Data storage for file updates

• US 8,301,588 B2
• Filed: 03/07/2008
• Issued: 10/30/2012
• Est. Priority Date: 03/07/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Kannan, Naresh, Clarke, Simon Peter, Skaria, Simon, Watson, Andrew Sean, Patel, Mitesh Pankaj, Bose, Miko Arnab Sakya Singha
• Primary Examiner(s): Perveen, Rehana
• Assistant Examiner(s): BUI, THUY T
• Application Number:
  US12/044,744
• Publication Number:
  US 20090228473A1
• Time in Patent Office: 1,698 Days
• Field of Search: 707/608, 707/609
• US Class Current: 707/609
• CPC Class Codes: G06F 16/1865 Transactional file systems

Abstract

A storage system according to one embodiment for managing a first file includes a blob storage and a handler. The blob storage may be configured to store data related to the first file temporarily. The handler may be configured to store in the blob storage any data related to the first file and received from an application. The handler also may be configured to transfer at least a portion of the data from the blob storage to persistent storage in response to a commit instruction received from an application or generated by the handler. The handler also may be configured to retrieve any portion of the data stored in the blob storage and to send the same to a requesting application in response to an access request.

Claims

1. A method for updating a file maintained in cold storage, the method comprising:

receiving at a first computing device a plurality of updates for the file from a plurality of computing devices for a first application;

storing the updates from the plurality of computing devices temporarily at a blob storage on the first computing device, the updates being stored at the blob storage independently of an underlying file format of data in the updates, at least one of the updates being stored as an arbitrary length binary string, the arbitrary length binary string being different than the underlying file format of data in the at least one of the updates, the updates from the plurality of computing devices being stored temporarily at the blob storage until a commit instruction is received at the first computing device;

receiving the commit instruction;

retrieving at the first computing device at least a portion of the file from the cold storage in response to the commit instruction;

retrieving at the first computing device at least a first update of the plurality of updates from the blob storage;

integrating at the first computing device the first update with the portion of the file to generate an updated file, the integrating of the first update with the portion of the file being performed at the first computing device before the updated file is stored in the cold storage;

performing at least one pre-storage process on the updated file, wherein the pre-storage process was not performed on the first update prior to storing the first update at the blob storage;

storing the updated file in the cold storage;

receiving a request at the first computing device from a second application to access the first update from the blob storage before the first update is integrated and stored in the cold storage;

reading from the blob storage to obtain a copy of the first update; and

sending the copy of the first update to the second application.

2. The method of claim 1, wherein performing at least one pre-storage process on the updated file further comprises performing an expensive pre-storage process on the updated file.

3. The method of claim 1, wherein retrieving at least the portion of the file from the cold storage comprises:

sending a request for at least the portion of the file to a remote storage device on which the cold storage is maintained; and

receiving at least the portion of the file from the remote storage device.

4. The method of claim 1, wherein integrating at least the first update with at least the portion of the file to generate the updated file comprises:

identifying a file format of the file;

identifying a relationship between the first update and at least the portion of the file; and

integrating the first update with at least the portion of the file based on the file format of the file and the relationship between the first update and at least the portion of the file.

5. The method of claim 1, wherein storing the updated file in the cold storage comprises:

identifying a file format of the updated file;

determining an appropriate storage API from a plurality of storage APIs for storing the updated file based on the file format of the updated file; and

storing the updated file in the cold storage using the appropriate storage API.

6. The method of claim 1, further comprising:

receiving a removal request at the first computing device from the second application to remove the first update from the blob storage without committing the first update; and

removing the first update from the blob storage.

7. A storage system for managing a first file, the storage system comprising:

at least one computing device and memory;

a blob storage, on a first computing device, configured to store data related to the first file temporarily and independently of an underlying file format of the data, the blob storage being configured to store a plurality of data entries from a plurality of computing devices, the blob storage storing at least one of the data entries as an arbitrary length binary string, the arbitrary length binary string being different than the underlying file format of the at least one of the data entries;

a cold storage, on the first computing device, in which the first file is stored;

a handler, on the first computing device, communicatively coupled to a plurality of applications, the handler being configured to store in the blob storage any data received from any of the applications and related to the first file, the handler being configured to store the data in the blob storage until a commit instruction is received, the handler also being configured to coalesce and process at least a portion of the data from the blob storage in response to receipt of the commit instruction, the coalescing and processing of the at least a portion of the data from the blob storage comprising integrating the at least a portion of the data from the blob storage with data obtained from the cold storage, the handler also being configured to retrieve any portion of the data from the blob storage and to send the portion of the data to a first one of the applications when the first application requests access to the portion of the data, wherein the handler is configured to coalesce and process at least the portion of the data from the blob storage by selecting an appropriate storage API from the plurality of storage APIs based on a file format of the first file, and

a file format metadata storage containing file format metadata associated with the first file, the file format metadata indicating a file structure of the first file to enable incremental access to the first file.

8. The storage system of claim 7, wherein the cold storage is associated with at least one pre-storage process that is performed on any data being stored in the cold storage.

9. The storage system of claim 7, wherein the cold storage is maintained at a remote location from the blob storage.

10. The storage system of claim 7, wherein the cold storage is associated with a plurality of storage APIs, each storage API being configured to store a certain type of data in the cold storage.

11. The storage system of claim 7, wherein the handler is configured to search the blob storage for data in response to access requests received from one of the applications.

12. A computer readable storage memory encoding instructions for implementing a computer method of updating a master copy of a document, the method comprising:

receiving at a first storage device from a plurality of computing devices a plurality of incremental updates for the document, each incremental update being received from one of a plurality of applications, wherein at least two of the incremental updates are received from different applications;

storing the incremental updates from the plurality of computing devices at a blob storage on the first storage device, the incremental updates being stored at the blob storage independently of an underlying file format of data in the incremental updates, at least one of the incremental updates being stored as an arbitrary length binary string, the arbitrary length binary string being different than the underlying format of the at least one incremental update, and the incremental updates being stored at a same database table in the blob storage, the updates from the plurality of computing devices being stored temporarily at the blob storage until a commit instruction is received at the first computing device;

receiving at the first storage device the commit instruction;

retrieving at the first storage device the master copy of the document from a cold storage in response to receiving the commit instruction;

coalescing at the first storage device the incremental updates with the retrieved master copy of the document to create an updated master copy of the document, the coalescing including integrating the incremental updates with the retrieved master copy of the document, the integrating of the incremental updates with the retrieved master copy of the document being performed before the updated master copy of the document is stored in the cold storage; and

storing the updated master copy of the document in the cold storage in place of the master copy of the document

receiving an access request from one of the applications to read from the master copy of the document prior to receiving the commit instruction;

analyzing the incremental updates stored in the blob storage to determine whether any of the incremental updates satisfy the access request;

sending a copy of any incremental update satisfying the access request;

retrieving at least a relevant portion of the master copy of the document from the cold storage if none of the incremental updates satisfies the access request;

receiving from one of the applications a tag associated with one of the incremental updates; and

storing the tag in the blob storage in association with the incremental update, the incremental update being searchable within the blob storage by the tag.

13. The computer readable storage memory of claim 12, further comprising:

emptying the blob storage when the updated master copy is stored in the cold storage.