Hi,

This chart shows the mean number of successful logons (active sessions) over time. Here's how to interpret it in the context of your connection pool configuration.
Baseline behavior: The chart shows a steady baseline of approximately 120–130 concurrent sessions during off-peak hours (evenings and nights). This aligns with your jdbc.MinLimit=25 setting — the minimum connections are kept alive, along with other persistent connections such as monitoring agents, background jobs, and batch processes.

• Peak 1 (~06:00–07:30 on Apr 7): Logons spike to roughly 370, which is notably close to your jdbc.MaxLimit=350. This likely represents the morning login rush when users and application threads start competing for connections. Because the count exceeds 350, some sessions may be coming from sources outside the JDBC pool (e.g., direct DB links, admin tools, or scheduled jobs). If users experienced slowness during this window, connection pool exhaustion is a likely cause.
  
• Peak 2 (~13:00–14:00): A secondary spike to around 270, probably corresponding to post-lunch activity. This is well within your pool limits and should not cause issues.
  Evening decline: After 18:00, sessions gradually return to baseline. Your InactivityTimeout=1200 (20 minutes) is working correctly — idle connections are being released back to the pool and eventually closed.
  When this chart is useful:
  

You should monitor this chart for capacity planning (if peaks consistently approach 350, consider raising MaxLimit and the database PROCESSES/SESSIONS parameters), performance troubleshooting (correlate slow response times with logon spikes to identify pool exhaustion), and leak detection (if the count climbs steadily without dropping, it may indicate application code that fails to close connections properly).

Recommendation: Check your application server logs for connection wait timeouts during the morning peak. If they exist, increasing MaxLimit or optimizing connection usage in your application would be advisable.

Regards,