Client-side, variable drive health determination

• US 9,792,192 B1
• Filed: 03/29/2012
• Issued: 10/17/2017
• Est. Priority Date: 03/29/2012
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Brooker, Marc J., Holgers, Tobias L., Wei, Danny, Parthasarathy, Madhuvanesh
• Primary Examiner(s): Madamba, Glenford
• Application Number:
  US13/433,658
• Time in Patent Office: 2,028 Days
• Field of Search: 709224
• CPC Class Codes: G06F 11/008 Reliability or availability analysis
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 11/3485 for I/O devices

Abstract

The relative health of data storage drives may be determined based, at least in some aspects, on data access information and/or other drive operation information. In some examples, upon receiving the operation information from a computing device, a health level of a drive may be determined. The health level determination may be based at least in part on operating information received from a client entity. Additionally, a storage space allocation instruction or operation may be determined for execution. The allocation instruction or operation determined to be performed may be based at least in part on the determined health level.

Claims

1. A computer-implemented method, comprising:

obtaining, from a customer of a computing resource service provider, performance criteria associated with at least a subset of data stored in a first data volume, the first data volume being operatively attached to a client computing device hosted by the computing resource service provider on behalf of the customer;

obtaining, from the client computing device, an operating parameter associated with a first data volume indicating an ability to perform data operations to the data stored in the first data volume;

determining, based at least in part on the operating parameter, a health level of the first data volume from a plurality of possible health levels; and

as a result of determining that the health level is below a predefined health level, the predefined health level being between a first health level and a second health level of the plurality of possible health levels;

selecting a second data volume based at least in part on performance characteristics of the second data volume meeting the set of performance criteria;

based at least in part on the performance criteria, causing an operation to be performed on at least the subset of the data, the operation resulting in a change in storage location associated with the at least the subset of the data from a first storage location provided by hardware used to maintain the data of the first data volume to a second storage location provided by hardware used to maintain data of the second data volume; and

associating the second data volume with the client computing device.

2. The computer-implemented method of claim 1, wherein the operating parameter is based at least in part on a number of packets lost, as determined by the client computing device, during at least one of a client computing device read from the first data volume or a client computing device write to the first data volume.

3. The computer-implemented method of claim 1, wherein determining the health level of the first data volume is based at least in part on comparing the received operating parameter against a previous operating parameter associated with the first data volume.

4. The computer-implemented method of claim 1, wherein the health level of the first data volume indicates at least one of:

whether the first data volume is failing, whether the first data volume will fail within a predefined time, or whether the first data volume has failed prior to the determination.

5. The computer-implemented method of claim 1, further comprising updating a memory controller mapping to indicate that the health of the first data volume is below the predefined level.

6. The computer-implemented method of claim 1, wherein the first data volume is virtually attached to the client computing device.

7. A computer-implemented method, comprising:

obtaining, from a customer of a computing resource service provider, performance criteria associated with a first subset of a dataset located on a remote storage device operatively attached to a client entity of one or more computer systems of a distributed system of a computing resource service provider, the client entity hosted by the computing resource service provider on behalf of the customer;

receiving, from the client entity, operational parameter information associated with operation of a remote storage device;

as a result of determining, based at least in part on the operational parameter information received, that a failure risk of the remote storage device is above a predefined level, causing an operation to be performed by the remote storage server on the dataset of the remote storage device such that;

the first subset of the dataset is maintained on storage media that performs in conformity with the obtained performance criteria; and

a second subset of the dataset is maintained on the remote storage device.

8. The computer-implemented method of claim 7, wherein the remote storage server is virtually attached to the client entity.

9. The computer-implemented method of claim 7, wherein the information associated with operation of the remote storage device is based at least in part on at least one of:

a data read latency between the client entity and the remote storage server or

a data write latency between the client entity and the remote storage server.

10. The computer-implemented method of claim 9, wherein the information associated with operation of the remote storage device is based at least in part on a difference between the data read latency between the client entity and the remote storage server and the data write latency between the client entity and the remote storage server.

11. The computer-implemented method of claim 7, wherein the information associated with operation of the remote storage device is based at least in part on a variance in at least one of a data read latency between the client entity and the remote storage server or a data write latency between the client entity and the remote storage server.

12. The computer-implemented method of claim 7, wherein the client entity comprises a computing device hosting a virtual machine instance.

13. A system of a computing resource service provider, comprising:

at least one memory that stores computer-executable instructions; and

at least one processor that accesses the at least one memory to execute the computer-executable instructions to;

receive, based at least in part on accessing an operationally attached remote storage system, health information associated with operation of the operationally attached, remote storage system, the remote storage system implemented by a memory other than the at least one memory;

determine, based at least in part on the received health information, a level of health of the operationally attached, remote storage system, the level of health selected from multiple health levels above at least a predefined health level;

obtain an operational parameter value specified for a first subset of a dataset by a customer of the computing resource service provider;

determine, based at least in part on the level of health, an associated storage management instruction set to be executed on the operationally attached, remote storage system, the corresponding storage management instruction set specifying that the first subset of the dataset, exclusive of a second subset of the dataset, is to be stored on a storage device of the remote storage system that performs in conformity with the specified operational parameter value; and

cause transmission of the corresponding storage management instruction set to a computing device associated with the operationally attached, remote storage system, the computing device associated with the operationally attached, remote storage system implemented by a processor other than the at least one processor and configured to cause execution of the corresponding storage management instruction on the operationally attached, remote storage system.

14. The system of claim 13, wherein the operationally attached, remote storage system comprises at least a first storage device and wherein the at least one processor is further configured to execute the computer executable instructions to cause transmission of a request to store data of the first storage device, that is associated with a client entity operationally attached to the first storage device, in a second storage device with a higher level of health than the first storage device.

15. The system of claim 13, wherein the information associated with operation of the operationally attached, remote storage system is based at least in part on at least one of packet loss information or latency information.

16. The system of claim 13, wherein the at least one processor is further configured to execute the computer executable instructions to:

determine an amount of collective data that is stored, on behalf of a client entity operationally attached to the remote storage system, in at least one storage device with a health level below a predefined amount; and

as a result of a percentage of the amount of collective data that is stored in the at least one storage device with a health level below a predefined amount being greater than a predefined percentage, cause transmission of a request to store the collective data in a second storage device.

17. The system of claim 13, wherein the computing device associated with the operationally attached, remote storage system comprises at least one processor associated with controlling the operationally attached, remote storage system.

18. The system of claim 13, wherein the computing device associated with the operationally attached, remote storage server comprises a monitoring system configured to monitor health of the remote storage system.

19. One or more non-transitory computer-readable media collectively storing computer-executable instructions that, as a result of execution by one or more processors, causes the one or more processors to perform operations comprising:

receiving, from a client entity of a first computing system, an operation parameter associated with operation of a first subset of a dataset in first storage media of a second computing system;

determining, based at least in part on the operation parameter, a performance level of the first storage media, the performance level selected from a plurality of healthy performance levels; and

transmitting, to a control system of the second computing system that is operationally attached to the first storage media, one or more data storage management instructions associated with the performance level determined, the control system of the second computing system configured to at least perform a data storage operation based at least in part on the one or more data storage management instructions such that;

the data storage operation at least moves the first subset of the dataset to second storage media associated with a healthier performance level than the storage media, the healthier performance level based at least in part on an amount of risk of failure associated with the first storage media; and

a second subset of the dataset remains on the first storage media.

20. The one or more non-transitory computer-readable media of claim 19, wherein the operating parameter is based at least in part on at least one of packet loss information or latency information.

21. The one or more non-transitory computer-readable media of claim 19, wherein the plurality of healthy performance levels does not include an unhealthy level associated with non-performing storage media.

22. The one or more non-transitory computer-readable media of claim 19, wherein the plurality of healthy performance levels includes a sliding scale of levels associated with storage media performing as intended by a provider of the storage media.