Oracle Fusion Cloud is rolling out a new architecture for bulk data extraction, and if you're running BICC pipelines today, it's worth understanding where this is headed.
RODS (Read-Optimized Data Store) is a continuously refreshed replica of Fusion's transactional data, built on Autonomous AI Lakehouse technology. Oracle GoldenGate replicates changes from the live Fusion database into RODS in near real-time. The point is separation: extraction workloads run against this isolated copy instead of competing with live business transactions on the production database.
The Data Extraction Tool is the customer-facing application built on top of RODS. Think of it as BICC's eventual successor — not a reporting tool like OTBI or BI Publisher, but a bulk extraction framework for feeding downstream systems (data warehouses, integrations, lakehouses). Key things to know:
- Query language: BQL (Business Object Query Language) replaces BICC-style extract definitions. An AI-assisted Query Transformer Agent can help convert existing SQL/BIP logic into BQL.
- Output: CSV (default) or JSON, plus a metadata file, delivered to UCM — not natively pushed to an OCI bucket the way BICC can be.
- Execution: synchronous for small/immediate requests, asynchronous/batch for larger scheduled extracts.
- Access: only through Fusion's Redwood UI or REST APIs (IDCS OAuth-secured) — there's no raw database connection to RODS.
- Coverage: growing quickly each release (1,100+ objects and counting), but not yet a 1:1 replacement for every BICC extract.
For teams migrating from BICC: the swap only replaces the extraction hop. Getting files into your own OCI bucket for downstream ADW loads currently requires a small custom pipeline (Batch API → download → OCI SDK upload) rather than native storage cloud delivery. Everything downstream — staging, transforms, dimensional models — stays unchanged since output is still CSV.
Bottom line: still early-release and opt-in (requires enabling via Setup Manager + a Support Request), so a phased pilot — not a wholesale cutover — is the sensible path for now.
For the full deep dive — architecture diagrams, execution flow, and a walkthrough of migrating a real SCM/Financials BICC pipeline — see the companion long-form post ( https://tinyurl.com/57tfnxve)
