You have implemented a continuous deployment pipeline that automatically deploys source code changes to your cloud infrastructure within instance groups designed for self-healing. After deploying a recent change, you notice a significant negative impact on a key performance indicator (KPI). Resolving the issue through investigation could take up to a week. What should you do to quickly mitigate the negative impact?