Method and apparatus for dynamic striping

• US 7,424,574 B1
• Filed: 04/21/2004
• Issued: 09/09/2008
• Est. Priority Date: 04/21/2004
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Ahrens, Matthew A., Bonwick, Jeffrey S.
• Primary Examiner(s): Shah; Sanjiv
• Assistant Examiner(s): Campos; Yaima
• Application Number:
  US10/828,677
• Time in Patent Office: 1,602 Days
• Field of Search: 711/6, 711/114, 711/170, 711202-203, 711/154, 707/10, 707/205
• US Class Current: 711/114
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 21/6218 to a system of files or objects, e.g. local or distributed file system or database
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0644 Management of space entities, e.g. partitions, extents, pools
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A method for dynamic striping, involving receiving a request to write a data block into a storage pool, determining a physical disk location in the storage pool to store the data block using a dynamic striping policy, storing the data block at the physical disk location, and storing a first indirect block in the storage pool, wherein the first indirect block comprises the data block location and the data block checksum.

Claims

1. A method for dynamic striping, comprising:

receiving a request to write a file into a storage pool, wherein the file comprises a first data block, a second data block, and a first indirect block;

determining a physical disk location in the storage pool to store the first data block using a first dynamic striping policy, wherein the first dynamic striping policy comprises one selected from a group consisting of a dynamic striping policy based on physical disk speed, a dynamic striping policy based on free space available on physical disks, a dynamic striping policy based on load on physical disks, and a round robin policy;

storing the first data block using the first dynamic striping policy at the physical disk location in the storage pool;

storing the second data block in the storage pool using the first dynamic striping policy;

selecting a second dynamic striping policy, wherein the second dynamic striping policy comprises one selected from the group consisting of the dynamic striping policy based on physical disk speed, the dynamic striping policy based on free space available on physical disks, the dynamic striping policy based on load on physical disks, and the round robin policy, and wherein the first dynamic striping policy is distinct from the second dynamic striping policy; and

storing the first indirect block using the second dynamic stripping policy, wherein the indirect block includes a first block pointer referencing the first data block and a second block pointer referencing the second data block.

2. The method of claim 1, further comprising:

retrieving the first data block using the first indirect block.

3. The method of claim 1, further comprising:

assembling the first indirect block, wherein assembling the first indirect block comprises populating a block pointer.

4. The method of claim 3, wherein populating the block pointer comprises:

storing the first data block checksum in a checksum field within the block pointer; and

storing the first data block location in the block pointer, wherein storing the data block location comprises storing a metaslab ID and offset.

5. The method of claim 4, further comprising:

storing a birth value in a birth field within the block pointer.

6. The method of claim 3, wherein the first indirect block is assembled using a data management unit.

7. The method of claim 1, wherein the storage pool comprises at least one storage device.

8. The method of claim 1, wherein the storage pool is divided into a plurality of metaslabs.

9. The method of claim 8, wherein each of the plurality of metaslabs is associated with a metaslab ID.

10. The method of claim 9, wherein the first data block location comprises the metaslab ID and an offset.

11. The method of claim 1, wherein storing the first data block comprises using a storage pool allocator.

12. A system for storing a first data block, comprising:

a storage pool comprising a file, wherein the file comprises;

a first data block and a second data block;

a first indirect block; and

a parent block referencing the first indirect block, wherein the first indirect block comprises a first block pointer that references the first data block and a second block pointer that references the second data block, wherein the first block pointer comprises the first data block location and the first data block checksum and the second block pointer comprises the second data block location and the second data block checksum; and

a storage pool allocator configured to;

store the first data block and the second data block in the storage pool using a first dynamic striping policy, wherein the first dynamic striping policy comprises one selected from a group consisting of a dynamic string policy based on physical disk speed, a dynamic striping policy based on free space available on physical disks, a dynamic striping policy based on load on physical disks, and a round robin policy,

select a second dynamic striping policy after storing the first and second data block, wherein the second dynamic striping policy comprises one selected from the group consisting of the dynamic striping policy based on physical disk speed, the dynamic striping policy based on free space available on physical disks, the dynamic striping policy based on load on physical disks, and the round robin policy, and wherein the first dynamic striping policy is distinct from the second dynamic striping policy, and

store the first indirect block and the parent block in the storage pool using the second dynamic striping policy.

13. The system of claim 12, further comprising:

a second indirect block, comprising a first indirect block checksum and a first indirect block location,

wherein the storage pool allocator is further configured to store the second indirect block in the storage pool using the second dynamic striping policy.

14. The system of claim 12, further comprising:

a data management unit configured to assemble the first indirect block and request the storage pool allocator to store the first indirect block.

15. The system of claim 12, wherein the storage pool comprises at least one storage device.

16. The system of claim 12, wherein the storage pool is divided into a plurality of metaslabs.

17. The system of claim 16, wherein each of the plurality of metaslabs is associated with a metaslab ID.

18. The system of claim 17, wherein the first data block location comprises the metaslab ID and an offset.

19. A computer system for dynamic striping, comprising:

a processor;

a memory;

a storage pool; and

software instructions stored in the memory for enabling the computer system under control of the processor, to;

receive a request to write a file into the storage pool, wherein the file comprises a first data block, a second data block, and a first indirect block;

determine a physical disk location in the storage pool to store the first data block using a first dynamic striping policy, wherein the dynamic striping policy comprises one selected from a group consisting of a dynamic striping policy based on physical disk speed, a dynamic striping policy based on free space available on physical disks, a dynamic striping policy based on load on physical disks, and a round robin policy;

store the first data block using the first dynamic striping policy at the physical disk location in the storage pool;

store the second data block in the storage pool using the first dynamic striping policy;

select a second dynamic striping policy, wherein the second dynamic striping policy comprises one selected from the group consisting of the dynamic striping policy based on physical disk speed, the dynamic striping policy based on free space available on physical disks, the dynamic striping policy based on load on physical disks, and the round robin policy, and wherein the first dynamic striping policy is distinct from the second dynamic striping policy; and

store the first indirect block using the second dynamic striping policy, wherein the indirect block includes a first block pointer referencing the first data block and a second block pointer referencing the second data block.

20. A network system having a plurality of nodes, comprising:

a storage pool comprising a file, wherein the file comprises;

a first data block and a second data block;

a first indirect block; and

a parent block referencing the first indirect block, wherein the first indirect block comprises a first block pointer that references the first data block and a second block pointer that references the second data block, wherein the first block pointer comprises the first data block location and the first data block checksum and the second block pointer comprises the second data block location and the second data block checksum; and

a storage pool allocator configured to;

store the first data block and the second data block in the storage pool using a first dynamic striping policy, wherein the first dynamic striping policy comprises one selected from a group consisting of a dynamic striping policy based on physical disk speed, a dynamic striping policy based on free space available on physical disks, a dynamic striping policy based on load on physical disks, and a round robin policy,

store the first indirect block and the parent block using a second dynamic striping policy, wherein the second dynamic striping policy comprises one selected from the group consisting of the dynamic striping policy based on physical disk speed, the dynamic striping policy based on free space available on physical disks, the dynamic striping policy based on load on physical disks, and the round robin policy, and wherein the first dynamic striping policy is distinct from the second dynamic striping policy,

wherein the storage pool is located on any one of the plurality of nodes,

wherein the storage pool allocator is located on any one of the plurality of nodes.