Granted Patent B2

US 11,593,327 · App. 17/038,814 · Granted Feb 28, 2023

Inventor: Nickolay Dalmatov (Saint Petersburg, RU)

Assignee: EMC IP Holding Company LLC

G06F16/215G06F16/22

Abstract

Claims (58)

1. A method comprising:

generating a deduplication index having multiple entries, each entry storing a digest of a data block that was previously stored in non-volatile data storage and a pointer to a location in the non-volatile data storage at which the data block was previously stored, wherein the entries of the deduplication index are divided into a plurality of deduplication index segments, wherein each deduplication index segment is made up of entries containing digests generated from data blocks received and stored in the non-volatile data storage during a corresponding one of multiple preceding time periods, together with pointers to locations in the non-volatile data storage at which those data blocks were stored;

storing a resident subset of the deduplication index segments in memory;

storing a non-resident subset of the deduplication index segments in the non-volatile data storage;

performing data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests stored in the entries of the deduplication index segments that are contained in the resident subset of the deduplication index segments;

storing digests generated from subsequently received data blocks that do not match any of the digests contained in the entries of the resident subset of the deduplication index segments into entries of an open deduplication index segment also stored in the memory of a data storage system;

performing data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests in the entries of the open deduplication index segment, wherein performing data deduplication for each subsequently received data block having a digest matching any one of the digests stored in the entries in the deduplication index segments contained in the resident subset of the deduplication index segments or in the open deduplication index segment includes storing, in the non-volatile data storage of the data storage system, a pointer to the location in the non-volatile data storage at which a previously received copy of the subsequently received data block was stored;

detecting that the open deduplication index segment has reached a predetermined maximum deduplication index segment size; and

in response to detecting that the open deduplication index segment has reached the predetermined maximum deduplication index segment size:

making the open deduplication index segment one of the deduplication index segments in the resident subset of the deduplication index segments, and

allocating a new open deduplication index segment within the memory of the data storage system.

2. The method of claim 1 , further comprising:

maintaining usage ratios for the deduplication index segments, such that each deduplication index segment has a usage ratio indicating how often digests in the deduplication index segment match digests of data blocks received by the data storage system; and

in response to the usage ratios for the deduplication index segments, selecting as the resident subset of the deduplication index segments one or more deduplication index segments having relatively higher usage ratios than the other deduplication index segments.

3. The method of claim 2 , further comprising:

detecting that a deduplication index segment in the non-resident subset of the deduplication index segments has a higher usage ratio than a usage ratio of one of the deduplication index segments in the resident subset of the deduplication index segments; and

in response to detecting that that the deduplication index segment in the non-resident subset of the deduplication index segments has a higher usage ratio than the usage ratio of the one of the deduplication index segments in the resident subset of the deduplication index segments:

moving the deduplication index segment in the non-resident subset of the deduplication index segments having the higher usage ratio to the resident subset of the deduplication index segments, and

moving the deduplication index segment in the resident subset of the deduplication index segments to the non-resident subset of the deduplication index segments.

4. The method of claim 3 , further comprising:

in response to identifying a digest that is contained in both i) an entry in a deduplication index segment in the non-resident subset of the deduplication index segments, and ii) an entry contained in a deduplication index segment in the resident subset of the deduplication index segments, updating a pointer stored in the entry contained in the deduplication index segment contained in the non-resident subset of the deduplication index segments to match a pointer in the entry contained in the deduplication index segment contained in the resident subset of the deduplication index segments.

5. The method of claim 3 , further comprising:

in response to identifying a digest that is contained in both i) an entry in a deduplication index segment in the non-resident subset of the deduplication index segments, and ii) an entry contained in the open deduplication index segment, removing the entry contained in the open deduplication index segment.

6. The method of claim 3 , further comprising:

in response to detecting deletion of at least one data block that was previously stored in the non-volatile data storage, removing at least one entry storing at least one pointer to the deleted data block from at least one of the deduplication index segments.

7. A data storage system comprising:

at least one storage processor including processing circuitry and a memory;

a plurality of data storage drives communicably coupled to the storage processor; and

wherein the memory has program code stored thereon, wherein the program code, when executed by the processing circuitry, causes the processing circuitry to:

generate a deduplication index having multiple entries, each entry storing a digest of a data block that was previously stored in non-volatile data storage and a pointer to a location in the non-volatile data storage at which the data block was previously stored, wherein the entries of the deduplication index are divided into a plurality of deduplication index segments, wherein each deduplication index segment is made up of entries containing digests generated from data blocks received and stored in the non-volatile data storage during a corresponding one of multiple preceding time periods, together with pointers to locations in the non-volatile data storage at which those data blocks were stored,

store a resident subset of the deduplication index segments in memory,

store a non-resident subset of the deduplication index segments in the non-volatile data storage,

perform data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests stored in the entries of the deduplication index segments that are contained in the resident subset of the deduplication index segments,

store digests generated from subsequently received data blocks that do not match any of the digests contained in the entries of the resident subset of the deduplication index segments into entries of an open deduplication index segment also stored in the memory of the data storage system;

perform data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests in the entries of the open deduplication index segment, wherein performing data deduplication for each subsequently received data block having a digest matching any one of the digests stored in the entries in the deduplication index segments contained in the resident subset of the deduplication index segments or in the open deduplication index segment includes storing, in the non-volatile data storage of the data storage system, a pointer to the location in the non-volatile data storage at which a previously received copy of the subsequently received data block was stored;

detect that the open deduplication index segment has reached a predetermined maximum deduplication index segment size; and

in response to detecting that the open deduplication index segment has reached the predetermined maximum deduplication index segment size:

make the open deduplication index segment one of the deduplication index segments in the resident subset of the deduplication index segments, and

allocate a new open deduplication index segment within the memory of the data storage system.

8. The data storage system of claim 7 , wherein the program code, when executed by the processing circuitry, further causes the processing circuitry to:

maintain usage ratios for the deduplication index segments, such that each deduplication index segment has a usage ratio indicating how often digests in the deduplication index segment match digests of data blocks received by the data storage system; and

in response to the usage ratios for the deduplication index segments, select as the resident subset of the deduplication index segments one or more deduplication index segments having relatively higher usage ratios than the other deduplication index segments.

9. The data storage system of claim 8 , wherein the program code, when executed by the processing circuitry, further causes the processing circuitry to:

detect that a deduplication index segment in the non-resident subset of the deduplication index segments has a higher usage ratio than a usage ratio of one of the deduplication index segments in the resident subset of the deduplication index segments; and

in response to detecting that that the deduplication index segment in the non-resident subset of the deduplication index segments has a higher usage ratio than the usage ratio of the one of the deduplication index segments in the resident subset of the deduplication index segments:

move the deduplication index segment in the non-resident subset of the deduplication index segments having the higher usage ratio to the resident subset of the deduplication index segments, and

move the deduplication index segment in the resident subset of the deduplication index segments to the non-resident subset of the deduplication index segments.

10. A computer program product including a non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed on processing circuitry, cause the processing circuitry to perform the steps of:

generating a deduplication index having multiple entries, each entry storing a digest of a data block that was previously stored in non-volatile data storage and a pointer to a location in the non-volatile data storage at which the data block was previously stored, wherein the entries of the deduplication index are divided into a plurality of deduplication index segments, wherein each deduplication index segment is made up of entries containing digests generated from data blocks received and stored in the non-volatile data storage during a corresponding one of multiple preceding time periods, together with pointers to locations in the non-volatile data storage at which those data blocks were stored;

storing a resident subset of the deduplication index segments in memory;

storing a non-resident subset of the deduplication index segments in the non-volatile data storage;

performing data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests stored in the entries of the deduplication index segments that are contained in the resident subset of the deduplication index segments;

storing digests generated from subsequently received data blocks that do not match any of the digests contained in the entries of the resident subset of the deduplication index segments into entries of an open deduplication index segment also stored in the memory of a data storage system;

performing data deduplication for each subsequently received data block for which a digest is generated that matches any one of the digests in the entries of the open deduplication index segment, wherein performing data deduplication for each subsequently received data block having a digest matching any one of the digests stored in the entries in the deduplication index segments contained in the resident subset of the deduplication index segments or in the open deduplication index segment includes storing, in the non-volatile data storage of the data storage system, a pointer to the location in the non-volatile data storage at which a previously received copy of the subsequently received data block was stored;

detecting that the open deduplication index segment has reached a predetermined maximum deduplication index segment size; and

in response to detecting that the open deduplication index segment has reached the predetermined maximum deduplication index segment size:

making the open deduplication index segment one of the deduplication index segments in the resident subset of the deduplication index segments, and

allocating a new open deduplication index segment within the memory of the data storage system.

Assignments (9)

Priority Claims (1)

Continuity (1)

Cited By (2)

US 12,411,810 US 12,468,466