My Skill Lab
Cloud & DevOpsAdvanced

Site Reliability Engineering Fundamentals

SLOs, error budgets, on-call, incident management and observability.

Taught by Suman Ghimire · DevOps Engineer

What you'll learn

  • Define user-centred SLIs and SLOs for a service and record them as Prometheus rules
  • Calculate an error budget and write an error budget policy the whole team agrees to
  • Build multi-window, multi-burn-rate alerts that page only when the SLO is genuinely at risk
  • Run an incident with clear roles, severity levels and a communication cadence
  • Write a blameless postmortem with actionable, owned follow-up items
  • Plan capacity with load tests and measure, prioritise and automate away toil

About this course

This course is for developers, DevOps engineers and operations staff who want to move into a site reliability engineering (SRE) role or bring SRE practice to their current team. You will learn to define SLIs and SLOs that reflect what users actually feel, turn them into error budgets, and write multi-window burn-rate alerts in Prometheus that page for real problems instead of noise. You will design a sane on-call rotation, run incidents with clear roles, write blameless postmortems, plan capacity with load tests, and measure and reduce toil. Hands-on lessons use real tools such as Prometheus, Alertmanager, OpenTelemetry, Grafana and k6. You should already be comfortable with Linux, HTTP services and at least one cloud or container platform; basic Kubernetes and monitoring experience will help. By the end you will have an SLO document, an alerting policy and an incident process you can take straight to work.

Course content

7 lessons
  1. 1What SRE Is and Why It ExistsThe origins of SRE, its core principles and how it differs from traditional operations and DevOps.Preview
  2. 2SLIs and SLOs: Measuring What Users FeelChoose service level indicators that reflect user experience and set objectives the business agrees to.
  3. 3Error Budgets and Burn-Rate AlertingTurn an SLO into a budget, agree a policy for spending it, and alert on how fast it burns.
  4. 4Observability and Alert DesignGolden signals, metrics, logs and traces with OpenTelemetry, and routing alerts to the right people.
  5. 5On-Call and Incident ManagementDesign a humane on-call rotation and run incidents with clear roles, severities and communication.
  6. 6Blameless Postmortems That Change ThingsTurn incidents into learning with blameless postmortems, contributing factors and owned action items.
  7. 7Capacity Planning, Toil and Safe ReleasesLoad test for capacity, measure and automate toil, and release changes with canaries and fast rollback.