Production Readiness Checklist

A production readiness checklist answers one question before a service takes real traffic: when it breaks, and it will, will someone find out quickly, know what to do and be able to fix it without making things worse? Most launch problems are not exotic. A call with no timeout ties up every thread. A retry loop with no backoff turns a small blip into an outage. An alert fires into a channel nobody reads. A backup exists but has never been restored. The fix for each is small if you catch it before launch, and painful if you catch it during an incident.

This checklist follows the shape of Google's SRE launch coordination checklist and the reliability chapters of the SRE book and workbook, adapted for a typical team shipping web services and APIs today. The chapters cover ownership and documentation, reliability and capacity, observability and alerting, deployment and rollback, data and recovery, security basics, and incident readiness.

Every check has a severity. A must is something that will turn an ordinary failure into a long or user-visible outage if it is missing. A should closes a common gap that tends to bite within the first few months. A nice adds depth for services that matter more. Every check also says how to verify it: a command to run, a setting to look at, a drill to perform. A box ticked without that evidence does not count.

Run it as a review with the people who will be on call for the service, not as a form one person fills in. It takes an hour or two for a small service, and the gaps it finds become tickets with owners and dates before the launch goes ahead.