You are managing a relational database system that experiences high traffic at certain times of the day. During one such peak traffic period, the primary database instance crashes, and the standby replica fails to be promoted to the primary role, leading to service disruption. What should you do to prevent this issue from occurring in the future?