Kepiye CPU scheduling mlaku, lan apa wae algoritma sing umum?

Question

IT Interview Practice · Accepted Answer

**CPU scheduling** iku keputusan kernel babagan thread siap *sing endi* sing entuk CPU *sabanjure*, lan *suwene sepira*. Amarga biasane thread sing bisa mlaku luwih akeh tinimbang core, **scheduler** multiplex dheweke, kanthi tujuan nyeimbangake target sing sulaya — **throughput**, **latency / wektu respon sing endhek**, lan **adil** — karo ngindhari **starvation**.

## Algoritma umum

```text
FCFS (first-come, first-served) → simple, but one long job blocks everyone (convoy effect)
SJF  (shortest job first)       → optimal avg wait, but needs to know job length; can starve long jobs
Round-robin (RR)                → each job a fixed TIME SLICE (quantum), cycle through → fair, good latency
Priority                        → highest priority first → can STARVE low priority (fix: aging)
MLFQ (multi-level feedback)     → several priority queues; jobs that use a full quantum sink,
                                  interactive jobs that yield stay high → approximates SJF without knowing lengths
```

**Round-robin** nggambarake tombol inti — **quantum**:

```text
quantum too SMALL → fair & responsive, but lots of context-switch overhead
quantum too LARGE → less overhead, but degrades toward FCFS (poor responsiveness)
```

## Scheduler nyata

Standar Linux sing wis suwe yaiku **CFS (Completely Fair Scheduler)**: tinimbang irisan tetep, dheweke nglacak **virtual runtime (vruntime)** saben task lan tansah mlaku thread sing nampa CPU paling sithik nganti saiki, ditimbang miturut nilai **nice**-ne — nyedhaki "saben wong entuk bagean sing adil." Dheweke nganggo red-black tree kanthi kunci vruntime, mula milih task sabanjure iku O(log n). (Kernel sing luwih anyar nganggo EEVDF, panyempurnaan kanthi target adil sing padha ditambah wates latency sing luwih ketat.)

**Preemptive vs cooperative:** OS modern iku **preemptive** — timer interrupt ngidini scheduler ngrebut CPU kanthi peksa, saengga siji thread ora bisa mborong core.

## Kenapa iki penting

Scheduling langsung mbentuk latency lan throughput sing dirasakake pangganggo. Pewawancara ndhedhes bab iki kanggo ndeleng apa kowe ngerti trade-off — kenapa beban kerja interaktif pengin quantum cendhak utawa boost prioritas, kenapa priority scheduling butuh **aging** kanggo ngindhari starvation, lan kepiye MLFQ/CFS entuk prilaku interaktif sing apik tanpa ngerti dawane job luwih dhisik. Iki uga model pikir ing mburi nyetel `nice`, prioritas real-time, lan ndiagnosa service sing keluwen utawa lag.

Algoritma umum

Scheduler nyata

Kenapa iki penting

Pitakon sing gegandhengan