Multi-tier storage environment with burst buffer middleware appliance for batch messaging

1. A multi-tier storage system, comprising:

at least first and second storage tiers comprising storage for storing data; and

at least one burst buffer appliance comprising a high-speed memory having a substantially lower access time than at least one of said first and second storage tiers, wherein said at least one burst buffer appliance further comprises an abstract storage interface comprising a software interface module that communicates with at least one persistent storage device in at least one of said first and second storage tiers, wherein said at least one persistent storage device provides a list application programming interface (API) that processes multiple input-output operations as a single function, wherein said software interface module of said at least one burst buffer appliance provides a batch operation that aggregates a plurality of individual input-output operations to said at least one persistent storage device using a single aggregated message, wherein at least one argument of said batch operation comprises an array having a plurality of elements, where each of said individual input-output operations corresponds to one element of said array, wherein said batch operation aggregates a plurality of metadata operations and wherein said batch operation processes metadata from said metadata operations using one or more key-value operations that process at least one key-value metadata element.

2. The multi-tier storage system of claim 1 , wherein said abstract storage interface provides a storage-specific interface for a plurality of storage destination types.

3. The multi-tier storage system of claim 1 , wherein said software interface module communicates with at least one persistent storage device that does not provide a list API using said plurality of individual operations within said batch operation.

4. The distributed storage system of claim 1 , wherein the high speed memory of said burst buffer appliance comprises the first storage tier or is used to access the first storage tier and the second storage tier.

5. The distributed storage system of claim 1 , wherein said software interface module of said at least one burst buffer appliance maintains semantic information associated with key-value data.

6. The multi-tier storage system of claim 1 , wherein said batch operation for processing metadata replicates said metadata on a plurality of storage nodes.

7. The distributed storage system of claim 6 , wherein the metadata is replicated on said first storage tier and said second storage tier.

8. A burst buffer appliance for use in a multi-tier storage system comprising at least first and second storage tiers comprising storage for storing data, said burst buffer appliance comprising:

a high-speed memory having a substantially lower access time than at least one of said first and second storage tiers; and

at least one hardware device configured to provide an abstract storage interface comprising a software interface module that communicates with at least one persistent storage device in said multi-tier storage system providing a list application programming interface (API)) that processes multiple input-output operations as a single function, wherein said software interface module of said at least one burst buffer appliance provides a batch operation that aggregates a plurality of individual input-output operations to said at least one persistent storage device using a single aggregated message, wherein at least one argument of said batch operation comprises an array having a plurality of elements, where each of said individual input-output operations corresponds to one element of said array, wherein said batch operation aggregates a plurality of metadata operations and wherein said batch operation processes metadata from said metadata operations using one or more key-value operations that process at least one key-value metadata element.

9. The burst buffer appliance of claim 8 , wherein said abstract storage interface provides a storage-specific interface for a plurality of storage destination types.

10. The burst buffer appliance of claim 8 , wherein said batch operation for processing metadata replicates said metadata on a plurality of storage nodes.

11. The burst buffer appliance of claim 8 , wherein said software interface module communicates with at least one persistent storage device that does not provide a list API using said plurality of individual operations within said batch operation.

12. The burst buffer appliance of claim 8 , wherein the high speed memory of said burst buffer appliance comprises the first storage tier or is used to access the first storage tier and the second storage tier.

13. The burst buffer appliance of claim 8 , wherein said software interface module of said at least one burst buffer appliance maintains semantic information associated with key-value data.

14. A method performed by a burst buffer appliance in a multi-tier storage system comprising at least first and second storage tiers comprising storage for storing data, wherein the first storage tier comprises a high-speed memory having a substantially lower access time than the second storage tier, comprising:

providing, by said burst buffer appliance, an abstract storage interface comprising a software interface module that communicates with at least one persistent storage device in said multi-tier storage system providing a list application programming interface (API) that processes multiple input-output operations as a single function, wherein said burst buffer appliance comprises a high-speed memory having a substantially lower access time than at least one of said first and second storage tiers;

obtaining a batch operation that aggregates a plurality of individual input-output operations; and

providing, by said software interface module of said at least one burst buffer appliance, said batch operation to said at least one persistent storage device using a single aggregated message, wherein at least one argument of said batch operation comprises an array having a plurality of elements, where each of said individual input-output operations corresponds to one element of said array, wherein said batch operation aggregates a plurality of metadata operations and wherein said batch operation processes metadata from said metadata operations using one or more key-value operations that process at least one key-value metadata element.

15. The method of claim 14 , wherein said abstract storage interface provides a storage-specific interface for a plurality of storage destination types.

16. The method of claim 14 , wherein said batch operation for processing metadata replicates said metadata on a plurality of storage nodes.

17. The method of claim 14 , wherein said software interface module communicates with at least one persistent storage device that does not provide a list API using said plurality of individual operations within said batch operation.

18. The method of claim 14 , wherein the high speed memory of said burst buffer appliance comprises the first storage tier or is used to access the first storage tier and the second storage tier.

19. The method of claim 14 , wherein said software interface module of said at least one burst buffer appliance maintains semantic information associated with key-value data.

20. A tangible machine-readable recordable storage medium, wherein one or more software programs when executed by one or more processing devices implement the steps of the method of claim 14 .