Accelerating data reduction through reinforcement learning
View Patent ↗Compressing data of a storage device includes selecting a portion of data for data compression that is predicted to be unlikely to be accessed, selecting a particular one of a plurality of data compression algorithms to apply to the portion according to a frequency value associated with each of the data compression algorithms, and adjusting the frequency value of the particular one of a plurality of data compression algorithms according to performance of the particular one of a plurality of data compression algorithms. The performance may vary according to a ratio of amount of compression achieved to processing cycles of a processor used to compress the portion of data. The processor may perform storage device functions that are separate from data compression. The portion of data may be selected from a plurality of logical devices used in connection with a single application that accesses the storage device.
1. A method of compressing data of a storage device, comprising:
selecting a portion of data for data compression based on past access patterns of the portion of data;
selecting a particular one of a plurality of data compression algorithms to apply to the portion according to a frequency value associated with each of the data compression algorithms; and
adjusting the frequency value of the particular one of a plurality of data compression algorithms according to performance of the particular one of a plurality of data compression algorithms, wherein the frequency value is decreased in response to the portion of data being accessed after being compressed.
2. The method, according to claim 1 , wherein the performance varies according to a ratio of amount of compression achieved to processing cycles of a processor used to compress the portion of data.
3. The method, according to claim 2 , wherein the processor performs storage device functions that are separate from data compression.
4. The method, according to claim 1 , wherein the portion of data is selected from a plurality of logical devices used in connection with a single application that accesses the storage device.
5. The method, according to claim 1 , wherein the portion is selected based on the portion not having been previously compressed, the portion not being busy, and the portion not having been accessed recently.
6. The method, according to claim 1 , wherein the storage device includes a hardware compression device that is used in connection with at least some of the compression algorithms.
7. The method, according to claim 6 , wherein the performance varies according to a ratio of an amount of compression achieved to processing cycles used by a processor to compress the portion of data, wherein the processor is separate from the hardware compression device.
8. The method, according to claim 7 , wherein the processor performs storage device functions that are separate from data compression.
9. The method, according to claim 1 , wherein the data is compressed using out-of-band data compression that is independent of external accesses to the storage device.
10. The method, according to claim 9 , wherein in-line data compression is performed in connection with external accesses of data on the storage device and wherein the in-line data compression is independent of the out-of-band data compression.
11. A non-transitory computer-readable medium that contains software that compresses data of a storage device, the software comprising:
executable code that selects a portion of data for data compression based on past access patterns of the portion of data;
executable code that selects a particular one of a plurality of data compression algorithms to apply to the portion according to a frequency value associated with each of the data compression algorithms; and
executable code that adjusts the frequency value of the particular one of a plurality of data compression algorithms according to performance of the particular one of a plurality of data compression algorithms, wherein the frequency value is decreased in response to the portion of data being accessed after being compressed.
12. The non-transitory computer-readable medium, according to claim 11 , wherein the performance varies according to a ratio of amount of compression achieved to processing cycles of a processor used to compress the portion of data.
13. The non-transitory computer-readable medium, according to claim 12 , wherein the processor performs storage device functions that are separate from data compression.
14. The non-transitory computer-readable medium, according to claim 11 , wherein the portion of data is selected from a plurality of logical devices used in connection with a single application that accesses the storage device.
15. The non-transitory computer-readable medium, according to claim 11 , wherein the portion is selected based on the portion not having been previously compressed, the portion not being busy, and the portion not having been accessed recently.
16. The non-transitory computer-readable medium, according to claim 11 , wherein the storage device includes a hardware compression device that is used in connection with at least some of the compression algorithms.
17. The non-transitory computer-readable medium, according to claim 16 , wherein the performance varies according to a ratio of an amount of compression achieved to processing cycles used by a processor to compress the portion of data, wherein the processor is separate from the hardware compression device.
18. The non-transitory computer-readable medium, according to claim 17 , wherein the processor performs storage device functions that are separate from data compression.
19. The non-transitory computer-readable medium, according to claim 11 , wherein the data is compressed using out-of-band data compression that is independent of external accesses to the storage device.
20. The non-transitory computer-readable medium, according to claim 19 , wherein in-line data compression is performed in connection with external accesses of data on the storage device and wherein the in-line data compression is independent of the out-of-band data compression.