Data storage system and method for operating a disk controller including allocating disk space for compressed data

• US 5,574,952 A
• Filed: 05/11/1994
• Issued: 11/12/1996
• Est. Priority Date: 05/11/1994
• Status: Expired due to Term

• Current Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Menon, Jaishankar, Brady, James T.
• Primary Examiner(s): Harvey, Jack B.
• Assistant Examiner(s): Etienne, Ario
• Application Number:
  US08/241,987
• Time in Patent Office: 916 Days
• Field of Search: 395/439, 395/442, 395/497.04, 395/888, 395/600, 395/885, 395/886, 369/54, 369/58, 360/48, 341/51, 341/95
• US Class Current: 710/68
• CPC Class Codes: G06F 2003/0697 device management, e.g. handlers, drivers, I/O schedulers
• CPC Class Codes: G06F 3/0601 Interfaces specially adapted for storage systems

Abstract

A data storage system and method for operating a disk controller, and also a disk controller operated in accordance with the method are disclosed. The method includes the steps of allocating a first amount of disk space for a compressed data unit as a first predetermined percentage of an uncompressed size of the data unit; and then increasing the allocation by a second predetermined percentage that is less than the first predetermined percentage to obtain a total amount of allocated disk space. The first predetermined percentage is a function of an expected compression ratio for the data unit, and the second predetermined percentage is a function of an expected change in the size of the compressed data unit as a result of an update operation performed on the data unit. The method further includes compressing the updated data unit; comparing the size of the compressed updated data unit to the total amount of allocated disk space; and if the size of the compressed updated data unit is equal to or less than the total amount of allocated disk space, storing the compressed updated data unit within the allocated disk space. Otherwise, if the size of the compressed updated data unit is greater than the total amount of allocated disk space, other disk space is allocated for storing the compressed data unit.

Claims

1. A data storage system, comprising:

a data compression unit for compressing a data unit prior to the data unit being written to a disk in said data storage unit; and

a disk controller coupled to said disk for allocating contiguous space in said disk for writing a compressed data unit therein, said disk controller including means for comparing a size of a compressed data unit to an amount of said contiguous space previously allocated for the data unit, and for managing said disk in a log-structured manner wherein modifications, instead of being written in place in a data unit being modified, are written to said disk in a sequential log-like file structure and only when said compressed data unit is larger than the contiguous space previously allocated for the data unit.

2. A data storage system as set forth in claim 1 and further comprising:

a first buffer for storing an uncompressed data unit; and

a second buffer for storing the compressed data unit;

wherein said disk controller is responsive to data indicative of a number of bytes in the compressed data unit and in the uncompressed data unit for selectively writing either the content of the first buffer or the second buffer to said disk so as to write the fewest number of bytes to said disk.

3. A method for operating a disk controller, comprising the steps of:

allocating a first amount of disk space for a compressed data unit as a first predetermined percentage of an uncompressed size of the data unit; and

increasing the allocation by a second predetermined percentage that is less than the first predetermined percentage to obtain a total amount of allocated disk space, wherein the first predetermined percentage is a function of an expected compression ratio for the data unit, and the second predetermined percentage is a function of an expected change in the size of the compressed data unit as a result of an update operation performed on the data unit.

4. A method as set forth in claim 3 wherein the first predetermined percentage is approximately 30 percent, and wherein the second predetermined percentage is approximately 5 percent.

5. A method as set forth in claim 3 and further including the steps of:

in response to an update operation performed on the data unit, compressing the updated data unit;

comparing the size of the compressed updated data unit to the total amount of allocated contiguous disk space for a compressed data unit; and

if the size of the compressed updated data unit is equal to or less than the total amount of said allocated contiguous disk space, storing the compressed updated data unit within the allocated contiguous disk space;

else

if the size of the compressed updated data unit is greater than the total amount of allocated contiguous disk space, allocating other disk space for storing the compressed data unit.

6. A method as set forth in claim 3 wherein the first predetermined percentage is based on a most probable data compression ratio for the data unit, and wherein the second predetermined percentage is based on a most probable change in the size of the compressed data unit as a result of an update operation performed on the data unit.

7. A method for operating a disk controller, comprising the steps of:

allocating a first amount of contiguous disk space for a compressed data unit, the first amount being a first percentage of an uncompressed size of the data unit, wherein the first percentage is a function of a most probable data compression ratio for the data unit; and

increasing the first amount of said contiguous disk space by a second percentage that is less than the first percentage to obtain a total amount of allocated contiguous disk space for said compressed data unit, wherein the second percentage is a function of a most probable change in the size of the compressed data unit as a result of an update operation performed on the data unit.