AdvancedObservability & Operations

Lab: diagnose in production

Choose the right signal and the right response when production misbehaves.

Lab · optionalFundamentalsAdvanced9 min

Recommended first

By the end of this lesson you will be able to:

Optional scenario lab. Operations is about seeing clearly and responding calmly. Practice both.

1.
You need to follow one slow request as it flows across several services. Which signal?
LogsMetricsA distributed traceAn alert
2.
Which is the best thing to page a human about at 2am?
CPU at 80%Checkout error rate spikingDisk 60% fullA deploy finished successfully
3.
Production is down and you've confirmed it's a recent deploy. First priority?
Write the postmortemMitigate — restore service (e.g. roll back)Find the exact root cause before doing anythingFigure out who to blame
4.
A good postmortem focuses on systemic gaps (missing alert, fragile deploy), not on who typed the command.
TrueFalse

See clearly, alert on what matters, mitigate before you diagnose, and learn without blame.

Finished reading? Mark it complete to track your progress.

Incident response

A calm, repeatable process for when production breaks — and learning from it after.

Thinking like an attacker

Threat modeling — systematically asking how something could be abused.