What indicates a TCP connectivity issue for a process?

A TCP connectivity issue for a process is often indicated by a significant rise in new connection failures. The threshold set at more than 3% new connection failures is critical because it suggests that the system might be experiencing instability or problems with network configuration, firewall settings, or underlying infrastructure issues. This percentage is a middle ground that reflects a significant enough anomaly in connection attempts that requires further investigation.

Monitoring new connection failures assists in assessing the health of the service and ensuring users can reliably access resources. If the failure rate exceeds this threshold, it's a signal that there may be systemic problems that need to be addressed before they lead to more severe operational issues. This enables proactive troubleshooting, ensuring that potential service disruptions can be mitigated early.

The other thresholds set at 1%, 2%, and 5% may not capture problems as effectively. A lower percentage may overlook issues that warrant immediate attention, while a higher one might indicate an already severe disruption before action is taken. Therefore, the 3% benchmark represents a balanced and insightful threshold for diagnosing TCP connectivity issues.