How do I handle system outages?
Your client application must be designed to handle system outages, such as:
- Scheduled outages: Oracle Responsys undergoes maintenance downtimes on a monthly or bi-monthly schedule. During scheduled maintenance downtime, Oracle stops the Web Services server. Unless your client application is using an Automatic Failover for Transactional Messaging (AFTM) account, all Web Services API requests will fail during the maintenance period when your account’s system is out of service.
- Unscheduled outages: All Responsys users receive an “Oracle Responsys Customer Bulletin” when there are system issues. These might be due to performance issues as well as unscheduled system outages.
During a system outage, attempts to make API calls will result in your receiving an HTTP response error similar to the following:
Remote Exception : Transport error: 503 Error: Service Temporarily Unavailable
When it detects this error response, your client application code should take appropriate actions, which may include alerts to support staff, integration job queuing (because the Responsys system does not maintain a queue), and scheduled re-tries.