Database Connection Pool Sizing Runbook Builder

You are a senior backend engineer and database reliability reviewer. Build a connection pool sizing runbook for:

Database engine and version: [Postgres, MySQL, SQL Server, Aurora, Cloud SQL, RDS, self-hosted]
Application stack: [Node, Rails, Django, Java, Go, serverless, workers, Kubernetes]
Traffic pattern: [steady OLTP, bursty API, cron jobs, background workers, multi-tenant SaaS]
Current pool settings: [pool size, timeout, idle timeout, max lifetime, per-process settings]
Deployment topology: [pods, instances, serverless concurrency, read replicas, PgBouncer, proxy layer]
Database limits: [max_connections, reserved connections, memory, CPU, connection overhead]
Symptoms: [timeouts, exhausted pool, idle connections, slow queries, lock waits, failover issues]
Observed metrics: [active connections, wait time, queue depth, query latency, CPU, memory, throughput]
Constraints: [no downtime, managed DB limits, peak events, compliance, low risk tolerance]

Create:
1. Current-state connection math across every app, worker, and replica
2. Safe sizing formula with assumptions and margin of safety
3. Decision tree for app pool tuning, DB max connection changes, PgBouncer/proxy use, and query optimization
4. Load test plan that validates wait time, saturation, and database health
5. Monitoring dashboard and alert thresholds
6. Rollout plan with staged config changes and rollback steps
7. Incident triage checklist for connection exhaustion
8. Follow-up actions for serverless concurrency, long transactions, and idle connection cleanup
9. Executive summary for reliability and capacity planning

Favor reducing uncontrolled concurrency before raising database limits.