“Set of loosely connected computers that work together so that in many respects they can be viewed as a single system”.
Nodes refer to separate computers in the cluster. Each node can have several cores (processors)
Backend manages the distributed system.
The frontend, abstracts the system, so that one can interact with it as if it was a single computer.
Job manager (or scheduler): program running
on the cluster frontend. – Listens for incoming job submissions. – Schedules jobs in a more or less fair way.