Module Review: File Systems

[!NOTE] This module explores the core principles of Module Review: File Systems, deriving solutions from first principles and hardware constraints to build world-class, production-ready expertise.

1. Key Takeaways

  1. Everything is a File: The Unix philosophy unifies access to documents, devices, and sockets via the same open/read/write/close API.
  2. Inodes are Identity: A file is defined by its Inode (Metadata), not its name. Hard links are multiple names pointing to the same Inode.
  3. Journaling Saves Time: Write-Ahead Logging (WAL) prevents file system corruption by logging intent before action, making recovery nearly instantaneous (vs FSCK).
  4. VFS is Polymorphism: The Virtual File System layer allows the Kernel to support multiple file systems (Ext4, NTFS, NFS) through a common interface (file_operations).
  5. Network State Matters: NFSv3 is stateless (robust but limited), while NFSv4 and SMB are stateful (support locking/caching but complex recovery).

2. Interactive Flashcards

Test your retention of the core concepts.

What is an Inode?

(Click to reveal)

Index Node

A data structure storing file metadata (Size, Permissions, Block Pointers). It does NOT store the filename.

Hard Link vs Soft Link?

The Difference

Hard Link: Another name for the *same* Inode.

Soft Link: A file containing the *path* to another file.

Purpose of VFS?

Abstraction

Allows the OS to use different file systems (Ext4, FAT, NFS) using a single, unified system call API.

What is WAL?

Write-Ahead Logging

A technique where modifications are written to a log (Journal) *before* being applied to the main data structures. Ensures crash consistency.

Stateless vs Stateful NFS?

The Trade-off

Stateless (v3): Simple recovery, no locking.

Stateful (v4): Supports locking/caching, complex recovery.

3. Cheat Sheet

Concept Description
File Descriptor (FD) Integer representing an open file session in a process.
Inode On-disk structure holding file metadata (Size, Permissions, Pointers).
Dentry In-memory structure mapping a Filename to an Inode.
Superblock Metadata about the entire File System (Block size, Total blocks).
Journaling Writing intent to a log to ensure crash consistency.
Extent A range of contiguous blocks [Start, Length]. Efficient for large files.
VFS Kernel abstraction layer for File Systems.
Mount Point Directory where a new File System is attached to the tree.

4. Next Steps

You have mastered the persistent storage layer. Now, let’s look at how the OS talks to the outside world.