Long-form writing on runtime, distributed systems, kernel internals, Go concurrency, memory models, eBPF, and assembly walkthroughs.
Most performance advice for Go is ritual — 'use sync.Pool,' 'avoid interface boxing,' 'preallocate slices.' Useful sometimes, hollow often. A production engineer's guide to profiling Go systems with pprof, reading escape analysis output, and understanding when the compiler actually inlines.
sync.Pool is the Go feature most likely to be used incorrectly. A working engineer's guide to when pooling buffers actually saves GC pressure, when it just adds complexity, and the benchmark methodology that tells the difference.
Why blind fail-fast during leader election causes retry storms, and how bounded retry budgets, failure boundaries, and error normalization create predictable distributed systems.
The most common microservices mistake isn't picking the wrong transport. It's misreading who is responsible for knowing the work finished. A field guide to completion ownership in RPC, message bus, and event-driven systems.
All three are 'messaging systems.' None of them is interchangeable with the others. A practical breakdown of NATS, Kafka, and MQTT — by the actual design axes that determine which one breaks when you misuse it.
Docker is not a virtual machine. Kubernetes is not a container tool. Fifteen years in, both are misunderstood — and misused — as a result. A working engineer's explanation of what they actually changed.
The 'which language is fastest' benchmark wars miss the real question. Rust, Go, Java, and Python aren't competing on the same axis. They're tuned for different scaling strategies — and picking the wrong one costs you years.
The unit-vs-integration framing is a junior lens. Production Go backends need a different taxonomy: deterministic tests, contract tests, race tests, and fidelity tests. The ones that actually catch production bugs.
Tracing systems optimize for signal-to-noise. Billing systems optimize for accuracy and auditability. They look similar and they are not the same thing. A dual-path architecture for running both without cross-contamination.
context.Context is not documentation, not a metadata bag, and not optional. A production-hardened guide to cancellation propagation, the background-goroutine trap, and the patterns that keep services alive when downstreams slow down.
© 2026 HarrisonSec. All rights reserved.