INTERMEDIATE · 8 MODULES · 64 LESSONS
SRE and Incident Response
Measure how reliable your system is in actual numbers, and respond to a production incident with clear roles, useful alerts, and a blameless postmortem.
- For you if
- Your AWS system already works, and nobody has defined what reliable means in numbers.
- By the end
- You'll know how to define SLIs, SLOs, and error budgets, alert on budget burn, run an incident, and write a runbook that works.
- You need
- An AWS project with its pipeline already in place, Docker, basic Python, and a LocalStack account.

Curriculum
01What Is Sre And Reliability As A Feature
- Guide introduction: from "it works, it's secure, it fits the budget" to "it's reliable"
- What SRE is (and isn't)
- Reliability as a feature, not an accident
- Hands-on: reading Andes Cargo like an SRE would
- Two real incidents, two different questions
- Hands-on: the error budget the Claude Code incident burned
- The vocabulary you'll use all guide long
- Project: Andes Cargo's reliability charter
02Slis Slos And The Error Budget
- Introduction: from vocabulary to math
- Choosing a good SLI: the four golden signals
- Hands-on: your first SLI definition
- Hands-on: the error budget calculator
- Choosing an SLO that means something
- Hands-on: burn rate, the consumption rate, not just the balance
- Hands-on: applying the calculator to a real Andes Cargo scenario
- Project: Andes Cargo's `SLO.md`
03Observability As Sli Input
- Introduction: observability in service of a question, not a dashboard
- The three pillars, with an SRE lens
- Hands-on: real metrics from the inherited Lambda
- Hands-on: real logs with `jq`
- Hands-on: traces with OpenTelemetry and Jaeger
- Hands-on: Prometheus and Grafana, the stack the market asks for
- From raw telemetry to a measured SLI
- Project: Andes Cargo's observability-to-SLI pipeline
04Alerting On Error Budget Burn Rate
- Introduction: why a static threshold isn't enough
- Multi-window, multi-burn-rate: the real Google SRE pattern
- Hands-on: the burn rate evaluator
- Hands-on: the real rule in Alertmanager
- Hands-on: a real CloudWatch alarm on the Lambda
- What AWS already automates: CloudWatch Application Signals, named
- Hands-on: routing the alert
- Project: Andes Cargo's alerting policy
05The Incident Lifecycle
- Introduction: the framework before the case
- The lifecycle of an incident
- Severity levels, with real Andes Cargo examples
- Roles during an incident
- Hands-on: Andes Cargo's severity matrix
- On-call, with honesty about its cost
- Hands-on: a deterministic on-call rotation
- Project: Andes Cargo's `INCIDENT-RESPONSE-PLAN.md`
06Operating The Claude Code Incident
- Introduction: operating, not narrating again
- Hands-on: reconstructing the exact timeline
- Hands-on: classifying the severity
- Hands-on: declaring the incident
- Communication during an incident: internal vs. external
- Hands-on: the mitigation decision tree
- This incident's error budget, with the full vocabulary
- Project: Andes Cargo's final `TIMELINE.md` for this case
07Blameless Postmortems And Runbooks
- Introduction: from the timeline to systemic learning
- What makes a postmortem "blameless"
- Hands-on: writing the Claude Code incident postmortem
- Action items that actually get done
- What a runbook is (and is not)
- Hands-on: Andes Cargo's first real runbook
- Hands-on: the honest backup/restore attempt
- Project: Andes Cargo's postmortem and runbooks package
08Capstone The Andes Cargo Reliability Package
- Introduction to the capstone
- Architecture review: the full reliability machine
- Hands-on: introducing a deterministic synthetic incident
- End-to-end: the alert fires, the incident gets declared
- End-to-end: the runbook mitigates, the postmortem closes
- What this guide left representative, final honesty
- What Andes Cargo still needs
- Final project: Andes Cargo's reliability package as a deliverable
Reviews
What students say
These reviews are from enrolled students who completed at least 50% of the course. We moderate reviews only on content grounds (spam, offensive language, personal data), never for being critical or negative.
No approved reviews yet.
Be the first to share your experience!