Error detection for files

• US 8,793,440 B2
• Filed: 06/17/2010
• Issued: 07/29/2014
• Est. Priority Date: 06/17/2010
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Miller, Thomas J., Cargille, Jonathan M., Tipton, William R., Verma, Surendra
• Primary Examiner(s): ELMORE, STEPHEN C
• Application Number:
  US12/817,219
• Publication Number:
  US 20110314229A1
• Time in Patent Office: 1,503 Days
• Field of Search: 711/141, 711/154, 711/156, 714/52, 714/54, 714/746, 714/764
• US Class Current: 711/141
• CPC Class Codes: G06F 11/004 Error avoidance G06F11/07 and subgroups take precedence
• CPC Class Codes: G06F 11/0766 Error or fault reporting or storing
• CPC Class Codes: G06F 11/0772 Means for error signaling, e.g. using interrupts, exception flags, dedicated error registers
• CPC Class Codes: G06F 11/1004 to protect a block of data words, e.g. CRC or checksum G06F11/1076 takes precedence; security arrangements for protecting computers or computer systems against unauthorized activity G06F21/00
• CPC Class Codes: G06F 11/1064 in cache or content addressable memories
• CPC Class Codes: G06F 11/1072 in multilevel memories
• CPC Class Codes: G06F 11/1402 Saving, restoring, recovering or retrying
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata
• CPC Class Codes: G06F 11/1441 Resetting or repowering
• CPC Class Codes: G06F 12/0826 Limited pointers directories; State-only directories without pointers
• CPC Class Codes: G06F 12/0828 with concurrent directory accessing, i.e. handling multiple concurrent coherency transactions
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: G06F 16/178 Techniques for file synchronisation in file systems
• CPC Class Codes: G06F 2201/82 Solving problems relating to consistency

Abstract

Aspects of the subject matter described herein relate to error detection for files. In aspects, before allowing updates to a clean file, a flag marking the file as dirty is written to non-volatile storage. Thereafter, the file may be updated as long as desired. Periodically or at some other time, the file may be marked as clean after all outstanding updates to the file and error codes associated with the file are written to storage. While waiting for outstanding updates and error codes to be written to storage, if additional requests to update the file are received, the file may be marked as dirty again prior to allowing the additional requests to update the file. The request to write a clean flag regarding the file may be done lazily.

Claims

1. A method performed by a computer, the method comprising:

receiving a request to update a file;

refraining from updating the file until after a dirty flag for the file has been written to non-volatile storage;

setting a dirty flag to indicate the file is potentially dirty;

sending a request to write the dirty flag to the non-volatile storage;

receiving an indication that the dirty flag has been written to the non-volatile storage; and

after receiving the indication, updating, by the computer, the file in accordance with the request to update the file.

2. The method of claim 1, further comprising:

determining that error codes and data associated with outstanding requests to update the file are to be written to the non-volatile storage;

in a data structure, holding an indication of one or more incoming requests, if any, to update the file;

receiving an indication that the data associated with the one or more outstanding requests to update the file has been written to the non-volatile storage;

receiving an indication that the error codes associated with the outstanding requests have been written to the non-volatile storage;

setting a clean flag to indicate that the file is no longer dirty; and

sending a request to write the clean flag to the non-volatile storage.

3. The method of claim 2, further comprising updating the file in accordance with the one or more incoming requests potentially concurrently with setting the clean flag.

4. The method of claim 3, wherein setting the clean flag is done in context of a transaction, and further comprising comparing first error codes of metadata written in the context of the transaction with second error codes computed from the file as updated and aborting the transaction if the first error codes do not match the second error codes.

5. The method of claim 2, further comprising sending a flush command to flush the clean flag to the non-volatile storage prior to allowing the file to be updated in accordance with the one or more incoming requests to update the file.

6. The method of claim 2, wherein the determining comprises determining that a pre-defined period has expired.

7. The method of claim 6, further comprising tuning the pre-defined period according to one or more characteristics of the non-volatile storage.

8. The method of claim 2, wherein the determining comprises receiving a request to close the file.

9. The method of claim 2, further comprising writing the error codes and the clean flag to the non-volatile storage in context of a transaction.

10. The method of claim 1, wherein the updating comprises updating the file in place without copying the file or any portion thereof to another location of the non-volatile storage.

11. At least one memory storage device storing computer-executable instructions that, when executed by a computing device, cause the computing device to perform actions, comprising:

in a file system in which at least some files are updated by writing in place, as part of a recovery for a file maintained on the file system, performing the actions, comprising;

reading a flag indicative of whether the file was marked as dirty, the flag previously written to non-volatile storage prior to allowing any updates to the file;

if the flag indicates that the file was marked as dirty, indicating that the file is suspect;

if the flag indicates that the file was not marked as dirty, indicating that the file is clean.

12. The at least one memory storage device of claim 11, wherein the indicating that the file is clean comprises indicating that all error codes associated with the file are stored on the non-volatile storage.

13. The at least one memory storage device of claim 11, wherein the indicating that the file is suspect comprises indicating a potential that an error code associated with the file and/or update data associated with the error code was/were not written to the non-volatile storage.

14. The at least one memory storage device of claim 11, further comprising:

prior to the recovery, performing further actions, comprising;

receiving a request to update the file; and

refraining from updating the file until after a dirty flag for the file has been written to the non-volatile storage.

15. The at least one memory storage device of claim 11, further comprising periodically writing all outstanding updates and error codes associated therewith to the non-volatile storage and thereafter lazily sending a request to write the flag to mark the file as clean.

16. A system comprising a computing device and at least one program module that are together configured for performing actions comprising:

providing access to files of a storage device;

determining error codes for files of the storage device, each error code including data regarding a file corresponding to the error code, the data usable to detect one or more errors regarding whether the file or a portion thereof was correctly written to the storage device;

writing a dirty flag to the storage device prior to allowing any update to the file;

receiving a request to update the file;

marking the file as dirty prior to allowing the update to the file in accordance with the request; and

determining an error code for the file in response to changes made to the file in response to the request to update the file.

17. The system of claim 16, the at least one program module comprising a clean flag manager operable to set a clean flag and write the clean flag to the storage device lazily or prior to allowing additional updates to the file in conjunction with or after all error codes associated with the file have been written to the storage device.

18. The system of claim 16, the at least one program module comprising a recovery manager operable to read data from the storage device that indicates whether the file is marked as dirty, and if the data indicates that the file is marked as dirty, indicating that the file is suspect;

otherwise indicating that the file is clean.

19. The system of claim 16, wherein an error code and data corresponding to a file are associated with the error code are non-atomically written to the storage device.