Database Connection Time rules fire frequent alarms (4296109)

Critical notifications from the Database Connection Time rule. Connections on the database are not interrupted or failed.

Example:

• High Database or host load - example: High Memory utilization by the DBMS or other services, or when batch or timed processes spike I/O on the host or instance.
• Poor network performance - example: Ping responses higher than 10ms; slow WAN links between components, specifically in a GEO or WAN clustering configuration where hosts are in different locations for Disaster Recovery or HA purposes.
• Missing or misconfigured Database permissions preventing efficient collection of metrics and causing timeouts on the connection.

Recommendations

• Increase RAM or CPUs on Host
• Decrease Memory usage by applications
• Add additional Network adapters for connections, and configure NIC teaming if possible
• Ensure the required permissions are set on the host and instance to  allow collections to occur unhindered.
• Increase Foglight agent connection timeouts(does not always help as this just provides more time for Foglight monitoring collections to complete)

1. Check Database agent(s) to modify on the DB dashboard
2. Click [Agent Administration]
3. Click the "Alarms" tab
4. Find the "Database Connection Time" Alarm
5. Modify the "Warning" and "Critical" values to the higher preferred response times which prevent or minimize alarms, but is within a reasonable range
6. Click Set and Save Changes
7. To verify, clear the current alarms and see if the new thresholds prevent unexpected alarms