Storage task scheduling method, device, equipment and system

Abstract

The invention discloses a storage task scheduling method, device, equipment and system, and relates to the technical field of storage, and the method comprises the following steps: adding an obtained service input / output task of a solid state disk into a preset input / output queue; according to the single-layer unit cache monitoring information of the solid state disk, determining a single batch processing quantity of a preset input and output queue; writing the business input and output tasks in the preset input and output queue into a preset disk input and output queue according to the single batch processing quantity, so as to issue the business input and output tasks in the preset disk input and output queue to the solid state disk; according to the invention, the added preset input / output queue is used for storing the business input / output task, so that the quantity of IO which is taken out and processed in batches each time can be adjusted according to the monitored release speed and / or residual storage space of the SLC cache, the adaptation of IO scheduling and the SLC cache condition of the SSD is realized, the performance of the SSD is optimized, and the stability of IO time delay and bandwidth is ensured.

Description

Technical Field

[0001] The present invention relates to the field of storage technology, and in particular to a storage task scheduling method, device, equipment and system. Background Art

[0002] With the continuous advancement of storage technology, solid-state drives (SSDs) have become a widely used storage device. However, SSD performance bottlenecks often arise in I / O (input / output) scheduling and cache management. SLC (Single-Level Cell) cache is a high-performance cache technology in SSDs, but its release speed often struggles to adapt to I / O scheduling policies, leading to unstable I / O latency and bandwidth.

[0003] In other words, the IO scheduling algorithm used in related technologies to submit IO tasks to SSDs does not match the release rate of the SSD's SLC cache. When the SLC cache is quickly exhausted, the disk IO latency is instantly amplified dozens of times by switching from SLC mode to TLC (Trinary-Level Cell) / QLC (Quad-Level Cell) mode. The overall IO latency and bandwidth become extremely unstable. This can cause drastic fluctuations in the cluster's throughput on cloud storage platforms, and in extreme cases, even result in a large number of failed IOs and timed IOs. Therefore, how to adapt IO scheduling to the SSD's SLC cache, optimize SSD performance, and ensure the stability of IO latency and bandwidth is an urgent problem that needs to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide a storage task scheduling method, device, equipment and system to achieve adaptation of IO scheduling to the SLC cache of SSD, optimize the performance of SSD, and ensure the stability of IO latency and bandwidth.

[0005] To solve the above technical problems, the present invention provides a storage task scheduling method, comprising:

[0006] Add the acquired business input and output tasks of the solid-state drive to the preset input and output queue;

[0007] Determining the single batch processing quantity of the preset input and output queue according to the single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes a release speed and / or remaining storage space of the single-layer unit cache;

[0008] According to the single batch processing quantity, the business input and output tasks in the preset input and output queue are written to the preset disk input and output queue, so that the business input and output tasks in the preset disk input and output queue are sent to the solid-state drive, and the solid-state drive is controlled to perform corresponding data reading and writing operations.

[0009] On the other hand, after writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue according to the single batch processing quantity, the method further includes:

[0010] Monitoring the return delay of business input and output tasks sent to the solid-state drive;

[0011] The single batch processing quantity is adjusted according to the return delay.

[0012] On the other hand, after writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue, the method further includes:

[0013] The target sequential write tasks in the preset disk input and output queue are merged into a merged input and output task; wherein the target sequential write tasks are business input and output tasks written to the solid state hard disk with continuous disk locations.

[0014] On the other hand, the amount of written data of the merged input and output task is less than or equal to a data amount threshold.

[0015] On the other hand, adding the acquired business input and output tasks of the solid-state drive to a preset input and output queue includes:

[0016] Adding a target business task among the acquired business input and output tasks of the solid-state hard disk to the preset input and output queue; wherein the target business task includes a data writing task.

[0017] On the other hand, after writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue, the method further includes:

[0018] The target random write task in the preset input and output queue is converted into a pre-write log file, and the pre-write log file is sent to the pre-write log partition storage of the solid-state drive; wherein, the target random write task is a business input and output task in which the disk location of the solid-state drive to be written is discontinuous and the amount of written data is less than or equal to the merged data amount threshold.

[0019] On the other hand, converting the target random write task in the preset input and output queue into a write-ahead log file includes:

[0020] The merged data amount threshold is adjusted according to the detected remaining storage space of the single-layer unit cache of the solid-state drive.

[0021] The present invention also provides a storage task scheduling device, comprising:

[0022] The first queue module is used to add the acquired business input and output tasks of the solid-state drive to a preset input and output queue;

[0023] a batch determination module, configured to determine the single batch processing quantity of the preset input and output queue based on the single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes the release speed and / or remaining storage space of the single-layer unit cache;

[0024] The second queue module is used to write the business input and output tasks in the preset input and output queue to the preset disk input and output queue according to the single batch processing quantity, so as to send the business input and output tasks in the preset disk input and output queue to the solid-state hard disk and control the solid-state hard disk to perform corresponding data reading and writing operations.

[0025] The present invention also provides a storage task scheduling device, comprising:

[0026] memory for storing computer programs;

[0027] The processor is configured to implement the steps of the storage task scheduling method described above when executing the computer program.

[0028] In addition, the present invention also provides a storage task scheduling system, comprising: a solid state hard disk and the storage task scheduling device as described above.

[0029] A storage task scheduling method provided by the present invention includes: adding acquired business input and output tasks of a solid-state drive to a preset input and output queue; determining a single batch processing quantity of the preset input and output queue based on single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes a release rate and / or remaining storage space of the single-layer unit cache; writing the business input and output tasks in the preset input and output queue to a preset disk input and output queue based on the single batch processing quantity, so as to dispatch the business input and output tasks in the preset disk input and output queue to the solid-state drive, thereby controlling the solid-state drive to perform corresponding data read and write operations;

[0030] As can be seen, the present invention utilizes additional preset input and output queues to store business input and output tasks. This allows the number of IOs processed in each batch to be adjusted based on the monitored SLC cache release rate and / or remaining storage space, adapting IO scheduling to the SSD's SLC cache. This minimizes SSD disk SLC cache consumption, avoids sudden drops in disk performance caused by SLC cache exhaustion, optimizes SSD performance, and ensures IO latency and bandwidth stability. Furthermore, the present invention provides a storage task scheduling device, equipment, and system, which also have the aforementioned beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0032] Figure 1 A flowchart of a storage task scheduling method provided by an embodiment of the present invention;

[0033] Figure 2 A queue architecture diagram of a storage task scheduling method provided by an embodiment of the present invention

[0034] Figure 3 A structural block diagram of a storage task scheduling device provided by an embodiment of the present invention;

[0035] Figure 4 A schematic diagram of the structure of a storage task scheduling device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0036] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0037] Please refer to Figure 1 , Figure 1 This is a flow chart of a storage task scheduling method provided by an embodiment of the present invention. The method may include:

[0038] Step 101: Add the acquired business input and output tasks of the solid-state drive to a preset input and output queue.

[0039] The business input and output tasks in this embodiment can be tasks (or requests) at the business level requesting data read and write operations on the solid-state drive, such as the IO tasks of various businesses obtained by the underlying storage engine of the cloud storage platform (such as Figure 2 The preset input and output queue in this embodiment may be a first-in, first-out queue pre-set for storing business input and output tasks.

[0040] Accordingly, the method provided in this embodiment can be applied to electronic devices connected to solid-state drives, such as computer devices or server host devices. For example, the host device of a distributed storage system can utilize the underlying storage engine of a cloud storage platform to execute the method provided in this embodiment to complete the IO processing process. Accordingly, this embodiment uses the IO task scheduling of a single solid-state drive as an example. The IO task scheduling of multiple solid-state drives can be implemented using the same or similar methods as the method provided in this embodiment, and this embodiment does not impose any restrictions on this.

[0041] It should be noted that the specific method of adding the business input and output tasks of the solid-state hard drive obtained in this step to the preset input and output queue, that is, the specific IO tasks stored in the preset input and output queue, can be set by the designer according to the practical scenario and user needs. For example, in this step, all the business input and output tasks of the solid-state hard drive obtained can be added to the preset input and output queue, that is, the preset input and output queue can be used to store all the business input and output tasks of the solid-state hard drive.

[0042] Correspondingly, this step can also add the target business task among all the business input and output tasks of the acquired solid-state hard disk to the preset input and output queue; wherein, the target business task can be a task that affects the storage space of the single-layer unit cache (SLC cache) of the solid-state hard disk, such as the target business task can include a data writing task; that is, the preset input and output queue can be used to store the target business tasks of the solid-state hard disk. Accordingly, for business input and output tasks other than the target business tasks of the acquired solid-state hard disk (i.e., non-target business tasks), it can be directly implemented in the same or similar manner as the business input and output task processing method of the solid-state hard disk in the related art, such as directly writing the business input and output tasks to the preset disk input and output queue, and then sending them to the solid-state hard disk for processing. This embodiment does not impose any restrictions on this.

[0043] Correspondingly, step 101 may also include a process for acquiring business input and output tasks of the solid-state drive, such as by adopting a method that is the same or similar to the task acquisition method for performing data reading and writing operations on the solid-state drive in the related art. This embodiment does not impose any restrictions on this.

[0044] Step 102: Determine the single batch processing quantity of the preset input and output queue according to the single-layer unit cache monitoring information of the solid state drive; wherein the single-layer unit cache monitoring information includes the release speed and / or remaining storage space of the single-layer unit cache.

[0045] It is understandable that in this embodiment, the number of single batch processing of the preset input and output queues (such as Figure 2 N in the input / output queue), so that each time the IO tasks (such as business input / output tasks) in the preset input / output queue are written to the preset disk input / output queue according to the single batch processing quantity.

[0046] Correspondingly, the number of single batch processing can be the number of IO tasks sent to the preset disk input and output queue each time. For example, in this embodiment, the number of IO tasks for single batch processing can be sent to the preset disk input and output queue at a preset time interval, that is, N (that is, the number of single batch processing) IO tasks are taken out from the queue of the preset input and output queue at a preset time interval and sent to the preset disk input and output queue.

[0047] Accordingly, the specific method of determining the single batch processing quantity of the preset input and output queue based on the single-layer unit cache monitoring information of the solid-state hard disk in this step can be set by the designer according to the usage scenario and user needs. For example, the single batch processing quantity of the preset input and output queue can be determined directly based on the single-layer unit cache monitoring information of the solid-state hard disk; for example, when the single-layer unit cache monitoring information includes the release speed and remaining storage space of the single-layer unit cache, the single batch processing quantity of the preset input and output queue can be determined based on the release speed and remaining storage space of the single-layer unit cache; for example, in this embodiment, the preset quantity corresponding to each preset remaining storage space range corresponding to each preset release speed interval can be pre-set to match a corresponding preset quantity as the current single batch processing quantity based on the release speed and remaining storage space in the single-layer unit cache monitoring information. In other embodiments, the preset quantities corresponding to each preset release speed interval and the preset quantity adjustment amounts corresponding to each preset remaining storage space range can also be pre-set to match a corresponding preset quantity and preset quantity adjustment amount according to the release speed and remaining storage space in the single-layer unit cache monitoring information, and the value obtained after adjusting the preset quantity by the preset quantity adjustment amount will be used as the current single batch processing quantity.

[0048] Correspondingly, in this step, the number of single batch processings of the preset input and output queue can also be determined based on the single-layer unit cache monitoring information of the solid-state drive and the monitored return delay of the IO task (such as the business input and output task or the target business task) sent to the solid-state drive. For example, after determining the single batch processing number of the preset input and output queue based on the single-layer unit cache monitoring information of the solid-state drive, the determined single batch processing number can also be adjusted based on the return delay of the IO task currently monitored and sent to the solid-state drive to obtain the final single batch processing number; if the return delay is greater, the single batch processing number is adjusted to be lowered. In other words, this embodiment can also include monitoring the return delay of the business input and output task sent to the solid-state drive; and adjusting the single batch processing number based on the return delay.

[0049] Step 103: Based on the single batch processing quantity, the business input and output tasks in the preset input and output queue are written to the preset disk input and output queue, so as to send the business input and output tasks in the preset disk input and output queue to the solid-state drive, and control the solid-state drive to perform corresponding data read and write operations.

[0050] It is understood that in this step, N (i.e., the number of single batches of processing) IO tasks (e.g., business IO tasks) can be taken from the queue of the preset IO queue each time and sent to the preset disk IO queue, so that the IO tasks can be sent to the solid-state drive via the preset disk IO queue. Accordingly, the method provided in this embodiment can also include a process of sending IO tasks (e.g., business IO tasks) in the preset disk IO queue to the solid-state drive.

[0051] Furthermore, due to the frequent writing of small blocks of data, the SLC cache of the SSD disk will be quickly exhausted, causing a decline in the performance of the SSD. In this embodiment, the business input and output tasks with continuous write disk positions (i.e., the disk positions of the solid-state drive) can also be merged to reduce the number of IO tasks that fall instantaneously on the disk. For example, after step 103, it can also include merging the target sequential write tasks in the preset disk input and output queue into merged input and output tasks; wherein the target sequential write tasks are business input and output tasks with continuous write disk positions on the solid-state drive; that is, in the process of sending the IO tasks in the preset disk input and output queue to the solid-state drive, the IO tasks with continuous write disk positions can be merged, and the merged IO tasks obtained by merging can be sent to the solid-state drive.

[0052] In other embodiments, the above-mentioned merging processing process can also be performed in the IO task sending process in the preset input and output queue in step 103 or the IO task writing process of the preset input and output queue in step 101. For example, in step 103, the N (i.e., the number of single batch processing) IO tasks sent to the preset disk input and output queue each time may include merged IO tasks and IO tasks other than merged IO tasks (i.e., non-merged IO tasks). The write disk positions of the merged IO tasks and the non-merged IO tasks are not continuous, that is, the target sequential write tasks can be directly merged into the corresponding merged input and output tasks in the process of taking out IO tasks from the preset input and output queue.

[0053] Correspondingly, the data structure of the data written by the merged IO task can be adapted to the storage structure of the SSD, such as the Page / Block structure of the SSD. Furthermore, the amount of data written by the merged IO task can be less than or equal to the data volume threshold to avoid the situation where a single IO task writes too much data.

[0054] Furthermore, the merging process of the above-mentioned IO tasks is suitable for scenarios involving large amounts of continuous IO testing. For random IO, since the disk locations to which it is written are not continuous, IO merging at the business level is not possible. Therefore, the method provided in this embodiment can also add a dedicated WAL (Write Ahead Log) partition within the solid-state drive, converting business input and output tasks with discontinuous write disk locations (i.e., SSD disk locations) and small amounts of written data (e.g., less than a merged data volume threshold) into write-ahead log (WAL) files, and writing them to the SSD's WAL partition to merge random small-block IO into sequential large-block IO.

[0055] WAL is a mechanism for ensuring transaction persistence in databases and storage systems. Its core concept is to persist data modifications in a file as a log before they are actually written to disk. This embodiment uses a method that writes the IO tasks received by the storage system as a log to a specific disk area (i.e., a WAL partition) of the solid-state drive (SSD) in a sequential IO manner. Subsequently, when the system or SSD is idle, the WAL partition's log (i.e., the WAL file) can be replayed, and data written to the corresponding area based on the log content. In other words, this embodiment utilizes a log (i.e., the WAL file) that records IO operations to implement a write-ahead log function, merging random small-block IOs into sequential large-block IOs. This reduces the number of IOs that instantaneously hit the disk, thereby reducing the instantaneous consumption of the SLC cache. This improves overall IO stability while maintaining the same SLC cache release rate. After the IO peak passes, the logs in the WAL partition can be flushed to the corresponding disk location based on the current system pressure, achieving peak shaving.

[0056] For example, the method step 103 provided in this embodiment may also include: converting the target random write task in the preset input and output queue into a pre-write log file, and sending the pre-write log file to the pre-write log partition storage of the solid-state hard disk; wherein the target random write task is a business input and output task in which the disk location of the solid-state hard disk to be written is discontinuous and the amount of written data is less than or equal to the merged data amount threshold. Correspondingly, the above-mentioned process of converting the target random write task in the preset input and output queue into a pre-write log file may include: converting the target random write task in the preset input and output queue into a pre-write log file according to the remaining storage space of the single-layer unit cache of the solid-state hard disk detected; for example, the merged data amount threshold can be adjusted according to the remaining storage space of the single-layer unit cache to realize the dynamic selection process of the target random write task, such as increasing the merged data amount threshold as the remaining storage space of the single-layer unit cache decreases. This embodiment does not impose any restrictions on this.

[0057] In other embodiments, the WAL file conversion process can also be performed during the IO task dispatching process in the preset input / output queue in step 103 or the IO task writing process in the preset input / output queue in step 101. For example, the N (i.e., the number of single batch processing) IO tasks dispatched to the preset disk input / output queue each time in step 103 can include IO tasks other than the target random write task, that is, the target random write task can be directly converted into a corresponding WAL file during the process of taking the IO task from the preset input / output queue. This embodiment does not impose any restrictions on this.

[0058] Accordingly, the amount of data written to each of the above-mentioned write-ahead log files may be less than or equal to a corresponding threshold (such as a data amount threshold).

[0059] In this embodiment, the embodiment of the present invention uses the added preset input and output queues to save business input and output tasks, so that the number of IOs processed in each batch can be adjusted according to the monitored SLC cache release speed and / or remaining storage space, so as to achieve adaptation of IO scheduling and the SLC cache situation of the SSD, so as to minimize the consumption of the SSD disk SLC cache, avoid the sudden drop in disk performance caused by the exhaustion of the SLC cache, optimize the performance of the SSD, and ensure the stability of IO latency and bandwidth.

[0060] Corresponding to the above method embodiment, an embodiment of the present invention further provides a storage task scheduling device. The storage task scheduling device described below and the storage task scheduling method described above can be referenced to each other.

[0061] Please refer to Figure 3 , Figure 3This is a structural block diagram of a storage task scheduling device provided by an embodiment of the present invention. The device may include:

[0062] The first queue module 10 is used to add the acquired business input and output tasks of the solid-state drive to a preset input and output queue;

[0063] A batch determination module 20 is configured to determine a single batch processing quantity for a preset input and output queue based on single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes a release rate and / or remaining storage space of the single-layer unit cache;

[0064] The second queue module 30 is used to write the business input and output tasks in the preset input and output queue to the preset disk input and output queue according to the number of single batch processing, so as to send the business input and output tasks in the preset disk input and output queue to the solid-state hard disk and control the solid-state hard disk to perform corresponding data reading and writing operations.

[0065] In another aspect, the apparatus may further comprise:

[0066] A monitoring module is used to monitor the return delay of business input and output tasks sent to the solid-state drive;

[0067] The adjustment module is used to adjust the number of single batch processing according to the return delay.

[0068] In another aspect, the apparatus may further comprise:

[0069] The sequential merging module is used to merge the target sequential writing tasks in the preset disk input and output queue into a merged input and output task; wherein the target sequential writing task is a business input and output task with continuous disk locations of the solid-state hard disk being written.

[0070] On the other hand, the amount of written data of the merged input and output tasks is less than or equal to the data amount threshold.

[0071] On the other hand, the first queue module 10 can be specifically configured to add a target business task among the acquired business input and output tasks of the solid-state drive to a preset input and output queue; wherein the target business task includes a data writing task.

[0072] In another aspect, the apparatus may further comprise:

[0073] The write-ahead log module is used to convert the target random write tasks in the preset input and output queues into write-ahead log files, and send the write-ahead log files to the write-ahead log partition storage of the solid-state drive; wherein, the target random write tasks are business input and output tasks in which the disk location of the solid-state drive to be written is discontinuous and the amount of written data is less than or equal to the merged data amount threshold.

[0074] In another aspect, the write-ahead log module may include:

[0075] The threshold adjustment submodule is used to adjust the merged data amount threshold according to the detected remaining storage space of the single-layer unit cache of the solid-state hard disk.

[0076] In this embodiment, the embodiment of the present invention utilizes an added preset input and output queue to save business input and output tasks, so that the batch determination module 20 can adjust the number of IOs processed in each batch according to the release speed and / or remaining storage space of the monitored SLC cache, thereby realizing the adaptation of IO scheduling to the SLC cache situation of the SSD, so as to minimize the consumption of the SSD disk SLC cache, avoid the sudden drop in disk performance caused by the exhaustion of the SLC cache, optimize the performance of the SSD, and ensure the stability of IO latency and bandwidth.

[0077] Corresponding to the above method embodiment, an embodiment of the present invention further provides a storage task scheduling device. The storage task scheduling device described below and the storage task scheduling method described above can refer to each other.

[0078] Please refer to Figure 4 , Figure 4 This is a schematic diagram of the structure of a storage task scheduling device provided by an embodiment of the present invention. The device may include:

[0079] Memory D1, for storing computer programs;

[0080] The processor D2 is configured to implement the steps of the storage task scheduling method provided in the above method embodiment when executing a computer program.

[0081] The storage task scheduling device provided in this embodiment may specifically be a host device, such as a server host of a distributed storage system.

[0082] Corresponding to the above device embodiment, an embodiment of the present invention further provides a storage task scheduling system. The storage task scheduling system described below and the storage task scheduling device described above can refer to each other.

[0083] A storage task scheduling system includes a solid-state hard disk and the storage task scheduling device provided in the above embodiment.

[0084] The solid state drive may be connected to the storage task scheduling device for communication.

[0085] Corresponding to the above method embodiment, an embodiment of the present invention further provides a computer program product. The computer program product described below and the storage task scheduling method described above can refer to each other.

[0086] A computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps of the storage task scheduling method provided by the above method embodiment.

[0087] Corresponding to the above method embodiment, an embodiment of the present invention further provides a computer-readable storage medium. The computer-readable storage medium described below and the storage task scheduling method described above can refer to each other.

[0088] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the storage task scheduling method of the above method embodiment.

[0089] The computer-readable storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, which may store program codes.

[0090] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. References to the common and similar parts between the various embodiments are sufficient. The devices, equipment, systems, computer program products, and computer-readable storage media disclosed in the embodiments are described briefly because they correspond to the methods disclosed in the embodiments. For relevant details, refer to the description of the methods.

[0091] The above is a detailed introduction to the storage task scheduling method, device, equipment and system provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only intended to help understand the method and core ideas of the present invention. It should be pointed out that, for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the present invention.

Claims

1. A storage task scheduling method, characterized in that: include: Add the acquired business input and output tasks of the solid-state drive to the preset input and output queue; Determining the single batch processing quantity of the preset input and output queue according to the single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes a release speed and / or remaining storage space of the single-layer unit cache; According to the single batch processing quantity, the business input and output tasks in the preset input and output queue are written to the preset disk input and output queue, so that the business input and output tasks in the preset disk input and output queue are sent to the solid-state drive, and the solid-state drive is controlled to perform corresponding data reading and writing operations.

2. The storage task scheduling method according to claim 1, characterized in that: After writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue according to the single batch processing quantity, the method further includes: Monitoring the return delay of business input and output tasks sent to the solid-state drive; The single batch processing quantity is adjusted according to the return delay.

3. The storage task scheduling method according to claim 1, wherein: After writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue, the method further includes: The target sequential write tasks in the preset disk input and output queue are merged into a merged input and output task; wherein the target sequential write tasks are business input and output tasks written to the solid state hard disk with continuous disk locations.

4. The storage task scheduling method according to claim 3, wherein: The amount of data written by the merged input and output task is less than or equal to a data amount threshold.

5. The storage task scheduling method according to claim 1, wherein: The step of adding the acquired business input and output tasks of the solid-state drive to a preset input and output queue includes: Adding a target business task among the acquired business input and output tasks of the solid-state hard disk to the preset input and output queue; wherein the target business task includes a data writing task.

6. The storage task scheduling method according to any one of claims 1 to 5, characterized in that: After writing the business input and output tasks in the preset input and output queue to the preset disk input and output queue, the method further includes: The target random write task in the preset input and output queue is converted into a pre-write log file, and the pre-write log file is sent to the pre-write log partition storage of the solid-state drive; wherein, the target random write task is a business input and output task in which the disk location of the solid-state drive to be written is discontinuous and the amount of written data is less than or equal to the merged data amount threshold.

7. The storage task scheduling method according to claim 6, characterized in that: The converting the target random write task in the preset input and output queue into a pre-write log file includes: The merged data amount threshold is adjusted according to the detected remaining storage space of the single-layer unit cache of the solid-state drive.

8. A storage task scheduling device, characterized in that: include: The first queue module is used to add the acquired business input and output tasks of the solid-state drive to a preset input and output queue; a batch determination module, configured to determine the single batch processing quantity of the preset input and output queue based on the single-layer unit cache monitoring information of the solid-state drive; wherein the single-layer unit cache monitoring information includes the release speed and / or remaining storage space of the single-layer unit cache; The second queue module is used to write the business input and output tasks in the preset input and output queue to the preset disk input and output queue according to the single batch processing quantity, so as to send the business input and output tasks in the preset disk input and output queue to the solid-state hard disk and control the solid-state hard disk to perform corresponding data reading and writing operations.

9. A storage task scheduling device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the storage task scheduling method according to any one of claims 1 to 7 when executing the computer program.

10. A storage task scheduling system, characterized in that: include: A solid state drive and a storage task scheduling device as claimed in claim 9.

Images