Architecture for supporting sparse volumes

Abstract

An architecture, including a file-level protocol, for supporting sparse volumes on a storage system is provided. The file-level protocol provides coherency checking for use in retrieving data stored on a backing store remote from a storage system.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS[0001]The present application is a continuation of U.S. patent application Ser. No. 11 / 409,624, filed on Apr. 24, 2006, which claims the benefit of U.S. Provisional Patent Application Ser. No. 60 / 674,641, which was filed on Apr. 25, 2005, by Jason Ansel Lango for an Architecture For Supporting Sparse Volumes and is hereby incorporated by reference.RELATED APPLICATION[0002]This application is related to U.S. patent application Ser. No. 11 / 409,887, filed on Apr. 24, 2006, entitled SYSTEM AND METHOD FOR SPARSE VOLUMES, by Jason Lango, et al, the contents of which are hereby incorporated by reference.FIELD OF THE INVENTION[0003]The present invention relates to file systems and, more specifically, to a protocol for use with a file system that includes volumes having one or more files with blocks that require a special operation to retrieve data associated therewith from a remote backing store.BACKGROUND OF THE INVENTION[0004]A storage system typical...

AI Technical Summary

Benefits of technology

[0018]is The present invention overcomes the disadvantages of the prior art by providing a system and method for supporting a sparse volume within a file system of a storage system. As used herein, a sparse volume contains one or more files with at least one data block (i.e., an absent block) that is not stored locally on disk coupled to the storage system. By not storing the data block (or a block of zeros as in a hole environment), the sparse volume may be generated and exported quickly with minimal write operations required. The “missing” data of an absent block is stored on an alternate, possibly remote, source (e.g., a backing store) and is illustratively retrieved using a remote fetch operation.

Problems solved by technology

During storage system operation, a volume (or other data container, such as a file or directory) may become corrupted due to, e.g., physical damage to the underlying storage devices, software errors in the storage operating system executing on the storage system or an improperly executing application program that modifies data in the volume.

Here, reconstruction may occur “on-the-fly”, resulting in virtually no discernable s time where the data is not accessible.

In other situations, reconstruction of the data may not be possible.

However, such “brute force” data copying is generally inefficient, as the time required to transfer substantial amounts of data, e.g., terabytes, may be on the order of days.

Similar disadvantages are associated with restoring data from a tape device or other offline data storage.

However, a noted disadvantage of such a hole-based technique is that repeated write operations are needed to generate the appropriate number of zero-filled blocks on disk for the volume.

Thus, the creation of a hole-filled volume is oftentimes impractical due to the need for quick data access to a volume.

As a result, the near line storage server is typically optimized to perform bulk data restore operations, but suffers reduced performance when serving individual client data access requests.

This latter situation may arise where a primary storage system encounters a failure that damages its volume in such a manner that a client must send its data access requests to the server in order to access data in the volume.

Images