
Cover service ownership, testing, security, monitoring, deployment, backups and recovery, adapted to the service.
No. It helps the team identify and manage risks, but cannot eliminate all failure modes.
Start during development and revisit it before major releases and after incidents or changes to the architecture.