On-call burnout: strategies for sustainable incident management

on-call burnout is a serious issue in our team. we're a small team of 8 engineers, and we average about 5-10 pages per week, with 2-3 of those requiring outside-hours intervention. we've already lost two engineers, and more are expressing fatigue. we need to make our incident management more sustainable. we're focusing on a few areas: drastically reducing alert noise (about 85% of our alerts are non-actionable or self-correcting), improving our runbooks to enable quicker resolution, and investing in self-healing mechanisms for common issues. the goal is to shift from reactive firefighting to proactive prevention and automation. it's a marathon, not a sprint, but we need to see progress quickly. what strategies have you found most effective for combating on-call burnout in small or medium-sized teams? beyond the obvious (better alerting, runbooks), are there specific tools, cultural shifts, or process changes that made a significant difference in improving on-call sustainability and engineer well-being?