Method and system for adaptive wear leveling in solid state memory
A method for controlling wear level operations in solid state memory. The method includes receiving a request to write to a memory location identified by a write address of the solid state memory, and making a first determination that the writing to the memory location results in a duplicate write to the memory location. The method further includes, based on the first determination, making a second determination that a number of encountered duplicate writes has reached an adaptively controlled maximum number of duplicate writes, and based on the second determination, performing a wear level operation.
1. A method for controlling wear level operations in solid state memory, the method comprising:
receiving a request to write to a first memory location identified by a write address of the solid state memory;
making a first determination that a duplicate write to the first memory location occurred;
making a second determination, based on the first determination, that a number of encountered duplicate writes has reached an adaptively controlled maximum number of duplicate writes, wherein making the second determination comprises:
generating a first hash of the write address using a first hash function;
generating a second hash of the write address using a second hash function;
matching the first hash to a first location in a compact representation of memory locations of the solid state memory;
matching the second hash to a second location in the compact representation;
making an assessment, using a probabilistic scale, that the matching of the first hash and the matching of the second hash results in a positive match that is more likely to be true than false; and
decrementing, based on the assessment, a divider count by a quantity less than 1, wherein the difference between the quantity and 1 represents a probability that the positive match is false; and
performing, based on the second determination, a wear level operation.
2. The method of claim 1 , wherein the number of encountered duplicate writes is a total number of duplicate writes to the first memory location.
3. The method of claim 2 , wherein a frequency of encountered duplicate writes is determined in a time window selected from a group consisting of a sliding time window and a time window with a fixed beginning.
4. The method of claim 3 , wherein the time window has a fixed duration.
5. The method of claim 2 , wherein a frequency of encountered duplicate writes is obtained from one selected from a group consisting of:
a ratio of a number of duplicate writes to memory locations to a number of non-duplicate writes to memory locations, and
a ratio of a number of duplicate writes to memory locations to a number of writes to memory locations.
6. The method of claim 2 , wherein the adaptively controlled maximum number of duplicate writes is adjusted based on one selected from the group consisting of:
linearly modulating the adaptively controlled maximum number of duplicate writes, based on a frequency of encountered duplicate writes, and
stepwise modulating the adaptively controlled maximum number of encountered duplicate writes, based on the frequency of observed duplicate writes.
7. A storage appliance comprising: storage modules comprising solid state memory, wherein the storage appliance is configured to: receive a request to write to a first memory location identified by a write address of the solid state memory; make a first determination that a duplicate write to the first memory location occurred; make a second determination, based on the first determination, that a number of encountered duplicate writes has reached an adaptively controlled maximum number of duplicate writes, wherein making the second determination comprises: generating a first hash of the write address using a first hash function; generating a second hash of the write address using a second hash function; matching the first hash to a first location in a compact representation of memory locations of the solid state memory; and matching the second hash to a second location in the compact representation; make an assessment, using a probabilistic scale, that the matching of the first hash and the matching of the second hash results in a positive match that is more likely to be true than false, decrementing, based on the assessment, a divider count by a quantity less than 1, wherein the difference between the quantity and 1 represents a probability that the positive match is false; and perform, based on the second determination 1 a wear level operation.
8. The storage appliance of claim 7 , wherein the number of encountered duplicate writes is a total of number of duplicate writes to the first memory location.
9. The storage appliance of claim 8 , wherein a frequency of encountered duplicate writes is determined in a time window selected from a group consisting of a sliding time window and a time window with a fixed beginning.
10. The storage appliance of claim 9 , wherein the time window has a fixed duration.
11. The storage appliance of claim 8 , wherein a frequency of encountered duplicate writes is obtained from one selected from a group consisting of:
a ratio of a number of duplicate writes to memory locations to a number of non-duplicate writes to memory locations, and
a ratio of a number of duplicate writes to memory locations to a number of writes to memory locations.
12. The storage appliance of claim 8 , wherein the adaptively controlled maximum number of duplicate writes is adjusted based on one selected from the group consisting of:
linearly modulating the adaptively controlled maximum number of duplicate writes, based on a frequency of encountered duplicate writes, and
stepwise modulating the adaptively controlled maximum number of encountered duplicate writes, based on the frequency of observed duplicate writes.
13. A non-transitory computer readable medium (CRM) comprising instructions that enable a storage appliance to: receive a request to write to a first memory location identified by a write address of solid state memory; make a first determination that a duplicate write to the first memory location occurred; make a second determination that a number of encountered duplicate writes has reached an adaptively controlled maximum number of duplicate writes, wherein making the second determination comprises: generating a first hash of the write address using a first hash function; generating a second hash of the write address using a second hash function; matching the first hash to a first location in a compact representation of memory locations of the solid state memory; and matching the second hash to a second location in the compact representation; making an assessment, using a probabilistic scale, that the matching of the first hash and the matching of the second hash results in a positive match that is more to be true than false, decrementing, based on the assessment, a divider count by a quantity less than 1, wherein the difference between the quantity and 1 represents a probability that the positive match is false; and perform, based on the second determination a wear level operation.
14. The non-transitory CRM of claim 13 , wherein the number of encountered duplicate writes is a total of number of duplicate writes to the first memory location.
15. The non-transitory CRM of claim 14 , wherein a frequency of encountered duplicate writes is determined in a time window selected from a group consisting of a sliding time window and a time window with a fixed beginning.
16. The non-transitory CRM of claim 15 , wherein the time window has a fixed duration.
17. The non-transitory CRM of claim 14 , wherein a frequency of encountered duplicate writes is obtained from one selected from a group consisting of:
a ratio of a number of duplicate writes to memory locations to a number of non-duplicate writes to memory locations, and
a ratio of a number of duplicate writes to memory locations to a number of writes to memory locations.