Question

After a two-hour outage caused by an engineer applying an untested configuration change directly to production, your SRE team conducts a postmortem. During the review, several managers push to identify who made the mistake and include a formal reprimand as an action item. As the DevOps lead responsible for the site reliability practice, what should you advocate to best align the postmortem process with SRE principles and reduce the likelihood of recurrence?

Accepted Answer

Keep the postmortem blameless, focusing on systemic gaps such as the lack of a required peer-review and automated validation gate for production config changes, and track fixes as action items.

Answer

Document the individual responsible in the postmortem so accountability is clear, and require them to complete additional training before touching production again.

Answer

Skip the postmortem for outages under four hours and instead rely on monitoring alerts to catch similar issues faster next time.

Answer

Freeze all production changes indefinitely until every engineer signs a formal agreement accepting personal responsibility for outages they cause.