Transcript Types of OS
Operating System Classification Single User, Multi-User, Simple Batch Processing, Multiprogramming, Multi tasking, Real Time Systems, Parallel Systems, Distributed systems Single User • This OS is designed to manage the computer so that one user can effectively do one thing at a time. • It is aimed at maximum user convenience and responsiveness • Provides good interface to a single user • Example: PCs running MS Windows, and apple Macintosh, MS DOS, OS/2, Linux Multi-User • Allows many different users to take advantage of computer’s resources simultaneously • It is used as a single server and multiple number of clients • The time shared system allows multiple users to share the system simultaneously • The system switches from one user to another rapidly, giving the impression that each user has full control of the system Multi-User cont… • Store the information about the number of user’s connected to the server and provides security to data from unauthorized access • Each user has a private set of programs and access techniques for accessing the data. • Example : Windows 2000, Novell Netware, Windows NT, Unix, VMS (Virtual Memory System), MVS (Multiple virtual Storage- for mainframe systems) Simple Batch Processing • Batch is a sequence of user jobs formed for the purpose of processing by a Bath Processing OS • Batch is not the computational unit, Each job in a batch is independent of other jobs and may belong to different users. • A job consists of program, data and some control information about the program. Simple Batch Processing cont.. • The primary function of the batch processing system is to service the jobs in a batch one after another without requiring the operator’s intervention. • Achieved by automating the transition from execution of one job to that of the next job in the batch Simple Batch Processing cont.. • Batch processing is implemented by the kernel or batch monitor which resides in one part of computer’s memory • The remaining memory is used for servicing a user job- the current job in the batch • When the operator gives the command to initiate the processing of a batch, monitor sets up the processing of first job in the batch Simple Batch Processing cont.. • At the end of the job, it performs job termination processing and initiates the next job in the batch • At the end of the batch , it terminates the batch termination processing and waits for the next batch initiation by the operator (only at start and end of a batch) System area User area Batch Monitor Current Job of the Batch Start Job1 Job2 Job n End Simple Batch Processing cont.. • Uses Notion of Virtual Devices to conserve the CPU time • Uses virtual devices like tapes/disks instead of punched cards/printers • A program first records a batch of jobs on a magnetic tape • Batch system processes the jobs and save the results on another magnetic tape. • Contents of this magnetic tape are then printed by another program • These two spooling operations called inspooling and outspooling were performed on smaller systems with slow CPU- resulted in lesser CPU idle times Disadvantages • CPU idle during I/O • I/O devices idle when CPU busy Multiprogramming Systems Definition • The OS can put many programs in the memory and let the CPU execute instructions of one program while the I/O subsystem is busy for I/O operation for another program MultiProgramming Kernel MultiProgramming Kernel MultiProgramming Kernel Program 1 Program 1 Program 1 I/O CPU I/O Program 2 Program 3 I/O CPU CPU Program 2 Program 3 I/O Program 2 Program 3 Cont… • Multiprogramming kernel performs – Scheduling (simple scheduling policy and performs simple partitioned or pool based allocation of memory and I/O devices) – Memory Management – I/O Management Cont.. • The CPU and the I/O subsystem could operate on same program • Thus the program must explicitly synchronize the activities of the CPU and the I/O subsystem to ensure that the program executes correctly • Done by allocating CPU to program only when the program is not performing I/O Architectural support for Multiprogramming • DMA(Direct Memory Access) • Memory Protection • Privileged mode of CPU DMA(Direct Memory Access) • Makes Multiprogramming feasible by permitting concurrent operation of CPU and I/O devices – In this mode a block of data can be transferred between memory and I/O device without involving the CPU – An I/O instruction indicates the I/O operations to be performed an also number of bytes to be transferred – I/O operation starts when the instruction is executed – Data transfer between the device and memory takes place over the system bus – CPU is not involved in this transfer , an interrupt is raised at the end of transfer of all bytes Memory Protection • Prevents mutual interference between programs – Ensures that the programs does not access or destroy contents of memory areas occupied by other programs or the OS Privileged mode of CPU • Provides a method of implementing memory protection and other measures that avoid interference between programs • A program interrupt is raised if the program tries to execute privileged instruction when CPU is in user mode User service • The turnaround time of a job is affected by the amount of CPU attention devoted to other jobs executing concurrently with it • It depends upon the number of jobs in the system and priorities assigned to different jobs by the scheduler Functions of Kernel-Scheduling • Scheduling is performed after every interrupt using simple priority based preemptive scheduling scheme • Priority is a tie breaking notion used in a scheduler to decide which request should be scheduled on the server when many requests await service • Preemption is the forced deallocation of the CPU from a program (from a low priority program to high priority program) Functions of Kernel-Memory Management • To protect two programs from interference , memory protection hardware is used and CPU is put in the Non-Privileged mode while executing user programs. • An effort by a user program to access memory locations situated outside its memory area, or use a privileged instruction leads to an interrupt • Interrupt processing routines for these interrupts terminate the program that caused and interrupt Multitasking Systems or Time Sharing systems Definition Logical extension of multiprogramming, where CPU executes multiple jobs by switching between them so frequently that users can interact with each program while it is running • for immediate results – Require low response time • Allow many users to share the computer simultaneously, users have the impression that they have their own machine • CPU is multiplexed among several jobs that are kept in memory and on disk User Service • Characterized in terms of time taken to service a sub-request i.e the response time • E.g – A typical request from a user to compile a statement or execution of program on given data, the response consists of message from compiler or results computed. User issues next request receiving the response. Good response time would lead to productivity of the user Scheduling • To give good response time to all users, they must have equal opportunity to present their computational requests and have them serviced. • Two provisions to ensure this– No priorities assigned to programs, but are executed by turn (Round robin scheduling) – Program is prevented form consuming unreasonable amount of CPU time (Time Slicing) Preempted program Time Slice over Scheduler Scheduling List Selected Program CPU Computation over Round Robin Scheduling • When a user makes a computational unit to his program, the program is added to the end of a Scheduling List • A small unit of time, called timeslice or quantum, is defined which is the largest amount of CPU time any program can consumed when scheduled to execute. • The CPU scheduler goes around this queue, allocating the CPU to each process for a time interval of one quantum. • The CPU scheduler picks the first process from the scheduling list, sets a timer to interrupt after one quantum, and dispatches the process. • If the process is still running at the end of the quantum, the CPU is preempted and the process is added to the tail of the scheduling list. • If the process finishes before the end of the quantum, the process itself releases the CPU voluntarily. • In either case, the CPU scheduler assigns the CPU to the next process in the ready queue. Memory Management • Swapping is the technique of temporarily removing inactive programs from the memory of a computer system • Three kinds of programs – Active Programs – Programs being swapped out of the memory – Programs being swapped into the memory • When an active program becomes inactive, the OS swaps it out by copying its instructions and data onto a disk. A new program is loaded in its place • Use of swapping is feasible in time sharing systems because the time sharing kernel can estimate when the program is likely to be scheduled next. Real Time Operating Systems Definition A real time application is a program that responds to activities in an external system within a maximum time determined by the external system Applications- missile guidance, command and control applications like process control and air traffic control, data sampling and applications like railway reservation and banking system Hard & Real Time Systems • Hard real time system – dedicated to processing real time applications and provably meets the response requirements of an application under all conditions. • Soft real time system – makes best effort to meet the response real time application but cannot guarantee that it will be able to meet it under all conditions Features • Permits creation of multiple processes within an application • Permits priorities to be assigned to processes. • Permits a programmer to define interrupts and interrupt processing routines • Uses priority driven or deadline oriented scheduling • Provides fault tolerance and graceful degradation capabilities Distributed Operating System Definition A distributed OS exploits the multiplicity of resources and the presence of a network to provide the advantages of resource sharing across computers, reliability of operation,speed up of applications and communication between users Features • Resource sharing – improves resource utilization across boundaries of individual computer systems. • Reliability- Availability of resources and services despite failures. • Computation speed-up- parts of a computation can be executed in different computer systems to speed-up the computation • Communication-Provides means of communication between remote entities. • Incremental growth- capabilities of a system can be enhanced at a price proportional the nature and size of the enhancement Key concepts and techniques used • Distributed control- a function is performed through participation of several nodes, possibly all nodes, in a distributed system. • Transparency-A resource or service can be accessed without having to know its location in the distributed system. • Remote Procedure Call (RPC)-A process calls a procedure that is located in a different computer system. The procedure call is analogous to a procedure or function call in programming language, except that is the OS that passes parameters to the remote procedure and returns its results. Operation of the process making the call is resumed when results are returned to it. Parallel Systems/ Multiprocessor System Definition Systems that have two or more processors in close communication, sharing the computer bus and sometimes the clock, memory and peripheral devices Types of Multiprocessor systems • Asymmetric multiprocessing – each processor is assigned a specific task. A master processor controls the system, the other processors either look to the master for instruction or have predefined task. Defines a master-slave relationship. The master processor schedules and allocates work to the slave processors. SunOS version 4 • Symmetric Multiprocessing- Each processor performs all tasks within the operating system.. No master-slave relationship. SunOS Version 5 Advantages • Increased throughput- By increasing the number of processors, get more work done in less time. A certain amount of overhead incurred in keeping all the parts working correctly and contention for shared resources lowers the expected gain. • Economy of scale -Cost less than equivalent multiple single processor systems as they share peripherals, mass storage and power supplies. • Increased reliability – If the functions can be distributed properly among several processor, then the failure of one processor will not affect the system, only slows it down (graceful degradation and fault tolerance)