intermediate
Observability: Metrics to Decisions
Prometheus-style metrics, structured logging, and tracing hooks are wired into a sample microservice. You will define SLIs, craft alert routes that avoid noise, and facilitate a tabletop drill.
Logistics
5 weeks · 6h/week · Hybrid · ¥108,000 JPY (informational)
Included focus areas
- RED/USE metrics with exemplar-friendly histograms
- Log schema design with correlation IDs
- Tracing spans for critical user journeys
- Alert routing with on-call empathy in mind
- Dashboards that answer "so what" in one screen
- Post-incident templates with blameless prompts
- Load tests to validate burn rates
Outcomes
- Ship a dashboard your PM can interpret without jargon.
- Draft an alert policy with explicit owner and rollback.
- Run a 45-minute drill and capture three follow-ups.
Responsible instructor
Ami Fujita
SRE coach for cross-border product squads.
FAQ
Which stack?
Labs use Prometheus + Grafana; tracing uses OpenTelemetry-friendly exporters.
Is on-call simulation included?
We run a facilitated tabletop; live pager rotations are not part of tuition.
What is not included?
Vendor-specific APM agents beyond OTel basics.
Experience notes
“Burn-rate exercise made our error budget meetings shorter — pulled straight from the Observability track.”
“Correlation ID module is in prod now. Drill format was a bit intense but fair.”