CN118093315A

CN118093315A - Monitoring control method and device of hard disk device, storage medium and electronic device

Info

Publication number: CN118093315A
Application number: CN202410231804.6A
Authority: CN
Inventors: 刘畅; 刘帅
Original assignee: Suzhou Metabrain Intelligent Technology Co Ltd
Current assignee: Suzhou Metabrain Intelligent Technology Co Ltd
Priority date: 2024-02-29
Filing date: 2024-02-29
Publication date: 2024-05-28

Abstract

The embodiment of the application provides a monitoring control method and device of hard disk equipment, a storage medium and electronic equipment, wherein the method comprises the following steps: receiving a hard disk monitoring request initiated by a server; responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device; creating a target monitoring list according to a target monitoring item corresponding to a target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data and a monitoring type to which the target monitoring item belongs; and sending the target monitoring list to the server. The application solves the problem of lower monitoring control efficiency of the hard disk device, thereby achieving the effect of improving the monitoring control efficiency of the hard disk device.

Description

Monitoring control method and device of hard disk device, storage medium and electronic device

Technical Field

The embodiment of the application relates to the field of computers, in particular to a monitoring control method and device of hard disk equipment, a storage medium and electronic equipment.

Background

In the related art, when the existing data center management software monitors and manages the hard disk device, the hard disk device is monitored mainly through a preset fixed and universal monitoring template, and fixed monitoring items are arranged in the monitoring template, wherein the monitoring items comprise key indexes of the hard disk device, performance of the hard disk device, faults of the hard disk device and the like.

It can be understood that in the related art, when monitoring different hard disk devices, the same monitoring items of the different hard disk devices are monitored through the same monitoring template. In this way, the same monitoring item is monitored on different hard disk devices, which may result in lower monitoring control efficiency of the hard disk devices.

Disclosure of Invention

The embodiment of the application provides a monitoring control method and device of hard disk equipment, a storage medium and electronic equipment, which are used for at least solving the problem of low monitoring control efficiency of the hard disk equipment in the related technology.

According to an embodiment of the present application, there is provided a monitoring control method of a hard disk device, including: the method is applied to a monitoring controller, the monitoring controller is connected with a server, hard disk equipment is deployed on the server, and the method comprises the following steps: receiving a hard disk monitoring request initiated by the server, wherein the hard disk monitoring request is used for requesting to control a monitoring process of a target hard disk device in the hard disk devices deployed on the server; responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device, wherein the target operation data comprises monitoring values of each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scenes are used for indicating the performance requirement of the service operated by the hard disk device on the hard disk device; creating a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data and a monitoring type to which the target monitoring item belongs, wherein the target monitoring item and a target monitoring mode with a corresponding relation are recorded in the target monitoring list; and sending the target monitoring list to the server, wherein the server is used for monitoring the operation of the target hard disk device according to the target monitoring list.

In an exemplary embodiment, the identifying, according to the target operation data of the target hard disk device, a target operation scene corresponding to the target hard disk device from a plurality of operation scenes includes: according to the target operation data, calculating the probability that the target hard disk device corresponds to each of the plurality of operation scenes; and screening N operation scenes corresponding to the probability meeting the target screening condition from the operation scenes according to the operation scenes and the probability with the corresponding relation, wherein N is a positive integer.

In an exemplary embodiment, the selecting, according to the operation scenes and probabilities having the correspondence, N operation scenes from the plurality of operation scenes, where the corresponding probabilities satisfy a target screening condition, as the target operation scene includes: extracting operation scenes from the operation scenes one by one according to the corresponding probability from high to low; detecting the relation between the probability corresponding to the current extracted operation scene, the probability corresponding to the extracted operation scene and the target threshold value; determining the current extracted operation scene and the extracted operation scene as the N operation scenes as the target operation scenes under the condition that the probability sum is greater than or equal to the target threshold value; and under the condition that the sum of the probabilities is smaller than the target threshold value, continuously extracting the next operation scene from high to low according to the corresponding probability.

In an exemplary embodiment, the identifying, according to the target operation data of the target hard disk device, a target operation scene corresponding to the target hard disk device from a plurality of operation scenes includes: inputting the target operation data into a target scene recognition model, wherein the target scene recognition model is obtained by training an initial scene recognition model by using an operation data sample marked with an operation scene label; and acquiring the target operation scene output by the target scene identification model.

In an exemplary embodiment, the creating a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data, and a monitoring type to which the target monitoring item belongs includes: creating an initial monitoring list by using a target monitoring item corresponding to the target operation scene; according to the monitoring type of the target monitoring item, the target monitoring mode is allocated to the target monitoring item to obtain a reference monitoring list, wherein the target monitoring mode comprises the following steps: a mode of monitoring according to the change of the data and a mode of monitoring according to the threshold value of the data; and under the condition that the target monitoring mode is a mode of monitoring according to the threshold value of the data, distributing a target monitoring threshold value for the target monitoring item according to the reference operation data to obtain the target monitoring list.

In an exemplary embodiment, the assigning the target monitoring threshold to the target monitoring item according to the reference operation data includes: the method comprises the steps that an ith monitoring threshold value is distributed to an ith monitoring item according to ith operation data, wherein the reference operation data comprise M pieces of operation data, the M pieces of operation data comprise the ith operation data, the target monitoring item comprises M monitoring items, the M monitoring items comprise the ith monitoring item, the target monitoring threshold value comprises M monitoring threshold values, the M monitoring threshold values comprise the ith monitoring threshold value, i is a positive integer smaller than or equal to M, and M is a positive integer; selecting the ith candidate operation data used for representing that the target hard disk device is in a normal operation state from the ith operation data; performing average value operation on the ith candidate running data to obtain an average value, and determining the standard deviation of the ith candidate running data; determining a first candidate upper threshold as a sum of the average value and the standard deviation of the target number, and determining a first candidate lower threshold as a value obtained by subtracting the standard deviation of the target number from the average value; determining a larger upper limit threshold from a second candidate upper limit threshold and the first candidate upper limit threshold to obtain a target upper limit threshold, and determining a smaller lower limit threshold from a second candidate lower limit threshold and the first candidate lower limit threshold to obtain a target lower limit threshold, wherein the target monitoring threshold comprises the target upper limit threshold and the target lower limit threshold, the second candidate upper limit threshold is the data with the largest value in the i candidate operation data, and the second candidate lower limit threshold is the data with the smallest value in the i candidate operation data.

In an exemplary embodiment, before said sending the target monitor list to the server, the method further comprises: transmitting a data reporting instruction to the server, wherein the data reporting instruction is used for instructing the server to report alarm data belonging to the target operation scene to the monitoring controller; receiving target alarm data reported by the server in response to the data reporting instruction; extracting keywords in each alarm data in the P alarm data to obtain P groups of target keywords under the condition that the target alarm data comprises P alarm data and the target hard disk device comprises T hard disk devices, wherein a w-th group of target keywords in the P groups of target keywords comprise identifications of w-th hard disk devices in the T hard disk devices, positions of the w-th hard disk devices deployed in the server and device types of the w-th hard disk devices, P is a positive integer, T is a positive integer, and w is a positive integer smaller than or equal to P; vectorizing the P groups of target keywords to obtain P groups of word vectors; determining the similarity between the s ₁ th word vector in the s th word vector in the P group word vectors and the r ₁ th word vector in each group word vector in the P-1 group word vectors except the s th word vector to obtain the s ₁ th similarity, determining the similarity between the s ₂ th word vector in the s th group word vector and the r ₂ th word vector in each group word vector in the P-1 group word vectors to obtain the s ₂ th similarity, determining the similarity between the s ₃ th word vector in the s th group word vector and the r ₃ th word vector in each group word vector in the P-1 group word vectors, obtaining an s ₃ th group similarity, wherein the s ₁ th word vector is a word vector used for representing the identifier of the s hard disk device in the s th group word vector, the s ₂ th word vector is a word vector used for representing the position of the s hard disk device deployed in a server in the s th group word vector, the s ₃ th word vector is a word vector used for representing the device type of the s hard disk device in the s th group word vector, the r ₁ th word vector is a word vector used for representing the identifier of the hard disk device in each group word vector in the P-1 group word vector, the r ₂ th word vector is a word vector used for representing a location of the hard disk device deployed in the server in each of the P-1 group of word vectors, the r ₃ th word vector is a word vector used for representing a device type of the hard disk device in each of the P-1 group of word vectors, the T hard disk devices include the s-th hard disk device, s is a positive integer less than or equal to P, s ₁、s₂, s ₃ are both positive integers, and r ₁、r₂, and r ₃ are both positive integers; performing average value operation on the s ₁ th group of similarity, the s ₂ th group of similarity and the s ₃ th group of similarity to obtain an s average similarity; under the condition that the s-th average similarity is larger than or equal to a preset average similarity threshold value, determining s-th alarm data corresponding to the s-th group word vector in the P alarm data as a target alarm monitoring item; and adding the target alarm monitoring item to the target monitoring list.

According to another embodiment of the present application, there is provided a monitoring control apparatus for a hard disk device, including: the device is applied to a monitoring controller, the monitoring controller is connected with a server, hard disk equipment is deployed on the server, and the device comprises: the first receiving module is used for receiving a hard disk monitoring request initiated by the server, wherein the hard disk monitoring request is used for requesting to control the monitoring process of a target hard disk device in the hard disk devices deployed on the server; the identification module is used for responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device, wherein the target operation data comprises monitoring values of each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scene is used for indicating the performance requirement of the service operated by the hard disk device on the hard disk device; the creating module is used for creating a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data and a monitoring type to which the target monitoring item belongs, wherein the target monitoring item and a target monitoring mode with a corresponding relation are recorded in the target monitoring list; the first sending module is used for sending the target monitoring list to the server, wherein the server is used for monitoring the operation of the target hard disk device according to the target monitoring list.

According to a further embodiment of the application, there is also provided a computer readable storage medium having stored therein a computer program, wherein the computer program is arranged to perform the steps of any of the method embodiments described above when run.

According to a further embodiment of the application there is also provided an electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.

According to a further embodiment of the application, there is also provided a computer program product comprising a computer program which, when executed by a processor, implements the steps of any of the method embodiments described above.

According to the application, according to the operation data of the hard disk device, the operation scene corresponding to the hard disk device is identified from a plurality of operation scenes, the operation scene can be used for indicating the performance requirement of the operation of the hard disk device on the hard disk device, and according to the monitoring item corresponding to the operation scene, the operation data corresponding to the monitoring item in the operation data and the monitoring type to which the monitoring item belongs, a monitoring list is created, wherein the monitoring item and the monitoring mode with the corresponding relation are recorded in the monitoring list, and it is understood that the monitoring item and the monitoring mode in the monitoring list can be changed under the condition that the performance requirements of the hard disk device are different, by adopting the mode, the monitoring list suitable for the performance requirement of the hard disk device is created according to the identified performance requirement of the operation of the hard disk device, therefore, the problem that the monitoring control efficiency of the hard disk device is lower can be solved, and the monitoring control efficiency effect of the hard disk device is improved can be achieved.

Drawings

Fig. 1 is a hardware block diagram of a server device of a monitoring control method of a hard disk device according to an embodiment of the present application;

fig. 2 is an application scenario schematic diagram of a monitoring control method of a hard disk device according to an embodiment of the present application;

fig. 3 is a flowchart of a monitoring control method of a hard disk device according to an embodiment of the present application;

FIG. 4 is a schematic diagram of an alternative target scene recognition model according to an embodiment of the application;

FIG. 5 is a schematic diagram of an alternative deterministic monitoring template in accordance with an embodiment of the present application;

FIG. 6 is a schematic diagram of an alternative method of monitoring and controlling a hard disk device according to an embodiment of the present application;

fig. 7 is a block diagram of a monitor control apparatus of a hard disk device according to an embodiment of the present application.

Detailed Description

Embodiments of the present application will be described in detail below with reference to the accompanying drawings in conjunction with the embodiments.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order.

The method embodiments provided in the embodiments of the present application may be executed in a server apparatus or similar computing device. Taking the operation on the server device as an example, fig. 1 is a hardware structure block diagram of the server device of a monitoring control method of the hard disk device according to an embodiment of the present application. As shown in fig. 1, the server device may include one or more (only one is shown in fig. 1) processors 102 (the processor 102 may include, but is not limited to, a microprocessor MCU, a programmable logic device FPGA, or the like processing means) and a memory 104 for storing data, wherein the server device may further include a transmission device 106 for communication functions and an input-output device 108. It will be appreciated by those of ordinary skill in the art that the architecture shown in fig. 1 is merely illustrative and is not intended to limit the architecture of the server apparatus described above. For example, the server device may also include more or fewer components than shown in FIG. 1, or have a different configuration than shown in FIG. 1.

The memory 104 may be used to store a computer program, for example, a software program of application software and a module, such as a computer program corresponding to a method for controlling monitoring of a hard disk device in an embodiment of the present application, and the processor 102 executes the computer program stored in the memory 104, thereby performing various functional applications and data processing, that is, implementing the above-mentioned method. Memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include memory remotely located with respect to the processor 102, which may be connected to the server device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmission device 106 is used to receive or transmit data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of a server device. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, simply referred to as a NIC) that can connect to other network devices through a base station to communicate with the internet. In one example, the transmission device 106 may be a Radio Frequency (RF) module, which is configured to communicate with the internet wirelessly.

First, the features involved in the embodiments of the present application may be, but are not limited to, applicable to the following explanation:

The Self-Monitoring analysis and reporting technology is called smart for short, which can monitor the temperature, internal circuit, medium material on the surface of the disk, etc. of the hard disk, so as to analyze the possible problems of the hard disk in time and send out warning, thereby protecting the data from loss.

SATA: SERIAL ATA, serial ATA.

SAS (SERIAL ATTACHED SCSI, serial attached SCSI interface), serial attached small computer system interface.

RAID (Redundant Arrays of INDEPENDENT DISKS, RAID) disk array is composed of a large number of independent disks, and a disk group with huge capacity is formed by combining a plurality of independent disks, so that the efficiency of the whole disk system is improved by utilizing the addition effect generated by providing data by individual disks. With this technique, data is cut into a number of sections, which are stored on individual hard disks.

W2V: the W2V algorithm is a neural network-based word vector calculation tool, which was developed by Google in 2013. The birth of W2V solves two main problems of the traditional word vector calculation method (such as one-hot coding): extremely high dimensions and no word-to-word correlation. There are two models for W2V, CBOW model and Skip-gram model, respectively. Both models vectorize words, differing in that: the CBOW model takes a word as an output target and takes words adjacent to the word as input; the Skip-gram model takes a word as input and takes the words adjacent to the word as output targets.

Out-of-band: it can be simply understood that the management channel provided by the hardware controller of the server itself is independent of the server operating system.

In-band: management channels associated with the server operating system.

Attention neural network: a neural network model emphasizes important information in generating an output by dynamically assigning different weights to different elements in an input sequence. This approach enables the neural network to better handle long sequences and tasks that focus on important information.

In this embodiment, a method for monitoring and controlling a hard disk device is provided, fig. 2 is a schematic view of an application scenario of a method for monitoring and controlling a hard disk device according to an embodiment of the present application, as shown in fig. 2, where a hard disk device 1, a hard disk device 2, and a hard disk device 3 are deployed on a server, and a monitoring controller is connected to the server, where the method for monitoring and controlling a hard disk device in the embodiment of the present application may be implemented by, but not limited to, the following steps.

In step S202, a server initiated hard disk monitoring request is received, where the hard disk monitoring request is used to request a monitoring process of a target hard disk device (e.g., hard disk device 1) among hard disk devices (e.g., hard disk device 1, hard disk device 2, and hard disk device 3) deployed on a control server.

In response to the hard disk monitoring request, step S204 identifies, from a plurality of operation scenes (e.g., operation scene 1, operation scene 2, and operation scene 3), a target operation scene (e.g., operation scene 1) corresponding to the target hard disk device (e.g., hard disk device 1) according to target operation data (e.g., operation data of hard disk device 1) of the target hard disk device (e.g., operation data of hard disk device 1), where the target operation data (e.g., operation data of hard disk device 1) includes a monitored value under each monitored item in a set of monitored items generated when the target hard disk device (e.g., hard disk device 1) is running on a server, and the operation scene is used to indicate a performance requirement of a service operated by the hard disk device for the hard disk device.

In step S206, a target monitoring list is created according to the target monitoring items corresponding to the target operation scene (e.g., operation scene 1) in the monitoring item set, the reference operation data corresponding to the target monitoring items in the target operation data (e.g., operation data of the hard disk device 1), and the monitoring types to which the target monitoring items belong, where the target monitoring items (e.g., monitoring item 1, monitoring item 2, and monitoring item 3) and the target monitoring modes (e.g., monitoring mode 1, monitoring mode 2, and monitoring mode 3) having the corresponding relationships are recorded in the target monitoring list.

In step S208, the target monitoring list is sent to the server, where the server is configured to monitor the operation of the target hard disk device (e.g., the hard disk device 1) according to the target monitoring list.

In this embodiment, a method for monitoring and controlling a hard disk device is provided, which is applied to a monitoring controller, where the monitoring controller is connected to a server, and the hard disk device is deployed on the server, and fig. 3 is a flowchart of a method for monitoring and controlling a hard disk device according to an embodiment of the present application, as shown in fig. 3, where the flowchart includes the following steps:

Step S302, receiving a hard disk monitoring request initiated by the server, wherein the hard disk monitoring request is used for requesting to control a monitoring process of a target hard disk device in hard disk devices deployed on the server;

Step S304, responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device, wherein the target operation data comprises monitoring values of each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scene is used for indicating the performance requirement of the service operated by the hard disk device on the hard disk device;

Step S306, a target monitoring list is created according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data, and a monitoring type to which the target monitoring item belongs, wherein the target monitoring item and a target monitoring mode with a corresponding relation are recorded in the target monitoring list;

and step 308, the target monitoring list is sent to the server, wherein the server is used for monitoring the operation of the target hard disk device according to the target monitoring list.

By the steps, according to the operation data of the hard disk device, the operation scene corresponding to the hard disk device is identified from a plurality of operation scenes, the operation scenes can be used for indicating the performance requirement of the operation of the hard disk device on the hard disk device, and according to the monitoring item corresponding to the operation scene, the operation data corresponding to the monitoring item in the operation data and the monitoring type to which the monitoring item belongs, a monitoring list is created, wherein the monitoring item and the monitoring mode with the corresponding relation are recorded in the monitoring list, and it is understood that the monitoring item and the monitoring mode in the monitoring list can be changed under the condition that the performance requirements of the hard disk device are different.

In the technical solution provided in step S302, one or more hard disk devices may be but not limited to be disposed in the server, where the hard disk devices may be but not limited to be used for storing data of the server, and as an optional example, the data storage hard disk devices may be but not limited to include a solid state hard disk device, a mechanical hard disk, a removable hard disk, or a RAID array, etc.

Optionally, in this embodiment, the monitoring controller may be but not limited to be deployed at the cloud end, and the monitoring control method of the hard disk device in the embodiment of the present application may be but not limited to include two parts, namely a client end and a cloud end (server end). The client is client data center infrastructure management software, and can collect various monitoring indexes (or called monitoring items) of the server hard disk. The client can periodically report the acquired data (corresponding to the operation data of the hard disk device) to the cloud, and can acquire a monitoring template customized based on the self scene from the cloud. And monitoring various indexes of the hard disk of the server through the monitoring templates.

The cloud can classify and summarize various hard disk indexes of the client, analyze big data in an artificial intelligence mode and train monitoring templates suitable for hard disks of different manufacturers and under different service scenes (equivalent to running scenes). And provided to clients as needed.

Compared with a closed local management software scheme in the related art, the embodiment of the application uses a scheme of interconnecting the client and the server, and the following design is made on the access and security of the client and the cloud. The client and the cloud end use unidirectional communication, namely the cloud end actively exposes the client and the cloud end on the public network, the client can actively request the cloud end, and the cloud end cannot actively request the client. The client pushes data to the cloud end at regular time, pulls task information (for example, whether a new monitoring template can be downloaded or not, if so, the cloud end is actively requested to download), and reports the data according to the task information.

In addition, the communication between the client and the cloud may be, but not limited to, communication based on the HTTPS protocol of digital signature and certificate mutual authentication, while using a token mechanism for authentication and authorization. For example, using a token mechanism such as JWT (JSON Web Token) or OAuth, the client may prove its identity to the server and obtain the corresponding rights.

Through the one-way communication between the client and the cloud, the cloud is prevented from actively acquiring the operation data of the hard disk device, so that the operation data of the hard disk device is prevented from being leaked, the safety of the operation data of the hard disk device is improved, the communication between the client and the cloud is verified and authorized through a token mechanism, the accuracy of the client communicating with the cloud is ensured, and the safety of the communication between the client and the cloud is improved.

In the technical solution provided in step S304, the target operation data of the target hard disk device may be obtained, but not limited to, by: acquiring initial operation data of a target hard disk device, wherein the initial operation data comprises operation data on a plurality of acquisition periods; and executing average value operation on the initial running data to obtain target running data.

Optionally, in this embodiment, the operation scenario may be, but not limited to, a performance requirement for the hard disk device, where the performance requirement for the hard disk device may be different for the service operated by the hard disk device, it may be understood that the performance requirement for the hard disk device may be different for the service operated by the hard disk device, and as an optional example, the performance requirements of the operation scenarios may be, but not limited to, a high performance computing and I/O (Input/Output) intensive scenario, a large-scale data storage and high availability scenario, a content distribution and network service scenario, a graphics and multimedia processing scenario, a data backup and recovery scenario, and the like, which may be as shown in table 1.

TABLE 1

Alternatively, in this embodiment, the target operation scenario may include, but is not limited to, one or more of multiple operation scenarios, for example, according to target operation data of the target hard disk device, a target operation scenario corresponding to the target hard disk device identified from the multiple operation scenarios includes a high performance computing and I/O intensive scenario and a data backup and recovery scenario in table 1.

In one exemplary embodiment, the target operation scene corresponding to the target hard disk device may be identified from a plurality of operation scenes according to the target operation data of the target hard disk device, but not limited to, by: according to the target operation data, calculating the probability that the target hard disk device corresponds to each of the plurality of operation scenes; and screening N operation scenes corresponding to the probability meeting the target screening condition from the operation scenes according to the operation scenes and the probability with the corresponding relation, wherein N is a positive integer.

Alternatively, in this embodiment, the probability of each of the plurality of operation scenes corresponding to the target hard disk device may be, but not limited to, equal to or greater than 0 and equal to or less than 1, and the sum of the probabilities of each of the plurality of operation scenes corresponding to the target hard disk device is equal to 1.

Optionally, in this embodiment, the target operation scenario that corresponds to the probability that meets the target screening condition may be selected from the multiple operation scenarios by the following method: screening candidate probability with the highest probability from probabilities of each of a plurality of operation scenes corresponding to the target hard disk device, and determining the operation scene corresponding to the candidate probability in the plurality of operation scenes as a target operation scene under the condition that the candidate probability is larger than or equal to a preset probability threshold.

By the method, the situation that the corresponding operation scene with the largest probability of the operation scenes is directly determined as the target operation scene under the condition that the probability of each scene in the plurality of operation scenes corresponding to the target hard disk device is smaller is avoided, and it can be understood that the corresponding operation scene with the largest probability cannot accurately indicate the performance requirement of the service operated by the hard disk device on the hard disk device, and the accuracy of the determined target operation scene is improved by determining the operation scene corresponding to the candidate probability as the target operation scene under the condition that the candidate probability is greater than or equal to a preset probability threshold (for example, 0.5, 0.6 or the like, and the like.

In one exemplary embodiment, N operation scenes whose corresponding probabilities satisfy a target screening condition may be screened out of the plurality of operation scenes as the target operation scenes according to the operation scenes and probabilities having a correspondence relation by, but not limited to: extracting operation scenes from the operation scenes one by one according to the corresponding probability from high to low; detecting the relation between the probability corresponding to the current extracted operation scene, the probability corresponding to the extracted operation scene and the target threshold value; determining the current extracted operation scene and the extracted operation scene as the N operation scenes as the target operation scenes under the condition that the probability sum is greater than or equal to the target threshold value; and under the condition that the sum of the probabilities is smaller than the target threshold value, continuously extracting the next operation scene from high to low according to the corresponding probability.

Optionally, in this embodiment, N operation scenes whose corresponding probabilities satisfy the target screening condition may be selected as target operation scenes from the multiple operation scenes according to the operation scenes and probabilities having the correspondence relation, and may include, but not limited to, 5 operation scenes (e.g., operation scene 1, operation scene 2, operation scene 3, operation scene 4, and operation scene 5), the probabilities of operation scene 1, operation scene 2, operation scene 3, operation scene 4, and operation scene 5 corresponding to the target hard disk device are 0.3, 0.4, 0.1, and 0.1, respectively, and the target threshold is 0.55, and may include, but not limited to, extracting operation scenes from the multiple operation scenes one by one according to the corresponding probabilities from high to low.

For example, firstly, extracting the corresponding operation scene 2 with the probability of 0.4; detecting a relationship between a probability (e.g., 0.4) corresponding to a currently extracted operation scene (e.g., operation scene 2) and a probability sum (e.g., 0.4) of a probability corresponding to an extracted operation scene (e.g., empty) and a target threshold (e.g., 0.55); in the case where the probability sum (e.g., 0.4) is smaller than the target threshold (e.g., 0.55), the next operation scene is continuously extracted from high to low with the corresponding probability, for example, operation scene 1 with the corresponding probability of 0.3 is continuously extracted.

In the case where the probability (e.g., 0.3) corresponding to the currently extracted operation scene (e.g., operation scene 1) and the probability (e.g., 0.4) sum (e.g., 0.7) corresponding to the extracted operation scene (e.g., operation scene 2) are greater than or equal to the target threshold (e.g., 0.55), the currently extracted operation scene (e.g., operation scene 1) and the extracted operation scene (e.g., operation scene 2) are determined as N operation scenes as target operation scenes.

In this way, when the probability corresponding to a certain operation scene in the plurality of operation scenes is smaller than the target threshold value, and the difference between the target threshold value and the probability corresponding to the operation scene is smaller than or equal to the preset difference threshold value, it can be understood that the probability corresponding to the operation scene is very close to the target threshold value, and in such a case, the operation scene may not accurately indicate the performance requirement of the service operated by the hard disk device on the hard disk device, and in such a case, by the embodiment of the application, one or more operation scenes indicating the performance requirement of the service operated by the hard disk device on the hard disk device can be determined as the target operation scene, so that the accuracy of the service operated by the target operation scene on the performance requirement of the hard disk device is improved.

In one exemplary embodiment, the target operation scene corresponding to the target hard disk device may be identified from a plurality of operation scenes according to the target operation data of the target hard disk device, but not limited to, by: inputting the target operation data into a target scene recognition model, wherein the target scene recognition model is obtained by training an initial scene recognition model by using an operation data sample marked with an operation scene label; and acquiring the target operation scene output by the target scene identification model.

Alternatively, in the present embodiment, the target scenario recognition model may be, but is not limited to, a target operation scenario corresponding to a target hard disk device from a plurality of operation scenarios, and as an optional example, the recognition includes a neural network model (for example, a convolutional neural network or a cyclic neural network or an artificial neural network, etc., to which the present application is not limited), an attention neural network or other recognition model, and so on.

As an optional example, the target operation scenario corresponding to the identified target hard disk device in the embodiment of the present application may be explained and illustrated by taking, but not limited to, that the target scenario identification model includes a neural network model as an example, and may be applicable to the embodiment of the present application.

FIG. 4 is a schematic diagram of an alternative target scene recognition model, as shown in FIG. 4, which may include, but is not limited to, a multi-layer neural network and an Attention (Attention) neural network, according to an embodiment of the present application. The terms described above are first defined. a. The dimensions and metrics are features in the neural network, where the dimensions (corresponding to the type of monitoring) are secondary characteristics and the metrics (corresponding to the monitoring terms) are primary characteristics, such as: "performance dimension" is a secondary feature and "I/O count per second index" is a primary feature. b. The operation scene is a label in the neural network. As an alternative example, tags may be, but are not limited to, classified into 5 categories.

The operational scenario to which the hard disk device corresponds may be, but is not limited to, identified by: a. setting a first layer of the neural network as a first-level feature, setting a second layer of the neural network as a second-level feature, setting the finally classified labels as 5 types, and using normalization processing as a basis for label classification. b. An Attention neural network is added to the secondary features to facilitate subsequent dimension visualization weight analysis to determine which dimensions (e.g., asset dimension, performance dimension, environment dimension, usage dimension, etc.) are needed for hard disk monitoring in this scenario. c. An Attention neural network is added to the first-level characteristics so as to facilitate the subsequent index visualization weight analysis, so as to determine which index (such as the number of times of disk I/O) is needed for monitoring the hard disk in the scene, and the I/O waiting time percentage of a CPU (Central Processing Unit ) is determined.

D. As the data input by the neural network model continues to increase, the model continues to be trained, associating the 5 defined scenarios described above (e.g., high performance computing and I/O intensive scenarios, or large scale data storage and high availability scenarios, etc.) for the 5 classes of labels output by the model based on the weight visualization analysis described for b, c.

E. when the neural network input is less than 2w and the continuous input is less than 7 days, the model is considered unusable without training.

According to the embodiment of the application, the operation scenes corresponding to the hard disk equipment are distinguished based on the target scene recognition model (such as a multi-layer neural network, a two-level Attention neural network and the like), the response speed of the target scene recognition model is higher, the operation scenes corresponding to the hard disk equipment can be rapidly determined, and the efficiency of determining the operation scenes corresponding to the hard disk equipment is improved.

In the technical solution provided in step S306, the monitoring types to which the target monitoring items belong may include a plurality of monitoring types, and may be, but not limited to, explained and illustrated by taking the example that the monitoring types to which the target monitoring items belong include 4 monitoring types (for example, an asset dimension, a performance dimension, an environment dimension, and a usage dimension).

The asset dimension may be, but is not limited to, an index item (equivalent to a monitoring item) for indicating that the asset attribute is included, and may be, but is not limited to, an asset attribute including the hard disk itself and an asset attribute including a server to which the hard disk belongs, where the attributes have a certain influence on the operation environment and the operation scene division of the hard disk, so that the data center scene of the hard disk can be analyzed from the viewpoint of an initial physical specification. Index items of asset dimensions (equivalent to monitoring items) may include, but are not limited to, the following:

a. server specifications (number, vendor, model, serial number, motherboard model).

B. server CPU specification (number, vendor, model number, serial number, master frequency, core number).

C. Server memory specifications (number, vendor, model, serial number, master frequency, capacity).

D. server RAiD card specifications (number, vendor, model number, serial number, interface type, cache capacity, maximum disk number, maximum data transfer rate, current RAID level, and hard disk array topology).

E. Hard disk specifications (quantity, vendor, model, serial number, media type, capacity, rated power consumption).

F. server network card specification (number, interface type, transfer rate).

G. Operating system specifications (system version, software version, etc.).

It should be noted that the data of the asset dimension may be, but not limited to, obtained by an out-of-band management controller (such as BMC (baseboard management controller, baseboard Management Controller)) of the server, an out-of-band management interface, and an in-band management tool.

The performance dimension may be, but is not limited to, an index item for representing the performance of the server, and may be, but is not limited to, time series class data, so that the current running condition of the server can be determined in real time, and the scene of the hard disk is analyzed from the running dimension. The monitoring items of the performance dimension may include, but are not limited to, the following:

a. server overall performance (current power consumption, air intake temperature, system run time, system average load, number of system runs, number of system dormancy processes, number of system zombie processes).

B. Server CPU performance (temperature, usage, percentage of CPU occupied by kernel space, percentage of CPU time waiting for input and output, total amount of time consumed by CPU to service hard interrupts, total amount of time consumed by CPU to service soft interrupts).

C. Server memory performance (free physical memory, kernel memory buffer usage, swap partition capacity, swap partition usage).

D. server hard disk performance (I/O read/write times per second, I/O read/write rate per second MB/s, storage bandwidth usage, hard disk temperature).

E. Server network card performance (transmit receive rate per second, bandwidth usage, packet loss).

The data of the performance dimension is usually obtained through an out-of-band management controller (such as BMC) of the server, an out-of-band management interface and an in-band management tool.

The index of the performance dimension includes the rate of change of the index, and additional calculation is required.

The environmental dimensions may be used, but are not limited to, to represent factors that may affect hard disk operation, such as data center temperature (20 ℃ to 25 ℃), humidity (40% to 50%), noise, smoke (particulate matter concentration, ozone concentration, nitrogen oxide concentration, carbon monoxide concentration, and carbon dioxide concentration), number of shocks, power supply (under-voltage and under-current, over-voltage and over-current), number of abnormal power outages, and the like. The system can be integrated with various dynamic environment systems of a data center to acquire related environment indexes.

It should be noted that some indicators of the environmental dimension (such as temperature, moderate, noise) also include the rate of change of the above indicators, and additional calculation is required.

The usage dimension may include, but is not limited to, factors that may create usage loss to the server hard disk. The monitoring items of the usage dimension may include, but are not limited to, the following:

a. Configuration counts (RAID firmware upgrade times, RAID configuration times, hard disk firmware upgrade times, hard disk formatting times).

B. migration count (number of server location changes, number of hard disk plug changes).

C. abnormal operation count (number of server abnormal power down).

It should be noted that, such data may be collected and maintained by conventional data center operation and maintenance software, and may be obtained by conventional means.

Optionally, in this embodiment, the monitoring items belonging to different monitoring types may correspond to different or identical monitoring manners, and as an optional example, in a case where the monitoring type is the above-mentioned property dimension, the corresponding monitoring manner includes whether the monitoring asset information is changed, and in a case where the monitoring type is the above-mentioned property dimension, the corresponding monitoring manner includes whether the monitoring value under the corresponding monitoring item is greater than or equal to the threshold value.

Optionally, in this embodiment, the target running scenario may, but not limited to, have a corresponding target monitoring item in the monitoring item set, for example, the above-mentioned high-performance computing monitoring item corresponding to the performance dimension (equivalent to the monitoring type) of the I/O intensive scenario, and as another optional example, may also, but not limited to, determine the target monitoring item by: inputting the target operation data into a target scene recognition model, wherein the target scene recognition model is obtained by training an initial scene recognition model by using an operation data sample marked with an operation scene label; acquiring the target operation scene output by the target scene recognition model, and inputting the target operation scene and the target operation data into a target monitoring item recognition model, wherein the target scene recognition model is obtained by training an initial monitoring item recognition model by using a reference operation data sample marked with a monitoring item label and an operation scene label; and acquiring the target monitoring item output by the target monitoring item identification model.

As an alternative example, the target monitoring item may also be, but is not limited to, determined by: according to the target operation data, calculating the probability of each candidate monitoring item in the candidate monitoring item set corresponding to the target hard disk device in the target operation scene in the monitoring item set; and screening one or more candidate monitoring items with the probability meeting the reference screening condition from the candidate monitoring item set according to the monitoring items and the probabilities with the corresponding relations, wherein the one or more candidate monitoring items are selected as the monitoring items.

As an alternative example, the reference screening condition may include, but is not limited to, extracting candidate monitoring items from the candidate monitoring item set one by one with a corresponding probability from high to low; detecting the relation between the probability corresponding to the currently extracted candidate monitoring item, the target sum value of the probability corresponding to the extracted candidate monitoring item and a preset probability threshold value; determining the currently extracted candidate monitoring item and the extracted candidate monitoring item as target monitoring items under the condition that the target sum value is greater than or equal to the probability threshold value; and under the condition that the target sum value is smaller than the probability threshold value, continuously extracting the next candidate monitoring item from high to low according to the corresponding probability.

As an alternative example, the probability of each candidate monitor is directly proportional to the impact on the performance requirements of the hard disk device of the service operated by the candidate monitor, e.g., the greater the probability of each candidate monitor, the greater the impact on the performance requirements of the hard disk device of the service operated by the candidate monitor.

By the mode, key monitoring is realized on the monitoring item with larger influence on the performance requirement of the hard disk device on the service operated by the hard disk device, the key monitoring item with larger influence on the performance requirement of the hard disk device is realized, and the stability of the service operated by the hard disk device is improved.

In one exemplary embodiment, the target monitoring list may be created according to, but not limited to, a target monitoring item corresponding to a target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data, and a monitoring type to which the target monitoring item belongs, by: creating an initial monitoring list by using a target monitoring item corresponding to the target operation scene; according to the monitoring type of the target monitoring item, the target monitoring mode is allocated to the target monitoring item to obtain a reference monitoring list, wherein the target monitoring mode comprises the following steps: a mode of monitoring according to the change of the data and a mode of monitoring according to the threshold value of the data; and under the condition that the target monitoring mode is a mode of monitoring according to the threshold value of the data, distributing a target monitoring threshold value for the target monitoring item according to the reference operation data to obtain the target monitoring list.

Optionally, in this embodiment, the monitoring manner according to the change of the data may, but is not limited to, including whether the monitored value under the target monitoring item changes, for example, in the case where the monitored type is the asset dimension described above, the corresponding monitoring manner includes whether the monitored asset information (for example, the identifier of the hard disk, or the identifier of the server where the hard disk is located) changes, which is usually that the hard disk or the server has a component replaced (for example, a RAID card model is replaced), or that the hard disk is plugged into another server.

In one exemplary embodiment, an ith monitoring threshold value may be assigned to an ith monitoring item according to an ith operating data, wherein the reference operating data includes M operating data including the ith operating data, the target monitoring item includes M monitoring items including the ith monitoring item, the target monitoring threshold value includes M monitoring threshold values including the ith monitoring threshold value, wherein i is a positive integer less than or equal to M, and M is a positive integer; selecting the ith candidate operation data used for representing that the target hard disk device is in a normal operation state from the ith operation data; performing average value operation on the ith candidate running data to obtain an average value, and determining the standard deviation of the ith candidate running data; determining a first candidate upper threshold as a sum of the average value and the standard deviation of the target number, and determining a first candidate lower threshold as a value obtained by subtracting the standard deviation of the target number from the average value; determining a larger upper limit threshold from a second candidate upper limit threshold and the first candidate upper limit threshold to obtain a target upper limit threshold, and determining a smaller lower limit threshold from a second candidate lower limit threshold and the first candidate lower limit threshold to obtain a target lower limit threshold, wherein the target monitoring threshold comprises the target upper limit threshold and the target lower limit threshold, the second candidate upper limit threshold is the data with the largest value in the i candidate operation data, and the second candidate lower limit threshold is the data with the smallest value in the i candidate operation data.

Alternatively, in the present embodiment, in the case where the second candidate upper limit threshold is greater than the first candidate upper limit threshold and the second candidate lower limit threshold is greater than the first candidate lower limit threshold, the target upper limit threshold is determined as the second candidate upper limit threshold and the target lower limit threshold is determined as the first candidate lower limit threshold; when the second candidate upper limit threshold is smaller than the first candidate upper limit threshold and the second candidate lower limit threshold is smaller than the first candidate lower limit threshold, determining the target upper limit threshold as the first candidate upper limit threshold and determining the target lower limit threshold as the second candidate lower limit threshold; when the second candidate upper limit threshold is greater than the first candidate upper limit threshold and the second candidate lower limit threshold is less than the first candidate lower limit threshold, determining the target upper limit threshold as the second candidate upper limit threshold and the target lower limit threshold as the second candidate lower limit threshold; when the second candidate upper limit threshold is smaller than the first candidate upper limit threshold and the second candidate lower limit threshold is larger than the first candidate lower limit threshold, determining the target upper limit threshold as the first candidate upper limit threshold and the target lower limit threshold as the first candidate lower limit threshold; the target monitoring threshold comprises a target upper limit threshold and a target lower limit threshold, wherein the second candidate upper limit threshold is the data with the largest value in the i candidate operation data, and the second candidate lower limit threshold is the data with the smallest value in the i candidate operation data.

By means of the method, all history data without alarm association are filtered, normal distribution methods are used for determining the upper and lower boundaries of the threshold value of normal operation, and accuracy of the threshold value corresponding to the monitoring item is improved.

In the technical solution provided in step S308, the target monitoring list may be, but not limited to, sent to the server, and the server may, but not limited to, monitor the operation of the hard disk device according to the monitoring items in the target monitoring list and the monitoring modes corresponding to the monitoring items, and it may be understood that, for the same hard disk device, it may correspond to different operation scenarios at different times, and in such a case, the monitoring items in the target monitoring list and the monitoring modes corresponding to the monitoring items may change.

In an exemplary embodiment, the method further comprises, before said sending said target monitor list to said server: transmitting a data reporting instruction to the server, wherein the data reporting instruction is used for instructing the server to report alarm data belonging to the target operation scene to the monitoring controller; receiving target alarm data reported by the server in response to the data reporting instruction; extracting keywords in each alarm data in the P alarm data to obtain P groups of target keywords under the condition that the target alarm data comprises P alarm data and the target hard disk device comprises T hard disk devices, wherein a w-th group of target keywords in the P groups of target keywords comprise identifications of w-th hard disk devices in the T hard disk devices, positions of the w-th hard disk devices deployed in the server and device types of the w-th hard disk devices, P is a positive integer, T is a positive integer, and w is a positive integer smaller than or equal to P; vectorizing the P groups of target keywords to obtain P groups of word vectors; determining the similarity between the s ₁ th word vector in the s th word vector in the P group word vectors and the r ₁ th word vector in each group word vector in the P-1 group word vectors except the s th word vector to obtain the s ₁ th similarity, determining the similarity between the s ₂ th word vector in the s th group word vector and the r ₂ th word vector in each group word vector in the P-1 group word vectors to obtain the s ₂ th similarity, determining the similarity between the s ₃ th word vector in the s th group word vector and the r ₃ th word vector in each group word vector in the P-1 group word vectors, obtaining an s ₃ th group similarity, wherein the s ₁ th word vector is a word vector used for representing the identifier of the s hard disk device in the s th group word vector, the s ₂ th word vector is a word vector used for representing the position of the s hard disk device deployed in a server in the s th group word vector, the s ₃ th word vector is a word vector used for representing the device type of the s hard disk device in the s th group word vector, the r ₁ th word vector is a word vector used for representing the identifier of the hard disk device in each group word vector in the P-1 group word vector, the r ₂ th word vector is a word vector used for representing the position of the hard disk device deployed in the server in each group of word vectors in the P-1 group of word vectors, the r ₃ th word vector is a word vector used for representing the device type of the hard disk device in each group of word vectors in the P-1 group of word vectors, the T hard disk devices comprise the s-th hard disk device, s is a positive integer less than or equal to P, s ₁、s₂, s ₃ are both positive integers, and r ₁、r₂, and r ₃ are both positive integers; performing average value operation on the s ₁ th group of similarity, the s ₂ th group of similarity and the s ₃ th group of similarity to obtain an s average similarity; under the condition that the s-th average similarity is larger than or equal to a preset average similarity threshold value, determining s-th alarm data corresponding to the s-th group word vector in the P alarm data as a target alarm monitoring item; and adding the target alarm monitoring item to the target monitoring list.

Alternatively, in this embodiment, the similarity between the s ₁ th word vector and the r ₁ th word vector may include, but is not limited to, cosine similarity or euclidean distance, etc., and it should be noted that the manner of determining the similarity between the s ₂ th word vector and the r ₂ th word vector may be the same as, but not limited to, the manner of determining the similarity between the s ₁ th word vector and the r ₁ th word vector, and the manner of determining the similarity between the s ₃ th word vector and the r ₃ th word vector may be the same as, but not limited to, the manner of determining the similarity between the s ₁ th word vector and the r ₁ th word vector.

Optionally, in this embodiment, in a case where the s-th average similarity is greater than or equal to a preset average similarity threshold, the s-th alarm data corresponding to the s-th word vector in the P-th alarm data is not determined to be the target alarm monitoring item.

Optionally, in this embodiment, in a case where the target alarm monitoring item is determined, the candidate monitoring threshold of the s-th monitoring item in the target monitoring list may be but not limited to be adjusted according to the alarm threshold in the s-th alarm data, as an optional example, the s-th alarm upper limit threshold and the s-th alarm lower limit threshold of the s-th monitoring item in the s-th alarm data may be but not limited to be acquired, in a case where the s-th alarm upper limit threshold is smaller than the candidate upper limit threshold of the s-th monitoring item, the candidate upper limit threshold of the s-th monitoring item is adjusted to be the s-th alarm upper limit threshold, and/or in a case where the s-th alarm lower limit threshold is larger than the candidate lower limit threshold of the s-th monitoring item, the candidate lower limit threshold of the s-th monitoring item is adjusted to be the s-th alarm lower limit threshold, where the target monitoring list includes the s-th alarm upper limit threshold and the s-th alarm lower limit threshold of the candidate monitoring item.

In this way, the alarm data corresponding to the word vector with higher average similarity in the alarm data is determined to be the alarm data needing to be focused, the alarm data needing to be focused may be the frequently-occurring alarm data or may influence the performance requirement of the service running on the hard disk device, and in this way, the commonality monitoring item under the scene is determined based on the W2V neural network word vector method and the normal distribution method by filtering out all the historical alarms. Meanwhile, the determined upper and lower threshold boundaries can be optimally corrected again based on the key monitoring items, but are not limited to. .

In order to better understand the control procedure of the monitoring control of the hard disk device in the embodiment of the present application, the following explanation and explanation are given with reference to the alternative embodiment, which is applicable to, but not limited to, the embodiment of the present application.

FIG. 5 is a schematic diagram of an alternative deterministic monitoring template according to an embodiment of the present application, as shown in FIG. 5, which can be illustrated and described by taking as an example a number of operational scenarios including the high performance compute and I/O intensive scenario shown in Table 1, the large-scale data storage and high availability scenario, the content distribution and web services scenario, the graphics and multimedia processing scenario, and the data backup and restore scenario, taking as an example the monitoring type to which the target monitoring item belongs including 4 (e.g., asset dimension, performance dimension, environmental dimension, and usage dimension). The method includes determining an operation scene (for example, scene a) corresponding to a hard disk device according to operation data reported by a client through a classification model (equivalent to a target scene recognition model), determining all monitoring items and corresponding threshold intervals according to a non-alarm sample data set in all the acquired client data sets of the scene a, screening key monitoring items according to alarm sets in all the acquired client data sets of the scene a, and further obtaining a monitoring template (equivalent to a target monitoring list).

As an alternative example, the operational scenario to which the hard disk device corresponds may be, but is not limited to, determined by: after the neural network model (equivalent to the target scene recognition model) is trained to a usable state. And acquiring a plurality of (e.g. 5) samples according to the time interval according to the latest historical data reported by the client, and using the average value (equivalent to the target operation data) of the samples as a model input. And according to model output, determining the weight (equivalent to the probability of each of a plurality of operation scenes corresponding to the hard disk device) of 5 types of operation scenes corresponding to the client, and taking all operation scenes (equivalent to the target operation scene) with the sum of the weights being more than 0.5 as a set C0. And the operation scene of the hard disk device finally belongs to.

All monitoring items and corresponding threshold intervals may be determined, but are not limited to, by: and determining a secondary characteristic set of each scene according to the secondary characteristic Attention neural network output of the scene set C0, and taking the union set C1 as a monitoring dimension (equivalent to the monitoring type of the monitoring item) mark of the monitoring template. Then, determining a first-level feature set of each scene according to the first-level feature Attention neural network output of the scene set C0, taking the union set C2 as a monitoring index (equivalent to a monitoring item) of a monitoring template, wherein the first-level features can comprise different monitoring modes according to the second-level feature types (equivalent to the monitoring types to which the monitoring item belongs):

a. the secondary features are asset dimensions: the main monitoring of asset information changes is that the server to which the hard disk belongs is replaced with a component (such as replacing a RAID card model), or the hard disk is plugged into another server.

B. the secondary features are performance dimensions, environmental dimensions, or usage dimensions: the monitoring of such dimensions is typically threshold monitoring, and the threshold value corresponding to each monitored item may be set, but is not limited to, by the following means.

For example, all sample data DA (data all) reported by clients belonging to such a scenario are acquired, excluding sample data DW (data warning) of the hard disk in an alert state. The threshold upper and lower bounds are determined for the remaining samples DN (data normal, equivalent to running data) based on normal distribution.

The mean (mean) and standard deviation (STANDARD DEVIATI/On) of the samples were calculated.

And calculating an upper critical value and a lower critical value of the normal distribution curve according to the characteristic of the normal distribution. As an alternative example, and without limitation, taking the target number of 3 as an example, in a normal distribution, approximately 99.7% of the data lies within 3 standard deviations of the mean, which is considered normal. Therefore, the upper critical value (corresponding to the first candidate upper limit threshold) may be set to the average value plus 3 standard deviations, and the lower critical value (corresponding to the first candidate lower limit threshold) may be set to the average value minus 3 standard deviations.

The upper and lower thresholds of the sample set may be determined, but are not limited to, by comparing the upper and lower thresholds to the actual range of sample data. If the actual range contains an upper critical value and a lower critical value, an upper limit (corresponding to a target upper limit threshold) and a lower limit (corresponding to a target lower limit threshold) are set as a minimum value (corresponding to a second candidate lower limit threshold) and a maximum value (corresponding to a second candidate upper limit threshold) of the actual range, respectively; otherwise, the upper and lower limits are set to an upper critical value (corresponding to the first candidate upper limit threshold) and a lower critical value (corresponding to the first candidate lower limit threshold), respectively.

The key monitoring items may be screened, but are not limited to, by:

And for the running scene union C0, acquiring all alarm data AA (alarm all) reported by clients belonging to the running scenes. The key monitor term with commonality can be screened out by, but is not limited to, a W2V algorithm (a neural network-based word vector calculation tool, or other algorithms, to which the present application is not limited). The method may include, but is not limited to, the following steps:

The key word in the alarm information is extracted, and the key word comprises a component name (corresponding to the identification of the hard disk device), a position (corresponding to the position where the hard disk device is deployed in the server) and a device type (corresponding to the device type of the hard disk device).

Words may be vectorized, but are not limited to, by the Skip-gram model of W2V. W2V trains the corpus to represent each word as a low-dimensional vector, and the vectors capture semantic information of the words, so that semantically similar words are close in distance in a vector space.

And then, calculating the similarity of each vector with other vectors, taking a mean value, and eliminating the vectors with low similarity based on normal distribution. The alarms represented by the vector are alarms needing to be focused, and then the important monitoring items are screened out by combining the primary characteristics C2 according to the component name, the position and the equipment type, and meanwhile, the threshold value corresponding to the monitoring items is screened and optimized.

According to the embodiment of the application, the data items (equivalent to the monitoring items) related to the monitoring of the hard disk equipment of the server are classified (such as asset dimension, performance dimension, environment dimension and use dimension), and the data items (equivalent to the monitoring items) are not the dimension of the traditional pure hard disk, but can comprehensively evaluate the operation scene and the characteristic corresponding to the hard disk equipment in the whole dimension of the data center (including, but not limited to, the scale of the data center, various characteristics of the server, the power environment, the operation use and the like). Meanwhile, the number of the use scenes of the hard disk is definitely defined in the embodiment of the application.

In addition, based on the classification method, the embodiment of the application provides a method for distinguishing hard disk scenes based on a multi-layer neural network and a secondary Attention neural network. Determination methods of different scene monitoring templates (e.g., model training, sampling input, and determining a running scene set from weights >0.5, etc.) are also presented. For the above-mentioned monitoring templates, the embodiment of the application provides a method for determining a monitoring item set of a template (for example, a union set is obtained according to a scene set, different monitoring rules for asset monitoring items and numerical value monitoring items, etc.). The embodiment of the application provides the hard disk monitoring system which can be rapidly put into use, performs optimization analysis by combining a large sample and dynamically optimizes the monitoring strategy based on the real-time dynamic of the user scene, can optimize the problems of inaccuracy and slow updating iteration possibly existing in the aspect of monitoring the hard disk of the traditional data center management platform, and improves the accuracy and updating iteration efficiency of monitoring the hard disk equipment.

Fig. 6 is a schematic diagram of an alternative monitoring control method for a hard disk device according to an embodiment of the present application, as shown in fig. 6, which can be explained and illustrated by taking a plurality of operation scenarios including the high performance computing and I/O intensive scenario, the large-scale data storage and high availability scenario, the content distribution and network service scenario, the graphics and multimedia processing scenario, and the data backup and recovery scenario shown in table 1 as examples, taking the monitoring type to which the target monitoring item belongs as 4 (e.g., asset dimension, performance dimension, environment dimension, and usage dimension).

The control process of the monitor control of the hard disk device in the embodiment of the present application can be realized, but is not limited to, in the following manner.

In step S601, the client acquires information of each associated index item of the hard disk device of the server according to the embodiment of the present application through the out-of-band BMC management interface, the in-band Agent northbound interface, the smart log, the ring system, the life cycle tracing subsystem, and the like of the server, and performs calculation classification and summarization on the index item (corresponding to the monitoring item) according to the dimension classification (for example, the asset dimension, the performance dimension, the environment dimension, and the usage dimension) described in the embodiment of the present application.

In step S602, the client may, but is not limited to, periodically initiate a unidirectional request to the cloud based on the HTTPS and the token, and report the information of each index item described in step S601. At the same time, based on the request return value, it is determined whether a new task request needs to be initiated (e.g., a new monitoring template is available for download).

In step S603, after the cloud receives the report from the client, the data is classified based on the target scene recognition model (e.g. neural network), where the number of classifications is limited to the number of running scenes described in the embodiment of the present application (e.g. 5, but this number may also be adjusted as required). As data reported by multiple clients increases, a target scene recognition model (e.g., a neural network model) of the cloud is trained and optimized. The cloud end can determine the operation scene of a certain client hard disk in a mode based on the Attention neural network, and determine the monitoring item and the threshold value of the monitoring scene in a mode based on the W2V neural network and normal distribution described in the application.

For example, the multi-dimensional data collection may be, but is not limited to, data including an asset dimension, a performance dimension, an environment dimension, and an operation dimension, wherein the asset dimension may be, but is not limited to, including asset specification information (e.g., servers, processors, memory, network cards, hard disk devices, and RAIDs, etc.), the performance dimension may be, but is not limited to, including performance timing class information (e.g., servers, processors, memory, network cards, hard disks, etc.), the environment dimension may be, but is not limited to, including ring status, timing class information (e.g., temperature, humidity, noise, smoke, vibration, power supply, etc.), and the operation dimension may be, but is not limited to, including operation maintenance class information (e.g., configuration, migration, upgrade, abnormal operation, etc.).

The data may be processed, for example, without limitation, with additional feature values such as change rates, state transitions, and durations, etc., and then an operational scenario corresponding to the hard disk device may be determined, for example, one or more of a high performance computing and I/O intensive scenario, a large-scale data storage and high availability scenario, a content distribution and web services scenario, a graphics and multimedia processing scenario, and a data backup and restore scenario.

In step S604, the cloud returns a message body with a new monitoring template available for downloading when the next client requests. The client initiates a new one-way request, acquires a monitoring template provided by the cloud, and then takes effect as required.

By means of the cloud architecture and the access mode of the unidirectional request of the client, the cloud architecture and the access mode can be safely accessed, the power of big data analysis and artificial intelligence is exerted, an optimal hard disk monitoring template is generated based on extraction of a large number of data center level scene features, monitoring threshold values and key monitoring items are determined, the client dynamically pulls the flow and the security strategy of the monitoring template from the cloud, the client can use the latest monitoring template which is most in line with the scene and is issued by the server as required, and the instantaneity of the monitoring template is improved.

It should be noted that, in the embodiments of the present application, the hard disk device surrounding the data center is exemplified, and because the manufacturers, models and mediums of the hard disk device are relatively more, the classification of the operation scene is also relatively clear, and the parameter indexes surrounding the hard disk device are relatively more. The embodiment of the application can be applied to the monitoring of the core components such as the processor, the memory, the network card and the like of the data center. For monitoring of different components, different data acquisition dimensions and index items are determined, possible influences caused by the data acquisition dimensions and index items are determined, and meanwhile different operation scene classifications are defined.

From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present application.

The embodiment also provides a monitoring control device for a hard disk device, which is used for implementing the foregoing embodiments and preferred embodiments, and is not described in detail. As used below, the term "module" may be a combination of software and/or hardware that implements a predetermined function. While the means described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

Fig. 7 is a block diagram of a monitoring control apparatus for a hard disk device according to an embodiment of the present application, which is applied to a monitoring controller, the monitoring controller being connected to a server on which the hard disk device is disposed, as shown in fig. 7, the apparatus comprising:

A first receiving module 702, configured to receive a hard disk monitoring request initiated by the server, where the hard disk monitoring request is used to request to control a monitoring process of a target hard disk device in hard disk devices deployed on the server;

The identifying module 704 is configured to respond to the hard disk monitoring request, and identify, from a plurality of operation scenarios, a target operation scenario corresponding to the target hard disk device according to target operation data of the target hard disk device, where the target operation data includes a monitoring value under each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scenario is used to indicate a performance requirement of a service operated by the hard disk device on the hard disk device;

A creating module 706, configured to create a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data, and a monitoring type to which the target monitoring item belongs, where the target monitoring list records the target monitoring item and a target monitoring mode that have a corresponding relationship;

And a first sending module 708, configured to send the target monitoring list to the server, where the server is configured to monitor, according to the target monitoring list, operation of the target hard disk device.

According to the embodiment of the application, according to the operation data of the hard disk device, the operation scene corresponding to the hard disk device is identified from a plurality of operation scenes, the operation scene can be used for indicating the performance requirement of the operation of the hard disk device on the hard disk device, and the monitoring list is created according to the monitoring item corresponding to the operation scene, the operation data corresponding to the monitoring item in the operation data and the monitoring type to which the monitoring item belongs, wherein the monitoring item and the monitoring mode with the corresponding relation are recorded in the monitoring list, and it can be understood that the monitoring item and the monitoring mode in the monitoring list can be changed under the condition that the performance requirements of the hard disk device are different, by adopting the mode, the monitoring list suitable for the performance requirement of the hard disk device is created according to the identified performance requirement of the operation of the hard disk device on the hard disk device, so that the problem that the monitoring control efficiency of the hard disk device is lower can be solved, and the monitoring control efficiency effect of the hard disk device is improved can be achieved.

In one exemplary embodiment, the identification module includes:

The computing unit is used for computing the probability that the target hard disk device corresponds to each of the plurality of operation scenes according to the target operation data;

And the screening unit is used for screening N operation scenes, corresponding to which the probability meets the target screening condition, from the operation scenes according to the operation scenes and the probabilities with the corresponding relation, wherein N is a positive integer.

In an exemplary embodiment, the screening unit is configured to:

extracting operation scenes from the operation scenes one by one according to the corresponding probability from high to low;

Detecting the relation between the probability corresponding to the current extracted operation scene, the probability corresponding to the extracted operation scene and the target threshold value;

determining the current extracted operation scene and the extracted operation scene as the N operation scenes as the target operation scenes under the condition that the probability sum is greater than or equal to the target threshold value;

and under the condition that the sum of the probabilities is smaller than the target threshold value, continuously extracting the next operation scene from high to low according to the corresponding probability.

In one exemplary embodiment, the identification module includes:

The input unit is used for inputting the target operation data into a target scene recognition model, wherein the target scene recognition model is obtained by training an initial scene recognition model by using an operation data sample marked with an operation scene label;

and the acquisition unit is used for acquiring the target operation scene output by the target scene identification model.

In one exemplary embodiment, the creation module includes:

The creating unit is used for creating an initial monitoring list by using the target monitoring items corresponding to the target operation scene;

The first allocation unit is configured to allocate the target monitoring mode to the target monitoring item according to the monitoring type to which the target monitoring item belongs, so as to obtain a reference monitoring list, where the target monitoring mode includes: a mode of monitoring according to the change of the data and a mode of monitoring according to the threshold value of the data;

and the second allocation unit is used for allocating a target monitoring threshold value to the target monitoring item according to the reference operation data under the condition that the target monitoring mode is a mode of monitoring according to the threshold value of the data, so as to obtain the target monitoring list.

In an exemplary embodiment, the second allocation unit is configured to:

the method comprises the steps that an ith monitoring threshold value is distributed to an ith monitoring item according to ith operation data, wherein the reference operation data comprise M pieces of operation data, the M pieces of operation data comprise the ith operation data, the target monitoring item comprises M monitoring items, the M monitoring items comprise the ith monitoring item, the target monitoring threshold value comprises M monitoring threshold values, the M monitoring threshold values comprise the ith monitoring threshold value, i is a positive integer smaller than or equal to M, and M is a positive integer;

Selecting the ith candidate operation data used for representing that the target hard disk device is in a normal operation state from the ith operation data;

performing average value operation on the ith candidate running data to obtain an average value, and determining the standard deviation of the ith candidate running data;

determining a first candidate upper threshold as a sum of the average value and the standard deviation of the target number, and determining a first candidate lower threshold as a value obtained by subtracting the standard deviation of the target number from the average value;

Determining a larger upper limit threshold from a second candidate upper limit threshold and the first candidate upper limit threshold to obtain a target upper limit threshold, and determining a smaller lower limit threshold from a second candidate lower limit threshold and the first candidate lower limit threshold to obtain a target lower limit threshold, wherein the target monitoring threshold comprises the target upper limit threshold and the target lower limit threshold, the second candidate upper limit threshold is the data with the largest value in the i candidate operation data, and the second candidate lower limit threshold is the data with the smallest value in the i candidate operation data.

In one exemplary embodiment, the apparatus further comprises:

the second sending module is used for sending a data reporting instruction to the server before the target monitoring list is sent to the server, wherein the data reporting instruction is used for indicating the server to report alarm data belonging to the target operation scene to the monitoring controller;

The second receiving module is used for receiving target alarm data reported by the server in response to the data reporting instruction;

The extraction module is used for extracting keywords in each alarm data in the P alarm data to obtain P groups of target keywords when the target alarm data comprise P alarm data and the target hard disk equipment comprises T hard disk equipment, wherein a w-th group of target keywords in the P groups of target keywords comprise an identifier of a w-th hard disk equipment in the T hard disk equipment, a deployed position of the w-th hard disk equipment in the server and an equipment type of the w-th hard disk equipment, P is a positive integer, T is a positive integer, and w is a positive integer smaller than or equal to P;

The first execution module is used for executing vectorization operation on the P groups of target keywords to obtain P groups of word vectors;

A first determining module, configured to determine a similarity between an s ₁ th word vector in the s th word vector in the P-th word vectors and an r ₁ th word vector in each of the P-1 th word vectors except for the s-th word vector, obtain an s ₁ th similarity, determine a similarity between an s ₂ th word vector in the s-th word vector and an r ₂ th word vector in each of the P-1 th word vectors, obtain an s ₂ th similarity, and determining the similarity between the (s ₃) th word vector in the(s) th word vector and the (r ₃) th word vector in each group of word vectors in the P-1 group of word vectors to obtain the(s) ₃ th group of similarity, wherein the (s ₁) th word vector is a word vector used for representing the identifier of the(s) th hard disk device in the(s) th group of word vectors, the (s ₂) th word vector is a word vector used for representing the position of the(s) th hard disk device deployed in a server in the(s) th group of word vectors, the (s ₃) th word vector is a word vector used for representing the device type of the(s) th hard disk device in the(s) th group of word vectors, the r ₁ th word vector is a word vector used for representing the identification of the hard disk device in each group of word vectors in the P-l group of word vectors, the r ₂ th word vector is a word vector used for representing the position of the hard disk device deployed in the server in each group of word vectors in the P-1 group of word vectors, the r ₃ th word vector is a word vector used for representing the device type of the hard disk device in each group of word vectors in the P-1 group of word vectors, the T hard disk devices comprise the s hard disk device, s is a positive integer less than or equal to P, s ₁、s₂, And s ₃ are positive integers, r ₁、r₂ and r ₃ are positive integers;

The second execution module is used for executing average value operation on the s ₁ th group of similarity, the s ₂ th group of similarity and the s ₃ th group of similarity to obtain an s average similarity;

The second determining module is used for determining the s-th alarm data corresponding to the s-th word vector in the P alarm data as a target alarm monitoring item under the condition that the s-th average similarity is greater than or equal to a preset average similarity threshold value;

and the adding module is used for adding the target alarm monitoring item to the target monitoring list.

It should be noted that each of the above modules may be implemented by software or hardware, and for the latter, it may be implemented by, but not limited to: the modules are all located in the same processor; or the above modules may be located in different processors in any combination.

Embodiments of the present application also provide a computer readable storage medium having a computer program stored therein, wherein the computer program is arranged to perform the steps of any of the method embodiments described above when run.

In one exemplary embodiment, the computer readable storage medium may include, but is not limited to: a usb disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory RAM), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing a computer program.

An embodiment of the application also provides an electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.

In an exemplary embodiment, the electronic device may further include a transmission device connected to the processor, and an input/output device connected to the processor.

Embodiments of the application also provide a computer program product comprising a computer program which, when executed by a processor, implements the steps of any of the method embodiments described above.

Embodiments of the present application also provide another computer program product comprising a non-volatile computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of any of the method embodiments described above.

Embodiments of the present application also provide a computer program comprising computer instructions stored in a computer-readable storage medium; the processor of the computer device reads the computer instructions from the computer readable storage medium and executes the computer instructions to cause the computer device to perform the steps of any of the method embodiments described above.

Specific examples in this embodiment may refer to the examples described in the foregoing embodiments and the exemplary implementation, and this embodiment is not described herein.

It will be appreciated by those skilled in the art that the modules or steps of the application described above may be implemented in a general purpose computing device, they may be concentrated on a single computing device, or distributed across a network of computing devices, they may be implemented in program code executable by computing devices, so that they may be stored in a storage device for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than that shown or described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple modules or steps of them may be fabricated into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

The above description is only of the preferred embodiments of the present application and is not intended to limit the present application, but various modifications and variations can be made to the present application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the principle of the present application should be included in the protection scope of the present application.

Claims

1. A monitoring control method of hard disk equipment is characterized in that,

The method is applied to a monitoring controller, the monitoring controller is connected with a server, hard disk equipment is deployed on the server, and the method comprises the following steps:

Receiving a hard disk monitoring request initiated by the server, wherein the hard disk monitoring request is used for requesting to control a monitoring process of a target hard disk device in the hard disk devices deployed on the server;

Responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device, wherein the target operation data comprises monitoring values of each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scenes are used for indicating the performance requirement of the service operated by the hard disk device on the hard disk device;

Creating a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data and a monitoring type to which the target monitoring item belongs, wherein the target monitoring item and a target monitoring mode with a corresponding relation are recorded in the target monitoring list; and sending the target monitoring list to the server, wherein the server is used for monitoring the operation of the target hard disk device according to the target monitoring list.

2. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The identifying, according to the target operation data of the target hard disk device, a target operation scene corresponding to the target hard disk device from a plurality of operation scenes includes:

According to the target operation data, calculating the probability that the target hard disk device corresponds to each of the plurality of operation scenes;

And screening N operation scenes corresponding to the probability meeting the target screening condition from the operation scenes according to the operation scenes and the probability with the corresponding relation, wherein N is a positive integer.

3. The method of claim 2, wherein the step of determining the position of the substrate comprises,

According to the operation scenes and the probabilities with the corresponding relation, N operation scenes with the probabilities meeting the target screening conditions are screened out from the operation scenes to serve as the target operation scenes, and the method comprises the following steps:

4. The method of claim 1, wherein the step of determining the position of the substrate comprises,

Inputting the target operation data into a target scene recognition model, wherein the target scene recognition model is obtained by training an initial scene recognition model by using an operation data sample marked with an operation scene label;

and acquiring the target operation scene output by the target scene identification model.

5. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The creating a target monitoring list according to the target monitoring item corresponding to the target operation scene in the monitoring item set, the reference operation data corresponding to the target monitoring item in the target operation data, and the monitoring type to which the target monitoring item belongs, includes:

Creating an initial monitoring list by using a target monitoring item corresponding to the target operation scene;

According to the monitoring type of the target monitoring item, the target monitoring mode is allocated to the target monitoring item to obtain a reference monitoring list, wherein the target monitoring mode comprises the following steps: a mode of monitoring according to the change of the data and a mode of monitoring according to the threshold value of the data;

And under the condition that the target monitoring mode is a mode of monitoring according to the threshold value of the data, distributing a target monitoring threshold value for the target monitoring item according to the reference operation data to obtain the target monitoring list.

6. The method of claim 5, wherein the step of determining the position of the probe is performed,

The allocating a target monitoring threshold value for the target monitoring item according to the reference operation data comprises the following steps:

7. The method of claim 1, wherein the step of determining the position of the substrate comprises,

Before said sending the target monitor list to the server, the method further comprises:

transmitting a data reporting instruction to the server, wherein the data reporting instruction is used for instructing the server to report alarm data belonging to the target operation scene to the monitoring controller;

receiving target alarm data reported by the server in response to the data reporting instruction;

Extracting keywords in each alarm data in the P alarm data to obtain P groups of target keywords under the condition that the target alarm data comprises P alarm data and the target hard disk device comprises T hard disk devices, wherein a w-th group of target keywords in the P groups of target keywords comprise identifications of w-th hard disk devices in the T hard disk devices, positions of the w-th hard disk devices deployed in the server and device types of the w-th hard disk devices, P is a positive integer, T is a positive integer, and w is a positive integer smaller than or equal to P;

Vectorizing the P groups of target keywords to obtain P groups of word vectors;

Determining the similarity between the s ₁ th word vector in the s th word vector in the P group word vectors and the r ₁ th word vector in each group word vector in the P-1 group word vectors except the s th word vector to obtain the s ₁ th similarity, determining the similarity between the s ₂ th word vector in the s th group word vector and the r ₂ th word vector in each group word vector in the P-1 group word vectors to obtain the s ₂ th similarity, determining the similarity between the s ₃ th word vector in the s group word vector and the r ₃ th word vector in each group word vector in the P-1 group word vector to obtain the s ₃ th similarity,

The s ₁ th word vector is a word vector used for representing an identifier of an s hard disk device in the s th word vector, the s ₂ th word vector is a word vector used for representing a location of the s hard disk device deployed in a server in the s th word vector, the s ₃ th word vector is a word vector used for representing a device type of the s hard disk device in the s th word vector, the r ₁ th word vector is a word vector used for representing an identifier of a hard disk device in each word vector in the P-1 word vector, the r ₂ th word vector is a word vector used for representing a location of a hard disk device deployed in the server in each word vector in the P-1 word vector, the r ₃ th word vector is a hard disk vector used for representing a device type of a device in each word vector in the P-1 word vector, the T hard disk device comprises the s hard disk device, the T hard disk device comprises a positive integer of P ₁、s₂ s, a positive integer of P ₁、s₂ s and a positive integer of r 35, and a positive integer of each integer of P-35 s;

Performing average value operation on the s ₁ th group of similarity, the s ₂ th group of similarity and the s ₃ th group of similarity to obtain an s average similarity;

under the condition that the s-th average similarity is larger than or equal to a preset average similarity threshold value, determining s-th alarm data corresponding to the s-th group word vector in the P alarm data as a target alarm monitoring item;

and adding the target alarm monitoring item to the target monitoring list.

8. A monitoring control device of hard disk equipment is characterized in that,

The device is applied to a monitoring controller, the monitoring controller is connected with a server, hard disk equipment is deployed on the server, and the device comprises:

The first receiving module is used for receiving a hard disk monitoring request initiated by the server, wherein the hard disk monitoring request is used for requesting to control the monitoring process of a target hard disk device in the hard disk devices deployed on the server;

The identification module is used for responding to the hard disk monitoring request, and identifying a target operation scene corresponding to the target hard disk device from a plurality of operation scenes according to target operation data of the target hard disk device, wherein the target operation data comprises monitoring values of each monitoring item in a monitoring item set generated when the target hard disk device operates on the server, and the operation scene is used for indicating the performance requirement of the service operated by the hard disk device on the hard disk device;

The creating module is used for creating a target monitoring list according to a target monitoring item corresponding to the target operation scene in the monitoring item set, reference operation data corresponding to the target monitoring item in the target operation data and a monitoring type to which the target monitoring item belongs, wherein the target monitoring item and a target monitoring mode with a corresponding relation are recorded in the target monitoring list;

the first sending module is used for sending the target monitoring list to the server, wherein the server is used for monitoring the operation of the target hard disk device according to the target monitoring list.

9. A computer-readable storage medium comprising,

The computer readable storage medium has stored therein a computer program, wherein the computer program when executed by a processor realizes the steps of the method as claimed in any of claims 1 to 7.

10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that,

The processor, when executing the computer program, implements the steps of the method as claimed in any one of claims 1 to 7.