Why is it that none of the traditional scheduling algorithms can be considered fair scheduling algorithms

A fair scheduler organizes jobs by resource pools and divides resources fairly among these pools. By default, each user has a separate pool so that each user gets an equal share of cluster resources regardless of how many jobs they submit. It is also possible to set a job's resource pool by the user's Unix group or jobconf attribute. Within each pool, capacity is shared between running jobs using fair sharing. Users can also give resource pools appropriate weights to ****share clusters in a non-proportional manner.

In addition to providing a fair ****sharing method, the fair scheduler allows for a guaranteed minimum ****sharing resource to be assigned to a resource pool, which is useful for ensuring that a particular user, group, or production application always has access to enough resources. When a resource pool contains jobs, it gets at least its guaranteed minimum resources, but when a resource pool doesn't need all the guaranteed resources it has, the extra portion is sliced and diced among the other resource pools.

The main features are as follows:

Multi-user multi-queue support

Fair ****-enjoyment of resources (fair ****-enjoyment is determined by priority)

Guaranteed minimum ****-enjoyment

Time-slice preemption

Limit the amount of concurrency of jobs in order to prevent the intermediate data from filling up the disk

3. Analysis of fair scheduling algorithm

3.1 Variables Definitions