Deduplicated and compressed non-volatile memory cache

1. A method for applying deduplication, compression, or both, to data stored or to be stored at a non-volatile memory (NVM) cache, comprising:

identifying the NVM cache;

determining whether deduplication or compression, or both, is to be applied to data stored or to be stored at the NVM cache; and

applying either deduplication or compression, or both, to the data stored or to be stored at the NVM cache,

wherein containers are created by concatenating data directly in the NVM cache without first creating the containers in a dynamic random access memory (DRAM) when compression is applied,

wherein whether data sections are identical is determined based on cryptographic hashes when deduplication is applied, wherein in response to deduplication being applied, a first index and a second index are maintained in the NVM cache to map client keys to locations of the NVM cache, the first index mapping client keys to hashes, the second index mapping the hashes to locations of the NVM cache and including, in each entry comprising a hash-NVM cache location pair, a reference count indicative of a quantity of live client keys referencing the respective hash, wherein when a reference count reaches zero, NVM cache space associated with the respective NVM cache location is recovered for future reuse, wherein in response to deduplication not being applied, a third index is maintained to map client keys directly to locations of the NVM cache without use of hashes, each entry of the third index including a size of a corresponding data chunk, and

wherein based on a level of system computational loads at a time of data storage, deduplication, compression, or both, are applied either inline as the data is being written to the NVM cache or at a later time after the data has been written to the NVM cache.

2. The method of claim 1 , wherein deduplication, compression, or both, are applied inline as the data is being written to the NVM cache.

3. The method of claim 1 , wherein deduplication, compression, or both, are applied after the data has been written to the NVM cache.

4. The method of claim 1 , wherein deduplication is applied, and wherein whether data sections are identical is determined using byte-by-byte comparison.

5. The method of claim 1 , wherein compression is applied, and wherein log-structured storage is used to manage space of the NVM cache.

6. The method of claim 1 , wherein both deduplication and compression are applied, and deduplication is applied before compression.

7. A data processing system, comprising:

a processor; and

a memory coupled to the processor storing instructions which, when executed by the processor, cause the processor to perform testing operations, the operations including

identifying the NVM cache;

determining whether deduplication or compression, or both, is to be applied to data stored or to be stored at the NVM cache; and

applying either deduplication or compression, or both, to the data stored or to be stored at the NVM cache,

wherein containers are created by concatenating data directly in the NVM cache without first creating the containers in a dynamic random access memory (DRAM) when compression is applied,

wherein whether data sections are identical is determined based on cryptographic hashes when deduplication is applied, wherein in response to deduplication being applied, a first index and a second index are maintained in the NVM cache to map client keys to locations of the NVM cache, the first index mapping client keys to hashes, the second index mapping the hashes to locations of the NVM cache and including, in each entry comprising a hash-NVM cache location pair, a reference count indicative of a quantity of live client keys referencing the respective hash, wherein when a reference count reaches zero, NVM cache space associated with the respective NVM cache location is recovered for future reuse, wherein in response to deduplication not being applied, a third index is maintained to map client keys directly to locations of the NVM cache without use of hashes, each entry of the third index including a size of a corresponding data chunk, and

wherein based on a level of system computational loads at a time of data storage, deduplication, compression, or both, are applied either inline as the data is being written to the NVM cache or at a later time after the data has been written to the NVM cache.

8. The data processing system of claim 7 , wherein deduplication, compression, or both, are applied inline as the data is being written to the NVM cache.

9. The data processing system of claim 7 , wherein deduplication, compression, or both, are applied after the data has been written to the NVM cache.

10. The data processing system of claim 7 , wherein deduplication is applied, and wherein whether data sections are identical is determined using byte-by-byte comparison.

11. The data processing system of claim 7 , wherein compression is applied, and wherein log-structured storage is used to manage space of the NVM cache.

12. The data processing system of claim 7 , wherein both deduplication and compression are applied, and deduplication is applied before compression.

13. A non-transitory machine-readable medium having instructions stored therein which, when executed by a processor, cause the processor to perform testing operations, the operations comprising:

identifying the NVM cache;

determining whether deduplication or compression, or both, is to be applied to data stored or to be stored at the NVM cache; and

applying either deduplication or compression, or both, to the data stored or to be stored at the NVM cache,

wherein containers are created by concatenating data directly in the NVM cache without first creating the containers in a dynamic random access memory (DRAM) when compression is applied,

wherein whether data sections are identical is determined based on cryptographic hashes when deduplication is applied, wherein in response to deduplication being applied, a first index and a second index are maintained in the NVM cache to map client keys to locations of the NVM cache, the first index mapping client keys to hashes, the second index mapping the hashes to locations of the NVM cache and including, in each entry comprising a hash-NVM cache location pair, a reference count indicative of a quantity of live client keys referencing the respective hash, wherein when a reference count reaches zero, NVM cache space associated with the respective NVM cache location is recovered for future reuse, wherein in response to deduplication not being applied, a third index is maintained to map client keys directly to locations of the NVM cache without use of hashes, each entry of the third index including a size of a corresponding data chunk, and

wherein based on a level of system computational loads at a time of data storage, deduplication, compression, or both, are applied either inline as the data is being written to the NVM cache or at a later time after the data has been written to the NVM cache.

14. The non-transitory machine-readable medium of claim 13 , wherein deduplication, compression, or both, are applied inline as the data is being written to the NVM cache.

15. The non-transitory machine-readable medium of claim 13 , wherein deduplication, compression, or both, are applied after the data has been written to the NVM cache.

16. The non-transitory machine-readable medium of claim 13 , wherein deduplication is applied, and wherein whether data sections are identical is determined using byte-by-byte comparison.

17. The non-transitory machine-readable medium of claim 13 , wherein compression is applied, and wherein log-structured storage is used to manage space of the NVM cache.

18. The non-transitory machine-readable medium of claim 13 , wherein both deduplication and compression are applied, and deduplication is applied before compression.