You are a Hadoop administrator for a healthcare company that owns a Hadoop cluster where the company runs different data processes. You want to allocate resources to each of the processes.

The healthcare company might have two data processes that run simultaneously. One process might handle inventories while another process might handle prescriptions. The following table describes how you might divide the data processes into queues:

Queue Name	Weight	Description
Inventory	0.33	Queue for jobs that process inventory data from hospital pharmacies.
Prescriptions	0.67	Queue for jobs that process data on the prescriptions that doctors write.

At a point in time, the queue

Inventory

might be running 2 jobs, while the queue

Prescriptions

might be running 3 jobs. The following image shows the proportion of cluster resources that each job can use:

Note that the resources allocated to the queue

Inventory

are divided in half. Each halve is allocated to one of the two jobs in the queue. The resources allocated to the queue

Prescriptions

are divided into three parts. Each part is allocated to one of the three jobs in the queue.

The first job in the queue

Inventory

completes. The following image shows the proportion of cluster resources that are available to the remaining jobs: