Performance Metrics or CPU Scheduling Criteria

 

Scheduling Criteria or Scheduling Methodology:

So many CPU scheduling algorithms are available. Different CPU scheduling algorithms have different properties. In choosing which algorithm to use in a particular situation, we must consider the properties of the various algorithms. The performance metrics are used for the same. We can determine which algorithm is the best algorithm and which one is the worst using the performance metrics.  These performance metrics are as follows:

 

1. Throughput:

 

Throughput means how many jobs are completed by the CPU within a time period. If the bigger process, this rate may be a process per hour; for short process throughput might be process per minute.

 

 2. Turnaround time:

The time interval between the submission of the process and time of the completion is called the turnaround time. Turnaround time is sum of the periods spent waiting to get into memory, waiting in ready queue, executing on the CPU and doing IO.

 

Turnaround time = waiting time in ready queue + executing time + waiting time in        waiting queue for IO

 

 3. Waiting time:

 Waiting time is the sum of the periods spent waiting by a process in the ready queue. Generally, a CPU scheduling algorithm having the least average waiting time is said to be the best algorithm. It may be expressed as turnaround time less the actual execution time.

 

4. Response time:

 Response time is the time duration between the submission and first response. For ex: a process entered in the ready queue at 10:05 a.m., but the process got the first response from the CPU is 10:10 a.m. Then the interval between 10:05 and 10:10 is said to be the response time.

5. CPU utilization:

 This is the percentage of time that the processor is busy. CPU utilization may range from 0 - 100%. In single user systems and real time systems this criteria is less important than the time sharing systems. The load on the system affects the level of utilization that can be achieved. On large and expensive systems, it may be the primary criteria.

 

6. Priority:

 Give superior treatment to processes having higher priorities.

 

7. Balanced utilization:

 Apart from CPU utilization, the utilization of other resources like memory, IO devices are also considered.

 

8. Fairness:

 Avoid the process from starvation. All the processes must be given equal opportunity to execute.