Knowledgebase

Responding to an Incident Print

  • devopsinfrastructure, devops, restore, woocommerce, troubleshooting, guide, howto, solution
  • 0

The first hour.

WHAT TO DO FIRST

Establish what is broken, and for whom.

WHAT NOT TO DO

Start changing things before understanding the scope.

WHAT TO ESTABLISH

Is it everyone or some people? Is it everything or one function? When did it start? What changed beforehand?

THAT LAST QUESTION

Most incidents follow a change. It is the fastest route to a cause.

WHAT TO DO ABOUT AN INCIDENT FOLLOWING A DEPLOYMENT

Revert first. Investigate afterwards, with service restored.

WHAT TO ASSIGN

Someone leading, someone communicating, someone investigating.

WHY SEPARATE

The person fixing cannot also be answering questions.

WHAT TO COMMUNICATE

That you are aware What is affected When you will next update

WHAT TO AVOID

Speculating about causes publicly Promising a resolution time you cannot meet Silence

WHAT TO RECORD AS YOU GO

Times, observations, and actions taken.

WHY

Memory is unreliable, and the record is what you review afterwards.


Was this answer helpful?
Back

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


Trustpilot