Who writes here
Pascal Snijers. I’ve been working on systems reliability for longer than most of the tools discussed here have existed.
The short version: first professional steps in IT in 1987, from technical support to founding a small business-software publisher (LogiBox, 1988). Then software QA, LAN/WAN networking and CTI telephony — before diving into Telecom BSS/OSS for twenty years: billing lead engineer at BASE, release manager and N2 support for the European Commission at Atos, then EMEA Tier 3 at Comverse/Amdocs — P1 incidents on 24/7 billing platforms, under sub-4-hour resolution SLAs, for operators like Orange Luxembourg and Telenet.
Since 2018 I’ve been on the banking side: first end-to-end monitoring (Elasticsearch, AIOps, NPM), then the SRE role — SLI/SLO definition, alerting strategy, and coaching development squads toward observability maturity.
ITIL v3 and PRINCE2 certified. I was scripting Shell and PL/SQL on Informix and Oracle well before it was vintage, and writing Clipper before you were born, probably. These days: PHP, MariaDB, and Prometheus.
Why this blog
My homelab is where I apply what I practice in production: Proxmox on heighliner, a full observability stack (Prometheus, Grafana, Loki, Alloy) on hubble, all systemd-native on Debian. Yes, the machines are named after Dune. No, I don’t put Docker everywhere — when I do, it’s justified.
What you’ll find here: real field reports, with the failures, the numbers, and the complete configs. No rehashed tutorials, no news commentary. The articles are written in French — if you don’t read it, browser translation does a decent job, and the configs speak for themselves.
I also build nines.arewel.com, a Prometheus SLO generator.
Opinions published here are my own, not my employer’s.
— Pascal · Last updated: July 2026