method as described in any one of claims 1-7.

10. A computer program product, the computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the information determination method according to any one of claims 1-7.

Images