The reading room
Articles
One system, one mechanism, one carefully followed thread at a time.
5 articles
Published first; undated drafts alphabetical OpenTelemetry / Observability / Performance
The Hidden Costs of OpenTelemetry Instrumentation
What really happens inside an instrumented application, from recording and aggregation to context propagation, batching, and shutdown.
Linux / Rust / Memory
NUMA and eBPF: getting incoming traffic onto the right worker
From Linux socket selection to a NUMA-aware telemetry receiver, with OpenTelemetry Arrow as a practical case study.
AI / Python / Observability
Building an AI Agent: Learning the Concepts Through Incident Investigation
Learn how a model, tools, a loop, state, and limits work together to investigate why checkout became slow.
Linux / Rust / Observability
Linux user_events, explained: from application events to OpenTelemetry
How applications publish events through Linux tracing, with a practical look at OpenTelemetry Arrow's Rust engine.
Linux / Rust / Memory
NUMA, explained: from CPU cores to a Rust telemetry engine
How memory placement affects performance, with a practical look at OpenTelemetry Arrow.