Object Storage as Infrastructure

Building Databases, Queues, Git Systems, and Distributed Applications on S3-Compatible Storage

Introduction

Chapter 1: The Object Storage Primitive

1. What Object Storage Is and Is Not
2. The S3 API Surface: Objects, Keys, Prefixes, Metadata
3. Core Operations: PUT, GET, HEAD, DELETE, LIST, Multipart Upload
4. Consistency Guarantees and Their Evolution
5. Durability, Availability, and Replication Strategies
6. Authentication, Authorization, and Security Primitives
7. Object Storage Versus Block File Database and Queue Abstractions

Chapter 2: Designing Around Constraints

1. Latency Throughput and Cost Realities
2. Immutable Objects and Update Semantics
3. Expensive Listing Operations and Namespace Design
4. Conditional Writes and Optimistic Concurrency
5. Metadata Layers and Indexing Strategies
6. Handling Partial Failures and Network Partitions
7. The Compute Separation Principle

Chapter 3: S3-Compatible Implementations and Ecosystem

1. AWS S3 Architecture and Behavior
2. Open Source Implementations: MinIO Ceph RGW
3. Cloud Provider Variants: GCS Azure Blob DigitalOcean Spaces
4. On-Premise and Air-Gapped Deployments
5. API Compatibility Gaps and Gotchas
6. Choosing and Evaluating an Implementation

Chapter 4: Content-Addressable Storage and Immutable Data Structures

1. Hash-Based Object Identification
2. Deduplication and Size Savings
3. Object Layout and Key Design
4. Implementation: A Content-Addressable Blob Store in Go
5. Testing and Verification
6. Performance Characteristics and Tuning

Chapter 5: Append-Only Event Logs on Object Storage

1. Log Semantics and Ordering Guarantees
2. Segment Design and Batching
3. Sequence Numbers and Fencing
4. Implementation: A Partitioned Event Log in Go
5. Compaction and Retention
6. Recovery from Partial Writes and Corruption

Chapter 6: Durable Queues and Message Streams on Object Storage

1. Queue Semantics and Delivery Guarantees
2. Visibility Tokens and Acknowledgement Patterns
3. Implementation: A Durable Queue on Object Storage
4. Partitioned Streams and Consumer Groups
5. Exactly-Once Considerations and Idempotency
6. Performance Comparison with Kafka and Traditional Brokers

Chapter 7: Key-Value Stores and Metadata Databases on Object Storage

1. Key-Value Semantics on Immutable Objects
2. SSTable-Inspired Layout and LSM Concepts
3. Indexes and Bloom Filters for Efficient Lookup
4. Implementation: A Lightweight KV Store in Go
5. Read Path Optimization and Caching Strategies
6. Compaction Snapshots and Garbage Collection

Chapter 8: Analytical Data Stores and Query Execution on Object Storage

1. Columnar Formats and Parquet on Object Storage
2. Partitioning Strategies for Query Performance
3. Predicate Pushdown and Filter Semantics
4. Implementation: A Simple Analytical Query Engine in Go
5. Schema Evolution and Compatibility
6. Metadata Catalogs and Table Management

Chapter 9: Git-Compatible Storage Backends on Object Storage

1. Git Objects References and the DAG
2. Immutable Storage and Deduplication in Git
3. Packfiles and Delta Compression
4. Designing an S3-Compatible Git Backend
5. Ref Consistency and Concurrent Updates
6. Implementation: A Git Service with Object Storage Backend

Chapter 10: Distributed Coordination on Object Storage

1. Leases and Leader Election Patterns
2. Distributed Locking with Conditional Writes
3. Fencing Tokens and Session Management
4. Configuration Stores and Feature Flags
5. When to Use Object Storage Versus Dedicated Coordination Services
6. Implementation: A Lease-Based Coordinator in Go

Chapter 11: Performance Engineering at Scale

1. Measuring What Matters: RPS Throughput Tail Latency Cost
2. Request Batching and Parallel Upload Strategies
3. Connection Management and Asynchronous I/O
4. Caching Architectures: Memory Local Disk CDN Layers
5. Compression Serialization and Object Sizing
6. Benchmarking Methodology and Realistic Tests

Chapter 12: Reliability Failure Handling and Recovery

1. Taxonomy of Failures in Object Storage Systems
2. Idempotent Operations and Safe Retries
3. Handling Corrupted Objects and Integrity Verification
4. Recovery Procedures for Each Component Type
5. Fault Injection and Property-Based Testing
6. Disaster Recovery and Multi-Region Architectures

Chapter 13: Security Access Control and Compliance

1. Identity and Access Management Patterns
2. Temporary Credentials and Workload Identity
3. Encryption in Transit and at Rest
4. Object Locking and Immutability Controls
5. Tenant Isolation in Multi-Tenant Systems
6. Security Auditing and Common Misconfigurations

Chapter 14: Observability Operations and Production Deployment

1. Structured Logging and Request Tracing
2. Metrics SLIs and SLOs for Object Storage Systems
3. Capacity Planning and Lifecycle Management
4. Backup Restore and Migration Procedures
5. Incident Response and Operational Runbooks

Chapter 15: Case Studies from the Field

1. Cloud-Native Databases Built on Object Storage
2. Open-Source Projects Using S3 as Primary Storage
3. Streaming and Queue Systems with Object Backends
4. Git Hosting Platforms and Large-Scale Repositories
5. Lessons Learned and Common Pitfalls
6. What These Architectures Teach Us

Chapter 16: A Production Reference Architecture

1. Architecture Overview and Component Interactions
2. Storage Layout Across Services
3. Implementation: A Multi-Service System in Go
4. Deployment Configuration and Infrastructure Setup
5. Operational Procedures and Monitoring
6. Scaling Upgrading and Evolving the System

Conclusion

References