Search

CN-121996167-A - Input/output scheduling method and device, electronic equipment and storage medium

CN121996167ACN 121996167 ACN121996167 ACN 121996167ACN-121996167-A

Abstract

The application discloses a scheduling method, a device, an electronic device and a storage medium for input and output, which relate to the technical field of computers. Therefore, the mapping relation between the load prediction and the storage medium characteristic is established, and the technical problems of uncontrollable key request delay and system throughput fluctuation during load sudden increase in a mixed read-write load scene can be solved, and the technical effects of ensuring low-delay response of the key request, stabilizing the system throughput and improving the overall performance and stability of the system are achieved.

Inventors

  • ZHANG ZIQI

Assignees

  • 济南浪潮数据技术有限公司

Dates

Publication Date
20260508
Application Date
20260127

Claims (10)

  1. 1. The input/output scheduling method is characterized by comprising the following steps: Acquiring historical input and output request data of each storage node in a storage system; Calculating to obtain the load state of the storage system in a future period based on the historical input and output request data; according to the load state, adjusting response sequences of input and output requests with different priorities; And adjusting the resource allocation strategy of the storage node based on the response sequence.
  2. 2. The method of claim 1, wherein the historical input output request data comprises at least one of request type, request priority, logical address range for request access, amount of data requested, response time for request.
  3. 3. The method of claim 1, wherein calculating a load state of the storage system at a future time period based on the historical input output request data comprises: And processing the historical input and output request data based on a time sequence prediction model to obtain a load change trend of a future period, wherein the load state comprises at least one of load level, read-write request proportion, input and output operand per second or bandwidth requirement.
  4. 4. The method of claim 1, wherein adjusting the order of response of the different priority input-output requests according to the load status comprises: When the load state is load aggravated, the response sequence of the high-priority request is promoted; And when the load state is load relief, promoting the response sequence of the low-priority requests.
  5. 5. The method of claim 1, wherein said adjusting the resource allocation policy of the storage node based on the response order comprises: Adjusting the cache strategies of different storage media; Adjusting the bandwidth allocation proportion of the input/output channels; coordinating garbage collection operations among a plurality of storage nodes; data with access frequency higher than a preset threshold value is migrated.
  6. 6. The method according to claim 1, wherein the method further comprises: when data is transmitted between storage nodes, integrity checking is performed on the transmitted data blocks.
  7. 7. An input/output scheduling apparatus, comprising: The acquisition unit is used for acquiring historical input and output request data of each storage node in the storage system; the calculating unit is used for calculating the load state of the storage system in a future period based on the historical input and output request data; the adjusting unit is used for adjusting the response sequence of the input/output requests with different priorities according to the load state; And the adjusting unit is also used for adjusting the resource allocation strategy of the storage node based on the response sequence.
  8. 8. An electronic device, comprising: A memory for storing a computer program; A processor for implementing the steps of the input-output scheduling method according to any one of claims 1 to 6 when executing said computer program.
  9. 9. A computer readable storage medium, wherein a computer program is stored in the computer readable storage medium, wherein the computer program, when executed by a processor, implements the steps of the input/output scheduling method according to any one of claims 1 to 6.
  10. 10. A computer program product comprising a computer program which, when executed by a processor, implements the steps of the input-output scheduling method of any one of claims 1 to 6.

Description

Input/output scheduling method and device, electronic equipment and storage medium Technical Field The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for scheduling input and output, an electronic device, and a storage medium. Background Distributed full flash memory is used as a core technology of a modern storage system and is widely applied to scenes such as cloud storage, databases and the like with severe delay sensitivity and throughput requirements. In the prior art, under a mixed read-write load scene, because the mapping relation between the prediction model and the storage medium characteristic is not established, the key request delay is uncontrollable when the load suddenly increases, the throughput of the system fluctuates, and the overall performance and the stability of the system are affected. Disclosure of Invention The application provides an input and output scheduling method, an input and output scheduling device, electronic equipment and a storage medium, which at least solve the problem that the overall performance and stability of a system are affected due to fluctuation of system throughput in the related technology. The application provides an input/output scheduling method, which comprises the following steps: Acquiring historical input and output request data of each storage node in a storage system; Calculating to obtain the load state of the storage system in a future period based on the historical input and output request data; according to the load state, adjusting response sequences of input and output requests with different priorities; And adjusting the resource allocation strategy of the storage node based on the response sequence. Optionally, the historical input/output request data includes at least one of request type, request priority, logical address range of request access, data amount of request, response time of request. Optionally, the calculating, based on the historical input/output request data, a load state of the storage system in a future period includes: And processing the historical input and output request data based on a time sequence prediction model to obtain a load change trend of a future period, wherein the load state comprises at least one of load level, read-write request proportion, input and output operand per second or bandwidth requirement. Optionally, the adjusting the response sequence of the input/output requests with different priorities according to the load state includes: When the load state is load aggravated, the response sequence of the high-priority request is promoted; And when the load state is load relief, promoting the response sequence of the low-priority requests. Optionally, the adjusting the resource allocation policy of the storage node based on the response sequence includes: Adjusting the cache strategies of different storage media; Adjusting the bandwidth allocation proportion of the input/output channels; coordinating garbage collection operations among a plurality of storage nodes; data with access frequency higher than a preset threshold value is migrated. Optionally, the method further comprises: when data is transmitted between storage nodes, integrity checking is performed on the transmitted data blocks. The application also provides an input/output scheduling device, which comprises: The acquisition unit is used for acquiring historical input and output request data of each storage node in the storage system; the calculating unit is used for calculating the load state of the storage system in a future period based on the historical input and output request data; the adjusting unit is used for adjusting the response sequence of the input/output requests with different priorities according to the load state; And the adjusting unit is also used for adjusting the resource allocation strategy of the storage node based on the response sequence. Optionally, the historical input/output request data includes at least one of request type, request priority, logical address range of request access, data amount of request, response time of request. Optionally, the computing unit is further configured to: And processing the historical input and output request data based on a time sequence prediction model to obtain a load change trend of a future period, wherein the load state comprises at least one of load level, read-write request proportion, input and output operand per second or bandwidth requirement. Optionally, the adjusting unit is further configured to: When the load state is load aggravated, the response sequence of the high-priority request is promoted; And when the load state is load relief, promoting the response sequence of the low-priority requests. Optionally, the adjusting unit is further configured to: Adjusting the cache strategies of different storage media; Adjusting the bandwidth allocation proportion of the input/output channels; coordinat