Data dynamic migration storage method and device, equipment and medium

Abstract

The invention discloses a data dynamic migration and storage method, device and equipment and a medium, and relates to the technical field of computers. The method comprises the steps that environment parameters of a current mechanical hard disk in a distributed cluster are collected; wherein the environment parameters comprise a vibration parameter, a temperature parameter and a noise parameter; quantifying the input and output delay performance of the current mechanical hard disk by using the environmental parameters of the current mechanical hard disk; if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold value, searching a target mechanical hard disk combination in the distributed cluster; and migrating and storing the target data block of the current mechanical hard disk into the target mechanical hard disk combination based on an erasure code and a multi-copy mechanism. Through the scheme, the reliability of storing the data in the mechanical hard disk in the distributed cluster is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment and medium for dynamic data migration and storage. Background Art

[0002] With the explosive growth of data center storage demand, hard disk drives (HDDs) remain the mainstream storage medium due to their high capacity and low cost. However, they are sensitive to the physical environment, and their performance is significantly affected by fan vibration, temperature fluctuations, and external environmental interference. When system fan speeds exceed 10,000 RPM, high-frequency vibrations above 20kHz can cause head positioning jitter, leading to performance degradation. Bit error rates increase exponentially when disk cavity temperatures exceed 50°C, and surface static electricity can damage printed circuit boards (PCBs) below 30°C.

[0003] Although existing distributed storage systems support data redundancy and load balancing, they lack the ability to dynamically respond to the physical environment. The data sharding strategy is based on the initial state of the hardware and cannot be adjusted according to real-time environmental changes. It only monitors the health status of the HDD in a single dimension, ignoring the coupling effect of environmental factors on performance.

[0004] It can be seen that how to improve the reliability of data stored on mechanical hard disks in a distributed cluster is a problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of the embodiments of the present invention is to provide a method, apparatus, device, and medium for dynamic data migration and storage to improve the reliability of data stored on mechanical hard disks in a distributed cluster. The specific solution is as follows:

[0006] In a first aspect, the present invention discloses a method for dynamic data migration and storage, comprising:

[0007] Collect environmental parameters of the current mechanical hard disks in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters, and noise parameters;

[0008] quantifying the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk;

[0009] If the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, searching for a target mechanical hard disk combination in the distributed cluster;

[0010] The target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination based on the erasure code and multiple copy mechanism.

[0011] Optionally, the quantifying the input and output delay performance of the current mechanical hard disk by using the environmental parameters of the current mechanical hard disk includes:

[0012] Using the vibration parameter, the temperature parameter, and the noise parameter, respectively, to obtain a first negative impact degree in the vibration dimension, a second negative impact degree in the temperature dimension, and a third negative impact degree in the noise dimension;

[0013] A weighted sum of the first negative impact degree, the second negative impact degree, and the third negative impact degree is determined as the input and output delay performance of the current mechanical hard disk.

[0014] Optionally, the using the vibration parameter, the temperature parameter, and the noise parameter to respectively obtain a first negative impact degree in a vibration dimension, a second negative impact degree in a temperature dimension, and a third negative impact degree in a noise dimension includes:

[0015] Performing a fast Fourier transform on the vibration parameter to obtain a frequency domain signal; wherein the vibration parameter is a vibration acceleration time domain signal collected by a vibration sensor of the current mechanical hard disk;

[0016] Obtaining a vibration energy distribution of the frequency domain signal within a unit frequency using a periodogram estimation algorithm, and determining the vibration energy distribution within the unit frequency as a vibration power spectral density;

[0017] Obtaining the resonance sensitivity of the current mechanical hard disk using the vibration parameter;

[0018] summing the vibration power spectrum density and the resonance sensitivity within a preset frequency range to obtain a first negative impact degree of the vibration dimension;

[0019] Obtaining a magnitude of temperature deviation from the reference value based on a difference between the temperature parameter and a preset temperature reference value, and determining the magnitude of temperature deviation from the reference value as a second negative impact degree of the temperature dimension; wherein the temperature parameter is a disk cavity temperature collected by a temperature sensor of the current mechanical hard disk;

[0020] Obtaining a target fan speed corresponding to the noise parameter; wherein the target fan speed is the speed of a fan in the storage server to which the current mechanical hard disk belongs;

[0021] The turbulence intensity obtained by performing airflow simulation on the target fan speed is determined as the third negative impact degree of the noise dimension.

[0022] Optionally, determining the target data block of the current mechanical hard disk includes:

[0023] Determining the access frequency of each data block in the current mechanical hard disk, and determining the data blocks with access frequencies higher than a first preset access frequency threshold as hot data blocks;

[0024] The hot data block is determined as the target data block of the current mechanical hard disk.

[0025] Optionally, the migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism includes:

[0026] Splitting the target data block of the current mechanical hard disk into target sub-data blocks, and generating a check block corresponding to the target data block by an erasure coding algorithm to obtain a coding block including the target sub-data block and the check block;

[0027] Migrating and storing the coding block to each mechanical hard disk in the target mechanical hard disk combination;

[0028] Filtering out a target coding block having an access frequency higher than a second preset access frequency threshold from each of the coding blocks, and generating a copy of the target coding block;

[0029] A copy of the target coding block is stored in a target mechanical hard disk in the target mechanical hard disk combination.

[0030] Optionally, the target mechanical hard disk combination includes each mechanical hard disk in the distributed cluster whose input and output delay performance is less than the preset threshold, the input and output delay performance of the target mechanical hard disk meets the preset performance optimal level condition, and the second preset access frequency threshold is greater than the first preset access frequency threshold.

[0031] Optionally, after migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism, the method further includes:

[0032] A mapping relationship is established among the target data block, the coding block, the copy of the target coding block, and the target mechanical hard disk, so as to read and / or migrate the target data block based on the mapping relationship.

[0033] In a second aspect, the present invention discloses a data dynamic migration storage device, comprising:

[0034] An environmental parameter acquisition module is used to collect environmental parameters of the current mechanical hard disk in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters and noise parameters;

[0035] a performance quantification module, configured to quantify the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk;

[0036] a hard disk search module, configured to search for a target mechanical hard disk combination in the distributed cluster if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold;

[0037] The migration storage module is used to migrate and store the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multiple copy mechanism.

[0038] In a third aspect, the present invention discloses an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] The processor is used to execute a computer program to implement the steps of the aforementioned disclosed method for dynamic data migration and storage.

[0041] In a fourth aspect, the present invention discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned disclosed method for dynamic data migration and storage are implemented.

[0042] It can be seen that the present invention collects the environmental parameters of the current mechanical hard disk in the distributed cluster; wherein, the environmental parameters include vibration parameters, temperature parameters and noise parameters; the environmental parameters of the current mechanical hard disk are used to quantify the input and output delay performance of the current mechanical hard disk; if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, the target mechanical hard disk combination is searched in the distributed cluster; based on the erasure code and multi-copy mechanism, the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination.

[0043] The beneficial effect is: the present invention collects the vibration parameters, temperature parameters and noise parameters of the current mechanical hard disk in the distributed cluster, that is, it collects multi-dimensional physical environment parameters, and uses these environmental parameters to quantify the input and output delay performance of the current mechanical hard disk, that is, to determine the input and output delay performance of the current mechanical hard disk under the current physical environment. If the input and output delay performance is greater than the preset threshold, it means that the current physical environment has a negative impact on the current mechanical hard disk and the impact is not small, so the input and output delay performance of the current mechanical hard disk is high, that is, the stability is poor, and it is not suitable to continue to store important data, so the target mechanical hard disk combination is searched in the distributed cluster, so that the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination, that is, the data is stored in a mechanical hard disk with better performance, and the migration storage based on the erasure code and multi-copy mechanism can make the data in the target data block more reliably stored in the distributed cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0045] Figure 1 A flow chart of a data dynamic migration and storage method provided by an embodiment of the present invention;

[0046] Figure 2 A specific data storage framework diagram provided by an embodiment of the present invention;

[0047] Figure 3 A flow chart of a specific data dynamic migration and storage method provided by an embodiment of the present invention;

[0048] Figure 4 A schematic diagram of dynamic data migration and storage for multi-dimensional environment awareness provided by an embodiment of the present invention;

[0049] Figure 5 A schematic diagram of a specific distributed data dynamic migration and storage based on multi-dimensional environment perception provided by an embodiment of the present invention;

[0050] Figure 6 A schematic structural diagram of a data dynamic migration storage device provided by an embodiment of the present invention;

[0051] Figure 7 A structural diagram of an electronic device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0053] With the explosive growth of data center storage demand, mechanical hard drives remain the mainstream storage medium due to their high capacity and low cost. However, they are sensitive to the physical environment, and their performance is significantly affected by fan vibration, temperature fluctuations, and external environmental interference. When the system fan speed exceeds 10,000 RPM, high-frequency vibrations above 20kHz can cause head positioning jitter, resulting in performance degradation. When the disk cavity temperature exceeds 50°C, the bit error rate increases exponentially, and below 30°C, surface static electricity can damage the PCB.

[0054] Although existing distributed storage systems support data redundancy and load balancing, they lack the ability to dynamically respond to the physical environment. The data sharding strategy is based on the initial state of the hardware and cannot be adjusted according to real-time environmental changes. It only monitors the health status of the HDD in a single dimension, ignoring the coupling effect of environmental factors on performance.

[0055] The terms "including" and "having," as used in the present description and accompanying drawings, and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements and may include steps or elements that are not listed.

[0056] In order to enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0057] Next, a data dynamic migration storage solution provided by an embodiment of the present invention is introduced in detail. Figure 1 A data dynamic migration storage method provided by an embodiment of the present invention includes:

[0058] Step S11: collecting environmental parameters of the current mechanical hard disks in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters, and noise parameters.

[0059] HDDs are sensitive to the physical environment, and their performance is significantly affected by fan vibration, temperature fluctuations, and external environmental interference. In the design of storage services, HDDs are often physically placed near the fan or at the back end of the machine. The actual working environment of the HDD often faces fan vibration, high temperature, and fine dust particles. Regarding fan vibration: When the system fan speed exceeds 10,000 RPM, the high-frequency vibration above 20kHz generated will cause the head positioning error rate (PER) to exceed the standard, resulting in an average of 1.2-3.5 retry anomalies per thousand I / O operations, causing IOPS performance to decrease by 10%-30%; Regarding temperature influence: The HDD operating temperature range has a strong nonlinear characteristic. When the disk cavity temperature exceeds 50°C, the lubricant viscosity decreases at a rate of 40%, and the disk's thermal expansion coefficient (Coefficient of thermal expansion) increases. The 3D scanning electron microscope (SEM) shows that the magnetic field strength (MSE) of the magnetic head and slider is 100 nm, which is equivalent to the width of the magnetic field. The 3D scanning electron microscope (SEM) shows that the 3D scanning electron microscope (SEM) ...

[0060] Considering that HDDs in storage servers are significantly affected by the machine's fan noise and ambient operating temperature, typical server designs incorporate HDD vibration suppression technologies. For example, these technologies utilize an acoustic proxy (HDD AcousticSurrogate) to quantify the vibration spectrum and guide chassis design (e.g., honeycomb waveguides and sound-absorbing materials, commonly used in storage servers). These technologies enhance the vibration damping capabilities of the HDD brackets and employ air bearings or rubber pads to isolate vibrations. However, these technologies cannot eliminate the resonant effects of high-frequency sound waves on the disk heads. Furthermore, due to storage server density requirements, inherent structural design constraints, and space limitations, server designers often balance the impact on HDDs with the need for higher density designs. Sometimes, they even consider sacrificing the performance of a few HDDs in specific scenarios (such as high temperatures) to meet the overall storage server specifications. Once the storage server specifications are confirmed, the corresponding fan quantity, specifications, and other parameters are also locked in. Storage server cooling strategies primarily utilize dynamic fan speed control, adjusting speed based on temperature thresholds. However, high loads can still trigger vibration thresholds, impacting the performance and reliability of some HDDs under high loads and abnormal ambient temperatures.

[0061] In this embodiment, sensors are arranged in the storage server where the mechanical hard disk is located, and the sensors are used to collect the environmental parameters of the current mechanical hard disk in the distributed cluster. Specifically, the environmental parameters include vibration parameters, temperature parameters and noise parameters, that is, multi-dimensional environmental parameters are collected.

[0062] Step S12: quantifying the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk.

[0063] Changes in environmental parameters can also cause changes in the input and output latency performance of a mechanical hard drive. Therefore, it is necessary to use the current environmental parameters of the mechanical hard drive to quantify the input and output latency performance of the current mechanical hard drive. That is, when the environmental parameters of the current mechanical hard drive are determined, the input and output latency performance of the current mechanical hard drive can be accurately determined. In this way, the input and output latency performance of the current mechanical hard drive can be more intuitively understood, thereby providing strong support for whether subsequent data migration is needed.

[0064] Excessive fan vibration, high temperature, and excessive noise in the server will have a negative impact on input and output performance, causing the time required for the computer system to complete input / output operations (such as reading or writing data from the hard disk) to increase from issuing a request to obtaining a response. Therefore, it is necessary to quantify the degree of negative impact of different vibration parameters, temperature parameters, and noise parameters on input and output performance. Specifically, the current environmental parameters are first used to obtain the first negative impact degree of the vibration dimension, the second negative impact degree of the temperature dimension, and the third negative impact degree of the noise dimension. Next, the weighted sum of the first negative impact degree, the second negative impact degree, and the third negative impact degree is determined as the input and output delay performance of the current mechanical hard disk.

[0065] Step S13: If the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, searching for a target mechanical hard disk combination in the distributed cluster.

[0066] When the input / output latency of a mechanical hard drive exceeds a preset threshold, it indicates that the drive's current operating environment (such as vibration, temperature, noise, etc.) has significantly negatively impacted its performance, potentially leading to problems such as head positioning jitter, increased data read errors, and frequent retries, thereby reducing the overall system's I / O efficiency and reliability. Without data migration, high-latency hard drives can become system bottlenecks, slowing data access and potentially even causing hard drive failures due to persistently harsh environments, increasing the risk of data loss. Therefore, it's necessary to identify a target mechanical hard drive combination for data migration to improve the data storage environment, ensure input / output performance for the data, and improve data input / output efficiency.

[0067] In this embodiment, the target mechanical hard disk combination includes mechanical hard disks in the distributed cluster whose input and output delay performance is less than the preset threshold.

[0068] It can be understood that the target mechanical hard disk combination is used to store the data blocks of the current mechanical hard disk, and the target mechanical hard disk combination includes multiple mechanical hard disks. These mechanical hard disks in the target mechanical hard disk combination are mechanical hard disks in the distributed cluster whose input and output delay performance is less than a preset threshold. In other words, the storage environment of these mechanical hard disks in the target mechanical hard disk combination is better than the current mechanical hard disk, and data blocks stored in the mechanical hard disks in the target mechanical hard disk combination are more reliable than those stored in the current mechanical hard disk, and the input and output efficiency is also higher.

[0069] Step S14: migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism.

[0070] In this embodiment, determining the target data block of the current mechanical hard disk includes: determining the access frequency of each data block in the current mechanical hard disk, and determining the data block whose access frequency is higher than a first preset access frequency threshold as a hot data block; determining the hot data block as the target data block of the current mechanical hard disk.

[0071] It is understandable that the current mechanical hard disk includes multiple data blocks for storing data. The data utilization rate of some data blocks is very low, while the data utilization rate of some data blocks is very high. Therefore, some data blocks will be frequently accessed, while the access frequency of some data blocks is very low. The access frequency of each data block in the current mechanical hard disk is determined, and the data blocks with access frequency higher than the first preset access frequency threshold are determined as hot data blocks, because hot data blocks have the most significant impact on the overall performance of the system, and their I / O delay directly determines the user experience and the response speed of key services. In harsh environments (such as high temperature and vibration), the continuous reading and writing of hot data will be aggravated. Hard disk load leads to problems such as head jitter and a surge in retry operations, further amplifying delays and increasing the risk of failure. Therefore, the hot data block is determined as the target data block of the current mechanical hard disk, and then the target data block is migrated and stored in the target mechanical hard disk combination based on the erasure code and multi-copy mechanism. By migrating hot data to a stable environment, not only can the number of high-latency operations be significantly reduced, but also data errors or hardware losses caused by environmental interference can be avoided. At the same time, hard disks in a stable environment have higher effective bandwidth and lower error rate, which can better support high-load access requirements, while cold data is not sensitive to delay fluctuations and can be temporarily retained on the original hard disk.

[0072] In this embodiment, the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination based on the erasure code and multiple copies mechanism, including: dividing the target data block of the current mechanical hard disk into target sub-data blocks, and generating a check block corresponding to the target data block through an erasure code algorithm to obtain a coding block containing the target sub-data block and the check block; migrating and storing the coding block to each mechanical hard disk in the target mechanical hard disk combination; screening out a target coding block with an access frequency higher than a second preset access frequency threshold from each coding block, and generating a copy of the target coding block; storing the copy of the target coding block to the target mechanical hard disk in the target mechanical hard disk combination.

[0073] Based on the erasure code and multi-copy mechanism, the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination. First, data segmentation and erasure code encoding are performed: the target data block to be migrated in the current mechanical hard disk is divided into multiple target sub-data blocks, and then the corresponding check blocks are generated by the erasure code (EC) algorithm. The erasure code technology divides the original data into several data blocks and check blocks through mathematical coding to ensure that even if some data blocks are lost or damaged, the original data can still be completely restored through the remaining data blocks and check blocks. For example, a 4+2 erasure code strategy (4 data blocks + 2 check blocks) is adopted to achieve data reconstruction when any 2 blocks of data are lost. This step not only improves storage efficiency (saves space compared to the pure copy mechanism), but also provides redundancy for data migration. Next, code block distribution and storage optimization are performed: the generated code blocks (including target sub-data blocks and check blocks) are evenly distributed to different hard disks in the target mechanical hard disk combination. This distribution strategy combines load balancing and fault tolerance requirements. On the one hand, it prevents a single hard disk from becoming a performance bottleneck due to centralized storage. On the other hand, it utilizes the redundant characteristics of erasure codes to ensure that data can still be recovered even if some hard disks fail. At the same time, the access frequency of each code block is monitored in real time, and the target code blocks with high-frequency access (i.e., hot data with an access frequency exceeding a second preset threshold) are screened out to provide a basis for subsequent replica generation. Finally, hot data copies are generated and stored: For target coding blocks that are accessed frequently, the system dynamically generates multiple copies and stores them on hard drives with better performance or more stable environments in the target mechanical hard drive combination, that is, the target mechanical hard drives. For example, hot data copies are preferentially stored on hard drives in low-vibration, low-temperature areas to reduce latency and improve response speed. This hybrid strategy (erasure coding + multiple copies) retains the storage efficiency advantage of erasure coding and provides additional performance redundancy for hot data through the multiple copy mechanism, thereby achieving a balance between cost, reliability and performance. It is particularly suitable for the efficient management of sensitive data in distributed storage.

[0074] For example Figure 2The following diagram shows a specific data flushing framework. In distributed storage systems, the data flushing strategy is a core mechanism that determines data distribution, redundancy, and reliability, and must balance performance, disaster recovery, and scalability. When writing distributed storage data to HDDs, data is split into fixed-size shards (e.g., 128MB) and mapped to different nodes using a hashing algorithm (e.g., consistent hashing) to achieve load balancing. For example, files are distributed to multiple nodes based on their hash values ​​to avoid excessive pressure on a single point. Multiple replicas (usually three) are generated for each shard and stored on different physical nodes or racks to ensure data accessibility in the event of a node failure. Combining a replication strategy with erasure coding (e.g., 4+2) can reduce storage overhead. This principle applies when flushing data to disk: data is evenly distributed across nodes based on the number of nodes, and then randomly flushed to different HDDs. Because the volume of data written is large enough, it can be assumed that the data is flushed to the HDDs randomly after being sharded.

[0075] In this embodiment, the input and output delay performance of the target mechanical hard disk meets the preset optimal performance level condition, and the second preset access frequency threshold is greater than the first preset access frequency threshold.

[0076] It should be noted that the second preset access frequency threshold is greater than the first preset access frequency threshold. For example, the target data blocks are target data block A and target data block B, and the target data block A and the target data block B are respectively divided into two sub-data blocks to obtain sub-data block A1, sub-data block A2 and sub-data block B1, sub-data block B2). Next, the four sub-data blocks are calculated using an erasure code algorithm (such as Reed-Solomon) to generate two check blocks (P1, P2) respectively. In this way, six coding blocks are obtained, and the six coding blocks are A1, A2, B1, B2, P1, and P2 respectively. The coding blocks are migrated and stored in each mechanical hard disk in the target mechanical hard disk combination; further, the coding blocks with an access frequency higher than the second preset access frequency threshold are screened out. A target coding block is stored on the target mechanical hard drive in the target mechanical hard drive combination, and a second preset access frequency threshold is set to be greater than the first preset access frequency threshold. The input and output latency performance of the target mechanical hard drive meets the preset optimal performance level conditions. In this way, hierarchical screening of hot data and precise resource allocation are achieved, thereby optimizing system performance and storage efficiency. Specifically, the first preset access frequency threshold (lower) is used to initially identify all hot data that needs to be migrated (e.g., access frequency > 100 times / second), ensuring that it is migrated from high-latency hard drives to a stable environment; while the higher second preset access frequency threshold (e.g., > 500 times / second) further filters out extremely high-frequency access data, which is extremely sensitive to latency and requires additional copies to ensure performance. This tiered design avoids the storage waste caused by blindly generating copies of all hot data. Instead, a copy strategy is implemented only for the most critical data (such as hot spots under bursty traffic), controlling costs while ensuring responsiveness in high-load scenarios. At the same time, the first preset access frequency threshold mainly responds to data migration needs triggered by environmental deterioration (such as vibration or excessive temperature) and belongs to the basic fault tolerance layer; the second preset access frequency threshold is aimed at the real-time performance requirements of business peaks, and provides elastic support through dynamic replicas. The two work together to form a complete optimization chain from basic protection to performance enhancement, and ultimately achieve a refined balance between reliability, latency and storage overhead.

[0077] In this embodiment, after migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism, it also includes: establishing a mapping relationship between the target data block, the coding block, the copy of the target coding block and the target mechanical hard disk, so as to read and / or migrate the target data block based on the mapping relationship.

[0078] After migrating and storing the target data blocks to the target mechanical hard disk combination based on erasure coding and multi-copy mechanisms, establishing and maintaining the mapping relationship between the target data blocks, coding blocks (including sub-data blocks and check blocks), target coding block copies, and the target mechanical hard disk is a core technical link to ensure efficient data management and reliable access. This mapping relationship is implemented through a metadata table or distributed index, which records the original logical address of each target data block, the corresponding coding block distribution location (specifically, the specific hard disk number where the sub-data blocks and check blocks are stored), the storage node of the copy, and the physical topology information of the target mechanical hard disk (specifically, the environmental status and load conditions of the target mechanical hard disk). In other words, the mapping relationship includes the original logical address of the target data block, the corresponding coding block distribution location, the storage node of the copy, and the environmental status and load conditions of the target mechanical hard disk. Among them, the coding block distribution location includes the specific hard disk number where the sub-data blocks and check blocks are stored.

[0079] When the system needs to read data, it can prioritize quickly obtaining it from the hard drive where the low-latency replica is located. If the replica is unavailable, the data is recovered from the distributed encoding blocks through the erasure code decoding process. When data is re-migrated or expanded, the mapping relationship can accurately locate all distributed nodes of the data block, ensuring that the migration process does not affect data integrity and access continuity. The technical advantages of this mechanism are mainly reflected in three aspects: First, through fine-grained mapping management, efficient data positioning is achieved, which significantly reduces access latency and especially optimizes the reading performance of hot data. Second, by combining the redundant characteristics of erasure codes with the parallel access capabilities of multiple replicas, system throughput is improved while ensuring data reliability. Third, dynamic mapping relationship maintenance enables the system to flexibly respond to scenarios such as hard drive failures, load balancing, or environmental changes. Local data adjustments can be completed without global data reorganization, greatly reducing operation and maintenance complexity and improving the elastic scalability of the storage system. Overall, the establishment and maintenance of this mapping relationship is the key technical foundation for distributed storage systems to achieve high performance, high reliability, and adaptive optimization.

[0080] It can be seen that the present invention collects the environmental parameters of the current mechanical hard disk in the distributed cluster; wherein, the environmental parameters include vibration parameters, temperature parameters and noise parameters; the environmental parameters of the current mechanical hard disk are used to quantify the input and output delay performance of the current mechanical hard disk; if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, the target mechanical hard disk combination is searched in the distributed cluster; based on the erasure code and multi-copy mechanism, the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination.

[0081] The beneficial effect is: the present invention collects the vibration parameters, temperature parameters and noise parameters of the current mechanical hard disk in the distributed cluster, that is, it collects multi-dimensional physical environment parameters, and uses these environmental parameters to quantify the input and output delay performance of the current mechanical hard disk, that is, to determine the input and output delay performance of the current mechanical hard disk under the current physical environment. If the input and output delay performance is greater than the preset threshold, it means that the current physical environment has a negative impact on the current mechanical hard disk and the impact is not small, so the input and output delay performance of the current mechanical hard disk is high, that is, the stability is poor, and it is not suitable to continue to store important data, so the target mechanical hard disk combination is searched in the distributed cluster, so that the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination, that is, the data is stored in a mechanical hard disk with better performance, and the migration storage based on the erasure code and multi-copy mechanism can make the data in the target data block more reliably stored in the distributed cluster.

[0082] See also Figure 3 The embodiment of the present invention discloses a specific method for dynamic data migration and storage. Compared with the previous embodiment, this embodiment further illustrates and optimizes the technical solution. It includes:

[0083] Step S21: collecting environmental parameters of the current mechanical hard disks in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters, and noise parameters.

[0084] For example Figure 4 The diagram shows a specific dynamic migration and storage diagram of multi-dimensional environmental perception data. The environmental parameter acquisition module uses the vibration sensor, temperature sensor, and noise sensor in the storage server to collect the vibration parameters, temperature parameters, and noise parameters of the current mechanical hard disk. The noise sensor can be a sound pressure meter, that is, the vibration spectrum (50Hz-20kHz), noise decibel (A-weighted), etc. The temperature of the HDD can be obtained through the sensor inside the HDD. The temperature gradient can be set in the environmental data acquisition module ( Accuracy)

[0085] Step S22: using the vibration parameter, the temperature parameter, and the noise parameter, respectively obtaining a first negative impact degree of the vibration dimension, a second negative impact degree of the temperature dimension, and a third negative impact degree of the noise dimension.

[0086] In this embodiment, the use of the vibration parameters, the temperature parameters and the noise parameters to respectively obtain the first negative influence degree of the vibration dimension, the second negative influence degree of the temperature dimension and the third negative influence degree of the noise dimension includes: performing a fast Fourier transform on the vibration parameters to obtain a frequency domain signal; wherein the vibration parameters are vibration acceleration time domain signals collected by the vibration sensor of the current mechanical hard disk; using a periodogram estimation algorithm to obtain the vibration energy distribution of the frequency domain signal within a unit frequency, and determining the vibration energy distribution within the unit frequency as a vibration power spectrum density; using the vibration parameters to obtain the resonance sensitivity of the current mechanical hard disk; The power spectral density and the resonance sensitivity are summed within a preset frequency range to obtain a first negative impact degree of the vibration dimension; the amplitude of the temperature deviation from the reference value is obtained based on the difference between the temperature parameter and the preset temperature reference value, and the amplitude of the temperature deviation from the reference value is determined as the second negative impact degree of the temperature dimension; wherein, the temperature parameter is the disk cavity temperature collected by the temperature sensor of the current mechanical hard disk; the target fan speed corresponding to the noise parameter is obtained; wherein, the target fan speed is the speed of the fan in the storage server to which the current mechanical hard disk belongs; the turbulence intensity obtained by performing airflow simulation on the target fan speed is determined as the third negative impact degree of the noise dimension.

[0087] The process for obtaining the first negative impact level is as follows: The vibration parameters (specifically, vibration acceleration) collected by the vibration sensor are converted into time-domain signals through a Fast Fourier Transform (FFT) to convert them into frequency-domain signals. The vibration energy distribution of the frequency-domain signal per unit frequency is calculated using a periodogram estimation algorithm to obtain the vibration power spectral density (PSD). The resonance sensitivity function of the current mechanical hard drive is obtained through OCP standard testing. The vibration parameters and the resonance sensitivity function are used to obtain the resonance sensitivity of the current mechanical hard drive. The vibration power spectral density and resonance sensitivity are summed within a preset frequency range (e.g., 80Hz–500Hz) to output the first negative impact level ND1. FFT and PSD analysis decompose the time-domain vibration signal into a frequency-domain energy distribution, avoiding the limitations of a single amplitude metric and better reflecting the sensitivity of the hard drive head to high-frequency vibration. The introduction of the resonance sensitivity function identifies the frequency bands most susceptible to damage to the hard drive (e.g., fan vibration above 20kHz) and allows for targeted optimization of data migration strategies. The formula for obtaining ND1 is:

[0088] ;

[0089] Where, Represents the vibration power spectral density, and its unit is Grms² / Hz, represents the resonance sensitivity function.

[0090] The process of obtaining the second negative impact level is as follows: directly collect the current mechanical hard disk cavity temperature parameters through the temperature sensor. Calculate the difference between the current temperature and the preset reference value (such as 20°C) to obtain the temperature deviation amplitude. , as the second negative impact level ND2. Directly using temperature differences to reflect environmental anomalies eliminates the need for complex calculations and offers high real-time performance. High temperatures (>50°C) and low temperatures (<30°C) correspond to decreased lubricant viscosity and the risk of static electricity accumulation, respectively. The magnitude of the difference indirectly indicates the hazard level of these two extreme conditions.

[0091] The process of obtaining the third degree of negative impact is as follows: historical noise parameters are obtained through the noise sensor, and the historical speed of the fan in the storage server to which the current mechanical hard disk belongs is obtained, thereby obtaining a mapping relationship between the historical noise parameters and the historical fan speed. In this way, when the noise parameters are obtained, the corresponding target fan speed can be obtained. Next, computational fluid dynamics simulation is performed based on the fan speed to derive the airflow turbulence intensity. , using airflow turbulence intensity as the third negative impact level, ND3. Fan speed is strongly correlated with turbulence intensity, which directly affects air disturbance and dust deposition around the hard drive. Simulation results better reflect long-term reliability risks than simple decibel values. High turbulence intensity can provide an early warning of fan aging or thermal design flaws, triggering proactive maintenance rather than reactive response.

[0092] Step S23: determining a weighted sum of the first negative impact degree, the second negative impact degree, and the third negative impact degree as the input / output delay performance of the current mechanical hard disk.

[0093] It should be noted that environmental parameters in different dimensions have different degrees of impact on the performance of mechanical hard disks. That is, the first, second, and third negative impact levels have different effects on the input and output latency performance of the current mechanical hard disk. Therefore, it is necessary to determine the weighting coefficients corresponding to the first, second, and third negative impact levels, and use the weighting coefficients to perform a weighted summation of the first, second, and third negative impact levels. The weighted summation result is determined as the input and output latency performance of the current mechanical hard disk. The specific formula is as follows:

[0094] ;

[0095] Where, are the weighted coefficients of the first negative impact degree, the second negative impact degree, and the third negative impact degree, respectively. is the first negative impact degree, the second negative impact degree and the third negative impact degree, Input and output delay performance.

[0096] Step S24: If the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, searching for a target mechanical hard disk combination in the distributed cluster.

[0097] Furthermore, a target prediction model is constructed. The target prediction model is a deep learning model for predicting the environmental parameters of the mechanical hard disk. If the input and output delay performance of the current mechanical hard disk is not greater than the preset threshold, the target prediction model is used to predict the environmental parameters of the current mechanical hard disk to obtain the environmental prediction result of the preset future time node. Then, the environmental parameters in the environmental prediction result are used to predict the input and output delay performance of the current mechanical hard disk to obtain the performance prediction result. If the performance prediction result indicates that the input and output delay performance of the current mechanical hard disk is not greater than the preset threshold at the preset future time node, no warning information is generated. If the performance prediction result indicates that the input and output delay performance of the current mechanical hard disk is greater than the preset threshold at the preset future time node, a warning information is generated. The warning information is at the preset future time node. That is, in this embodiment, if the target prediction model is used to predict the environmental parameters of the current mechanical hard disk, then even if the input and output delay performance at the current moment is not greater than the preset threshold, and the input and output delay performance of the current mechanical hard disk is greater than the preset threshold at the preset future time node, then the user can be recommended to actively issue a migration storage instruction to perform the data migration storage process, so as to avoid performing data migration storage only when the subsequent delay performance is high. If the target prediction model is not used to predict the environmental parameters of the current mechanical hard disk, then data migration storage can be performed only when the input and output delay performance of the current mechanical hard disk is greater than the preset threshold, thereby avoiding unnecessary data migration storage. Through the deep learning model, forward-looking prediction and proactive intervention can be achieved, which significantly improves the intelligence level and operation and maintenance efficiency of the storage system. Its core value is reflected in three aspects: First, by using the target prediction model to predict the future environmental parameters and I / O latency performance of mechanical hard drives, it can provide early warning and trigger data migration before actual performance degradation, avoid the risk of sudden performance degradation, and ensure business continuity; second, by distinguishing differentiated processing strategies between prediction scenarios (model enabled) and real-time monitoring scenarios (model not enabled), it can not only avoid risks in advance when necessary, but also avoid resource waste caused by excessive migration when there is no prediction demand, and achieve the optimal balance between cost and performance; finally, this mechanism deeply combines deep learning with storage hardware management, and through the correlation modeling of environmental parameters and performance, it breaks through the limitations of traditional static management based on threshold triggering, forming a dynamic perception-prediction-decision-making-optimization closed-loop control system, which not only extends the service life of hard drives, but also reduces the frequency of manual intervention, providing innovative solutions for the automated operation and maintenance of high-density data centers.

[0098] Step S25: migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism.

[0099] By integrating environmental perception with intelligent decision-making, the system incorporates multiple physical quantities, such as vibration, temperature, and noise, into data migration decisions. This addresses the performance bottlenecks of traditional distributed storage under physical environmental interference, improves the reliability of mechanical hard drives, and provides an innovative solution for high-density data centers. Its multi-dimensional dynamic optimization capabilities are widely applicable in scenarios such as cloud computing and AI training, offering significant economic benefits and technological foresight. For example, predictive scheduling uses AI models to anticipate environmental risks, enabling proactive defense rather than reactive response. Compatibility design supports mainstream distributed storage architectures such as OCP and OpenStack.

[0100] The key benefit of identifying and migrating data to a high-performing hard drive combination is a significant improvement in the overall performance, reliability, and resource utilization of the distributed storage system. Through multi-dimensional environmental sensing (vibration, temperature, and noise) and real-time performance evaluation, the system accurately identifies hard drives currently experiencing low vibration, low temperature, and minimal noise interference. These drives, due to their stable environment, offer lower I / O latency, higher throughput, and longer lifespan. Migrating data (especially frequently accessed, hot data) to these hard drives effectively reduces issues such as data retries and increased error rates caused by environmental interference, thereby improving user experience and business continuity. Furthermore, dynamic migration avoids local performance bottlenecks caused by static data distribution and optimizes overall cluster resource utilization through load balancing. Furthermore, combining erasure coding with multiple replication mechanisms, the system ensures data redundancy and fault tolerance during migration while providing additional replicas for hot data to further reduce access latency. This intelligent data distribution management not only extends the reliability and lifespan of hard drives but also reduces the risk of sudden failures caused by environmental degradation, providing an innovative solution for high-density data centers that balances performance, cost, and reliability.

[0101] Below is Figure 5 Taking a specific distributed data dynamic migration storage diagram based on multi-dimensional environment perception as an example, the present application is explained accordingly.

[0102] Vibration sensors, temperature sensors, and sound pressure meters deployed on storage servers collect real-time environmental parameters of mechanical hard drives (such as vibration spectrum, drive cavity temperature, and fan turbulence intensity). Using a predefined environment-performance coupling model (a weighted calculation of vibration power spectrum density and resonance sensitivity), the system quantifies the negative impact of each drive. When a drive's environmental parameters exceed preset thresholds (such as excessive vibration energy or temperature deviation from baseline), its performance is deemed significantly degraded, automatically triggering a data migration process to ensure data is moved from high-risk drives to more stable environments.

[0103] Based on real-time monitoring data, the target mechanical hard drive combination (HDDs with low vibration, low temperature, and low noise interference) is selected from the cluster. This selection process takes into account remaining capacity, load status, and network topology location. This selection process incorporates the distribution requirements of erasure codes to ensure that the target combination meets redundancy requirements while providing low-latency access paths for hot data. For example, hard drives with stable environments within the same rack are prioritized to reduce network overhead during migration.

[0104] The target data block to be migrated is split into sub-blocks, and a parity block is generated using an erasure coding algorithm. These coded blocks (sub-block + parity block) are then distributed across the target hard drive combination based on load balancing. For frequently accessed hot data blocks, additional copies are generated and stored on the highest-performing drives in the target combination (such as those in low-vibration zones) to further improve read speeds. The migration process is asynchronous to avoid blocking normal I / O operations. A metadata table records the mapping between data blocks and physical locations, ensuring rapid location after migration.

[0105] After the migration is complete, the system continuously monitors the target drive's environmental status and performance indicators. If the target mechanical drive's environmental parameters deteriorate (e.g., temperature rises or vibration intensifies), a second migration or replica adjustment is triggered. If the original drive's environmental parameters improve (e.g., vibration decreases after fan speed adjustment), cold data is gradually migrated back to balance the storage load. This closed-loop control mechanism dynamically responds to environmental changes, achieving optimal long-term allocation of storage resources, while extending drive life and reducing the frequency of maintenance interventions.

[0106] Figure 6 A schematic structural diagram of a data dynamic migration storage device provided in an embodiment of the present invention includes:

[0107] The environmental parameter acquisition module 11 is used to acquire the environmental parameters of the current mechanical hard disk in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters and noise parameters;

[0108] A performance quantification module 12 is configured to quantify the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk;

[0109] A hard disk search module 13 is configured to search for a target mechanical hard disk combination in the distributed cluster if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold;

[0110] The migration storage module 14 is configured to migrate and store the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multiple copy mechanism.

[0111] It can be seen that the present invention collects the environmental parameters of the current mechanical hard disk in the distributed cluster; wherein, the environmental parameters include vibration parameters, temperature parameters and noise parameters; the environmental parameters of the current mechanical hard disk are used to quantify the input and output delay performance of the current mechanical hard disk; if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, the target mechanical hard disk combination is searched in the distributed cluster; based on the erasure code and multi-copy mechanism, the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination.

[0112] The beneficial effect is: the present invention collects the vibration parameters, temperature parameters and noise parameters of the current mechanical hard disk in the distributed cluster, that is, it collects multi-dimensional physical environment parameters, and uses these environmental parameters to quantify the input and output delay performance of the current mechanical hard disk, that is, to determine the input and output delay performance of the current mechanical hard disk under the current physical environment. If the input and output delay performance is greater than the preset threshold, it means that the current physical environment has a negative impact on the current mechanical hard disk and the impact is not small, so the input and output delay performance of the current mechanical hard disk is high, that is, the stability is poor, and it is not suitable to continue to store important data, so the target mechanical hard disk combination is searched in the distributed cluster, so that the target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination, that is, the data is stored in a mechanical hard disk with better performance, and the migration storage based on the erasure code and multi-copy mechanism can make the data in the target data block more reliably stored in the distributed cluster.

[0113] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 7 This is a structural diagram of an electronic device according to an exemplary embodiment. The content in the diagram should not be considered as any limitation on the scope of use of this application. The electronic device may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps of the data dynamic migration and storage method disclosed in any of the aforementioned embodiments. In addition, the electronic device in this embodiment may specifically be an electronic computer.

[0114] In this embodiment, the power supply 23 is used to provide operating voltage for various hardware devices on the electronic device; the communication interface 24 can create a data transmission channel between the electronic device and external devices. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0115] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0116] The operating system 221 is used to manage and control the hardware devices on the electronic device and the computer program 222, which can be Windows Server, NetWare, Unix, Linux, etc. In addition to including a computer program capable of implementing the data dynamic migration and storage method performed by the electronic device disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program capable of performing other specific tasks.

[0117] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned method for dynamic data migration and storage is implemented. The specific steps of this method can be referred to the corresponding contents disclosed in the aforementioned embodiments and will not be repeated here.

[0118] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0119] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0121] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0122] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A data dynamic migration and storage method, characterized in that: include: Collect environmental parameters of the current mechanical hard disks in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters, and noise parameters; quantifying the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk; If the input and output delay performance of the current mechanical hard disk is greater than a preset threshold, searching for a target mechanical hard disk combination in the distributed cluster; The target data block of the current mechanical hard disk is migrated and stored in the target mechanical hard disk combination based on the erasure code and multiple copy mechanism.

2. The data dynamic migration storage method according to claim 1, characterized in that: The quantifying the input and output delay performance of the current mechanical hard disk by using the environmental parameters of the current mechanical hard disk includes: Using the vibration parameter, the temperature parameter, and the noise parameter, respectively, to obtain a first negative impact degree in the vibration dimension, a second negative impact degree in the temperature dimension, and a third negative impact degree in the noise dimension; A weighted sum of the first negative impact degree, the second negative impact degree, and the third negative impact degree is determined as the input and output delay performance of the current mechanical hard disk.

3. The data dynamic migration storage method according to claim 2, characterized in that: The obtaining, by using the vibration parameter, the temperature parameter, and the noise parameter, respectively, a first negative impact degree in the vibration dimension, a second negative impact degree in the temperature dimension, and a third negative impact degree in the noise dimension, comprises: Performing a fast Fourier transform on the vibration parameter to obtain a frequency domain signal; wherein the vibration parameter is a vibration acceleration time domain signal collected by a vibration sensor of the current mechanical hard disk; Obtaining a vibration energy distribution of the frequency domain signal within a unit frequency using a periodogram estimation algorithm, and determining the vibration energy distribution within the unit frequency as a vibration power spectral density; Obtaining the resonance sensitivity of the current mechanical hard disk using the vibration parameter; summing the vibration power spectrum density and the resonance sensitivity within a preset frequency range to obtain a first negative impact degree of the vibration dimension; Obtaining a magnitude of temperature deviation from the reference value based on a difference between the temperature parameter and a preset temperature reference value, and determining the magnitude of temperature deviation from the reference value as a second negative impact degree of the temperature dimension; wherein the temperature parameter is a disk cavity temperature collected by a temperature sensor of the current mechanical hard disk; Obtaining a target fan speed corresponding to the noise parameter; wherein the target fan speed is the speed of a fan in the storage server to which the current mechanical hard disk belongs; The turbulence intensity obtained by performing airflow simulation on the target fan speed is determined as the third negative impact degree of the noise dimension.

4. The data dynamic migration and storage method according to any one of claims 1 to 3, characterized in that: Determining the target data block of the current mechanical hard disk includes: Determining the access frequency of each data block in the current mechanical hard disk, and determining the data blocks with access frequencies higher than a first preset access frequency threshold as hot data blocks; The hot data block is determined as the target data block of the current mechanical hard disk.

5. The data dynamic migration storage method according to claim 4, characterized in that: The step of migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multiple copies mechanism includes: Splitting the target data block of the current mechanical hard disk into target sub-data blocks, and generating a check block corresponding to the target data block by an erasure coding algorithm to obtain a coding block including the target sub-data block and the check block; Migrating and storing the coding block to each mechanical hard disk in the target mechanical hard disk combination; Filtering out a target coding block having an access frequency higher than a second preset access frequency threshold from each of the coding blocks, and generating a copy of the target coding block; A copy of the target coding block is stored in a target mechanical hard disk in the target mechanical hard disk combination.

6. The data dynamic migration storage method according to claim 5, characterized in that: The target mechanical hard disk combination includes each mechanical hard disk in the distributed cluster whose input and output delay performance is less than the preset threshold, the input and output delay performance of the target mechanical hard disk meets the preset performance optimal level condition, and the second preset access frequency threshold is greater than the first preset access frequency threshold.

7. The data dynamic migration storage method according to claim 5, characterized in that: After migrating and storing the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multi-copy mechanism, the method further includes: A mapping relationship is established among the target data block, the coding block, the copy of the target coding block, and the target mechanical hard disk, so as to read and / or migrate the target data block based on the mapping relationship.

8. A data dynamic migration storage device, characterized in that: include: An environmental parameter acquisition module is used to collect environmental parameters of the current mechanical hard disk in the distributed cluster; wherein the environmental parameters include vibration parameters, temperature parameters and noise parameters; a performance quantification module, configured to quantify the input and output delay performance of the current mechanical hard disk using the environmental parameters of the current mechanical hard disk; a hard disk search module, configured to search for a target mechanical hard disk combination in the distributed cluster if the input and output delay performance of the current mechanical hard disk is greater than a preset threshold; The migration storage module is used to migrate and store the target data block of the current mechanical hard disk to the target mechanical hard disk combination based on the erasure code and multiple copy mechanism.

9. An electronic device, characterized in that: include: memory for storing computer programs; A processor, configured to execute the computer program to implement the steps of the data dynamic migration storage method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the data dynamic migration storage method according to any one of claims 1 to 7.

Images