Optimizing Database Connection Pool

Use this skill to reason about connection-pool sizing and behavior before changing configuration blindly.

Clarify First

• Application concurrency and request shape.
• Database connection limits and competing clients.
• Typical query latency and long-running outliers.
• Whether the workload is bursty, steady, or background-heavy.
• Current symptoms such as queueing, timeouts, saturation, or idle waste.

Tuning Principles

• Pool size should reflect concurrent useful work, not peak thread count.
• Long-running queries often matter more than raw pool size.
• Timeouts should fail noisy callers before the whole system backs up.
• Shared databases need headroom for maintenance and non-app traffic.
• Fixing query quality can outperform raising connection counts.

Review Output

• Likely bottleneck explanation.
• Pool sizing guidance with assumptions.
• Timeout or backpressure suggestions.
• Query-pattern or workload issues worth investigating.
• Risks of over-allocating connections.

Common Mistakes

• Setting pool size equal to server worker count by habit.
• Ignoring background jobs and admin traffic.
• Treating connection exhaustion as purely a config problem.
• Hiding slow queries with ever-larger pools.