Write-back cache protection

• US 8,327,080 B1
• Filed: 09/28/2010
• Issued: 12/04/2012
• Est. Priority Date: 09/28/2010
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Der, Kenneth P.
• Primary Examiner(s): CHOE, YONG J
• Application Number:
  US12/892,081
• Time in Patent Office: 798 Days
• Field of Search: 711/143, 711/141, 711/144, 711/E12.017
• US Class Current: 711/143
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 12/0804 with main memory updating G06F12/0806 takes precedence
• CPC Class Codes: G06F 12/0868 Data transfer between cache memory and other subsystems, e.g. storage devices or host systems
• CPC Class Codes: G06F 12/0873 Mapping of cache memory to specific storage devices or parts thereof
• CPC Class Codes: G06F 12/0886 Variable-length word access

Abstract

A technique protects host data. The technique involves receiving, at a local computerized node, a block of data from a host computer, the block of data including data sub-blocks. The technique further involves storing the block of data, as a dirty cache block, in a local cache of the local computerized node. The technique further involves performing a set of external caching operations to cache a set of sub-blocks in a set of external computerized nodes in communication with the local computerized node. Each external caching operation caches a respective sub-block of the set of sub-blocks in a cache of a respective external computerized node. The set of sub-blocks includes (i) the data sub-blocks of the block of data from the host and (ii) a set of checksums derived from the data sub-blocks of the block of data from the host.

Claims

1. A local computerized node, comprising:

a communications interface to connect to a host computer and a set of external computerized nodes;

a local cache; and

a controller coupled to the communications interface and the local cache, the controller being constructed and arranged to;

receive, through the communications interface, a block of data from the host computer, the block of data including data sub-blocks,

store the block of data, as a dirty cache block, in the local cache, and

perform, through the communications interface, a set of external caching operations to cache a set of sub-blocks in the set of external computerized nodes, each external caching operation caching a respective sub-block of the set of sub-blocks in a cache of a respective external computerized node, the set of sub-blocks including (i) the data sub-blocks of the block of data from the host computer and (ii) a set of checksums derived from the data sub-blocks of the block of data from the host computer.

2. A local computerized node as in claim 1, further comprising:

a storage interface coupled to the controller, the storage interface being constructed and arranged to connect to a set of non-volatile storage devices; and

wherein the controller is constructed and arranged to store, through the storage interface, the block of data in the set of non-volatile storage devices after performing the set of external caching operations.

3. A local computerized node as in claim 2 wherein the set of non-volatile storage devices includes magnetic disk drives arranged to implement a particular RAID (redundant array of inexpensive disks) level;

and

wherein the controller, when storing the block of data, is constructed and arranged to write the block of data to the magnetic disk drives in accordance with the particular RAID level.

4. A local computerized node as in claim 2 wherein the controller is further constructed and arranged to:

generate a set of Reed-Solomon error correcting codes as the set of checksums prior to performing the set of external caching operations.

5. A computer program product including a non-transitory computer readable storage medium which stores a set of instructions to protect host data, the set of instructions, when run on a processor of a local computerized node, causing the local computerized node to:

receive, at the local computerized node, a block of data from a host computer, the block of data including data sub-blocks;

store the block of data, as a dirty cache block, in a local cache of the local computerized node; and

perform a set of external caching operations to cache a set of sub-blocks in a set of external computerized nodes in communication with the local computerized node, each external caching operation caching a respective sub-block of the set of sub-blocks in a cache of a respective external computerized node, the set of sub-blocks including (i) the data sub-blocks of the block of data from the host computer and (ii) a set of checksums computed from the data sub-blocks of the block of data from the host computer.

6. A computer program product as in claim 5 wherein the local computerized node connects to a set of non-volatile storage devices;

and wherein the set of instructions, when run on the processor of the local computerized node, further cause the local computerized node to;

store the block of data in the set of non-volatile storage devices after performing the set of external caching operations.

7. A method of protecting host data, the method comprising:

receiving, at a local computerized node, a block of data from a host computer, the block of data including data sub-blocks;

storing the block of data, as a dirty cache block, in a local cache of the local computerized node; and

performing a set of external caching operations to cache a set of sub-blocks in a set of external computerized nodes in communication with the local computerized node, each external caching operation caching a respective sub-block of the set of sub-blocks in a cache of a respective external computerized node, the set of sub-blocks including (i) the data sub-blocks of the block of data from the host computer and (ii) a set of checksums generated from the data sub-blocks of the block of data from the host computer.

8. A method as in claim 7 wherein the local computerized node connects to a set of non-volatile storage devices;

and wherein the method further comprises;

storing the block of data in the set of non-volatile storage devices after performing the set of external caching operations.

9. A method as in claim 8 wherein the set of non-volatile storage devices includes magnetic disk drives arranged to implement a particular RAID (redundant array of inexpensive disks) level;

and wherein storing the block of data includes;

writing the block of data to the magnetic disk drives in accordance with the particular RAID level.

10. A method as in claim 9 wherein the local computerized node receives the block of data from the host computer as part of a write operation;

and wherein the method further comprises;

acknowledging completion of the write operation to the host computer after performing the set of external caching operations and prior to writing the block of data to the magnetic disk drives in accordance with the particular RAID level.

11. A method as in claim 8, further comprising:

after the local computerized node has failed, recovering the block of data from respective sub-blocks in the caches of the respective external computerized nodes.

12. A method as in claim 11 wherein recovering the block of data includes:

restoring the block of data from less than all of the data sub-blocks of the and at least one checksums derived from the data sub-blocks.

13. A method as in claim 8, further comprising:

generating, at the local computerized node, a set of Reed-Solomon error correcting codes as the set of checksums prior to performing the set of external caching operations.

14. A method as in claim 13 wherein each data sub-block has a predefined size which is an integer multiple of a size of the block of data;

and wherein generating the set of Reed-Solomon error correcting codes includes;

applying a set of functions to the data sub-blocks to derive the set of Reed-Solomon error correcting codes, each Reed-Solomon error correcting code having the predefined size which is the integer multiple of the size of the block of data.

15. A method as in claim 7 wherein each data sub-block is less than a complete copy of the block of data from the host computer.

16. A method as in claim 15 wherein the local computerized node and the set of external computerized nodes forms a data storage cluster which performs read and write operations using write-back caching on behalf of a set of host computers.

17. A method as in claim 16 wherein the block of data includes N data sub-blocks;

wherein the set of checksums includes M checksums, N and M being different integers, M being less than N, M being greater than zero, and N being greater than zero;

wherein the data storage cluster includes N+M computerized nodes; and

wherein the method further comprises;

after a failure of at least the local computerized node of the data storage cluster, recovering the block of data from sub-blocks cached by remaining non-failed computerized nodes of the data storage cluster.

18. A method as in claim 17 wherein recovering the block of data from the sub-blocks cached by the remaining non-failed computerized nodes of the data storage cluster includes:

encountering a failure of M+1 computerized nodes, and restoring the block of data from N remaining non-failed computerized nodes of the data storage cluster.

19. A method as in claim 18 wherein N+M is at least three (3);

wherein N is at least two (2); and

wherein M is at least one (1).

20. A method as in claim 18 wherein restoring the block of data from the N remaining non-failed computerized nodes includes reconstructing the block of data from any combination of N data or checksum sub-blocks.