Path 03 / Learning syllabus
SRE fundamentals
Follow a report from the user’s request to a usable result, then decide what to measure, when to respond and which reliability improvement to fund first.
Explore the syllabus
Your course project
A service reliability plan
Build a reliability plan for Dispatch, a fictional report-export service. Start with the promise made to a user, then add a dependency map, measurable objective, response policy and prioritized engineering proposal. Finish by defending one investment against a competing use of the same time.
Before you begin
Start where it makes sense.
For engineers and technical leaders new to SRE. Start with a basic understanding of an application and its users; no SRE background or production access is needed. The exercises use fictional service records and small calculations.
Already know the foundations? Jump to any lesson in the syllabus.
Start with the supplied examples. Each exercise explains what you need, what to try and how to check the result; software-based exercises use a disposable practice environment.
The ordered path
What you will learn
Follow the sequence to build the course project, or open the lesson that answers your question.
- 01BeginnerLesson · 7 min read
What makes a service reliable?
Start with an export a user can download, rather than a healthy process. Define the service promise and who is affected when it fails.
What you will try
Write Dispatch’s service brief and describe an unfinished export that a successful health check would miss.
- 02BeginnerLesson · 6 min read
Trace the journey from request to result
Map submission, queueing, processing, storage and download. Identify dependency owners and where a failure changes the user’s experience.
What you will try
Draw the export journey, mark its ownership boundaries and explain what happens when storage slows down.
- 03BeginnerLesson · 5 min read
Measure the outcome users care about
Define a service-level indicator using eligible exports, successful completion, a deadline and a measurement window. Account for work that disappears from the records.
What you will try
Write an SLI specification and classify successful, late, failed and missing exports against it.
- 04BeginnerLesson · 5 min read
Choose an SLO and explain the tradeoff
Use user expectations and observed service behavior to propose a target. Distinguish an internal objective from a customer commitment and make the cost of improving it visible.
What you will try
Compare two proposed targets, identify who must agree and record the evidence still needed before adoption.
- 05BeginnerLesson · 5 min read
Turn the error budget into a working policy
Calculate the allowance in the SLI’s units, then decide what budget consumption changes about releases and reliability work.
What you will try
Calculate remaining allowance from a sample window and apply the same policy to a feature release, an urgent repair and missing telemetry.
- 06IntermediateLesson · 5 min read
Page for a problem someone can act on
Connect the speed of user impact to response urgency. Separate a page requiring immediate action from a ticket that can wait.
What you will try
Draft a page with impact, owner, first check and escalation; explain why a noisy component signal belongs in a ticket instead.
- 07BeginnerLesson · 5 min read
Make the first on-call shift workable
Put access, ownership, escalation and recovery checks into a usable runbook. Consider coverage and workload alongside the technical procedure.
What you will try
Have another reader walk the runbook, then fix one missing permission, decision or escalation route they uncover.
- 08BeginnerLesson · 4 min read
Reduce toil without automating the confusion
Identify repetitive operational work, measure its burden and decide whether to remove, simplify or automate it. Include exceptions and maintenance in the comparison.
What you will try
Compare one manual task with a proposed automation, including effort saved, failure handling and ongoing ownership.
- 09IntermediateLesson · 5 min read
Release a change you can evaluate and stop
Choose a small initial exposure, signals that establish success and a stopping point. Check data and interface compatibility before relying on rollback.
What you will try
Write a staged release plan with a baseline, progression decision, stop criteria and a recovery path for an irreversible data change.
- 10IntermediateLesson · 4 min read
Keep a slow dependency from taking everything down
Follow queued work and retries through a storage slowdown. Use capacity limits, deadlines and degraded behavior to keep recovery possible.
What you will try
Draw the overload sequence and propose where to bound work; explain what users experience while the limit is active.
- 11IntermediateLesson · 4 min read
Turn a failure into an engineering decision
Connect a post-incident action to the condition it should change. Separate delivering the fix from demonstrating its effect on reliability.
What you will try
Specify one follow-up test, an owner and the exposure that remains if the test covers only part of the service.
- 12IntermediateLesson · 5 min read
Defend the next reliability investment
Bring the service promise, SLO evidence, operational burden and failure scenarios together. Choose work for its expected effect and cost.
What you will try
Present a one-page plan with one priority, a rejected alternative, the expected benefit and a measurable review point.
