Tiered data storage system using mobility scoring
View Patent ↗The system and methods disclosed herein relate to an improvement in automated data tiering technology. In these embodiments, users are able to weigh the importance of read, write, and pre-fetch operations in terms of tier placement within the data storage system. Data relocation evaluations are performed by an off-load engine, e.g., a graphics processing unit, which utilizes parallel processing on data arrays/vectors/extents. The data relocation candidates are identified via calculating a mobility score on an extent-by-extent basis. Data are then relocated within the various tiers of the data storage system.
1. A method for tiering data in a data storage device using an offload engine to calculate a mobility score comprising:
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of read operations performed on a first extent stored in the data storage device;
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of pre-fetch operations performed on the first extent; and
evaluating a tier location of the first extent based on the weighted, short-term moving exponential averages for the plurality of read, write, and pre-fetch operations to determine if the first extent should be moved to a new tier location.
2. The method of claim 1 further comprising moving the first extent to the new tier location.
3. The method of claim 1 wherein the offload engine is a graphics processing unit.
4. The method of claim 1 further comprising:
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of read operations performed on the first extent stored in the data storage device;
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of pre-fetch operations performed on the first extent; and
wherein the evaluating a tier location of the first extent is further based on the weighted, long-term moving exponential averages for the plurality of read, write, and pre fetch operations to determine if the first extent should be moved to a new tier location.
5. The method of claim 4 wherein the offload engine is a graphics processing unit.
6. A method for tiering data in a data storage device using, an offload engine to calculate a mobility score comprising:
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of read operations performed on a first extent stored in the data storage device;
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of pre-fetch operations performed on the first extent; and
evaluating a tier location of the first extent based on the weighted, long-term moving exponential averages for the plurality of read, write, and pre-fetch operations to determine if the first extent should be moved to a new tier location.
7. The method of claim 6 further comprising moving the first extent to the new tier location.
8. The method of claim 6 wherein the offload engine is a graphics processing unit.
9. A tiered data storage system comprising:
a processor;
a storage resource;
a memory; and
an offload engine configured for:
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of read operations performed on a first extent stored in the data storage device;
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, short-term moving exponential average for a plurality of pre-fetch operations performed on the first extent; and
evaluating a tier location of the first extent based on the weighted, short-term moving exponential averages for the plurality of read, write, and pre-fetch operations to determine if the first extent should be moved to a new tier location.
10. The tiered data storage system of claim 9 further configured for moving the first extent to the new tier location.
11. The tiered data storage system of claim 9 wherein the offload engine is a graphics processing unit.
12. The tiered data storage system of claim 9 further comprising:
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of read operations performed on the first extent stored in the data storage device;
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, long-term moving exponential average for the plurality of pre-fetch operations performed on the first extent; and
wherein the evaluating a tier location of the first extent is further based on the weighted, long-terms moving exponential averages for the plurality of read, write, and pre-fetch operations to determine if the first extent should be moved to anew tier location.
13. The tiered data storage system of claim 12 wherein the offload engine is a graphics processing, unit.
14. A tiered data storage system comprising:
a processor;
a storage resource;
a memory; and
an offload engine configured for:
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of read operations performed on a first extent stored in the data storage device;
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of write operations performed on the first extent;
using the offload engine to calculate a weighted, long-term moving exponential average for a plurality of pre-fetch operations performed on the first extent; and
evaluating a tier location of the first extent based on the weighted, long-term moving exponential averages for the plurality of read, write, and pre-fetch operations to determine if the first extent should be moved to a new tier location.
15. The tiered data storage system of claim 14 further configured for moving the first extent to the new tier location.
16. The tiered data storage system of claim 14 wherein the offload engine is a graphics processing unit.