Operating Systems: Evolution by Generation

• Operating systems have evolved through generations, each marked by advancements in hardware and software capabilities
• The earliest computers lacked operating systems, requiring direct interaction with hardware

OS Evolution: By Generation

• First Generation (1945-1955): Used vacuum tubes, lacked an OS, and relied on serial processing
• Second Generation (1955-1965): Transitioned to transistor-based computers, introducing batch systems
• Third Generation (1965-1980): Integrated circuits enabled multiprogramming, general-purpose computing, spooling, time-sharing systems, minicomputers, and UNIX
• Fourth Generation (1980-Present): Characterized by LSI and VLSI circuits, GUIs, distributed OS, multiprocessor OS, and real-time systems

Serial Processing

• Early computers operated without an OS, demanding direct hardware interaction
• Programmers used consoles with lights, switches, input devices, and printers
• Drawbacks included no scheduling and significant setup time

Types of Operating Systems

• Single-user
• Multiuser
• Batch processing
• Multiprogramming
• Multitasking
• Multiprocessing
• Real-time
• Parallel
• Distributed

Single-User Operating Systems

• Designed for one user to access the computer system at a time
• Utilize a single processor to execute one program at any given time
• Dedicate all resources like CPU and I/O devices to a single user

Types of Single-User Operating Systems

• Single-user, single-tasking: Allows one user to execute only one program at a time (e.g., MS-DOS, Palm OS)
• Single-user, multitasking: Allows one user to execute multiple programs concurrently (e.g., Windows, MAC-OS)

Advantages of Single-User Operating Systems

• Support for one user at a time
• Easy maintenance
• Lower risk of damage
• Less resource requirements

Disadvantages of Single-User Operating Systems

• High idle time
• Poor resource utilization

Multiuser Operating Systems

• Enable multiple users to access the computer system simultaneously
• Allow diverse users to leverage the computer's resources concurrently
• Facilitate access through a network of personal computers connected to a mainframe

Multiuser Operating Systems: Functionality

• Personal computers can exchange information with the mainframe
• Mainframes function as servers, while personal computers act as clients
• The OS manages user requests, preventing interference

Enhanced User Management Features

• Multiuser systems offer user authentication
• They include policies for managing user accounts
• They provide access protection for user privacy and data integrity
• Modern, general-purpose OSs are typically multiuser (e.g., Unix, Windows NT/2000/2003 Server)
• CPU switches between jobs to provide immediate response, enabled by scheduling and multiprogramming
• Time-sharing systems give each user a small time slice, creating the illusion of exclusive CPU use

Advantages of Multiuser Operating Systems

• Facilitate data and information sharing among users
• Enable sharing of hardware resources like printers and modems

Disadvantages of Multiuser Operating Systems

• Require expensive mainframe hardware
• Can reduce system performance due to multiple users
• Are complex

Single vs. Multiuser OS Comparison

• Single-user OS manages the computer for one user doing one task at a time, while multiuser OS allows many users to use computer resources simultaneously

Single vs. Multiuser OS: User Support

• Single-user platforms support one user at a time
• Multiuser systems regulate access for multiple users through databases

Single vs. Multiuser OS: Resource Utilization

• Single-user OS has poor resource utilization
• Multiuser OS has better utilization

Single vs. Multiuser OS: Usage

• Single-user OS is common on desktops and laptops
• Multiuser OS is common on mainframes

Single vs. Multiuser OS: Attributes

• Single-user OS is less complex and easier to maintain
• Multiuser OS is more complex and harder to maintain

Single vs. Multiuser OS: Cost

• Single-user OS has lower costs
• Multiuser OS requires expensive hardware

Single vs. Multiuser OS: Data Sharing

• Single-user OS does not allow data, information, or resource sharing
• Multiuser OS allows the sharing of data, information, and resources

Single vs. Multiuser OS: Types

• Single-user OS includes single-task and multi-task types
• Multiuser OS includes time-sharing systems and some mainframe batch processing systems

Simple Batch Systems

• Represent one of the earliest OS types
• Users do not interact directly; they submit jobs offline (e.g., via punch cards) to a computer operator

Batch Processing: Process

• Jobs with similar needs are grouped into batches for efficient processing
• Programmers submit programs to the operator, who sorts them into batches
• Batches are executed sequentially, and the operator returns the output to users

Simple Batch Systems: Automation

• Involves hiring an operator
• It used card readers
• It reduces setup time via batching
• It spools various user jobs in a queue
• This improves processing by automatic job sequencing

Simple Batch Systems: Resident Monitor

• There is a resident monitor, which is the part of main memory where the OS resides
• The resident monitor initializes batches, manages processing, terminates batches, and transfers control between programs

Simple Batch Systems: Memory Layout

• Only one program runs at a time

Functions Within Batch Systems

• Process scheduling: FCFS (First-Come, First-Served)
• Memory management: Divided into Resident Monitor and User Program areas
• File management: Serial access without access or concurrency control

Advantages of Batch Systems

• Increased performance
• Automated job execution
• Computer resource sharing among many users

Disadvantages of Batch Systems

• Lack of user interaction
• Slow performance because I/O and CPU operations do not overlap
• Poor CPU utilization because the processor is idle
• It is difficult to provide priority
• It is difficult to debug
• There is poor memory protection
• There is potential for jobs to enter infinite loops

Multiprogramming Operating Systems:

• Enables users to run multiple programs concurrently using a single CPU
• Several jobs reside in main memory, with the CPU rapidly switching between them

Multiprogramming Operating Systems: Operations

• The OS selects and starts a job from memory
• When a job requires I/O, the OS switches to another job
• There are fewer jobs in memory than on disk (job pool)
• CPU scheduling selects jobs to run

OS Features Needed for Multiprogramming

• Memory management allocates memory to jobs
• CPU scheduling selects which job to run next
• Allocation of devices
• Proper Job Mix: A balance of I/O-bound and CPU-bound processes is needed
• Job Status Preservation

Advantages of Multiprogramming

• Increased CPU and memory utilization
• Increased throughput of programs
• Ability to assign priorities to jobs

Disadvantages of Multiprogramming

• Greater OS overhead
• Sophistication and complexity
• Memory management for varied job sizes

Multitasking Operating Systems

• Enables a computer to perform multiple tasks concurrently using time sharing with a single CPU
• Multi-programming is used for multi-user systems, multitasking for single-user systems
• Only one program is executed at any moment, with the processor rapidly switching between tasks
• Multitasking offers reduced execution time

Multitasking Operating Systems: Processes

• Divides processes into foreground (accepting input) and background (processing data or in a waiting state)
• Examples are windows, Linux, mac, Unix, etc.

Types of Multitasking

• Pre-emptive: OS allocates CPU time slices
• Co-operative: A program holds the CPU as long as needed, ceding it for other programs

Multiprogramming vs. Multitasking

• Multiprogramming is for multi-user systems executing multiple users' jobs
• Multitasking is for single-user systems executing multiple jobs

Multiprogramming vs. Multitasking: Response Time

• Multiprogramming has a longer response time
• Multitasking has a shorter response time

Multiprocessing Operating Systems

• Systems with two or more CPUs to execute multiple programs simultaneously
• Multiple processors share clock, bus, memory, and peripherals

Multiprocessing Operating Systems: Functionality

• Processors either work on different parts of the same task or on different tasks
• Used to process vast data at high speeds
• Examples include satellite control and weather forecasting systems

Multiprocessing Operating Systems: Examples

• Examples include Linux, Unix, Windows 2000, Windows XP, Mac OS X

Multiprocessing Types

• Symmetric Multiprocessing (shared OS)
• Asymmetric Multiprocessing (master-slave)

Symmetric Multiprocessing (SMP)

• Processors share the same memory and run identical copies of the OS

Asymmetric Multiprocessing (AMP)

• Different processors handle different tasks
• A master processor controls the system's data structure
• It also schedules work for slave processors

Symmetric vs. Asymmetric Multiprocessing: Basic

• In SMP, each processor runs OS tasks
• In AMP, only the master processor runs OS tasks

Symmetric vs. Asymmetric Multiprocessing: Process

• SMP processors take processes from a common queue
• AMP master processor assigns processes

Symmetric vs. Asymmetric Multiprocessing: Architecture

• SMP processors have the same architecture
• AMP processors may have different architectures

Symmetric vs. Asymmetric Multiprocessing: Communication

• SMP processors communicate via shared memory
• AMP processors do not require communication

Symmetric vs. Asymmetric Multiprocessing: Failure

• In SMP, system computing capacity is reduced
• In AMP, a slave becomes the master

Symmetric vs. Asymmetric Multiprocessing: Ease

• SMP is complex due to processor synchronization needs
• AMP is simple as the master processor accesses data structures

Advantages of Multiprocessing OS

• Increased throughput
• Cost savings
• Increased reliability

Disadvantages of Multiprocessing OS

• Expensive
• Complex
• Large main memory requirement

Multiprocessing vs. Multiprogramming

• Multiprocessing processes multiple processes simultaneously with multiple CPUs
• Multiprogramming executes programs concurrently using a single CPU

Multiprocessing vs. Multiprogramming: CPU Usage

• Multiprocessing uses multiple CPUs
• Multiprogramming uses a single CPU

Multiprocessing vs. Multiprogramming: Parallel Processing

• Multiprocessing permits parallel processing
• Multiprogramming involves context switching

Multiprocessing vs. Multiprogramming: Time

• Multiprocessing takes less time to process jobs
• Multiprogramming takes more time

Multiprocessing vs. Multiprogramming: Efficiency

• Multiprocessing facilitates efficient device utilization
• Multiprogramming is less efficient

Real-Time Operating Systems (RTOS)

• Multitasking OS designed for real-time applications
• Guarantees result correctness and meeting deadlines

Characteristics of Real-Time OS

• The time taken by the system to respond to an input and display of required updated information is termed as the response time.
• In RTOS, the response time is very low
• Examples include embedded systems, robots, and engine controllers

Real-Time OS: Usage

• Controls instruments, machinery, industrial systems, and scientific instruments
• Resources are managed to ensure operations execute in a consistent time frame

Types of Real-Time OS

• Hard real-time OS: Guarantees critical task completion within a set time
• Soft real-time OS: Prioritizes critical tasks but allows flexibility

Hard Real-Time OS

• Completes tasks in guaranteed time
• May suspend less critical commands
• Has limited or no secondary storage
• If there is a delay then it fails
• Examples: Flight and satellite control systems

Soft Real-Time OS

• Prioritizes tasks into critical and non-critical tasks
• Allows the critical tasks enough time to be performed
• If a process completes correctly but takes longer than its given amount of time, the results may still useful
• Examples: Multimedia, Virtual Reality, Advanced Scientific Projects

RTOS Functions

• Process scheduling: Priority-based preemptive scheduling
• Memory management: Less demanding, with high-priority processes in memory
• I/O management: Includes interrupt management and I/O buffering
• File management: Fast access is prioritized

Parallel Systems

• Multiprocessor systems with multiple CPUs in close communication
• Processors share memory and a clock
• Communication occurs through shared memory

Advantages of Parallel Systems

• Increased throughput
• Economical
• Increased reliability through graceful degradation

Parallel Operating Systems: Functionality

• Interface multiple networked computers
• Manage resources such as memory, caches, storage, and processing power
• Divide calculations into smaller parts

Parallel Operating Systems: Inter-Core Communication

• Facilitates communication between processing cores
• Uses memory-sharing or memory distribution

Parallel Systems: Classification

• On the basis of Processor
  • Symmetric multiprocessing (SMP)
  • Asymmetric multiprocessing
• On the basis of instruction and Data (Flynn’s Classification)
  • Single Instruction Stream, Single Data Stream (SISD)
  • Single Instruction Stream, Multiple Data Stream (SIMD)
  • Multiple Instruction Stream, Single Data Stream (MISD)
  • Multiple Instruction Stream, Multiple Data Stream (MIMD)

Distributed Operating Systems

• Manages independent computers to appear as a single system
• Uses multiple processors to serve applications and users

Distributed Operating Systems: Processors

• Processors communicate through communication lines
• Processors vary in size and function

Advantages of Distributed OS

• Resource sharing
• Communications improvements
• Load sharing
• Reliability
• Improved customer service

Disadvantages of Distributed OS

• Limited software
• Slow networking
• Security risks
• Privacy concerns

Types of Distributed OS

• Client-server system
• Peer-to-peer system

Client-Server System

• Example: Railway reservation systems

Peer-to-Peer System

• Each computer acts as both client and server
• Computers can grant and request resources
• There is no central authority
• There is also no single point of failure

Time-Sharing Systems vs Multiprogrammed OS

• Commands in time-sharing systems are entered at the terminal
• Job control language commands are provided with the job in multi-programmed systems
• The principal objective for time-sharing is to minimize response time
• The principal objective for multi-programmed is to Maximize processor use and is batch

Parallel System v/s Distributed Systems

• In parallel processing the subdivided parts of a job which can be done at the same time will be carried out by different processors
• In distributed system various functions and different computational requirements are distributed among different processors located at various places.
• The aim of parallel processing is to accept a given job, break it down into subparts, divide it among all the available processors and complete it shortest possible time.
• Distributed computing involves multiple computer systems working together as a single system.
• The parallel systems are called tightly coupled system
• The distributed systems are called loosely coupled system
• All systems run a common operating system in parallel systems
• The operating system running on each processor may be different in distributed systems
• Any processor can execute any job and hence all processors are similar in parallel systems
• The processes can be executing on dissimilar processors in distributed systems
• Advantages of parallel systems are:
  • (a) Better performance
  • (b) Fault tolerance
  • (c) Incremental Growth
  • (d) Cost Performance
• Advantages of distributed systems
  • (a) Sharing of Resources
  • (b) Reliability
  • (c) Communication
  • (d) Computational speedup
• Parallel system have more than one processor in close communication, sharing the computer bus, clock, memory and peripheral devices
• Distributed systems do not share bus or clock. The processors communicate with one another through various communication lines such as high speed bus or telephone lines.
• The Application area of Parallel system are:
  • Medical community : for analyzing MRI images and study bone implant systems.
  • Airlines : for processing customer information to forecast demand and decide fare.
• The Application areas of Distributed system are:
  • Banking Services : facilitate query processing for local customer of a branch and also for global customer.

Spooling

• Spooling refers to putting jobs in a buffer(an area in memory or disk). This buffer is a special area in memory or hard disk which is accessible to I/O devices.
• Spooling helps when devices access data at different rate because buffer provides an area where data can rest until a slow device is ready to receive that data.
• Spooling is an acronym for (Simultaneous Peripheral Operations On Line).

Spooling and OS Functions

• An operating system does the following activities related to distributed environment −
  • Handles I/O device data spooling as devices have different data access rates.
  • Maintains the spooling buffer which provides a waiting station where data can rest while the slower device catches up.
  • Maintains parallel computation because of spooling process as a computer can perform I/O in parallel fashion. It becomes possible to have the computer read data from a tape, write data to disk and to write out to a tape printer while it is doing its computing task.