Service Level Objectives as Code: Bootstrapping SLOs with Terraform
summary
Shows how Sotheby’s implements SLOs as Terraform code using the Datadog Terraform provider to create SLO resources, dashboards and burn‑rate alerts; measures per‑route latency via an NGINX ingress because Linkerd and ELB lacked needed granularity; adds a Cron job that commits a known‑good change as a synthetic pipeline SLO for CI; publishes a reusable Terraform module that emits dashboards including an error budget burn down and burn rate alerting. Workshops with two teams produce working SLOs (place‑bid proxy and 99.9% front‑end render) and reduce alert noise while speeding iteration.
AI-written summary · about
chapters (13)
- 0:00Introduction to SLOs
- 1:30Why Sotheby's needs SLOs
- 4:21Understanding SLO concepts
- 7:17Challenges in implementation
- 9:39Prioritizing reliability
- 12:06Implementing CD pipeline SLOs
- 16:08CI pipeline SLOs
- 19:43SLOs as code with Terraform
- 22:56Alerting and burn rates
- 26:34Workshop-based adoption
- 30:55Future SLO initiatives
- 33:40Conclusion
- 35:43Q&A session
related talks
40mScaling SLOs with Kubernetes and Cloud Native Observability
26mUsing OpenTelemetry to Improve Service Level Objectives
31mServerless Observability: a case study of SLOs
28mLean Product Development Through SLOs
33mHow and why we lowered our SLO
32m9 Things You Should Do When Starting to Use SLOs
43mOur Experience Tracking and Driving SLO Adoption at Goldman Sachs
35mAvoiding Goodhart's Law - Use SLO's as Tools Not Cudgels