Unit 2.3
Software
Operating systems vary widely in their capabilities, from relatively simple single-user microcomputer systems, to sophisticated mainframe machines.
The modes of operation are:
Only one user can use the operating system at a time. Windows 9x, ME and XP allow different users to have different profiles on a single computer, but only one of those users can use the computer at one time. The difference between this and multi-user systems is that multi-user systems allow several users to access one computer at the same time. Windows NT and 2000 are Network OS, and are often run on servers which require simultaneous access from several users.
This allows several people to be running programs simultaneously from different workstations. This is also called a time sharing system, with each user in turn being allocated a time-slice.
Could be, for example, a single user standalone machine running an operating system, e.g. Windows 98, which allows several programs to be run simultaneously. When the term is applied to mini or mainframe computers, it implies the concurrent execution of two or more related tasks between which communication is possible. In other words, multitasking is taking place when a single job is broken down into stages called tasks that can be simultaneously executed often this will happen where a computer has multiple processors, so that each processor can run part of a program.
This means that two or more programs are being run at the same time. The operating system has to allocate resources to each program, allowing each one a small amount of processor time according to it's priority before moving on to the next one. The amount of processor time allocated to each job is not fixed, but is constantly updated as new jobs with differing priorities are added to the queue. This requires an executive capable of allocating resources dynamically to each job. The priority given to jobs will vary according to the following factors:
High Priority will be given to jobs requiring a large amount of I/O and a small amount of processor time.
Low priority is given to jobs which require a large amount of processor time, giving these jobs high priority would result in their hogging the processor and preventing I/O jobs from receiving what little attention they require.
This is where there are few facilities for multi-user, interactive computing, but nevertheless several batch jobs may be processed simultaneously in order to optimise the computer's resources. There is no interaction between the computer and the user during batch processing. One or more programs are submitted to the computer as a job and put in a batch queue until it is their turn to be run. An Executive assigns the jobs to the queue, then schedules them according to the priorities they have been assigned, and executes them, one after another. Some jobs will be required more urgently than others and will be given a higher priority. Jobs are loaded and executed according to their priority, not their position in the queue, so additional jobs can be added to the queue at any time.
A Job Control Language is used by the operating system to instruct the
processor which program to run and when. Batch processing is often used at night for routine back-ups or payroll operations that use serial or sequential access and require no input from the user.
Real-time processing systems must respond to the user within an acceptable time. The time limit will depend on the application e.g. an airline pilot will need very fast responses from his flight control OS. Real-time systems can be of different types:
Real-time Transaction processing is the only one of these which is required in the specification. A transaction is where some sort of exchange takes place. A common example of transaction processing is an ATM machine, where the user enters his card, PIN and amount required before the ATM retrieves his balance information from the bank database and dispenses the cash.