The Outage Story

The Outage Story

Step 1 of 10

"The deploy took checkout down for forty minutes" is a headline, not a story. The story is the order you did things.

The Idea

An outage story is judged on sequence, not drama. Stop the bleeding, then find out why. Interviewers are listening for someone who reached for a rollback before a debugger, kept people informed while working, and left behind a change that makes the same failure boring.

Real-World Example

"The deploy took checkout down for forty minutes." Better: "Error rate alerted at 14:02. I rolled back at 14:06 and posted in the channel. The cause was a migration that dropped a column still read by the old pods; we now gate drops behind a two-release rule."

The Tradeoff

Own your part without performing guilt — an interviewer reading contrition as instability is a real risk. And never end on a person. The strongest version ends on the mechanism: a check, an alert, a rule that no longer depends on anyone remembering.

Your turn

Put the steps in the right order.

1. 1Find the cause once customers are served again↑↓
2. 2Restore service by the fastest safe route↑↓
3. 3Land a mechanism that makes the failure impossible or boring↑↓
4. 4Say how it was detected, and how long that took↑↓

Check order

Mini quiz

1 / 3

The first priority during an incident is:

• AFinding the root cause
• BWriting the timeline
• CNotifying everyone
• DRestoring service