System for optimization of input/output from a storage array

• US 9,037,826 B1
• Filed: 02/29/2012
• Issued: 05/19/2015
• Est. Priority Date: 02/29/2012
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Brooker, Marc J., Certain, Tate Andrew, Paterson-Jones, Roland
• Primary Examiner(s): Choe, Yong
• Application Number:
  US13/408,816
• Time in Patent Office: 1,175 Days
• Field of Search: 711/167
• US Class Current: 711/167
• CPC Class Codes: G06F 11/0727 in a storage system, e.g. in a DASD or network based storage system drivers for digital recording or reproducing units G06F3/06; circuits for error detection or correction within digital recording or reproducing units G11B20/18; for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS], H04L67/1097
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06
• CPC Class Codes: G06F 13/1689 Synchronisation and timing concerns synchronisation on a memory bus G06F13/4234
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• 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 3/0673 Single storage device
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD
• CPC Class Codes: G11C 7/22 Read-write [R-W] timing or clocking circuits; Read-write [R-W] control signal generators or management
• CPC Class Codes: H04L 43/08 Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters

Abstract

A system and method for optimizing the performance of one or more disks in a storage array in response to access requests by other computer devices and processes across a network. For requests to access to the storage disk from the network, such as a read/write request, a response monitor notes the service time for each access requests by the storage disk, and if the performance in handling the access requests fails to meet, at least, a predetermined time threshold, actions can be taken to migrate data, modify accesses permitted to the disk, or otherwise indicate underperformance.

Claims

1. A system for optimizing the access performance of a storage array, comprising:

a plurality of storage disks that are accessible across a network;

an access controller that receives requests for access to the plurality of storage disks from the network, the access requests being serviced by one or more of the plurality of storage disks; and

a response monitor that monitors the service of each access request by the one or more of the plurality of storage disks, and the response monitor further determines the performance of the plurality of disks in handling each access request based upon, at least, a predetermined time threshold for the handling of said access request upon determining that the performance of the plurality of storage disks does not meet the predetermined threshold, causes data stored on the plurality of storage disks to be migrated.

2. The system of claim 1, wherein:

the access requests are from a plurality of computer processes across the network, and access requests are comprised of at least two priority levels of service; and

the response monitor, upon determining that the performance of the plurality of storage disks does not meet the predetermined threshold, causes data stored in the plurality of storage disks for at least one priority level of service for access requests to be migrated to one or more other storage disks that provide access request service, at least, at the predetermined threshold for that priority level of service.

3. The system of claim 1, wherein the response monitor, upon determining that the performance of the plurality of storage disks does not meet the predetermined threshold, indicates impending disk failure within the plurality of storage disks.

4. The system of claim 1, wherein the predetermined time threshold is dynamically adjusted for each access request based upon the performance of the plurality of storage disks to access requests.

5. The system of claim 1, wherein the response monitor further determines the performance of the plurality of disks in handling each access request based upon multiple predetermined time thresholds for the handling of said access request.

6. A system for optimizing the access performance of at least one storage disk, comprising:

at least one storage disk that is accessible across a network;

an access controller that receives requests for access to the at least one storage disk from the network; and

a response monitor that monitors the service of each access request by the at least one storage disk, and the response monitor further determines the performance in handling each access request based upon, at least, a predetermined time threshold for the handling of said access request wherein the response monitor, upon determining that the performance of the at least one storage disk does not meet the predetermined threshold, causes data stored on the at least one storage disk to be migrated to another storage accessible across the network.

7. The system of claim 6, wherein:

the access requests are from a plurality of computer processes across the network, and the access requests comprised of at least two priority levels of service; and

the response monitor, upon determining the performance of the at least one storage disk does not meet the predetermined threshold for a priority level of service, limits access requests having that priority level of service to the at least one storage disk.

8. The system of claim 6, wherein the response monitor, upon determining the performance of the at least one storage disk does not meet the predetermined threshold, indicates disk failure.

9. The system of claim 6, wherein the access controller is resident on the at least one storage disk.

10. The system of claim 6, wherein the access controller is located remotely from the at least one storage disk.

11. The system of claim 6, wherein the response monitor is resident on the at least one storage disk.

12. The system of claim 6, wherein the response monitor is located remotely from the at least one storage disk.

13. The system of claim 6, wherein the access request includes either a read request or a write request.

14. The system of claim 6, further comprising a plurality of storage disks that are accessible across a network, and wherein the predetermined time threshold is dynamically adjusted for each access request to the at least one storage disk based upon the performance of the plurality of storage disks to access requests.

15. The system of claim 6, wherein the response monitor further determines the performance of the at least one storage disk in handling each access request based upon multiple predetermined time thresholds for the handling of said access request.

16. A response monitor, comprising:

at least one computing device configured to;

monitor the servicing of access requests to at least one storage disk made across a network; and

determine the performance of the at least one storage disk in handling each access request based upon, at least, a predetermined time threshold for the handling of said access request;

cause data stored on the at least one storage disk to be migrated to another storage accessible across the network upon determining that the performance of the at least one storage disk does not meet the predetermined time threshold.

17. The response monitor of claim 16, wherein:

the access requests are from a plurality of computer processes across the network, and the access requests comprised of at least two priority levels of service; and

the response monitor, upon determining the performance of the at least one storage disk does not meet the predetermined threshold for a priority level of service, limits access requests having that priority level of service to the at least one storage disk.

18. The response monitor of claim 16, wherein the response monitor, upon determining the performance of the at least one storage disk does not meet the predetermined threshold, indicates disk failure.

19. The response monitor of claim 16, wherein the access request includes a read request or a write request.

20. The response monitor of claim 16, wherein the response monitor is resident on the at least one storage disk.

21. The response monitor of claim 16, wherein the response monitor is located remotely from the at least one storage disk.

22. The response monitor of claim 21, wherein the response monitor is located at a node that controls the access requests to a plurality of storage disks, and the response monitor determines the performance of the plurality of storage disks in handling access requests.

23. The response monitor of claim 22, wherein:

the access requests are from a plurality of computer processes across the network, and the access requests comprised of at least two priority levels of service; and

the response monitor selectively migrates from the node to at least one of the plurality of storage disks upon the response monitor determining the performance of the plurality of storage disks does not meet the predetermined threshold for at least one priority level of service of access requests.

24. The response monitor of claim 22, further comprising a plurality of storage disks that are accessible across a network, and wherein the predetermined time threshold is dynamically adjusted for each access request to the at least one storage disk based upon the performance of the plurality of storage disks to access requests.

25. The response monitor of claim 16, wherein the response monitor further determines the performance of the at least one storage disk in handling each access request based upon multiple predetermined time thresholds for the handling of said access request.

26. A method for optimizing the access performance of at least one storage disk, comprising:

receiving one or more requests for access to the at least one storage disk from a network for at least one storage disk that is accessible across a network;

monitoring the service of each access request by the at least one storage disk;

determining the performance in handling each access request based upon, at least, a predetermined time threshold for the handling of said access request; and

migrating data stored on the at least one storage disk to another storage accessible across the network upon determining that the performance in handling each access request does not meet the predetermined threshold.

27. The method of claim 26, wherein:

the access requests are from a plurality of computer processes across the network, and the access requests comprised of at least two priority levels of service; and

upon determining the performance of the plurality of storage disks does not meet the predetermined threshold for a priority level of service, limiting access requests having that priority level of service to the at least one storage disk.

28. The method of claim 26, wherein upon determining the performance of the plurality of storage disks does not meet the predetermined threshold, indicating a disk failure.

29. The method of claim 26, wherein receiving one or more access requests includes receiving either a read request or a write request to the at least one storage disk.

30. The method of claim 26, further comprising:

receiving access requests at a node to a plurality of storage disks; and

determining the performance of the plurality of storage disks in handling access requests, such determination occurring at a response monitor resident at the node.

31. The method of claim 30, wherein the access requests are from a plurality of computer processes across the network, and the access requests comprised of at least two priority levels of service, and further comprising:

determining if the performance of the plurality of storage disks does not meet the predetermined threshold; and

migrating the response monitor from the node to at least one of the plurality of storage disks upon the determination that the performance does not meet the threshold.

32. The method of claim 26, wherein the at least one storage disk is within a plurality of storage disks receiving access requests, and further comprising dynamically adjusting the predetermined time threshold for each access request to the at least one storage disk based upon the performance of the plurality of storage disks to access requests.

33. The method of claim 26, further comprising determining the performance in handling each access request based upon multiple predetermined time thresholds for the handling of said access request.