Interactive edition
Operating Systems: Three Easy Pieces
An interactive study companion to the book by Remzi H. Arpaci-Dusseau and Andrea C. Arpaci-Dusseau (Arpaci-Dusseau Books, Version 1.00, August 2018, ostep.org). Every chapter is rebuilt as labeled diagrams, step-through animations, live simulators, and quizzes — the same ideas, made visual.
New here? Start at the top and read straight through, or jump to any chapter below. Use ← / → to move between sections; the sidebar tracks what you've read.
Introduction
What an operating system is, and the three problems it exists to solve — a warm-up dialogue, the big picture, and a first look at the process, memory, and persistence abstractions.
Part I — Virtualization
Turning one CPU into many, and one physical memory into many private address spaces — scheduling policies, address translation, segmentation, paging, the TLB, and swapping.
- 3 A Dialogue on Virtualization 💬 pp. 23–24
- 4 The Abstraction: The Process pp. 25–36
- 5 Interlude: Process API pp. 37–48
- 6 Mechanism: Limited Direct Execution pp. 49–64
- 7 Scheduling: Introduction pp. 65–76
- 8 Scheduling: The Multi-Level Feedback Queue pp. 77–88
- 9 Scheduling: Proportional Share pp. 89–102
- 10 Multiprocessor Scheduling (Advanced) pp. 103–116
- 11 Summary Dialogue on CPU Virtualization 💬 pp. 117–118
- 12 A Dialogue on Memory Virtualization 💬 pp. 119–120
- 13 The Abstraction: Address Spaces pp. 121–130
- 14 Interlude: Memory API pp. 131–140
- 15 Mechanism: Address Translation pp. 141–154
- 16 Segmentation pp. 155–166
- 17 Free-Space Management pp. 167–184
- 18 Paging: Introduction pp. 185–198
- 19 Paging: Faster Translations (TLBs) pp. 199–214
- 20 Paging: Smaller Tables pp. 215–230
- 21 Beyond Physical Memory: Mechanisms pp. 231–242
- 22 Beyond Physical Memory: Policies pp. 243–260
- 23 Complete Virtual Memory Systems pp. 261–278
- 24 Summary Dialogue on Memory Virtualization 💬 pp. 279–284
Part II — Concurrency
Threads that share memory without corrupting it — locks built from hardware atomics, condition variables, semaphores, and the deadlocks and races that lurk between them.
- 25 A Dialogue on Concurrency 💬 pp. 285–286
- 26 Concurrency: An Introduction pp. 287–302
- 27 Interlude: Thread API pp. 303–314
- 28 Locks pp. 315–336
- 29 Lock-based Concurrent Data Structures pp. 337–350
- 30 Condition Variables pp. 351–366
- 31 Semaphores pp. 367–384
- 32 Common Concurrency Problems pp. 385–400
- 33 Event-based Concurrency (Advanced) pp. 401–412
- 34 Summary Dialogue on Concurrency 💬 pp. 413–416
Part III — Persistence
Making data survive crashes, failing disks, and lying devices — I/O and disks, RAID, file systems, journaling, log-structured and flash storage, data integrity, and distributed file systems.
- 35 A Dialogue on Persistence 💬 pp. 417–418
- 36 I/O Devices pp. 419–432
- 37 Hard Disk Drives pp. 433–448
- 38 Redundant Arrays of Inexpensive Disks (RAIDs) pp. 449–466
- 39 Interlude: Files and Directories pp. 467–492
- 40 File System Implementation pp. 493–510
- 41 Locality and The Fast File System pp. 511–524
- 42 Crash Consistency: FSCK and Journaling pp. 525–546
- 43 Log-structured File Systems pp. 547–562
- 44 Flash-based SSDs pp. 563–586
- 45 Data Integrity and Protection pp. 587–600
- 46 Summary Dialogue on Persistence 💬 pp. 601–602
- 47 A Dialogue on Distribution 💬 pp. 603–604
- 48 Distributed Systems pp. 605–620
- 49 Sun’s Network File System (NFS) pp. 621–638
- 50 The Andrew File System (AFS) pp. 639–652
- 51 Summary Dialogue on Distribution 💬 pp. 653–655