METHOD AND APPARATUS FOR BOTTLENECK IDENTIFICATION IN HIGH-PERFORMANCE STORAGE SYSTEMS

• US 20200133541A1
• Filed: 10/31/2018
• Published: 04/30/2020
• Est. Priority Date: 10/31/2018
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Inventors: Kleiner, Vladimir, Shveidel, Vladimir
• Granted Patent: US 10,747,453 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 11/1456 Hardware arrangements for backup
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0632 by initialisation or re-initialisation of storage systems
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0673 Single storage device

Abstract

A method is disclosed for configuring a storage system, comprising: identifying at least one synchronous flow that is executed by the storage system, the synchronous flow being executed in response to a plurality of I/O requests that are received at the storage system; identifying a first synchronization object that is used by the synchronous flow, the first synchronization object being configured to control access to a first resource based on availability of one or more first credits for accessing the first resource; identifying at least one background flow that is configured to supply the first credits to the first synchronization object; detecting whether a first latency growth that is associated with the background flow corelates with a second latency growth that is associated with the storage system; and when the first latency growth correlates with the second latency growth, causing the storage system to increase a rate at Which the first credits are supplied to the first synchronization object by the background flow.

Claims

1. method for configuring a storage system to improve performance, comprising:

identifying at least one synchronous flow that is executed by the storage system, the synchronous flow being executed in response to a plurality of I/O requests that are received at the storage system;

identifying a first synchronization object that is used by the synchronous flow, the first synchronization object being configured to control access to a first resource based on availability of one or more first credits for accessing the first resource;

identifying at least one background flow that is configured to supply the first credits to the first synchronization object;

detecting whether a first latency growth that is associated with the background flow corelates with a second latency growth that is associated with the storage system; and

when the first latency growth correlates with the second latency growth, causing the storage system to increase a rate at which the first credits are supplied to the first synchronization object by the background flow.

2. The method of claim 1, wherein:

the first latency growth incudes growth in respective durations of periods spent by the background flow in a suspended state, and

the second latency growth includes an end-to-end latency growth of the storage system.

3. The method of claim 1, further comprising executing a load test on the storage system, wherein identifying at least one synchronous flow includes identifying at least one synchronous flow that is executed by the storage system when the storage system is experiencing an increasing load.

4. The method of claim 1, wherein identifying at least one synchronous flow includes identifying at least one synchronous flow that is executed by the storage system when the storage system is in a state in which:

(i) I/O requests are provided to the storage system at a rate that is approximately constant, and (ii) a latency of the I/O requests increases at a rate that is greater than a second threshold.

5. The method of claim 1, wherein the second latency growth that is associated with the storage system is calculated as a function of increasing external load on the storage system.

6. The method of claim 1, wherein:

the synchronous flow includes one or more threads that are configured to store data associated with the I/O request into temporary storage, and

the background flow includes one or more threads that are configured to transfer data from the temporary storage into permanent storage.

7. The method of claim 1, wherein causing the storage system to increase the rate at which the first credits are supplied to the first synchronization object includes at least one of:

(i) increasing a size of the background flow, and (ii) increasing a priority of at least one thread that is part of the background flow.

8. The method of claim 1, wherein the at least one synchronous flow includes a combined flow.

9. A system comprising:

a memory; and

one or more processors operatively coupled to the memory, configured to;

identify at least one synchronous flow that is executed by a storage system, the synchronous flow being executed in response to a plurality of I/O requests that are received at the storage system;

identify a first synchronization object that is used by the synchronous flow, the first synchronization object being configured to control access to a first resource based on availability of one or more first credits for accessing the first resource;

identify at least one background flow that is configured to supply the first credits to the first synchronization object;

detect whether a first latency growth that is associated with the background flow corelates with a second latency growth that is associated with the storage system; and

when the first latency growth correlates with the second latency growth, cause the storage system to increase a rate at which the first credits are supplied to the first synchronization object by the background flow,

10. The system of claim 9, wherein:

the first latency growth incudes growth in respective durations of periods spent by the background flow in a suspended state, and

the second latency growth includes an end-to-end latency growth of the storage system.

11. The system of claim 9, wherein:

the one or more processors are further configured to execute a load test on the storage system, and

identifying at least one synchronous flow includes identifying at least one synchronous flow that is executed by the storage system when the storage system is experiencing a load that is greater than a threshold load as a result of the load test being performed.

12. The system of claim 9, wherein identifying at least one synchronous flow includes identifying at least one synchronous flow that is executed by the storage system when the storage system is in a state in which:

I/O requests are provided to the storage system at a rate that is approximately constant, and (ii) a latency of the I/O requests increases at a rate that is greater than a second threshold.

13. The system of claim 9, wherein the second latency growth that is associated with the storage system is calculated as a function of increasing external load on the storage system.

14. The system of claim 9, wherein:

the synchronous flow includes one or more threads that are configured to store data associated with the I/O request into temporary storage, and

the background flow includes one or more threads that are configured to transfer data from the temporary storage into permanent storage.

15. The system of claim 9, wherein causing the storage system to increase the rate at which the first credits are supplied to the first synchronization object includes at least one of:

(i) increasing a size of the background flow, and (ii) increasing a priority of at least one thread that is part of the background flow.

16. The system of claim 9, Wherein the at least one synchronous flow includes a combined flow.

17. A non-transitory computer-readable medium storing one or more processor-executable instructions, which when executed by one or more processors cause the one or more processors to perform the operations of:

identifying at least one synchronous flow that is executed by a storage system, the synchronous flow being executed in response to a plurality of I/O requests that are received at the storage system;

identifying a first synchronization object that is used by the synchronous flow, the first synchronization object being configured to control access to a first resource based on availability of one or more first credits for accessing the first resource;

identifying at least one background flow that is configured to supply the first credits to the first synchronization object;

detecting whether a first latency growth that is associated with the background flow corelates with a second latency growth that is associated with the storage system; and

when the first latency growth correlates with the second latency growth, causing the storage system to increase a rate at which the first credits are supplied to the first synchronization object by the background flow.

18. The non-transitory computer readable medium of claim 17, wherein:

the first latency growth incudes growth in respective durations of periods spent by the background flow in a suspended state, and

the second latency growth includes an end-to-end latency growth of the storage system.

19. The non-transitory computer readable medium of claim 17, wherein identifying at least one synchronous flow includes identifying at least one synchronous flow that is executed by the storage system when the storage system is in a state in which:

(i) I/O requests are provided to the storage system at a rate that is approximately constant, and (ii) a latency of the I/O requests increases at a rate that is greater than a second threshold.

20. The non-transitory computer readable medium of claim 17, wherein the synchronous flow includes an average synchronous flow.