If a backend endpoint suddenly became slow in production, what steps would you take to investigate and prioritize the issue?

Based on the Backend Developer role and the intermediate difficulty setting, which should assess practical production judgment.