System and method for recovering file systems without a replica

• US 9,990,253 B1
• Filed: 03/31/2011
• Issued: 06/05/2018
• Est. Priority Date: 03/31/2011
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Inventors: Rajimwale, Abhishek, Hsu, Windsor W.
• Primary Examiner(s): Harper, Eliyah S
• Application Number:
  US13/077,781
• Time in Patent Office: 2,623 Days
• Field of Search: 707687, 707713, 707674, 707999102, 707690
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata
• CPC Class Codes: G06F 11/1451 by selection of backup contents

Abstract

According to one embodiment, recovering file systems without a replica is described. In one example, a method includes discovering an error in a first storage partition of the file system containing a current version of a namespace metadata, finding a second storage partition containing a most recent copy of the namespace metadata, and restoring the found copy of the namespace metadata from the second storage partition to the first storage partition for use as a current version of the namespace metadata.

Claims

1. A computer implemented method for recovering a file system, the method comprising:

discovering an error in a first storage partition of the file system containing a current version of a namespace metadata, the error indicating that the first storage partition having an unavailable portion, wherein the namespace metadata contains file names, directory names, attributes and hierarchy of the files and directories, and wherein the file system is a deduplicated file storage system that stores a reference to a previously stored data segment upon receiving a new data segment when the new data segment has already been stored in the file storage system;

finding a second storage partition containing a most recent copy of the namespace metadata, wherein the second storage partition is among a plurality of storage partitions to find the most recent copy of the namespace metadata from a plurality of copies of the namespace metadata in the plurality of storage partitions, wherein each of the plurality of storage partitions is an archive partition archived and sealed at a different point in time, wherein when an archive partition is archived and sealed, an entire copy of current namespace metadata describing a namespace of the file system at a point in time is captured and stored in the archive partition; and

restoring the found copy of the namespace metadata from the second storage partition to the first storage partition for use as a current version of the namespace metadata,

wherein prior to discovering the error in the first storage partition, data is moved from the first storage partition to the second storage partition, and a copy of the current version of the namespace metadata is stored in the second storage partition.

2. The method of claim 1, wherein the found copy of the namespace metadata includes namespace data for a portion of the large scale data store outside the second partition.

3. The method of claim 2, wherein the found copy of the namespace metadata includes the entire namespace data of the large scale data store at a time corresponding to the recentness of the most recent copy.

4. The method of claim 1, further comprising removing files from the restored namespace metadata that had data in the first storage partition.

5. The method of claim 1, wherein storing the copy of the current version of the namespace metadata is performed when the second storage partition is filled up to a predetermined threshold.

6. The method of claim 5, wherein the second storage partition is sealed against storing additional data after storing the copy of the current version of the namespace metadata.

7. The method of claim 1, wherein restoring comprises copying the namespace metadata to a replacement storage partition and designating the replacement storage partition as storing the current version of the namespace metadata.

8. The method of claim 1, wherein the most recent copy of the namespace metadata comprises a version of the namespace metadata that is prior to the namespace metadata of the first storage partition.

9. The method of claim 1, wherein the second storage partition is located at a replica site having a second replica file storage system coupled to the file storage system through a network and wherein restoring the found copy of the namespace metadata comprises comprise receiving the found namespace metadata over the network.

10. A non-transitory computer-readable storage medium having instructions stored therein, which when executed by a computer, cause the computer to perform a method for recovering a file storage system global namespace metadata, the method comprising:

discovering an error in a first storage partition of the file system containing a current version of a namespace metadata, the error indicating that the first storage partition having an unavailable portion, wherein the namespace metadata contains file names, directory names, attributes and hierarchy of the files and directories, and wherein the file system is a deduplicated file storage system that stores a reference to a previously stored data segment upon receiving a new data segment when the new data segment has already been stored in the file storage system;

finding a second storage partition containing a most recent copy of the namespace

metadata, wherein the second storage partition is among a plurality of storage partitions to find the most recent copy of the namespace metadata from a plurality of copies of the namespace metadata in the plurality of storage partitions, and

wherein each of the plurality of storage partitions is an archive partition archived and sealed at a different point in time, wherein when an archive partition is archived and sealed, an entire copy of current namespace metadata describing a namespace of the file system at a point in time is captured and stored in the archive partition; and

restoring the found copy of the namespace metadata from the second storage partition to the first storage partition for use as a current version of the namespace metadata,

wherein prior to discovering the error in the first storage partition, data is moved from the first storage partition to the second storage partition, and a copy of the current version of the namespace metadata is stored in the second storage partition.

11. The non-transitory computer-readable storage medium of claim 10, wherein restoring comprises copying the namespace data to a replacement storage partition and designating the replacement storage partition as an active partition.

12. The non-transitory computer-readable storage medium of claim 10, wherein finding a backup storage partition comprises finding a storage partition previously used to archive portions of the active partition and previously sealed against writing additional data.

13. The non-transitory computer-readable storage medium of claim 12, wherein the most recent copy of namespace data comprises a prior version of namespace data from the sealed partition.

14. A storage system comprising;

a system memory;

a processor;

a file system executed in the system memory,

the file system having a plurality of storage partitions including at least a first storage partition and a second storage partition, each storage partition including a version of global namespace metadata of the file system, wherein the version of global namespace metadata contains file names, directory names, attributes and hierarchy of the files and directories, and wherein the file system is a deduplicated file storage system that stores a reference to a previously stored data segment upon receiving a new data segment when the new data segment has already been stored in the file storage system;

a partition manager coupled to the file system and the system memory, wherein the partition manager is configured to

discover an error in an active storage partition of the file storage system, the error indicating that the first storage partition having an unavailable portion;

find a backup storage partition containing the most recent global namespace metadata, wherein the second storage partition is among a plurality of storage partitions to find the most recent copy of the namespace metadata from a plurality of copies of the namespace metadata in the plurality of storage partitions, and

wherein each of the plurality of storage partitions is an archive partition archived and sealed at a different point in time, wherein when an archive partition is archived and sealed, an entire copy of current namespace metadata describing a namespace of the file system at a point in time is captured and stored in the archive partition;

restore the global namespace metadata in the active storage partition using the backup storage partition; and

prune the restored namespace to remove files that had data in the active partition,

wherein prior to discovering the error in the first storage partition, data is moved from the first storage partition to the second storage partition, and a copy of the current version of the namespace metadata is stored in the second storage partition.

15. The system of claim 14, wherein restoring comprises copying the global namespace data to a replacement storage partition and designating the replacement storage partition as an active partition.

16. The system of claim 14, wherein finding a backup storage partition comprises finding a backup storage partition previously used to back up the global namespace metadata of the active partition and previously sealed against writing additional data.

17. The system of claim 14, wherein the most recent copy of namespace data comprises a prior version of namespace data from the sealed partition.

18. The system of claim 16, wherein the global namespace metadata includes namespace metadata identifying files of the file storage system in the active storage partition and in other storage partitions.