Oracle Fusion BICC incremental extracts currently use the following selection logic: Last Update Date >= Last Extract Date
There is no upper-bound condition to restrict records updated after the extraction job has started. As confirmed by Oracle Support, this is the current product design and an Idea Lab enhancement is required if this behavior is to be changed.
Current Issue: When an incremental BICC extract is running, business users and integrations continue updating transactions. If a record is updated after the extraction job has started, it can still be picked up by the current extract. Since the next incremental extract starts from the previous job's submission timestamp, the same record satisfies the incremental condition again and is extracted a second time.
This results in:
- Duplicate records across consecutive incremental extracts.
- Additional downstream deduplication logic.
- Data reconciliation issues.
- Increased processing effort for integrations and data warehouses.
- Operational impact for customers running frequent incremental extracts in live environments.
This behavior is especially problematic for high-volume objects such as JournalLineExtractPVO and JournalHeaderExtractPVO, where transactions are continuously updated throughout the day.
Requested Enhancement
Provide an optional enhancement to BICC incremental processing to limit extracted records to those updated before the extraction job began.
Current behavior:
Last Update Date >= Last Extract Date
Requested behavior:
Last Update Date >= Last Extract Date
AND
Last Update Date < Job Submission Time
Alternatively, introduce a configurable extract preference (for example, Use Job Submission Time as Upper Bound) that customers can enable for incremental extracts.
Business Impact:
Customers operating 24×7 production environments cannot realistically suspend user activity or integrations during BICC execution windows. The current behavior causes duplicate journal lines in downstream systems and requires custom workarounds. Adding an upper-bound filter based on job submission time would significantly improve the robustness and predictability of BICC incremental extracts while reducing operational overhead.
