In Oracle Integration Cloud (OIC), retry handling plays an important role in improving integration reliability when temporary failures occur during invoke calls. Oracle currently documents two different retry approaches that developers should understand while designing integrations.
1. Project-Level Retry Logic
Oracle Integration provides a configurable retry mechanism for invoke connections within Projects. This retry functionality is not automatically enabled for every integration by default. Once configured, developers can customize:
Number of retry attempts
Delay interval between retries
Default Configuration
By default, OIC provides:
3 retry attempts
5-second retry interval
This retry mechanism applies broadly to invoke connection faults, including:
Business faults
Runtime faults
This helps integrations recover automatically from temporary endpoint failures without immediately moving to faulted state.
Use Case
Suppose an external REST/SOAP endpoint becomes temporarily unavailable or experiences network instability. Instead of failing immediately, OIC retries the invoke operation automatically based on configured retry settings.
Oracle Documentation
2. REST Adapter Built-In Retry Handling
Apart from project-level retry logic, the REST Adapter itself contains built-in retry handling for specific error scenarios. This behavior is adapter-specific and implemented internally by Oracle.
The REST Adapter automatically retries for:
HTTP 504 errors
HTTP 502 errors (GET operations only)
HTTP 401 errors
SSL-related retryable exceptions
Adapter-classified retryable exceptions
For these scenarios, the REST Adapter performs up to 3 retry attempts automatically.
Important Note
This retry behavior is currently documented specifically for the REST Adapter. Oracle documentation does not confirm the same built-in retry implementation across all adapters.
Best Practice
Oracle recommends enabling instance retries when invoking REST endpoints that may encounter:
401 errors
429 throttling errors
50x server errors
This improves resiliency when working with unstable or rate-limited external services.
Oracle Documentation
Conclusion
OIC provides retry handling at multiple levels:
Project-Level configurable retry mechanism
REST Adapter built-in automatic retries
Understanding both mechanisms helps developers design more resilient integrations and reduce failures caused by temporary network or endpoint issues.
When designing enterprise integrations, retry configuration should always be planned carefully to balance:
Reliability
Performance
Duplicate transaction prevention
External API limitations
Proper retry handling can significantly improve integration stability in real-world production environments.
No comments:
Post a Comment