Site Reliability Engineer
Sky · Lisbon, Portugal
Oct 2023 – Present
Comcast NBCUniversal's Global Streaming Technology group, behind Sky UK, Sky Deutschland, Sky Italia, Peacock (US), SkyShowtime and NOW TV. My team sits in the middle: we run on the shared EKS clusters and Akamai CDN owned by other platform teams, and the application teams we support run on what we provide.
- Assessed how a load balancer change by the EKS platform team would affect our edge configuration, and updated origin settings (SNI) across more than 30 Akamai properties ahead of the cutover.
- Wrote a Python Prometheus exporter for GitHub service health (Git operations, Actions, pull requests, issues), with dashboards and alerts, after GitHub outages kept disrupting the team's work.
- Automated credentials for new RDS and DocumentDB instances: admin and user passwords are generated and stored in HashiCorp Vault, and Terragrunt reads them at creation time, so nobody handles a secret by hand.
- Support the applications through releases, incidents and live sports peaks such as NFL games, and share the on-call rotation.
- Own their observability and alerting with Prometheus, Thanos, Grafana and Alertmanager, including SLOs, certificate expiry checks, and alert routing to Slack and incident tooling.
- Manage the Akamai configuration they use: delivery properties, Global Traffic Management (GTM), Cloudlets, WAF security policies and TLS certificates.
- Deploy and run the AWS resources they need as code with Terraform and Terragrunt, maintain the GitHub Actions and Jenkins pipelines onto the shared EKS clusters, scale non-production down outside traffic windows to cut cost, and patch vulnerabilities on hosts and in code.