Chapter 3

Can you trust your own judgment at 2 a.m.?

Lúc hai giờ sáng, đầu óc mình còn tin được không?

The chapter's question

Does the on-call person know what to do when a familiar error happens?

Big idea

At two in the morning nobody thinks clearly; a prewritten checklist thinks for you.

In this chapter

  1. Part 3.1

    The Five Common Failures

    Which five kinds of incident happened most over the past year?

  2. Part 3.2

    Step by Step

    Can a newcomer follow the runbook?

  3. Part 3.3

    Access When Needed

    Does the on-call person have enough permissions to follow the runbook?

  4. Part 3.4

    Testing the Runbook

    Has the runbook ever been tested by following it?

The reading for this chapter is inside Conan School

This chapter has 4 readings, with case stories and exercises. Chapter 1 of The Two A.M. Incident is free to read right here.

Read in Conan School

Chapters of The Two A.M. Incident

  1. 1.Why do customers find our outages before we do?
  2. 2.Mid-outage, should we hunt for the cause first?
  3. 3.Can you trust your own judgment at 2 a.m.?
  4. 4.Does finding who's to blame make systems safer?
  5. 5.Is this a new incident, or an unpaid old one?