Data reduction techniques for use with caching
Techniques for cache management may include: partitioning a cache into buckets of cache pages, wherein each bucket has an associated cache page size and each bucket includes cache pages of the associated cache page size for that bucket, wherein the cache includes compressed pages of data and uncompressed pages of data; and performing processing that stores a first page of data in the cache. The processing may include storing the first page of data in a first cache page of a selected bucket having a first associated cache page size determined in accordance with a first compressed size of the first page of data. The cache may be repartitioned among the buckets based on associated access frequencies of the buckets of cache pages.
1. A method for cache management comprising:
partitioning a cache into a plurality of buckets of cache pages, wherein each of the plurality of buckets has an associated cache page size and said each bucket includes a plurality of cache pages each of the associated cache page size for said each bucket, wherein the cache includes compressed pages of data and uncompressed pages of data; and
performing processing that stores a first page of data in the cache, said processing including:
storing the first page of data in a first cache page of a selected bucket of the plurality of buckets having a first associated cache page size determined in accordance with a first compressed size of the first page of data.
2. The method of claim 1 , further comprising:
receiving at least one access frequency metric for each of the plurality of buckets;
determining, in accordance with the at least one access frequency metric for each of the plurality of buckets, whether to repartition the cache among the plurality of buckets; and
responsive to determining to repartition the cache among the plurality of buckets, performing cache repartitioning, wherein said cache repartitioning includes reallocating at least a first portion of the cache from a first of the plurality of buckets to a second of the plurality of buckets in accordance with the at least one access frequency metric for each of the plurality of buckets.
3. The method of claim 2 , wherein said cache repartitioning comprises:
determining the first bucket as a donor bucket from which the first portion of the cache is selected; and
determining the second bucket as a receiving bucket to which the first portion of the cache is reallocated from the first bucket.
4. The method of claim 3 , wherein said determining whether to repartition the cache further comprises:
determining a first access frequency metric for the first bucket of cache pages exceeds a specified threshold of allowable variation or difference; and
responsive to determining the first access frequency metric for the first bucket of cache pages exceeds the specified threshold of allowable variation or difference, determining to repartition the cache among the plurality of buckets and performing said cache repartitioning.
5. The method of claim 4 , further comprising:
determining that the first access frequency metric indicates that the first bucket has a lower access frequency than any other of the plurality of buckets.
6. The method of claim 3 , wherein said determining whether to repartition the cache further comprises:
determining a second access frequency metric for the second bucket of cache pages exceeds a specified threshold of allowable variation or difference; and
responsive to determining the second access frequency metric for the first bucket of cache pages exceeds the specified threshold of allowable variation or difference, determining to repartition the cache among the plurality of buckets and performing said cache repartitioning.
7. The method of claim 6 , further comprising:
determining that the second access frequency metric indicates that the second bucket has a higher access frequency than any other of the plurality of buckets.
8. The method of claim 2 , wherein the first portion of the cache is determined as a least recently used cache portion of the first bucket.
9. The method of claim 2 , wherein each of the plurality of buckets has a different associated cache page size.
10. The method of claim 9 , wherein the plurality of buckets includes one bucket of cache pages including the uncompressed pages of data, wherein other ones of the plurality of buckets of cache pages are used to store the compressed pages of data, and wherein the other ones of the plurality of buckets include different sizes of compressed cache pages.
11. The method of claim 10 , wherein said processing that stores a first page of data in the cache further comprises:
determining whether an associated access frequency for the first page of data exceeds a threshold;
responsive to determining the associated access frequency for the first page of data exceeds the threshold, storing the first page of data in a cache page of said one bucket of cache pages used to store the uncompressed pages of data; and
responsive to determining the associated access frequency for the first page of data does not exceed the threshold, performing said storing the first page of data in the first cache page of the selected bucket of the plurality of buckets having the first associated cache page size determined in accordance with the first compressed size of the first page of data, wherein the selected bucket is one of the other ones of the plurality of buckets.
12. The method of claim 2 , wherein the at least one access frequency metric for each of the plurality of buckets is determined in accordance with cache hits of said each bucket of cache pages.
13. The method of claim 12 , wherein the at least one access frequency metric for each of the plurality of buckets is determined by dividing the cache hits of said each bucket by a size of said each bucket denoting an amount of the cache allocated for said each bucket.
14. The method of claim 1 , wherein the cache is a read cache that includes only read data for read I/O operations.
15. The method of claim 14 , wherein a write cache that is separate from the read cache is used to store write data for write I/O operations.
16. The method of claim 1 , wherein the cache includes data used in connection with read and write I/O operations.
17. The method of claim 2 , wherein the at least one access frequency metric for each of the plurality of buckets includes a first access frequency metric determined in accordance with a number of pages in said each bucket having a number of cache hits or accesses below a specified threshold number of cache hits or accesses, and wherein the first bucket of the plurality of buckets has a maximum value for the first access frequency metric with respect to said plurality of buckets and wherein the first bucket is selected as a donor bucket from which the first portion of the cache is selected, and wherein the second bucket of the plurality of buckets has a minimum value for the first access frequency metric with respect to said plurality of buckets and wherein the second bucket is selected as a receiving bucket to which the first portion of the cache is reallocated.
18. The method of claim 2 , wherein the cache includes a plurality of pages partitioned into the plurality of buckets, and wherein the method further comprises:
receiving at least one access frequency metric for each of the plurality of pages;
determining, in accordance with the at least one access frequency metric for each of the plurality of pages, a first portion of the plurality of pages to select as a plurality of donor pages;
determining one or more receiving buckets; and
moving the plurality of donor pages to the one or more receiving buckets.
19. A system comprising:
at least one processor; and
at least one memory comprising code stored thereon that, when executed, performs a method for cache management comprising:
partitioning a cache into a plurality of buckets of cache pages, wherein each of the plurality of buckets has an associated cache page size and said each bucket includes a plurality of cache pages each of the associated cache page size for said each bucket, wherein the cache includes compressed pages of data and uncompressed pages of data; and
performing processing that stores a first page of data in the cache, said processing including:
storing the first page of data in a first cache page of a selected bucket of the plurality of buckets having a first associated cache page size determined in accordance with a first compressed size of the first page of data.
20. A non-transitory computer readable medium comprising code stored thereon that, when executed, performs a method for cache management comprising:
partitioning a cache into a plurality of buckets of cache pages, wherein each of the plurality of buckets has an associated cache page size and said each bucket includes a plurality of cache pages each of the associated cache page size for said each bucket, wherein the cache includes compressed pages of data and uncompressed pages of data; and
performing processing that stores a first page of data in the cache, said processing including:
storing the first page of data in a first cache page of a selected bucket of the plurality of buckets having a first associated cache page size determined in accordance with a first compressed size of the first page of data.