Notes
A living collection of ideas, engineering deep dives, patterns, and practical insights learned while building real-world systems.
Sprout
Early concepts & raw ideas
Budding
Work in progress & growing
Evergreen
Mature, deep dives & polished
Auto-Increment ID vs UUID: The Hidden Cost of Distributed Primary Keys
The article analyzes and compares three popular primary key strategies—Auto-Increment ID, UUIDv4, and UUIDv7—from a system architecture perspective, aiming to help engineers choose the optimal option based on write topology and access patterns rather than choosing by default.
Architecture Decision: When to Introduce Asynchronous Communication (Kafka)?
Framework and decision-making process for an Engineering Leader when a team proposes integrating Kafka to resolve slow synchronous APIs, weighed against the risk of unnecessary complexity.
Troubleshooting Docker Containers: Works Locally, Fails in Production
Investigation and remediation process when a container runs perfectly in a local environment but fails in production despite using the exact same image.
Troubleshooting Backend API Memory Leaks
Step-by-step troubleshooting and remediation process for a memory leak causing RAM usage to spike to 95% after 6 hours of uptime in backend applications.
Why I Didn't Use Drupal Commerce for a Simple Donation Feature
Architectural decision record: When to use Drupal Commerce and why a one-time donation problem requires a much leaner domain model.
Building Deterministic Tool Calling for Local LLMs
How to structure structured JSON output schemas and self-correcting prompt loops with Ollama and Qwen 2.5.
Database Indexing Strategies for Multi-Tenant CRMs
Optimizing compound B-Tree indexes and partition strategies when scaling tables beyond millions of rows.