01INCIDENT RESPONSE
Incident response across ECS and EC2
Restoring confidence under pressure
A PHP2Shell response across multiple applications: runtime discovery, coordinated remediation and the limits of installing packages at boot.
- AWS
- ECS
- EC2
- WAF
CLOUD & PLATFORM ENGINEERING
Infrastructure, designed for the real world.
I'm Fabio Lopes. I work across cloud architecture, platform engineering and SRE, connecting systems design with the realities of running production.
20+ years in infrastructureBased in Brazil · Working in English & Portuguese
From architecture to operations
01
Design
Architecture, constraints & trade-offs
02
Build
Platforms, infrastructure & automation
03
Operate
Observability, recovery & reliability
Selected work
Decisions, constraints and lessons from hands-on engineering.
02OPERATIONAL AUTOMATION
Practical workflow
Making storage growth visible
From an urgent hardware-planning need to scheduled Ceph forecasts, charts and Teams reports, integrating Mimir metrics with Python.
- Ceph
- Python
- Grafana Mimir
- Teams
03PRODUCTION ENGINEERING
Operational lessons
Operating within real constraints
Running a community application on a single ARM host: consistent SQLite backups, controlled releases and recovery trade-offs.
- Linux ARM
- SQLite
- OCI
Field notes
Practical writing on infrastructure, automation and reliability.
16 min read
AWS / SRE
PHP2Shell on AWS: Incident Response and the Limits of Boot-Time Patching
Lessons from a multi-application PHP2Shell response involving lightweight discovery, AWS WAF containment, Amazon ECS image rebuilds, and EC2 bootstrap reliability.
10 min read
terraform / SRE
Beyond Day-1 Terraform: Rethinking Platform Operations in My Environment
Reflections on infrastructure ownership, ECS deployments, and Terraform lifecycle management in a multi-account AWS environment.
11 min read
SRE / DevOps
Production Engineering for Single-Node Workloads: Controlled Releases, SQLite Backups, and Native ARM Builds
Lessons from operating a Node/SQLite workload on Oracle Linux ARM: versioned releases, consistent backups, OCI Instance Principals, and the limits of a single-host design.