Automated Database Failover
2. Some customers have implemented a manual failover scheme requiring an application restart. Upon a database outage, application downtime could last from 20 minutes to 3 hours, resulting in a loss of potential revenue for the Enterprise.
The Heimdall’s automated database failover solution avoids the above scenarios for maximum uptime. The key benefits of the Heimdall approach are the following:
1. Persisted Connection Failover: Most solutions today (e.g. F5 Networks, AWS Elastic Load balancer) are at the TCP/DNS layer, resulting in connection timeouts and reconnects. Upon a detected database failure, the Heimdall Proxy queues up front end connections and gracefully connects to the standby database instance, reducing application errors and exceptions. Failover is transparent to the end user.
2. No application or database code changes: Scaling out the database tier always requires rewriting the application; not anymore with Heimdall Data.
3. SQL-aware read/write split for complex routing to different geographies or data sources.