Hi All,
Today, two Additional Polling Engines (APEs) that belong to the same HA pool experienced an issue where they became inaccessible via RDP. Although both APEs had the SolarWinds agent installed and were being monitored from the Main Polling Engine, their status continued to show Up throughout the incident.
During this period, polling from the affected APEs was not functioning, resulting in a significant monitoring impact. However, no SolarWinds alert was generated to indicate the loss of polling capability or the inaccessible state of the polling engines.
As an immediate workaround, a force reboot was performed on both APEs, after which polling resumed normally. We are currently investigating the RCA.
The main concern is that the outage lasted for more than 8 hours without detection, highlighting a gap in our current self-monitoring configuration.
Has anyone encountered similar behavior in the past? Also, could you share recommendations or best practices to strengthen SolarWinds self-monitoring and proactively detect scenarios such as:
- Polling Engine services stopped or hung
- Polling Engine not processing polls despite node status showing Up
- Server inaccessible (RDP/OS issues) while SolarWinds Agent continues reporting
- HA pool members becoming unhealthy
- Polling completion rate degradation
