Operating systems — Multilevel feedback queues: In CPU scheduling, what is the main characteristic of a multilevel feedback queue (MLFQ) scheduler, often informally called a “feedback queue”?

Introduction / Context:
Multilevel feedback queue (MLFQ) scheduling is a classic operating-system technique that adapts CPU allocation based on how processes behave. It is widely taught because it illustrates responsive time-sharing and priority aging without manual tuning once configured.

Given Data / Assumptions:

• The question refers to “feedback queue,” i.e., multilevel feedback queues (MLFQ).
• MLFQ typically uses several ready queues with different priorities and time quanta.
• Processes may move between queues based on observed execution behavior (CPU burst length, I/O wait).

Concept / Approach:

MLFQ adapts priorities dynamically. Interactive or I/O-bound tasks that frequently yield the CPU are rewarded with higher priority and shorter quanta to improve responsiveness. CPU-bound tasks that use their full quantum are gradually demoted to lower-priority queues, receiving longer quanta but less frequent dispatch. Thus, the scheduler “dispatches according to execution characteristics.”

Step-by-Step Solution:

Verification / Alternative check:

Standard OS texts describe MLFQ rules: if a job uses up its quantum, move it to a lower-priority queue; if it yields early or waits for I/O, move it up. This is precisely dispatching by execution characteristics.

Why Other Options Are Wrong:

• Very easy to implement: MLFQ is more complex than single-queue algorithms.
• Used to favor real-time tasks: real-time needs dedicated policies (e.g., rate monotonic, EDF).
• Require manual intervention: MLFQ adapts automatically after initial configuration.

Common Pitfalls:

Confusing MLFQ with fixed multilevel queues (no feedback). Assuming it is primarily a real-time scheduler rather than a general-purpose, interactive one.

Final Answer:

dispatch tasks according to execution characteristics