Browse Prior Art Database

Cluster Size Control Recording

IP.com Disclosure Number: IPCOM000015177D
Original Publication Date: 2002-Apr-11
Included in the Prior Art Database: 2003-Jun-20
Document File: 7 page(s) / 99K

Publishing Venue

IBM

Abstract

Disclosed is a method to improve efficiency of hard disk drive (HDD) disk usage by controlling the size of cluster dynamically.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 66% of the total text.

Page 1 of 7

Cluster Size Control Recording

Disclosed is a method to improve efficiency of hard disk drive (HDD)
disk usage by controlling the size of cluster dynamically.

This method divides each file into several fixed size clusters and
one variable size cluster with software. The size of the variable
cluster is determined depend on the included data size. No modification
is required for data recording and recovering system of HDD.

In the present file system, file is recorded into several clusters
which is the fixed size unit of data for data recording and recovering.
Each cluster consists of several sectors. The sector is a minimum physical
unit of HDD disk system. The size of the sector is 512 Byte which is
determined physically in HDD. The size of the cluster, however, can
be selected by user once at disk format timing. As for the present file
system such as FAT16 or FAT32, the size of the cluster can be selected
from 2 kByte to 32 kByte. That is, user can decide the size of the cluster
at disk formatting by determining how many sectors compose one cluster.

In this file system, file is divided into several clusters of same size.
Final cluster may be occupied just a small part of total cluster area
because it may include small fragment of the data. Statistically, the
average usage of final cluster is 50 %. The number of final clusters is
equal to the number of all files because any files own final cluster.
Thus the wasted disk area is considerably large.

Proposed method saves disk area by adopting variable cluster size for
the final cluster and controlling its size. In the proposed file
system, file is recorded in the following ways.

First, the file size is divided by fixed cluster size 'C' which is
determined by user. The quotient (integer) of this division is the
number of fixed size clusters. The reminder of this division is
recorded in a variable size cluster. The size of fixed size cluster
'C' can be determined by user and this size 'C' is fixed after the
first decision like conventional method. The fixed size cluster is
managed by cluster unit. The variable size cluster is managed by sector
unit. The size of variable size cluster can be from 512 Byte (1 sector
size) to 'C' (1 fixed cluster size).

Second, physical location of the fixed and variable clusters are
proposed. When data is removed, cluster gaps are generated. To make
the size of cluster gap constant for fixed and variable clusters
respectively, the data space of disk is separated into two region.
In the first region (Region 1), only fixed size clusters are recorded.
In the second region (Region 2), only variable size clusters are recorded.
This space separation makes the cluster gap handling and new data
recording after data removal very easy. The Region 1 and Region 2
can be separated physically on the disk with no restriction. However,
there are two methods to separate the disk. Followings are two proposed
methods.

(1) The disk is separated with a certain track and the Region 2 and
the Region 1 are assigned t...