Knowledgebase

Writing a Post-Incident Review Print

  • devopsinfrastructure, devops, troubleshooting, performance, guide, howto, solution, zillionkinghost
  • 0

Learning from failure.

WHAT TO PRODUCE

A short written record.

WHAT IT SHOULD CONTAIN

What happened, with a timeline What the impact was What the cause was Why it was not prevented Why it was not detected sooner What will change

WHAT TO FOCUS ON

The system, not the person.

WHAT TO ASK

What made this possible? What would have caught it earlier? What made recovery slow?

WHAT NOT TO CONCLUDE

That someone should have been more careful.

WHY

That produces no change, and discourages reporting.

WHAT TO MEASURE

Time to detect Time to respond Time to resolve

WHAT THOSE REVEAL

Where the process is weak.

A long time to detect is a monitoring problem, not a cause problem.

WHAT TO PRODUCE

A small number of specific actions, each with an owner and a date.

WHAT NOT TO PRODUCE

A long list nobody completes.

WHAT TO DO AFTERWARDS

Check that the actions were taken.


Was this answer helpful?
Back

Are you happy with your experience? Leave us a review on Trustpilot.


Trustpilot