Volume 01 / The diagnosis workbook

The Incident Interview Workbook

Find the fault. Explain the recovery.

A rollout is healthy, but the service is down. Start with the evidence available to the on-call engineer, choose your next check and defend the recovery. Eight cases take you through Kubernetes, DNS, resource limits and failures that cross service boundaries.

Read a sample case Get Premium access

Full PDF + practice kit included with Premium. Keep your downloaded copy for personal study.

Already a member? Sign in to download.

September 2026 · Personal study edition

Illustrated pages: 36
Worked cases: 8
Technical diagrams: 8
Practice material: PDF + Python kit

A complete case / Free to read

The Pods are ready. The Service still has no backends.

Read the opening incident from its first signals through the investigation and follow-up exercise. These are three consecutive pages from the workbook.

Page 5 from The Incident Interview Workbook: A healthy rollout, an unavailable service — 01 / The brief
A healthy rollout, an unavailable service
PDF page 5

Page 6 from The Incident Interview Workbook: Readiness and reachability answer different questions — 02 / Worked reasoning
Readiness and reachability answer different questions
PDF page 6

Page 7 from The Incident Interview Workbook: Explain why the green dashboard was insufficient — 03 / Your next attempt
Explain why the green dashboard was insufficient
PDF page 7

Select a page to read at full size, zoom in or switch to text. On a phone, swipe to see the next sample.

Who should pick up this volume

Practice the part you need to explain.

Engineers who already operate Linux or cloud services and want to improve incident diagnosis and technical communication.

KubernetesLinux and DNSDatabase reliabilitySREIncident ownership

What you will work through

Separate symptoms from causes using a request path and a hypothesis table.
Calculate retry amplification, queue drain time and error-budget consumption.
Defend a recovery decision and describe the evidence needed to close an incident.

The full reading route

Eight cases. A decision in each one.

Brief → worked reasoning → changed conditions.
Follow the cases in order or revisit a weak topic.

How to rehearse
A diagnosis you can explain
01 / A healthy rollout, an unavailable service
Trace traffic from the Service to the selected Pods.
02 / DNS fails in one namespace
Separate a lookup failure from a connectivity failure.
03 / Slow requests on idle nodes
Investigate CPU limits without trusting host averages.
04 / The connection pool is the messenger
Follow blocked work beyond the application pool.
05 / Retries keep the outage alive
Choose a recovery that reduces retry pressure.
06 / Free space that never comes back
Find the storage still held by an open file.
07 / A reassuring average hides the damage
Recalculate impact for the population that matters.
08 / A region switch with uncertain writes
Establish write authority before moving traffic.
Lead the incident and tell the project story
Run a mock and review the recording
Diagnostic reference and practice kit
Technical sources and publication notes

From the page to your next attempt

A practice kit you can run on your laptop.

Use local request fixtures and Python calculations to check weighted error rates, retry amplification and recovery time. Decision worksheets give you a place to record the evidence behind your answer.

Python 3.10+ · Synthetic local fixtures · No cloud provisioning

Make your call

Write your decision and the observation that would change it.

Check the reasoning

Use the worked case and local fixtures to find gaps in your explanation.

Rehearse the follow-up

Change the assumption, revise the answer and record what still needs practice.

One Premium membership

Keep all four guides within reach.

Download the full books and companion kits while your access is active. Each includes a four-week routine and a self-review worksheet.

See Premium access →

Edition, sources and personal use

Reviewed 2026-09-24. These are fictional teaching cases with primary technical references, not reproduced employer assignments or an employer scoring rubric. AI tools assisted preparation and cover illustration. Local exercises test the stated models and calculations; they do not validate a production deployment.

The three preview pages are public. The complete PDF and practice kit require current Premium access. Copies are for personal study under the site terms.

Continue your collection