OPERATING SYSTEM

Lecture 10

File Management

Dr Victoria Mukami
INTRODUCTION
During this lecture, we will focus on file management. More specifically, we will review
the role of the file manager, the difference between files and file systems, and the
different types of physical storage

Learning objectives
By the end of this topic, you should be able to:
1. Describe concepts around files and file systems
2. Understand the role of the file manager.
3. Describe the different types of physical storage

OVERVIEW
So far, we have reviewed the input and output management from an operating
system's point of view. We now move to file management. Most of the work that users
do with a computer or with an operating system is working on files and storing the
same files. We start by reviewing the concepts around files and the file system.

FILES AND FILE SYSTEM

A file can be defined as a collection of related data, programs, and information that is
stored within a computer. A file system on the other hand can be defined as one that
allows users to create data collections [2]. File systems also have several desirable
properties

Long-term existence [1]: Files are stored permanently until when the user deletes
them. That is the system does not clear files when the user stops using the machine.

Sharable between processes [1]: Files are named appropriately and can include
access permissions that allows for sharing.

Structure [1]: A file can have an internal structure that is convenient for a specific
application. File organization is also key for any operating system.

Functions of a file system

A file system has several functions that can be performed on files. Some of these
functions are discussed below:

Create: A new file is created and placed within the existing structure of files
 Delete: A file is removed from the file system and destroyed

Open: An already created file is opened and edited or worked on.

Close: An open file is closed when an application or process no longer needs

it.

Read: This is when a process reads a section or all the contents of a file

Write: When a process updates a file and makes changes to the file.

FILE MANAGER

The file manager is responsible for creating, editing, and controlling access to files [1].
It also manages the resources used by those files. The file manager needs to perform
four tasks to carry out its responsibilities [1]

1. Keep track of where each file is stored [1]

2. Use policies that determine where and how each file will be stored. This is done
by ensuring that the available space is used efficiently.
3. Allocate each file to a user who has rights to the file and track the user's use of
the file
4. Deallocate a file when it is ready to be returned to storage. Additionally, indicate
its availability if it is needed by others.

The file manager keeps track of "its file with directories that contain the filename, its
physical location in secondary storage and information about it "[1].

Within the cloud, the file manager does not have a specific job as all the files are
managed locally on devices. Remember on the cloud, the files are accessible remotely
but stored within specific devices. Before we continue, let us review some definitions:

Field: This is the basic element of data [2]. A field represents a single piece of data. It
is identified by a field header.

Record: This is a group of related fields. One record represents several related fields.

File: When several records are placed together, it becomes a file. In other words, a
file is a group of related records.

Database: This is a group of related files. Within the database, several relationships
exist between files making the structure complex.
Programs files contain instructions and data files contain data and the file manager
treats them the same way [1]. Directories, on the other hand, are special files that
contain lists of filenames and their attributes [1].

Files in a computer system reside within secondary storage. Each storage unit is
considered a volume with each volume having several files [1]. Each volume gets a
name and this name and other details concerning the volume are written by the file
manager to a place that is accessible. Once the volume descriptor (the common area
with the volume information) is created a master file directory (MDF) is stored and lists
the names and characteristics of every file within the volume [1]. The names within the
MDF refer to program files, data files, and system files [1].

A subdirectory is created when a user opens an account to access the computer

system [1]. This entry within MDF appears as a subdirectory even though the user’s
directory is treated like a file. Files are named depending on the application. File
names are either absolute or relative.

Relative filename

A relative file name includes only the name given by the user. If a user created a file
called Document_One, that would be the file name.

Absolute filename

An absolute filename includes all information including where the path is found. With
the example before where our relative file name was Document_One, the absolute file
name on a Windows machine would be C:\Users\Admin\Document_One.

Different operating systems have different ways in which file names can be done. The
following table is a summary of different operating systems and the requirements for
files [1].

Table 1: File Names Summary (Adapted) [1]

Operating Special Not Allowed Extension Maximum

System Case Characters Required Character
Sensitive Length
Android Yes *?$&[]/\ No 126/255*
Linux Yes *?$&[]/\ No 256
Mac OS X Yes Colon No 255
 MS-DOS No Only hyphens and Yes 8.3
underlines allowed

UNIX Yes *?$&[]/\ No 256

Windows No Most special Yes 255/256
characters are not
allowed

FILE ORGANIZATION

When we talk about file organization, we mainly look at the arrangement of records
within a file. Files are composed of records [1]. The format for any record is either fixed
length or variable length.

Fixed-length records: these are the easiest to access directly [1]. They are stored in
a fixed size, therefore, any part of the records that is larger than the fixed size is
truncated [1]. If the record is smaller than the fixed length, then the extra storage space
will be wasted.

Variable-length records: these do not leave any empty storage and do not cause any
truncation. They are not easily accessible directly because it is not easy to locate
where all the files are stored [1]. When accessed sequentially they are easy to locate.

PHYSICAL FILE ORGANIZATION

This concerns the way records are arranged and the characteristics used to store
them. There are several file organization methods.

Sequential record organization: this is the easiest to implement since variable-

length records are retrieved one after the other. To access the files, the file manager
just needs to search from the beginning until it finds it.

Direct file organization: this uses direct access files. This gives users the flexibility
to access any record in any order. Unlike the sequential record organization that
requires a sequence to retrieve files, this is known as a random organization [1].

Indexed sequential record organization: this takes the good of sequential and direct
access. It uses the indexed sequential access method application that handles
overflows while maintaining the record order [1].
PHYSICAL STORAGE ALLOCATION

Remember that records are subdivided into fields. For instance, a record may contain
several fields. Most records are managed by application programs apart from those
that are oriented toward database applications are managed by the file manager [1].
Physical storage is either contiguous, non-contiguous, or indexed.

Contiguous Storage
Contiguous means one after the other. In this regard, contiguous storage refers to
records being stored one after the other [1]. This is like a sequence which is very easy
to implement. Think of how music was stored on a CD or DVD. One after the other.
The primary advantage of this is the ease of implementation. In addition, there is the
ease of direct access due to each part of a file being stored in the same compact area
[1].

The biggest drawback is the lack of flexibility. A file cannot expand as it is not possible
unless there is a space next to the file. If the file needs to expand, then it needs to be
moved to a larger section every time it is updated [1]. A second drawback is a
fragmentation. Fragmentation occurs because space is reserved for files and this
space may be unused especially if the file is smaller. This space needs to be
compacted to minimize wasting space. During fragmentation, files are inaccessible.

Non-Contiguous Storage

This is like random file storage. Files are stored within any available space within the
disk. Any additional sections to a file are stored within other free sections of the hard
disk. These additional sections are known as extents [1]. Linking of the extents
happens by having a directory entry that consists of the file name, the storage location
of the first and last extent and the total number of extents [1]. The system can locate
the middle extents by following the pointers of the first extent [1]. Each extent then
holds a pointer to the next extent. A different method involves all the file extents at the
directory level. The drawback to this method is that it does not allow direct access.
This method eliminates the need for defragmentation.
Indexed Storage
This is a method that allows for direct record access by bringing together all the links
of the extent into an index block [1]. Each file has its index block that consists of the
addresses of each disk sector that make up the file. When a file has been created the
pointers to each index block are set to null [1]. As each sector is filled the pointer is
set to the appropriate address. Indexed storage supports both sequential and direct
access. It does not, however, improve storage space use due to the index block.

SUMMARY

During this lecture, we focused on file management. More specifically, we have

reviewed the role of the file manager, the difference between files and file systems,
and the different types of physical storage.

DISCUSSION TOPIC

Windows 11 is one of the newer operating systems by Microsoft. Find out the type of
physical storage that Windows 11 supports. What are the drawbacks and benefits of
the same?

REFERENCES

[1] McHoes, A., & Flynn, I., Understanding Operating Systems. Boston: Cengage
Learning, 2018

[2] Stallings, W., Operating Systems: Internals and Design Principles. Harlow: Pearson
Education Limited, 2018.