GENERALIZED STORAGE VIRTUALIZATION INTERFACE

• US 20160070652A1
• Filed: 12/12/2014
• Published: 03/10/2016
• Est. Priority Date: 09/04/2014
• Status: Active Grant

• Current Assignee: SanDisk Technologies LLC (Western Digital Corporation)
• Original Assignee: Fusion-IO Incorporated (Western Digital Corporation)
• Inventors: Sundararaman, Swaminathan, Talagala, Nisha, Wipfel, Robert, Subramanian, Sriram, Bolkhovitin, Vladislav
• Granted Patent: US 10,380,026 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 12/10 Address translation
• CPC Class Codes: G06F 12/109 for multiple virtual address spaces, e.g. segmentation G06F12/1036 takes precedence
• CPC Class Codes: G06F 2212/1016 Performance improvement
• CPC Class Codes: G06F 2212/657 Virtual address space management
• CPC Class Codes: G06F 3/0617 in relation to availability
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0665 at area level, e.g. provisioning of virtual or logical volumes
• CPC Class Codes: G06F 3/0667 at data level, e.g. file, record or object virtualisation
• 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 3/0679 Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]
• CPC Class Codes: G06F 3/0685 Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays

Abstract

A storage system implements a sparse, thinly provisioned logical-to-physical translation layer. The storage system may perform operations to modify logical-to-physical mappings, including creating, removing, and/or modifying any-to-any and/or many-to-one mappings between logical identifiers and stored data (logical manipulation operations). The storage system records persistent metadata to render the logical manipulation (LM) operations persistent and crash-safe. The storage system may provide access to LM functionality through a generalized LM interface. Clients may leverage the LM interface to efficiently implement higher-level functionality and/or offload LM operations to the storage system.

Claims

1. An apparatus, comprising:

a storage manager that stores data on a non-transitory storage medium in response to storage requests of a client;

a translation layer that maintains a logical address space corresponding to the non-transitory storage medium comprising associations between data stored on the non-transitory storage medium and respective logical addresses in the logical address space; and

a data virtualization interface having a plurality of functions available to the client, the functions configured to implement storage management operations by changing associations between data stored on the non-transitory storage medium and logical addresses of the logical address space.

2. The apparatus of claim 1, wherein the storage management operations change the logical address associated with a data block stored on the non-volatile storage medium without modifying the stored data block.

3. The apparatus of claim 1, wherein the data is stored on the non-transitory storage medium in response to requests issued to a block storage interface.

4. The apparatus of claim 1, wherein the data virtualization interface comprises one of a library, an application programming interface, a user-level module, a software development kit, and a kernel-level module.

5. The apparatus of claim 1, wherein the data virtualization interface comprises a range clone function that specifies a destination logical address range and a source logical address range, wherein the source logical address range is mapped to a data segment stored on the non-transitory storage medium, and wherein implementing the range clone function comprises associating the destination logical address range with the data segment stored on the non-transitory storage medium.

6. The apparatus of claim 1, wherein the data virtualization interface comprises a range move function that determines a destination logical address and a source logical address, wherein the source logical address corresponds to particular storage address on the non-transitory storage medium, and wherein the range move function maps the destination logical address to the particular storage address and unmaps the source logical address from the particular storage address.

7. The apparatus of claim 1, wherein the client implements a zero-write file copy operation by issuing a logical copy request to the data virtualization interface configured to associate data of a first file stored on the non-transitory storage medium with a second file, such that logical addresses assigned to the second file reference the same stored file data as logical addresses assigned to the first file.

8. The apparatus of claim 7, wherein the client is configured to identify files corresponding to duplicate data, and wherein the client issues the logical copy request to the data virtualization interface in response to determining that the first file and the second file correspond to duplicate data.

9. The apparatus claim 1, wherein the client is configured to leverage the data virtualization interface to implement a journaled storage transaction pertaining to a plurality of target logical addresses by:

storing a plurality of data blocks pertaining to the journaled storage transaction, such that the stored data blocks are mapped to logical addresses within a designated region of the logical address space, and

issuing a range move request to the data virtualization interface configured to associate the plurality of stored data blocks with the target logical addresses of the journaled storage transaction.

10. The apparatus of claim 9, wherein the client is configured to rollback a failed journaled storage transaction by invalidating stored data associated with addresses within the designated region of the logical address space.

11. The apparatus of claim 9, wherein the storage manager implements the range move request by a single atomic write to a persistent storage device.

12. A computer-implemented storage system, comprising:

a storage interface having a plurality of API functions for storing data blocks on a physical storage resource, such that the stored data blocks are mapped to respective identifiers of a logical address space; and

a uniform logical manipulation interface available to a client, the uniform logical manipulation interface having a plurality of API functions for modifying logical-to-physical mappings pertaining to the stored data blocks, the uniform logical manipulation interface comprising an API function to map a stored data block mapped to a first identifier of the logical address space to a second identifier of the logical address space.

13. The computer-implemented storage system of claim 12, wherein the API function of the uniform logical manipulation interface is configured to map the first identifier and the second identifier to the same stored data block on the physical storage resource.

14. The computer-implemented storage system of claim 12, wherein the API function of the uniform logical manipulation interface removes the mapping between the first identifier and the stored data block.

15. The computer-implemented storage system of claim 12, wherein the API function is defined in terms of identifier vectors comprising two or more identifiers in the logical address space, and wherein the API function is configured to map stored data blocks mapped to identifiers in a first vector to identifiers in a second, different vector.

16. The computer-implemented storage system of claim 15, further comprising a namespace manager that stores persistent metadata indicating that the stored data blocks are mapped to the second, different set of two or more identifiers on a non-volatile storage medium in a single, atomic write operation.

17. The computer-implemented storage system of claim 12, wherein the uniform logical manipulation interface comprises a library that is accessible to a user-level application operating on a computing device.

18. A method, comprising:

maintaining metadata in a persistent log to translate logical identifiers of a front-end address space of a storage resource to data segments stored on the storage resource; and

exposing a generalized interface to change translations between logical identifiers of the front-end address space and the data segments stored on the non-volatile storage device.

19. The method of claim 18, further comprising appending entries to the persistent log in response to a request to change a translation between a data segment stored on the storage resource and the front-end address space.

20. The method of claim 18, wherein the generalized interface is configured to receive one or more of:

a) requests to clone a stored data segment that translates to a first logical identifier to a second logical identifier, such that the first logical identifier and the second logical identifier translate to the same stored data segment, and

b) requests to move a stored data segment that translates to a first logical identifier to a second logical identifier, such that the second logical identifier translates to the stored data segment and the first logical identifier is untranslated.

21. An apparatus, comprising:

a client configured to issue requests to a storage layer to store data associated with respective addresses of a logical address space of a non-volatile storage medium, the logical address space presented to the client,

wherein the storage layer makes primitive storage operations for data stored on the non-volatile storage medium available to the client, the primitive storage operations configured to modify associations between the logical address space and the data stored on the non-volatile storage medium, and

wherein the client is configured to implement storage management operations by use of the primitive storage operations made available by the storage layer.

22. The apparatus of claim 21, wherein the storage layer is configured to modify the associations between the logical address space and the data stored on the non-volatile storage medium without changing the data stored on the non-volatile storage medium.

23. The apparatus of claim 21, wherein the client is configured to manage snapshots pertaining to the logical address space by use of a logical copy primitive accessible using the storage layer.

24. The apparatus of claim 21, wherein the client issues requests to write data to a block storage interface, and wherein the storage layer exposes the primitive storage operations configured to modify the associations between the logical address space and the data stored on the non-volatile storage medium though an interface that is separate from the block storage interface.