- Newest
- Most votes
- Most comments
🚨 The timeouts you experienced every 300 seconds could be related to your application's connection handling, or possibly some other configuration or infrastructure issue. 🤯 It's difficult to pinpoint the exact cause without more information about your application, infrastructure, and connection settings.
It is normal for the RDS listener to resolve to two IP addresses when you have configured a highly available (HA) SQL Server DB cluster. The RDS listener is designed to provide a single, reliable endpoint that can transparently fail over between the primary and secondary database instances in the HA setup.
💡 When your application connects to the RDS listener, it may initially connect to the primary node. However, if the primary instance becomes unavailable for any reason (e.g., maintenance, failover, etc.), the RDS listener will transparently redirect the connection to the secondary instance. This failover process can take up to
60-120
seconds, during which time your application may experience timeouts.
To handle the potential failover scenarios gracefully, your application should be designed to:
- Retry connections automatically if it encounters a timeout or connection error.
- Implement appropriate connection pooling and retries to mitigate the impact of failover events.
- Use connection strings that reference the RDS listener endpoint, rather than hard-coding the primary instance address.
Relevant content
- Accepted Answerasked 8 months ago
- Accepted Answerasked 5 years ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 10 months ago
- AWS OFFICIALUpdated 5 months ago