Network-aware caching

• US 10,686,905 B1
• Filed: 09/10/2014
• Issued: 06/16/2020
• Est. Priority Date: 09/10/2014
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Word, Jonathan Brian
• Primary Examiner(s): Hlaing, Soe
• Application Number:
  US14/482,647
• Time in Patent Office: 2,106 Days
• CPC Class Codes: H04L 67/1097 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]
• CPC Class Codes: H04L 67/568 Storing data temporarily at an intermediate stage, e.g. caching

Abstract

Methods and systems for implementing network-aware caching are disclosed. Respective performance characteristics are determined for accessing a plurality of storage locations from a client computing device. The storage locations comprise one or more shared network-access storage locations accessible to the client computing device over a network and optionally one or more dedicated storage locations of the client computing device. One of the storage locations is selected for storing data to which access is requested by the client computing device. The selected storage location is determined based on the respective performance characteristics of the storage locations. The data is cached in the selected storage location.

Claims

1. A system, comprising:

one or more computing devices configured to implement a network-aware caching system, wherein the one or more computing devices include a client computing device, and wherein the network-aware caching system is configured to;

determine performance characteristics comprising respective access characteristics of a plurality of storage locations as measured by the client computing device, wherein the plurality of storage locations comprise one or more dedicated storage locations of the client computing device and one or more shared network-access storage locations accessible to the client computing device over a network;

receive, from the client computing device, a request to access data wherein the requested access comprises read access, write access, or read/write access, and wherein the data is currently stored at one of the plurality of storage locations;

in response to the request or a user access pattern including the request, determine whether to cache the data based at least in part on differences between respective ones of the performance characteristics, and in response to a determination to cache the data;

select one of the plurality of storage locations for caching the data, wherein the selected storage location is determined based at least in part on the performance characteristics;

cache the data in the selected storage location, wherein to cache the data in the selected storage location the network-aware caching system is configured to;

copy the data to the selected storage location; and

maintain a copy of the data at the selected storage location such that updates to the data by a computing device other than the client computing device are reflected in the copy of the data at the selected storage location; and

implement the read access, write access, or read/write access based on the cached data.

2. The system as recited in claim 1, wherein the network-aware caching system is configured to:

select an additional one of the plurality of storage locations for storing an additional chunk of data to which access is requested by the client computing device, wherein the selected additional storage location is determined based on the performance characteristics and the one or more user access patterns, wherein the selected storage location and the selected additional storage location comprise one of the dedicated storage locations and one of the shared network-access storage locations.

3. The system as recited in claim 1, wherein the network-aware caching system is configured to:

update the performance characteristics based on performance metrics collected for the plurality of storage locations.

4. The system as recited in claim 1, wherein the requested access comprises read access, wherein the selected storage location comprises one of the dedicated storage locations, and wherein the selected storage location is determined by a caching layer on the client computing device.

5. A computer-implemented method, comprising:

determining performance characteristics comprising respective access characteristics of a plurality of storage locations as measured by a client computing device, wherein the plurality of storage locations comprise one or more dedicated storage locations managed by the client computing device and one or more shared network-access storage locations accessible to the client computing device over a network;

determining whether to cache data to which access is requested by the client computing device based at least in part on differences between respective ones of the performance characteristics, wherein the data is currently stored at one of the plurality of storage locations, and in response to a determination to cache the data;

selecting one of the plurality of storage locations for caching the data, wherein the selected storage location is determined based at least in part on the performance characteristics; and

caching the data in the selected storage location, comprising;

copying the data to the selected storage location; and

maintaining the copy of the data at the selected storage location such that updates to the data by a computing device other than the client computing device are reflected in the copy of the data at the selected storage location.

6. The computer-implemented method as recited in claim 5, further comprising:

selecting an additional one of the plurality of storage locations for storing additional data to which access is requested by the client computing device, wherein the selected additional storage location is determined based on the performance characteristics, wherein the selected storage location and the selected additional storage location comprise one of a plurality of dedicated storage locations for the client computing device and one of the shared network-access storage locations.

7. The computer-implemented method as recited in claim 5, further comprising:

updating the performance characteristics based on performance metrics collected for the plurality of storage locations.

8. The computer-implemented method as recited in claim 5, wherein the selected storage location is determined to provide a selected level of latency, throughput, reliability, or a combination thereof.

9. The computer-implemented method as recited in claim 5, wherein the performance characteristics further comprise storage costs, and wherein the method further comprises:

receiving, from a user, an indicator of a storage budget for the client computing device, wherein the selected storage location is determined to provide a selected storage cost, based on the storage budget.

10. The computer-implemented method as recited in claim 5, wherein the performance characteristics further comprise durability metrics, and wherein the method further comprises:

receiving, from a user, an indicator of an importance of the data, wherein the selected storage location is determined to provide a selected level of durability of the data, based on the importance of the data.

11. The computer-implemented method as recited in claim 5, wherein the performance characteristics further comprise reliability metrics, and wherein the method further comprises:

receiving, from a user, an indicator of a sensitivity of the data, wherein the selected storage location is determined to provide a selected level of security of the data, based on the sensitivity of the data.

12. The computer-implemented method as recited in claim 5, wherein the data is stored in the selected storage location using a caching layer, and wherein the caching layer emulates a local disk or local memory to an operating system of the client computing device.

13. The computer-implemented method as recited in claim 5, wherein the selected storage location is determined by a caching service external to the client computing device.

14. The computer-implemented method as recited in claim 5, wherein the selected storage location is determined by a caching layer resident on the client computing device.

15. A non-transitory computer-readable storage medium storing program instructions computer-executable to perform:

determining performance characteristics comprising respective access characteristics of a plurality of storage locations as measured by a client computing device, wherein the plurality of storage devices comprise one or more dedicated storage devices managed by the client computing device and one or more shared storage devices accessible to the client computing device over a network;

determining whether to cache data to which access is requested by the client computing device based at least in part on differences between respective ones of the performance characteristics, wherein the data is currently stored at one of the plurality of storage locations, and in response to a determination to cache the data;

selecting one of the plurality of storage locations for caching the data, wherein the selected storage location is determined based at least in part on the performance characteristics; and

causing the data to be cached on the selected storage device, comprising;

causing the data to be copied to the selected storage location; and

causing the copy of the data to be maintained at the selected storage location such that updates to the data by a computing device other than the client computing device are reflected in the copy of the data at the selected storage location.

16. The non-transitory computer-readable storage medium as recited in claim 15, wherein the program instructions are further computer-executable to perform:

updating the performance characteristics based on performance metrics collected for the plurality of storage devices.

17. The non-transitory computer-readable storage medium as recited in claim 15, wherein the selected storage device is determined to provide a selected level of latency, throughput, reliability, or a combination thereof.

18. The non-transitory computer-readable storage medium as recited in claim 15, wherein the performance characteristics further comprise storage costs, and wherein the program instructions are further computer-executable to perform:

receiving, from a user, an indicator of a storage budget for the client computing device, wherein the selected storage device is determined to provide a selected storage cost, based on the storage budget.

19. The non-transitory computer-readable storage medium as recited in claim 15, wherein the performance characteristics further comprise durability metrics, and wherein the program instructions are further computer-executable to perform:

receiving, from a user, an indicator of an importance of the data, wherein the selected storage device is determined to provide a selected level of durability of the data, based on the importance of the data.

20. The non-transitory computer-readable storage medium as recited in claim 15, wherein the performance characteristics further comprise reliability metrics, and wherein the program instructions are further computer-executable to perform:

receiving, from a user, an indicator of a sensitivity of the data, wherein the selected storage device is determined to provide a selected level of security of the data, based on the sensitivity of the data.