AGGREGATION OF READ REQUESTS REQUESTING COMMON DATA OBJECTS INTO A COMMON READ OPERATION IN A DATA STORAGE SYSTEM FOR IMPROVING THROUGHPUT

• US 20200133569A1
• Filed: 10/31/2018
• Published: 04/30/2020
• Est. Priority Date: 10/31/2018
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Damgar, James E., Calder, Errol J.
• Granted Patent: US 10,761,778 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0641 De-duplication techniques
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0673 Single storage device
• CPC Class Codes: G06F 3/0683 Plurality of storage devices
• CPC Class Codes: G06F 9/4881 Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues

Abstract

A computer-implemented method according to one embodiment includes suspending a parent/control thread of a data storage system, and receiving a plurality of read requests while the parent/control thread is suspended, each read request having storage information associated therewith. The computer-implemented method further includes resuming the parent/control thread, analyzing the read requests, aggregating read requests requesting common data objects into a common read operation, and dispatching the common read operation to a multi-threaded I/O layer of the data storage system for retrieving data associated with the read requests. The analyzing includes sorting the read requests using the associated storage information.

Claims

1. A computer-implemented method, comprising:

suspending a parent/control thread of a data storage system;

receiving a plurality of read requests while the parent/control thread is suspended, each read request having storage information associated therewith;

resuming the parent/control thread;

analyzing the read requests, wherein the analyzing includes sorting the read requests using the associated storage information;

aggregating read requests requesting common data objects into a common read operation; and

dispatching the common read operation to a multi-threaded I/O layer of the data storage system for retrieving data associated with the read requests.

2. The computer-implemented method of claim 1, comprising:

adding read requests that are received while the parent/control thread is suspended, to a hash/list structure.

3. The computer-implemented method of claim 2, comprising:

not adding read requests that are received while the parent/control thread is resumed, to the hash/list structure.

4. The computer-implemented method of claim 2, wherein aggregating the read requests requesting the common data objects into the common read operation is performed until a condition is no longer true, the condition being selected from the group consisting of:

at least one entry is present in the hash/list structure, a container value associated with a most previously aggregated read request has not changed, a predetermined read aggregation size threshold of an aggregation buffer used for retrieving data associated with the common read operation has not been exceeded, and an amount of data unnecessarily read when retrieving data associated with the common read operation does not exceed a predefined wasted space value.

5. The computer-implemented method of claim 4, wherein the dispatching of the common read operation is performed in response to determining that the condition is no longer true.

6. The computer-implemented method of claim 1, wherein the associated storage information of at least one read request is selected from the group consisting of:

an associated container-ID at which the data associated with one or more of the read requests is stored;

an associated offset, wherein the associated offset is a starting address at which the data associated with one or more of the read requests is read; and

an associated length, wherein the associated length specifies a length of data that is to be read past the associated offset for retrieving the data associated with one or more of the read requests.

7. The computer-implemented method of claim 1, wherein the data associated with at least some of the plurality of read requests include deduplicated data.

8. The computer-implemented method of claim 1, comprising:

consolidating the read requests into groups; and

in response to determining that a read request does not request common data objects as other read requests in the same group, dispatching the read request that does not request common data objects as other read requests in the same group to the multi-threaded I/O layer of the data storage system.

9. A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions readable and/or executable by a computer to cause the computer to perform a method comprising:

receiving, by the computer, a plurality of read requests, each read request having storage information associated therewith;

analyzing, by the computer, the read requests, wherein the analyzing includes sorting the read requests using the associated storage information;

aggregating, by the computer, read requests requesting common data objects into a common read operation; and

dispatching, by the computer, the common read operation to a multi-threaded I/O layer of the data storage system for retrieving data associated with the read requests.

10. The computer program product of claim 9, the program instructions readable and/or executable by the computer to cause the computer to perform the method comprising:

adding, by the computer, read requests that are received while a parent/control thread is suspended, to a hash/list structure.

11. The computer program product of claim 10, the program instructions readable and/or executable by the computer to cause the computer to perform the method comprising:

not adding read requests that are received while a parent/control thread is resumed, to the hash/list structure.

12. The computer program product of claim 10, wherein aggregating the read requests requesting the common data objects into the common read operation is performed until a condition is no longer true, the condition being selected from the group consisting of:

at least one entry is present in the hash/list structure, a container value associated with a most previously aggregated read request has not changed, a predetermined read aggregation size threshold of an aggregation buffer used for retrieving data associated with the common read operation has not been exceeded, and an amount of data unnecessarily read when retrieving data associated with the common read operation does not exceed a predefined wasted space value.

13. The computer program product of claim 12, wherein the dispatching of the common read operation is performed in response to determining that the condition is no longer true.

14. The computer program product of claim 9, wherein the associated storage information of at least one read request is selected from the group consisting of:

an associated container-ID at which the data associated with one or more of the read requests is stored;

an associated offset, wherein the associated offset is a starting address at which the data associated with one or more of the read requests is read; and

an associated length, wherein the associated length specifies a length of data that is to be read past the associated offset for retrieving the data associated with one or more of the read requests.

15. The computer program product of claim 9, wherein the data associated with at least some of the plurality of read requests include deduplicated data.

16. The computer program product of claim 9, the program instructions readable and/or executable by the computer to cause the computer to perform the method comprising:

consolidating the read requests into groups; and

in response to determining that a read request does not request common data objects as other read requests in the same group, dispatching, by the computer, the read request that does not request common data objects as other read requests in the same group to the multi-threaded I/O layer of the data storage system.

17. A system, comprising:

a processor; and

logic integrated with the processor, executable by the processor, or integrated with and executable by the processor, the logic being configured to;

suspend a parent/control thread of a data storage system;

receive a plurality of read requests while the parent/control thread is suspended, each read request having storage information associated therewith;

resume the parent/control thread;

analyze the read requests, wherein the analyzing includes sorting the read requests using the associated storage information;

aggregate read requests requesting common data objects into a common read operation; and

dispatch the common read operation to a multi-threaded I/O layer of the data storage system for retrieving data associated with the read requests.

18. The system of claim 17, the logic being configured to:

add read requests that are received while the parent/control thread is suspended, to a hash/list structure.

19. The system of claim 18, the logic being configured to:

not add read requests that are received while the parent/control thread is resumed, to the hash/list structure.

20. The system of claim 18, wherein aggregating the read requests requesting the common data objects into the common read operation is performed until a condition is no longer true, the condition being selected from the group consisting of:

at least one entry is present in the hash/list structure, a container value associated with a most previously aggregated read request has not changed, a predetermined read aggregation size threshold of an aggregation buffer used for retrieving data associated with the common read operation has not been exceeded, and an amount of data unnecessarily read when retrieving data associated with the common read operation does not exceed a predefined wasted space value.