Engineering investigations
Labs
Failure-oriented experiments with explicit assumptions, instrumentation, competing explanations, and measurable tradeoffs.
Oct 07, 2026
11 min
11 min
Why Your Backend Is Slow Even When CPU Looks Fine
A failure-oriented walkthrough of connection-pool saturation, database contention, queueing, and the misleading comfort of normal host-level CPU.
Oct 06, 2026
9 min
9 min
Kafka Under Load: When More Partitions Stop Helping
Partition count is a scaling mechanism, not a universal cure. This lab separates broker capacity, key skew, consumer concurrency, and rebalance overhead.
Oct 05, 2026
10 min
10 min
A RAG Demo Can Be Correct While the Production System Is Unreliable
A compact failure matrix for stale documents, retrieval misses, chunking errors, ranking drift, latency, and missing observability.