Blog
Notes on systems programming, open source, and things I've figured out the hard way.
-
RAG in Production: Stale Indexes, Broken Retrieval, and the Observability You Need
The notebook retrieves the right chunk and everyone claps. Then documents change, policies get retracted, chunk boundaries cut answers in half, and nobody can explain why the system said what it said.
ragllmretrieval -
Token Buckets, Concurrency Caps, and Load Shedders: Rate Limiting That Holds Under Pressure
A request limiter stops one bad script. A concurrency limiter stops resource contention. Two load shedders keep you alive during the incident. Here is what each one actually does, the Redis code behind them, and the failure modes nobody warns you about.
reliabilitydistributed-systemsredis -
Kafka Partition Expansion: Data Skew, Key Ordering Breaks, and Rebalance Storms
The alter command finishes in seconds. Then you find out about stranded data, broken key ordering, rebalance storms, and what it does to stateful stream apps.
kafkadistributed-systemsdevops -
PySpark in Production: Notes from Real Usage
Things I wish I had known before using PySpark at scale — partition management, schema handling, streaming gotchas, and performance patterns that actually matter.
pysparkdata-engineeringpythonstreaming -
Setting Up a VPS from Scratch: What I Actually Do
A walkthrough of how I set up a fresh Linux VPS — from locale and SSH hardening to nginx, Docker, fail2ban, and automated backups with rclone.
linuxself-hostingnginxdevops
No posts match your search.