Data storage system having mata bit maps for indicating whether data blocks are invalid in snapshot copies

• US 6,792,518 B2
• Filed: 08/06/2002
• Issued: 09/14/2004
• Est. Priority Date: 08/06/2002
• Status: Expired due to Term

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Armangau, Philippe, Tummala, Himabindu
• Primary Examiner(s): Moazzami, Nasser
• Assistant Examiner(s): CHACE, CHRISTIAN
• Application Number:
  US10/213,241
• Publication Number:
  US 20040030846A1
• Time in Patent Office: 770 Days
• Field of Search: 707/203, 707/204, 711/112, 711/156, 711/161, 711/162, 714/5
• US Class Current: 711/162
• CPC Class Codes: G06F 11/1451 by selection of backup contents
• CPC Class Codes: G06F 11/1469 Backup restoration techniques
• CPC Class Codes: G06F 16/128 Details of file system snapshots on the file-level, e.g. snapshot creation, administration, deletion use of snapshots for error detection or correction G06F11/14, G06F11/16
• CPC Class Codes: G06F 2201/84 Using snapshots, i.e. a logical point-in-time copy of the data
• CPC Class Codes: Y10S 707/99953 Recoverability
• CPC Class Codes: Y10S 707/99954 Version management

Abstract

In a data storage system providing access to a production dataset and snapshot copies of the production dataset, a production meta bit map identifies blocks of storage that are invalid in the production dataset. If a block in the production dataset is invalid when a snapshot copy is being made, then there is no need to copy the block to storage for the snapshot before writing to the block. Moreover, if a block in the production dataset supporting a snapshot copy is dynamically invalidated, it may be kept in the production dataset until it is written to. For this purpose, a respective snapshot copy of the meta bit map is made and kept with each of the snapshot datasets, and the snapshot copies of the meta bit map are merged in order to indicate the blocks that are invalid for all of the snapshots.

Claims

1. A data storage system for providing access to a production dataset and at least one snapshot dataset, said at least one snapshot dataset being the state of the production dataset at a point in time when said at least one snapshot dataset was created, the data storage system comprising storage for storing data of the production dataset and said at least one snapshot dataset,

wherein the data storage system is programmed for maintaining an indication of invalid blocks of the storage that are allocated to the production dataset, and

the data storage system is programmed for performing a write access upon a specified block of the storage allocated to the production dataset by checking whether or not the specified block is indicated to be invalid, and if the specified block is not indicated to be invalid, copying the specified block to said at least one snapshot dataset and then writing to the specified block, and if the specified block is indicated to be invalid, writing to the specified block without copying the specified block to said at least one snapshot dataset.

2. The data storage system as claimed in claim 1, wherein the data storage system is programmed for initially indicating each block of the production dataset as invalid when said each block is allocated to the production dataset but not yet written to, and for changing the indication for said each data block to indicate said each block as not invalid when said each block is written to.

3. The data storage system as claimed in claim 2, wherein the data storage system is programmed for responding to a request for invalidating at least one specified block of the production dataset concurrent with read/write access to the production dataset, and in response, setting the indication of invalid blocks of the storage that are allocated to the production dataset so as to indicate that said at least one specified block of the production dataset is invalid.

4. The data storage system as claimed in claim 1, wherein the data storage system is programmed for maintaining a meta bit map including the indication of invalid blocks of storage that are allocated to said at least one production dataset, the meta bit map being addressable with at least a portion of an address of the specified block to provide a respective bit for signaling whether the specified block is indicated as invalid.

5. The data storage system as claimed in claim 4, wherein each bit in the meta bit map provides an invalidity indication for a respective one and only one block of the storage allocated to the production dataset.

6. The data storage system as claimed in claim 4, wherein each bit in the meta bit map provides an invalidity indication for a respective group of a plurality of blocks of the storage allocated to the production dataset, said each bit indicating that the respective group is invalid only if all of the blocks in the respective group are invalid, and indicating that the respective group is valid if any one of the blocks in the respective the group is not invalid.

7. The data storage system as claimed in claim 1, wherein the data storage system is programmed for maintaining multiple snapshot datasets, each snapshot dataset being the state of the production dataset at a respective point in time when said each snapshot dataset was created, and for maintaining, for each snapshot dataset, a respective indication of the blocks of storage of the production dataset indicated as invalid at the respective time when said each snapshot dataset was created.

8. The data storage system as claimed in claim 7, wherein the data storage system is programmed for maintaining said respective indication for said each snapshot dataset by producing a respective snapshot copy of said indication of invalid blocks of the storage that are allocated to the production dataset as indicated at the respective point in time when the snapshot dataset was created.

9. The data storage system as claimed in claim 8, wherein the data storage system is programmed for copying the specified block to storage for said each snapshot dataset before writing to the specified block if said respective indication for said each snapshot dataset does not indicate that said specified block is invalid for said each snapshot dataset.

10. The data storage system as claimed in claim 7, wherein the data storage system is programmed for merging the respective indications for the snapshot datasets for determining whether or not the specified block is not indicated as invalid for any of the snapshot datasets.

11. The data storage system as claimed in claim 10, wherein the data storage system is programmed for copying the specified block to storage for snapshot datasets before writing to the specified block if the specified block is not indicated as invalid for any of the snapshot datasets.

12. A data storage system for providing access to a production dataset and a plurality of snapshot datasets, each snapshot dataset being the state of the production dataset at a point in time when the snapshot dataset was created, the data storage system comprising storage for storing data of the production dataset and the snapshot datasets,

wherein the data storage system is programmed for maintaining an indication of invalid blocks of the storage that are allocated to the production dataset, and for maintaining, for said each snapshot dataset, a snapshot copy of the indication of invalid blocks of the storage that were allocated to the production dataset at the point of time when said each snapshot dataset was created, and

the data storage system is programmed for performing a write access upon a specified block of the storage allocated to the production dataset by checking whether or not the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, and if the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, copying the specified block to at least the most recent snapshot dataset and then writing to the specified block, and if the specified block is indicated to be invalid in the production dataset and in all of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, writing to the specified block without copying the specified block to said at least the most recent snapshot dataset.

13. The data storage system as claimed in claim 12, wherein the data storage system is programmed for initially indicating each block of the production dataset as invalid when said each block is allocated to the production dataset but not yet written to, and for changing the indication for said each block to indicate said each block as not invalid when said each block is written to.

14. The data storage system as claimed in claim 13, wherein the data storage system is programmed for responding to a request for invalidating at least one specified block of the production dataset concurrent with read/write access to the production dataset, and in response, setting the indication of invalid blocks of the storage that are allocated to the production dataset so as to indicate that said at least one specified block of the production dataset is invalid.

15. The data storage system as claimed in claim 12, wherein the data storage system is programmed for maintaining a meta bit map for the production dataset, the meta bit map including the indication of invalid blocks of storage that are allocated to the production dataset, the meta bit map being addressable with at least a portion of an address of the specified block to provide a respective bit for signaling whether the specified block is indicated as invalid.

16. The data storage system as claimed in claim 15, wherein each bit in the meta bit map provides an invalidity indication for a respective one and only one block of the storage allocated to the production dataset.

17. The data storage system as claimed in claim 16, wherein each bit in the meta bit map provides an invalidity indication for a respective group of a plurality of blocks of the storage allocated to the production dataset, said each bit indicating that the respective group is invalid only if all of the blocks in the respective group are invalid, and indicating that the respective group is not invalid if any one of the blocks in the respective the group is valid.

18. The data storage system as claimed in claim 15, wherein the data storage system is programmed for keeping a respective snapshot copy of the meta bit map for the production dataset for said each snapshot dataset, and the data storage system is programmed for maintaining a merged meta bit map which is a merging of the respective snapshot copies of the meta bit map for the production dataset for the snapshot datasets, the merged meta bit map indicating whether the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created.

19. The data storage system as claimed in claim 18, wherein the data storage system is programmed for updating the merged meta bit map when any one of the snapshot datasets is deleted.

20. A data storage system for providing access to a production dataset and a plurality of snapshot datasets, each snapshot dataset being the state of the production dataset at a point in time when the snapshot dataset was created, the data storage system comprising storage for storing data of the production dataset and the snapshot datasets,

wherein the data storage system is programmed for maintaining a meta bit map indicating invalid blocks of the storage that are allocated to the production dataset, and for maintaining, for said each snapshot dataset, a snapshot copy of the meta bit map indicating invalid blocks of the storage that were allocated to the production dataset at the point of time when said each snapshot dataset was created, and

wherein the data storage system is programmed for using the snapshot copies of the meta bit map for deciding whether or not to copy blocks from the storage of the production dataset to storage of the snapshot datasets for saving the blocks to support the snapshot datasets.

21. The data storage system as claimed in claim 20, wherein the data storage system is programmed for merging the snapshot copies of the meta bit map in order to produce a merged meta bit map for indicating blocks that should be saved to support the snapshot datasets.

22. The data storage system as claimed in claim 21, wherein the data storage system is programmed to update the merged meta bit map when a snapshot dataset is deleted.

23. The data storage system as claimed in claim 21, wherein the data storage system is programmed for accessing the merged meta bit map for deciding whether or not to copy original data of a block of storage of the production dataset to storage of the snapshot datasets when writing new data to the block of storage of the production dataset.

24. A method of operating a data storage system for providing access to a production dataset and at least one snapshot dataset, said at least one snapshot dataset being the state of the production dataset at a point in time when said at least one snapshot dataset was created, the data storage system including storage for storing data of the production dataset and said at least one snapshot dataset, said method comprising:

maintaining an indication of invalid blocks of the storage that are allocated to the production dataset, and

performing a write access upon a specified block of the storage allocated to the production dataset by checking whether or not the specified block is indicated to be invalid, and if the specified block is not indicated to be invalid, copying the specified block to said at least one snapshot dataset and then writing to the specified block, and if the specified block is indicated to be invalid, writing to the specified block without copying the specified block to said at least one snapshot dataset.

25. The method as claimed in claim 24, which includes initially indicating each block of the production dataset as invalid when said each block is allocated to the production dataset but not yet written to, and changing the indication for said each data block to indicate said each block as not invalid when said each block is written to.

26. The method as claimed in claim 25, which includes responding to a request for invalidating at least one specified block of the production dataset concurrent with read/write access to the production dataset, and in response, setting the indication of invalid blocks of the storage that are allocated to the production dataset so as to indicate that said at least one specified block of the production dataset is invalid.

27. The method as claimed in claim 24, which includes maintaining a meta bit map including the indication of invalid blocks of storage that are allocated to the production dataset, and addressing the meta bit map with at least a portion of an address of the specified block to provide a respective bit for signaling whether the specified block is indicated as invalid.

28. The method as claimed in claim 27, wherein each bit in the meta bit map provides an invalidity indication for a respective one and only one block of the storage allocated to the production dataset.

29. The method as claimed in claim 27, wherein each bit in the meta bit map provides an invalidity indication for a respective group of a plurality of blocks of the storage allocated to the production dataset, said each bit indicating that the respective group is invalid only if all of the blocks in the respective group are invalid, and indicating that the respective group is valid if any one of the blocks in the respective the group is not invalid.

30. The method as claimed in claim 24, which includes maintaining multiple snapshot datasets, each snapshot dataset being the state of the production dataset at a respective point in time when said each snapshot dataset was created, and maintaining, for each snapshot dataset, a respective indication of the blocks of storage of the production dataset indicated as invalid at the respective time when said each snapshot dataset was created.

31. The method as claimed in claim 30, which includes maintaining said respective indication for said each snapshot dataset by producing a respective snapshot copy of said indication of invalid blocks of the storage that are allocated to the production dataset as indicated at the respective point in time when the snapshot dataset was created.

32. The method as claimed in claim 31, which includes copying the specified block to storage for said each snapshot dataset before writing to the specified block if said respective indication for said each snapshot dataset does not indicate that said data block is invalid for said each snapshot dataset.

33. The method as claimed in claim 30, which includes merging the respective indications for the snapshot datasets for determining whether or not the specified block is not indicated as invalid for any of the snapshot datasets.

34. The method as claimed in claim 33, which includes copying the specified block to storage for snapshot datasets before writing to the specified block if the specified block is not indicated as invalid for any of the snapshot datasets.

35. A method of operating a data storage system for providing access to a production dataset and a plurality of snapshot datasets, each snapshot dataset being the state of the production dataset at a point in time when the snapshot dataset was created, the data storage system including storage for storing data of the production dataset and the snapshot datasets, said method comprising:

maintaining an indication of invalid blocks of the storage that are allocated to the production dataset, and maintaining, for said each snapshot dataset, a snapshot copy of the indication of invalid blocks of the storage that were allocated to the production dataset at the point of time when said each snapshot dataset was created, and

performing a write access upon a specified block of the storage allocated to the production dataset by checking whether or not the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, and if the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, copying the specified block to at least the most recent snapshot dataset and then writing to the specified block, and if the specified block is indicated to be invalid in the production dataset and in all of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created, writing to the specified block without copying the specified block to said at least the most recent snapshot dataset.

36. The method as claimed in claim 35, which includes initially indicating each block of the production dataset as invalid when said each block is allocated to the production dataset but not yet written to, and changing the indication for said each block to indicate said each block as not invalid when said each block is written to.

37. The method as claimed in claim 36, which includes responding to a request for invalidating at least one specified block of the production dataset concurrent with read/write access to the production dataset, and in response, setting the indication of invalid blocks of the storage that are allocated to the production dataset so as to indicate that said at least one specified block of the production dataset is invalid.

38. The method as claimed in claim 35, which includes maintaining a meta bit map for the production dataset, the meta bit map including the indication of invalid blocks of storage that are allocated to the production dataset, the meta bit map being addressable with at least a portion of an address of the specified block to provide a respective bit for signaling whether the specified block is indicated as invalid.

39. The method as claimed in claim 38, wherein each bit in the meta bit map provides an invalidity indication for a respective one and only one block of the storage allocated to the production dataset.

40. The method as claimed in claim 39, wherein each bit in the meta bit map provides an invalidity indication for a respective group of a plurality of blocks of the storage allocated to the production dataset, said each bit indicating that the respective group is invalid only if all of the blocks in the respective group are invalid, and indicating that the respective group is not invalid if any one of the blocks in the respective the group is valid.

41. The method as claimed in claim 38, which includes keeping a respective snapshot copy of the meta bit map for the production dataset for said each snapshot dataset, and maintaining a merged meta bit map which is a merging of the respective snapshot copies of the meta bit map for the snapshot datasets, the merged meta bit map indicating whether the specified block is not indicated to be invalid in any of the snapshot copies of the indication of invalid blocks that were allocated to the production dataset at the point in time when said each snapshot dataset was created.

42. The method as claimed in claim 41, which includes updating the merged meta bit map when any one of the snapshot datasets is deleted.

43. A method of operating a data storage system for providing access to a production dataset and a plurality of snapshot datasets, each snapshot dataset being the state of the production dataset at a point in time when the snapshot dataset was created, the data storage system including storage for storing data of the production dataset and the snapshot datasets, said method comprising:

maintaining a meta bit map indicating invalid blocks of the storage that are allocated to the production dataset, and for maintaining, for said each snapshot dataset, a snapshot copy of the meta bit map indicating invalid blocks of the storage that were allocated to the production dataset at the point of time when said each snapshot dataset was created, and

using the snapshot copies of the meta bit map for deciding whether or not to copy blocks from the storage of the production dataset to storage of the snapshot datasets for saving the blocks to support the snapshot datasets.

44. The method as claimed in claim 43, which includes merging the snapshot copies of the meta bit map in order to produce a merged meta bit map for indicating blocks that should be saved to support the snapshot datasets.

45. The method as claimed in claim 44, which includes updating the merged meta bit map when a snapshot dataset is deleted.

46. The method as claimed in claim 44, which includes accessing the merged meta bit map for deciding whether or not to copy original data of a block of storage of the production dataset to storage of the snapshot datasets when writing new data to the block of storage of the production dataset.