Engineer-authored
technical content.
No marketing interns writing about DevOps. Real production experience, real trade-offs, real code. Written by the same senior engineering that builds and operates the systems described here.
Field notes from
production infrastructure
These aren't listicles or vendor recaps. Every post comes from a real deployment, a real migration, or a real incident. Where a claim is made, it's backed by a measurement. Where a trade-off exists, it's named honestly.
Open-Source TCO: The Full Breakdown
Real cost comparison from production deployments — line-itemed so you can run the math against your own workload. Licensing, per-node pricing, support contracts, and where the gap is wider than it looks.
Knowledge Management for DevOps Teams
Why a knowledge system was built, how it captures decisions and incidents, and how it makes delivery compound rather than depend on one engineer's memory.
Proxmox VE Hardening: What Production Actually Does
Production-tested hardening steps for Proxmox VE clusters. Network isolation, two-factor auth, encrypted storage, and update discipline — not theory.
Monitoring Stack: Zabbix + Graylog + Grafana
Three tools, one coherent observability stack. How they fit together, where they don't, and the tuning that turns alerts into signal.
CI/CD with Gitea Actions: What Actually Works
Pipelines that gate releases on real tests. Runner setup, secret management, promotion strategies, and the failure modes that bite at scale.
AI-Augmented DevOps: Patterns, Not Hype
How AI agents are actually used in production operations — log triage, dependency scanning, draft runbooks — and where the human boundary sits.
The Case of the Vanishing backdrop-filter
A CSS minifier silently dropped the unprefixed backdrop-filter from production builds. The debugging path from symptom to root cause, and the fix.
Policy-Based Routing Is Easy. Keeping It Alive Is the Job.
DNS-aware policy-based routing on OpenWRT for geo-restricted services, and the three production outages that taught a layered self-healing safety net.
What's coming next
New posts are published when there's something worth saying — not on a content calendar. Topics currently in the queue, from production work that is already running:
- OpenTofu patterns for Proxmox clusters. Reusable modules for cluster provisioning, with the failure modes that bite when you template a three-node cluster and one node is on different hardware.
- Wazuh deployment and rule tuning. SIEM that fits a mid-market environment — agent rollout, rule tuning for relevant threats, and the compliance reporting that comes out as a byproduct.
- Backup and restore drills that actually run. Proxmox Backup Server in production, with the monthly restore-test discipline that turns "we have backups" into "we have recoverable backups."
If a topic here matches a problem you have now, the fastest path is a call — the conversation will be more useful than waiting for the post. Schedule a review →
While you're here: the FAQ answers the most common questions in short form, and the TCO guide is the full math behind the open-source savings claim.
Want the ungated version?
The blog is public. The case studies, the actual TCO spreadsheets, and the architecture diagrams come out in a scoping call — under NDA, with context that doesn't fit in a post.