Optimized logging of data elements to a data storage device

• US 6,189,069 B1
• Filed: 02/17/1998
• Issued: 02/13/2001
• Est. Priority Date: 02/17/1998
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Parkes, Michael Andrew Brian, Nolte, Barry Michael, Boa, Douglas Stewart
• Primary Examiner(s): Yoo, Do Hyun
• Assistant Examiner(s): GOODARZI, NASSER MOAZZAMI
• Application Number:
  US09/024,248
• Time in Patent Office: 1,092 Days
• Field of Search: 711/154, 711/129, 711/100, 707/8, 707/205, 710/129, 370/449
• US Class Current: 711/100
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0656 Data buffering arrangements
• CPC Class Codes: G06F 3/0676 Magnetic disk device
• CPC Class Codes: G06F 8/4442 Reducing the number of cache misses; Data prefetching cache prefetching G06F12/0862
• CPC Class Codes: G06F 8/51 Source to source
• CPC Class Codes: Y10S 707/99938 Concurrency, e.g. lock management in shared database

Abstract

An invention for optimizing the logging of data elements to a hardware device is described. Using this invention, a large stream of data can be written to a hardware device at a rate that approaches the limits of the physical characteristics of the hardware device. To achieve this efficiency, the performance of a logging operation is divided between a data source and a data logging software processes which operate in different threads or processes. The data source collects pieces of data to be written to the hardware device into a larger buffer retrieved from a pool of empty buffers. When a buffer becomes full, the buffer is placed at the end of a full buffer queue. The data logger, operating asynchronously, retrieves a full buffer from the queue and then writes the data to the hardware. In this fashion, the hardware data logging device is able to perform optimally while creating or expanding the file.

Claims

1. In a computer system, a method for logging data produced by a data source to a hardware device by a data logger operating in a separate thread than the data source, the method comprising the steps of:

retrieving from a set of empty buffers, by the data source, a first buffer pointing to a first page in memory;

storing in the first page in memory, by the data source, data records to be logged to the hardware device;

placing the first buffer in a queue of full buffers;

retrieving from the set of empty buffers, by the data source, a second buffer pointing to a second page in memory;

storing in the second page in memory, by the data source, data records to be logged to the hardware device;

placing the second buffer in the queue of full buffers;

retrieving, by the data logger, the first and second buffers from the queue of full buffers;

packaging the first and second buffers into a data block; and

logging on the hardware device the data stored in the memory pages corresponding to the buffers contained in the data block.

2. The method of claim 1, wherein the logging step is performed using direct memory access (DMA).

3. The method of claim 1, wherein the logging step includes passing the data block to a hardware controller for logging on the hardware device.

4. The method of claim 3, wherein the logging step includes using a WriteFileGather operation.

5. The method of claim 3, wherein the hardware controller is a SCSI device controller.

6. The method of claim 3, further comprising the step of returning the first and second buffers to the set of empty buffers after the logging step is performed.

7. The method of claim 1, wherein the hardware device is a hard disk drive.

8. The method of claim 1, wherein the logging step is performed using a synchronous write operation to the hardware device.

9. The method of claim 1, wherein the logging step is performed using an asynchronous write operation to the hardware device.

10. The method of claim 1, wherein the data source and data logger operate in separate processes.

11. The method of claim 1, further comprising a second data source operating in a separate thread than the data logger, the method further comprising the steps of:

retrieving from the set of empty buffers, by the second data source, a third buffer pointing to a third page in memory;

storing in the third page in memory, by the second data source, data records to be logged to the hardware device; and

placing the third buffer in the queue of full buffers;

retrieving from the set of empty buffers, by the second data source, a fourth buffer pointing to a fourth page in memory;

storing in the fourth page in memory, by the second data source, data records to be logged to the hardware device; and

placing the fourth buffer in the queue of full buffers;

retrieving, by the data logger, the third and fourth buffers from the queue of full buffers;

packaging the third and fourth buffers into a second data block; and

logging on the hardware device the data stored in the memory pages corresponding to the third and fourth buffers contained in the second data block.

12. The method of claim 1, wherein the computer system further comprises a second hardware controller, a second hardware device, and a second data logger, the method further comprising the steps of:

retrieving from the set of empty buffers, by the data source, a third buffer pointing to a third page in memory;

storing in the third page in memory, by the data source, data records to be logged to the second hardware device;

placing the third buffer in the queue of full buffers;

retrieving from the set of empty buffers, by the data source, a fourth buffer pointing to a fourth page in memory;

storing in the fourth page in memory, by the data source, data records to be logged to the second hardware device;

placing the fourth buffer in a queue of full buffers;

retrieving, by the second data logger, the third and fourth buffers from the queue of full buffers;

packaging the third and fourth buffers into a second data block; and

logging on the second hardware device the data stored in the memory pages corresponding to the buffers contained in the second data block.

13. A computer-readable medium having computer-executable instructions representing the method of claim 1.

14. In a computer system, a method for logging a stream of data produced by a data source to a hardware device by a data logger operating in a separate thread than the data source, the method comprising the steps of:

retrieving from a set of empty buffers a first buffer pointing to a first block of memory;

filling the first block of memory with data records from the stream of data;

placing the first buffer in a queue of full buffers;

retrieving from the set of empty buffers a second buffer pointing to a second block of memory;

filling the second block of memory with data records from the stream of data;

placing the second buffer in the queue of full buffers;

retrieving the first and second buffers from the queue of full buffers;

packaging the contents of the first and second buffers into a larger data block;

passing the larger data block to a controller for logging the data stored in the first and second blocks of memory on the hardware device; and

logging the data stored in the first and second blocks of memory on the hardware device.

15. The method of claim 14, wherein the data source and data logger operate in separate processes.

16. The method of claim 14, wherein the first and second blocks of memory together do not comprise a contiguous block of memory.

17. The method of claim 14, wherein the logging step includes transferring the data using direct memory access (DMA).

18. The method of claim 17, wherein the hardware device is a disk.

19. The method of claim 18, wherein the logging step includes expanding the size of a data file stored on the disk with a single operation.

20. The method of claim 19, wherein the packaging step packages an amount of data equivalent to a multiple of the file allocation size of the disk.

21. The method of claim 19, wherein the size of the first and the size of second blocks of memory are both a multiple of the file allocation size of the disk.

22. The method of claim 19, wherein the passing the larger data block to the hardware controller is performed by the WriteFileGather operation.

23. The method of claim 14, further comprising the step of placing the first and second buffers in the set of empty buffers after the data records of the larger data block have been logged on the hardware device.

24. A computer-readable medium having computer-executable instructions representing the method of claim 14.

25. The method of claim 14, wherein the data source comprises a plurality of applications.

26. In a computer system, a method for logging data produced by a data source to a hard disk drive by a data logger operating in a separate thread than the data source, the method comprising the steps of:

retrieving from a set of empty buffers, by the data source, a first buffer pointing to a first page in memory;

storing in the first page in memory, by the data source, data records to be logged to the hard disk drive;

placing the first buffer in a queue of full buffers;

retrieving from the set of empty buffers, by the data source, a second buffer pointing to a second page in memory;

storing in the second page in memory, by the data source, data records to be logged to the hardware device;

placing the second buffer in the queue of full buffers;

retrieving the first and second buffers from the queue of full buffers;

packaging the addresses contained in the first and second buffers in a parameter array;

passing the parameter array to a hard drive controller;

logging on the hard disk drive the data stored in the memory pages corresponding to the addresses contained in the parameter array; and

returning the first and second buffers to the set of empty buffers.

27. The method of claim 26, wherein the data source and data logger operate in separate processes.

28. The method of claim 27, wherein the logging step includes transferring the data using direct memory access (DMA).

29. The method of claim 28, wherein the first and second memory pages together do not comprise a contiguous block of memory.

30. The method of claim 29, wherein the size of the first and second buffers are a multiple of the file allocation size of the disk.

31. A computer-readable medium having computer-executable instructions representing the method of claim 26.

32. The method of claim 26, wherein the passing of the parameter array step is performed using the WriteFileGather operation.