CN113111014B - Method, device, equipment and storage medium for cleaning non-hot data in cache - Google Patents

Abstract

The invention discloses a cleaning method, device, equipment and storage medium for non-hot data in the cache. The cleaning method for the non-hot data in the cache includes: step 1, partitioning the cache data; step 2, setting the execution frequency HZ and each time Execution time limit TIME; step 3, according to the execution frequency HZ set in step 2 and each execution time limit TIME, randomly extract the data in the partition of the step 1 to clean up non-hot data. The strategy of regular cleaning is used to clean the non-hot data in the storage system cache, and the friendliness to the cpu and storage space is ensured by setting the frequency and the maximum execution time, and the data block cleaning is ensured through data block partitioning and random selection of data blocks. efficiency and smoothness across the cache space.

Description

Method, device, equipment and storage medium for cleaning non-hot data in cache

technical field

The present invention relates to the technical field of clearing cached data, in particular to a method, device, device and storage medium for clearing non-hot data in a cache.

Background technique

There are many kinds of storage media in the storage system, and their read and write performance also has a large gap. In order to improve the access performance of the overall system, high-performance devices are often used as cache devices for low-performance devices. At the same time, due to the cost of high-tech equipment, the capacity of cache equipment is often relatively small. Therefore, it is often necessary to conduct heat statistics on cache data and use certain strategies to clean up non-hot data. Generally speaking, there are three different cleaning strategies, immediate cleanup, lazy cleanup, and scheduled cleanup.

Immediate cleaning is to create a callback condition when setting the data temperature. When the temperature drops to the threshold, the processing logic will automatically clean the data. This solution is not very cpu-friendly and will take up a lot of cpu time.

Lazy cleanup means that the data will not be processed if it expires, until the next access to this part of the data, if the data is found to be cold, the data will be cleaned up. This solution is not very friendly to storage and will take up a lot of storage space.

Scheduled cleaning is to check the data at regular intervals and clean and cool down the data. This will reduce the impact of cleaning on cpu and storage space. However, if it is executed too frequently or for too long, the regular cleaning strategy will degenerate into an immediate cleaning strategy, which will consume a lot of cpu time. If it is executed too little, or the execution time is too short, the scheduled cleanup strategy will degenerate into an inert cleanup strategy, resulting in a waste of storage space.

Contents of the invention

The object of the present invention is to solve the above problems, and provide a method, device, equipment and storage medium for cleaning non-hot data in the cache. Regular cleaning through data block partitioning and random selection of data blocks ensures the efficiency of data block cleaning and the smoothness of the entire cache space.

In order to achieve the above object, the present invention adopts the following technical solutions:

Methods for cleaning non-hot data in the cache, including:

Step 1, partition the cached data;

Step 2, set the execution frequency HZ and each execution time limit TIME;

Step 3: According to the execution frequency HZ set in step 2 and the execution time limit TIME for each execution, randomly extract the data in the partition of the step 1 to clean up the non-hot data.

In the first step, the cached data is partitioned, and the data in the partition is divided into multiple data blocks.

The time limit for each execution TIME=10000*the maximum acceptable number of data blocks detected each time/HZ, and the execution frequency is inversely proportional to the time limit for each execution. This ensures that each elimination will not block too many requests.

In said step 3, check one partition each time according to the set execution frequency HZ, each partition randomly selects a set data block for inspection, and cleans up if the heat is overdue; if the cleaning is completed or reaches the longest execution time, then Finish.

In the step 3, if there are data blocks larger than the set value among the randomly set number of data blocks in the partition to be cleaned this time, then the next cleaning cycle will continue to clean up the data blocks in the partition this time.

Each data block is checked up to 3 times in a row.

A device for cleaning non-hot data in a cache, comprising:

The data area division module is used to partition the cache data;

Limit setting module, used to set the execution frequency HZ and each execution time limit TIME;

The non-hot data cleaning module is used to randomly extract the data in the partition according to the set execution frequency HZ and each execution time limit TIME to clean the non-hot data.

A device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method for clearing non-hot data in the cache when executing the computer program.

A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for cleaning non-hot data in the cache are implemented.

Beneficial effects of the present invention:

The present invention adopts the strategy of timing cleaning to clean up non-hot data in the cache of the storage system, and ensures friendliness to the cpu and storage space by setting the frequency and the maximum execution time, and ensures the data block by partitioning the data block and randomly selecting the data block Cleanup efficiency and smoothness across the cache space.

By performing a cleaning operation at regular intervals, and by limiting the duration and frequency of cleaning operations, the impact of deletion operations on the CPU and the waste of storage space are reduced.

detailed description

The present invention will be further described below in conjunction with embodiment.

Methods for cleaning non-hot data in the cache, including:

(1) Set the execution frequency HZ.

(2) Through the formula: each execution time limit TIME=10000*acceptable maximum number of detected data blocks/HZ to obtain each execution time limit TIME, the execution frequency is inversely proportional to each execution time limit, like this It can be guaranteed that each elimination will not block too many requests.

(3) Divide the cached data into ten areas from 0 to 9, for example.

(4) Check the data block on time according to the execution frequency.

(5) Check one partition at a time, randomly select 20 data blocks from each partition for inspection, and clean up if the heat is overdue. If the cleanup is complete or the maximum execution time is reached, it will end.

(6) If more than 15 out of 20 random areas in this cleanup area (other values can be configured) need to be cleaned up, then the next cleanup cycle will continue to clean up this data block, and each data block will be checked up to 3 times in a row. In other cases check the next data block.

In the cache management of the storage system, the present invention uses a fixed-frequency, fixed-time, long-term, and on-time cleaning strategy of randomly extracting data by area for data partitioning to complete cache space management and reduce resource consumption.

A device for cleaning non-hot data in a cache, comprising:

The data area division module is used to partition the cache data;

Limit setting module, used to set the execution frequency HZ and each execution time limit TIME;

The non-hot data cleaning module is used to randomly extract the data in the partition according to the set execution frequency HZ and each execution time limit TIME to clean the non-hot data.

Using a timing cleaning strategy to clean up non-hot data in the cache device in the storage system by randomly extracting and cleaning the cache data partition and setting the frequency and maximum execution time can improve the efficiency of cleaning data and effectively utilize CPU performance and cache space.

A device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method for clearing non-hot data in the cache when executing the computer program.

A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for clearing non-hot data in the cache are realized.

Although the specific implementation of the present invention has been described above, it is not a limitation to the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, those skilled in the art can do it without creative work. Various modifications or deformations are still within the protection scope of the present invention.

Claims (7)

1. The method for cleaning the non-hot data in the cache is characterized by comprising the following steps:

partitioning cache data;

setting an execution frequency HZ and TIME limit of each execution TIME;

step three, randomly extracting the data in the partition in the step one according to the execution frequency HZ set in the step two and the TIME limit of each execution TIME to clean the non-hotspot data;

in the first step, the cache data is partitioned, and the data in the partition is divided into a plurality of data blocks;

the execution TIME per TIME limit TIME =10000 ×, the number of data blocks per maximum detection/HZ that can be accepted, and the execution frequency is inversely proportional to the execution TIME per TIME limit.

2. The method for clearing non-hot-spot data in cache of claim 1, wherein in the third step, one partition is checked at a time according to the set execution frequency HZ, and each partition randomly extracts a set number of data blocks for checking, and clears the data blocks if the heat degree is expired; and ending if the cleaning is completed or the longest execution time is reached.

3. The method according to claim 1, wherein in the third step, if any data block with a number greater than a predetermined value in the randomly set number of data blocks in the partition to be cleaned at this time needs to be cleaned, the next cleaning cycle continues to clean the data blocks in the current partition.

4. The method of claim 3, wherein each block is checked at most 3 times in succession.

5. A device for cleaning non-hot data in a cache by using the method for cleaning non-hot data in a cache according to claim 1, comprising:

the data area division module is used for dividing the cache data into partitions;

the limit setting module is used for setting an execution frequency HZ and a TIME limit for each execution TIME;

and the non-hotspot data cleaning module is used for randomly extracting data in the partition according to the set execution frequency HZ and the TIME limit of each execution TIME to clean the non-hotspot data.

6. An apparatus comprising a memory and a processor, the memory storing a computer program, wherein the processor when executing the computer program implements the steps of the method for cleaning up non-hot data in a cache according to any one of claims 1 to 4.

7. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for cleaning up non-hotspot data in a cache memory of any one of claims 1 to 4.