Site Reliability Engineering (SRE)

Discipline Track | 7 Modules | ~4 hours total

The practice of applying software engineering to operations. SRE provides concrete methods for measuring, maintaining, and improving the reliability of production systems.

---

Why SRE?

Traditional operations faces a fundamental tension: development wants to ship fast, operations wants stability. These goals seem opposed.

SRE resolves this tension through:

• Measurable reliability — SLOs replace vague “make it reliable” with concrete targets
• Error budgets — Calculated risk-taking, not reckless shipping or excessive caution
• Engineering mindset — Automate toil, don’t just do it
• Learning culture — Blameless postmortems turn failures into improvements

SRE isn’t just operations with a fancy name. It’s a fundamentally different approach to running production systems.

---

Modules

#	Module	Time	Description
1.1	What is SRE?	30-35 min	Origins, mindset, team structures, SRE vs DevOps
1.2	Service Level Objectives	35-40 min	SLI, SLO, SLA hierarchy, choosing and measuring
1.3	Error Budgets	30-35 min	Budget calculation, policies, balancing velocity
1.4	Toil and Automation	30-35 min	Identifying toil, 50% rule, automation strategies
1.5	Incident Management	35-40 min	Response roles, severity, on-call, runbooks
1.6	Postmortems and Learning	30-35 min	Blameless culture, postmortem structure, action items
1.7	Capacity Planning	35-40 min	Forecasting, provisioning, load testing, cost

---

Learning Path

START HERE
    │
    ▼
┌─────────────────────────────────────┐
│  Module 1.1                         │
│  What is SRE?                       │
│  └── SRE origins and mindset        │
│  └── SRE vs DevOps vs Platform      │
│  └── Team structures                │
│  └── The 50% rule                   │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.2                         │
│  Service Level Objectives           │
│  └── SLI, SLO, SLA hierarchy        │
│  └── Choosing good SLIs             │
│  └── Setting realistic SLOs         │
│  └── SLO-based alerting             │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.3                         │
│  Error Budgets                      │
│  └── Budget calculation             │
│  └── Error budget policies          │
│  └── Balancing reliability/velocity │
│  └── When to spend vs save          │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.4                         │
│  Toil and Automation                │
│  └── Defining and measuring toil    │
│  └── The 50% rule in practice       │
│  └── Automation hierarchy           │
│  └── ROI-based prioritization       │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.5                         │
│  Incident Management                │
│  └── Response roles (IC, Comms)     │
│  └── Severity levels                │
│  └── On-call best practices         │
│  └── Runbooks and playbooks         │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.6                         │
│  Postmortems and Learning           │
│  └── Blameless culture              │
│  └── The "second story"             │
│  └── Effective action items         │
│  └── Sharing learnings              │
└──────────────────┬──────────────────┘
                   │
                   ▼
┌─────────────────────────────────────┐
│  Module 1.7                         │
│  Capacity Planning                  │
│  └── Demand forecasting             │
│  └── Provisioning strategies        │
│  └── Load testing                   │
│  └── Cost optimization              │
└──────────────────┬──────────────────┘
                   │
                   ▼
            SRE COMPLETE
                   │
    ┌──────────────┼──────────────┐
    │              │              │
    ▼              ▼              ▼
Platform      GitOps       Observability
Engineering   Discipline      Toolkit

---

Key Concepts You’ll Learn

Concept	Module	What It Means
SLI	1.2	Service Level Indicator — what you measure
SLO	1.2	Service Level Objective — your target
SLA	1.2	Service Level Agreement — external promise
Error Budget	1.3	Allowed unreliability (1 - SLO)
Toil	1.4	Repetitive, automatable work
50% Rule	1.1, 1.4	Cap toil at 50% to ensure engineering time
Incident Commander	1.5	Person coordinating incident response
Blameless Postmortem	1.6	Learning from failure without blame
Burn Rate	1.2, 1.3	How fast error budget is being consumed
Headroom	1.7	Buffer capacity for traffic spikes

---

Prerequisites

• Required: Reliability Engineering Track — Failure modes and resilience
• Required: Systems Thinking Track — Understanding system dynamics
• Recommended: Observability Theory Track — Metrics and monitoring
• Helpful: Experience operating production systems
• Helpful: Some Kubernetes experience

---

Where This Leads

After completing the SRE Discipline, you’re ready for:

Track	Why
Platform Engineering Discipline	Build platforms using SRE principles
GitOps Discipline	Declarative infrastructure operations
IaC Discipline	Infrastructure as Code for reliable provisioning
Observability Toolkit	Implement Prometheus, Grafana, OpenTelemetry
IaC Tools Toolkit	Terraform, OpenTofu, Pulumi for automation
CKA Certification	Apply SRE to Kubernetes administration

---

Key Resources

Books (referenced throughout):

• “Site Reliability Engineering” — Google (free online, the original SRE book)
• “The Site Reliability Workbook” — Google (practical companion)
• “Implementing Service Level Objectives” — Alex Hidalgo

Articles:

• “What is SRE?” — Google Cloud
• “SRE vs DevOps” — Atlassian

Tools mentioned:

• Prometheus/Grafana: SLO monitoring
• PagerDuty/OpsGenie: Incident management
• k6/Locust: Load testing
• Kubernetes HPA: Auto-scaling

---

The SRE Mindset

Question to Ask	Why It Matters
”What’s our SLO?”	Can’t improve what you don’t measure
”How much error budget do we have?”	Know when to ship vs when to stabilize
”Is this toil?”	Automate if yes, protect engineering time
”What would prevent this incident?”	Systems fix, not human vigilance
”Who’s the IC?”	Clear roles in chaos
”Can we handle 2x traffic?”	Plan before you need it

---

Discipline Complete

After these 7 modules, you can:

• Define and measure reliability with SLOs
• Balance reliability and velocity with error budgets
• Identify and eliminate toil systematically
• Respond to incidents with clear roles and processes
• Learn from failures through blameless postmortems
• Plan for future capacity and growth

You’re now equipped to practice SRE, not just read about it.

---

“Hope is not a strategy.” — Traditional SRE saying