Using a migration threshold and a candidate list for cache management of sequential write storage

Patent metadata

Field	Value
Publication number	US-9348517-B2
Application number	US-201414471438-A
Country	US
Kind code	B2
Filing date	Aug 28, 2014
Priority date	Aug 28, 2014
Publication date	May 24, 2016
Grant date	May 24, 2016

Abstract

Embodiments of the invention relate to cache management of sequential write storage. Cache states of cache blocks in cache storage are tracked. The tracking incudes creating a migration candidate list that organizes the cache blocks into virtual volumes. Based on determining that a cache migration threshold has been reached, at least a subset of the cache blocks are selected from the migration candidate list for migration. The selected cache blocks are ordered into a sequential order based on the migration candidate list. At least a subset of the selected cache blocks are migrated in the sequential order, the migrating including writing contents of the selected cache blocks to the sequential write storage.

First claim

What is claimed is:

1. A method of cache management for sequential write storage, the method comprising:
tracking cache states of cache blocks in a cache storage, the tracking including creating a migration candidate list that organizes the cache blocks into virtual volumes;
determining that a migration threshold has been reached; and
based on reaching the migration threshold:
selecting a plurality of the cache blocks from the migration candidate list for migration;
ordering the selected cache blocks into a sequential order based on the migration candidate list, wherein the selected cache blocks that are located on the same virtual volume are grouped together in the sequential order; and
migrating, in the sequential order, at least a subset of the selected cache blocks, the migrating comprising, for each of the selected cache blocks in the subset, writing contents of the selected cache blocks to the sequential write storage.

2. The method of claim 1 , wherein the migrating further comprises:
periodically, based on the writing being completed for one or more of the selected cache blocks in the subset, updating the cache state of the one or more of the selected cache blocks in the subset to indicate that contents of the one or more of the selected cache blocks in the subset have completed migration to the sequential write storage.

3. The method of claim 2 , wherein:
the ordering includes assigning migration sequence numbers (MSNs) to the selected cache blocks to indicate the sequential order; and
the updating the cache state includes updating a commit sequence number (CSN) with an MSN associated with a selected cache block whose contents have completed migration to the sequential write storage.

4. The method of claim 3 , wherein:
a selected cache block has not completed migration to the sequential write storage and is in a cache state of dirty when the MSN assigned to the selected cache block is larger than the CSN; and
the selected cache block has completed migration to the sequential write storage and is in a cache state of clean when the MSN assigned to the selected cache block is less than or equal to the CSN.

5. The method of claim 1 , wherein the migration threshold is based on an amount of free space in the cache storage.

6. The method of claim 1 , wherein the migration threshold is based on at least one of a number of modified cache blocks belonging to one or more virtual volumes, an elapsed time since a write to one or more of the cache blocks, and a minimum amount of data stored in cache blocks having a cache state of dirty and modified.

7. The method of claim 1 , wherein only cache blocks that have a cache state of dirty and modified are selected for migration.

8. The method of claim 1 , wherein the selecting is based on at least one of an access sequence and a modification order.

9. The method of claim 1 , wherein the selecting is based on at least one of an elapsed time since a write to a cache block and a number of modified cache blocks in one or more virtual volumes.

10. The method of claim 1 , wherein the migration candidate list organizes the cache blocks into the virtual volumes based on application access.

11. The method of claim 1 , wherein the migration candidate list organizes the cache blocks within the virtual volumes based on at least one of a last write timestamp and a virtual extent number.

12. A memory system, comprising:
a cache storage comprising a cache storage device; and
a storage controller coupled to the cache storage and to a sequential write storage device, the memory system configured for:
tracking cache states of cache blocks in a cache storage, the tracking including creating a migration candidate list that organizes the cache blocks into virtual volumes;
determining that a migration threshold has been reached; and
based on reaching the migration threshold:
selecting a plurality of the cache blocks from the migration candidate list for migration;
ordering the selected cache blocks into a sequential order based on the migration candidate list, wherein the selected cache blocks that are located on the same virtual volume are grouped together in the sequential order; and
migrating, in the sequential order, at least a subset of the selected cache blocks, the migrating comprising, for each of the selected cache blocks in the subset, writing contents of the selected cache blocks to the sequential write storage.

13. The system of claim 12 , wherein the migrating further comprises:
periodically, based on the writing being completed for one or more of the selected cache blocks in the subset, updating the cache state of the one or more of the selected cache blocks in the subset to indicate that contents of the one or more of the selected cache blocks in the subset have completed migration to the sequential write storage.

14. The system of claim 13 , wherein:
the ordering includes assigning migration sequence numbers (MSNs) to the selected cache blocks to indicate the sequential order;
the updating the cache state includes updating a commit sequence number (CSN) with an MSN associated with a selected cache block whose contents have completed migration to the sequential write storage;
a selected cache block has not completed migration to the sequential write storage and is in a cache state of dirty when the MSN assigned to the selected cache block is larger than the CSN; and
the selected cache block has completed migration to the sequential write storage and is in a cache state of clean when the MSN assigned to the selected cache block is less than or equal to the CSN.

15. The system of claim 12 , wherein the migration threshold is based on an at least one of an amount of free space in the cache storage, a number of modified cache blocks belonging to one or more virtual volumes, an elapsed time since a write to one or more of the cache blocks, and a minimum amount of data stored in cache blocks having a cache state of dirty and modified.

16. The system of claim 12 , wherein the selecting is based on at least one of an access sequence, a modification order, an elapsed time since a write to a cache block, and a number of modified cache blocks in one or more virtual volumes.

17. The system of claim 12 , wherein the migration candidate list organizes the cache blocks into the virtual volumes based on application access.

18. A computer program product for cache management for sequential write storage, the computer program product comprising:
a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions readable by a processing circuit to cause the processing circuit to perform a method comprising:
tracking cache states of cache blocks in a cache storage, the tracking including creating a migration candidate list that organizes the cache blocks into virtual volumes;
determining that a migration threshold has been reached; and
based on reaching the migration threshold:
selecting a plurality of the cache blocks from the migration candidate list for migration;
ordering the selected cache blocks into a sequential order based on the migration candidate list, wherein the selected cache blocks that are located on the same virtual volume are grouped together in the sequential order; and
migrating, in the sequential order, at least a subset of the selected cache blocks, the migrating comprising, for each of the selected cache blocks in the subset, writing contents of the selected cache blocks to the sequential write storage.

Assignees

• IBM

Inventors

• Hineman Wayne C
• Rees Robert M
• Zhou Pin

Classifications

• G06F12/0871 Primary

  Allocation or management of cache space · CPC title

• G06F3/0665

  at area level, e.g. provisioning of virtual or logical volumes · CPC title

• G06F3/0619

  in relation to data integrity, e.g. data losses, bit errors · CPC title

• G06F2212/604

  Details relating to cache allocation · CPC title

• G06F3/061 Primary

  Improving I/O performance · CPC title