The Linux Virtual Filesystem

Architecture, Implementation and Administration from Kernel Internals to Cloud-Native Systems

1. About This Book
2. Introduction: The Abstraction That Makes Linux Possible
3. What Is a Virtual Filesystem?
4. The Scale of the Problem
5. Why VFS Matters to You
6. How This Book Is Organized

Chapter 1: Historical Evolution and Design Philosophy

1. Pre-VFS Unix Filesystem Landscape
2. Birth of VFS in Linux 1.3.x
3. Design Principles: Abstraction, Extensibility, Minimalism
4. The VFS as a Kernel Subsystem
5. Lessons from Other Operating Systems

Chapter 2: Core VFS Objects: The Building Blocks

1. The Superblock: Filesystem-Wide State
2. The Inode: Metadata Without a Name
3. The Dentry: Naming and Caching
4. The File Object: Open File Description
5. The Mount Structure: Binding Filesystems to Directories
6. How the Objects Fit Together

Chapter 3: Pathname Resolution and Caching Mechanisms

1. The Lookup Algorithm: Walking a Path Component by Component
2. The Dentry Cache (dcache): Structure and Hash Table Design
3. Negative Dentries and Caching Nonexistence
4. Inode Caching and the Icache
5. Follow Symlinks, Bind Mounts and Special Cases
6. Performance Characteristics: When Path Resolution Becomes a Bottleneck

Chapter 4: System Calls and the VFS Interface

1. The Syscall Entry Path: From User Space to VFS Dispatch
2. File Operations Vector (struct file_operations)
3. Inode Operations Vector (struct inode_operations)
4. Superblock Operations Vector (struct super_operations)
5. Dentry Operations Vector (struct dentry_operations)
6. Key System Calls Walked Through

Chapter 5: Security Model and Permissions

1. Discretionary Access Control: Traditional Unix Permissions
2. The Permission Checking Path in VFS
3. Linux Security Modules (LSM) Hooks in VFS
4. Capabilities and Fine-Grained Privilege Management
5. Extended Attributes, ACLs and Security Labels
6. SELinux, AppArmor and Smack Integration with VFS

Chapter 6: Pseudo-Filesystems I: /proc, /sys, tmpfs and devtmpfs

1. What Makes a Filesystem “Pseudo”?
2. /proc: The Process Information Filesystem
3. /sys: The Sysfs Hierarchy
4. tmpfs: Memory-Backed RAM Filesystem
5. devtmpfs: Automatic Device Node Management

Chapter 7: Pseudo-Filesystems II: cgroupfs, debugfs, tracefs, securityfs, configfs

1. cgroupfs: Control Groups Filesystem
2. debugfs: Kernel Debugging Interface
3. tracefs: Tracing Infrastructure Filesystem
4. securityfs: LSM Configuration Interface
5. configfs: Dynamic Object Configuration

Chapter 8: Advanced Virtual Filesystems: overlayfs and FUSE

1. overlayfs: Union Mount Architecture
2. overlayfs Implementation Details
3. FUSE: User-Space Filesystem Framework
4. FUSE Performance and Limitations
5. overlayfs Versus btrfs Subvolumes: A Structured Comparison
6. Real-World Case Studies

Chapter 9: Namespaces, Mounts and Containerization

1. Mount Namespaces: Isolating the View of the Filesystem Hierarchy
2. Private, Shared, Slave and Unbindable Mount Propagation
3. Bind Mounts and Their Role in Container Storage
4. The Mount Tree Data Structure
5. Containers and VFS: Docker, Podman and Kubernetes Perspectives
6. Container Storage Driver Comparison
7. Overlayfs as the Default Container Storage Driver

Chapter 10: Performance, Scalability and Optimization

1. The Page Cache: Where Most File I/O Actually Happens
2. Read-Ahead Algorithms and Prefetching Strategies
3. Writeback and the Dirty Page Management Subsystem
4. Direct I/O and Bypassing VFS Caching (O_DIRECT, io_uring)
5. Scalability Challenges: Large Directories, Namespace Storms, Lock Contention
6. Tuning Parameters and Performance Diagnostics

Chapter 11: Troubleshooting, Debugging, Tracing and Monitoring

1. Syscall Tracing with strace and Performance Analysis with perf
2. Kernel Tracing: ftrace, tracefs and Built-In VFS Traces
3. eBPF for VFS Diagnostics (kprobes, tracepoints, BCC tools)
4. Common Failure Patterns
5. Mount Options for Debugging
6. Monitoring Filesystem Health and Performance in Production

Chapter 12: Cloud-Native Use Cases and Virtualization

1. Container Storage Orchestration: CSI, Persistent Volumes and VFS
2. Network Filesystems Through the VFS Lens
3. Virtualization Filesystem Backends (virtiofs, 9p, VMware vmhgfs)
4. Ephemeral Storage Patterns in Kubernetes and Serverless
5. Cloud Provider Filesystems: EBS, GCE Persistent Disks, Azure Files
6. Distributed Filesystem Challenges at Scale

Chapter 13: Recent Kernel Developments and Future Directions

1. The New Mount API: fs_context, fs_parameter, open_tree, move_mount
2. io_uring and Asynchronous File I/O Through VFS
3. Landlock: Unprivileged Process Sandboxing via Pathnames
4. statx and Extended File Attributes
5. BPF Filesystem Hooks and Runtime Security
6. Where VFS Is Heading: Trends, Open Questions and Community Directions

Conclusion: The Living Abstraction

References