The SaaS Paradigm Shift

For decades, Oracle RMS was an on-premise behemoth. Retailers bought the software, installed it on their own servers (or hosted it in a private data center), and had absolute control over the database. DBAs could SSH into the server, run SQL*Plus against production, create indexes, add columns, and modify PL/SQL packages at will.

With Oracle Retail Merchandising Foundation Cloud Service (MFCS), Oracle shifted RMS to a true Software-as-a-Service (SaaS) model. You no longer buy the software; you rent access to it. Oracle manages the servers, the database, the upgrades, and the security.

This is not just a deployment change — it is a fundamental operating model change that affects every role in the retail IT organization.

MFCS Architecture

┌──────────────────────────────────────────────────────────────────┐
│                    ORACLE CLOUD INFRASTRUCTURE (OCI)             │
│                                                                  │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────────────┐  │
│  │  ORACLE JET  │  │    ORDS      │  │  AUTONOMOUS          │  │
│  │  Web UI      │  │  (REST API   │  │  DATABASE            │  │
│  │              │  │   Gateway)   │  │                      │  │
│  │  Managed by  │  │  Managed by  │  │  ┌────────────────┐  │  │
│  │  Oracle      │  │  Oracle      │  │  │  RMS Base      │  │  │
│  └──────────────┘  └──────────────┘  │  │  Schema        │  │  │
│                                       │  │  (READ-ONLY)   │  │  │
│  ┌──────────────┐  ┌──────────────┐  │  ├────────────────┤  │  │
│  │     POM      │  │     BDI      │  │  │  Custom Ext    │  │  │
│  │  (Batch      │  │  (Data       │  │  │  Schema        │  │  │
│  │  Scheduler)  │  │  Integration)│  │  │  (YOUR ACCESS) │  │  │
│  │  Managed by  │  │  Managed by  │  │  ├────────────────┤  │  │
│  │  Oracle      │  │  Oracle      │  │  │  APEX          │  │  │
│  └──────────────┘  └──────────────┘  │  │  Workspace     │  │  │
│                                       │  │  (YOUR ACCESS) │  │  │
│                                       │  └────────────────┘  │  │
│                                       └──────────────────────┘  │
└──────────────────────────────────────────────────────────────────┘

What You Lose

Moving to MFCS requires a massive mindset shift for legacy RMS developers, primarily because you lose direct access to the database.

The New Cloud Tools

To compensate for the locked-down environment, Oracle provides cloud-native alternatives:

Process Orchestration & Monitoring (POM)

POM replaces external batch schedulers like Control-M and Autosys. It provides a web-based interface for:

Key POM Features

• Job Scheduling: Define batch jobs with dependencies, execution windows, and retry policies
• Dependency Management: Job B runs only after Job A completes successfully — with automatic hold/release
• Monitoring Dashboard: Real-time status of all running, queued, and completed batch jobs
• Error Handling: Failed jobs are held for review with detailed error logs
• Parallel Execution: Configure multiple threads for parallel batch processing
• Calendar Integration: Schedule jobs based on business calendars (skip holidays, run on fiscal period boundaries)

POM vs. External Schedulers

The biggest adjustment for operations teams is losing the flexibility of external schedulers. In Control-M, you could create arbitrary job sequences mixing RMS batches with non-RMS processes. In POM, you are constrained to Oracle's predefined batch flows — though you can add custom jobs to the sequence.

Bulk Data Integration (BDI)

BDI (Bulk Data Integration) is the cloud replacement for direct database extracts and loads. Instead of running SELECT INTO against production or using SQL*Loader, BDI provides:

• Scheduled Extractions: Pull data from RMS base tables into flat files or JSON payloads
• Incremental (Delta) Support: Extract only changed records since the last run
• Pre-Built Extractors: Oracle ships BDI extractors for all major RMS entities (Items, Locations, POs, Inventory)
• Cloud Object Storage: Extracted data is deposited in Oracle Cloud Object Storage buckets for downstream consumption
• REST API Triggers: BDI jobs can be triggered via REST API for on-demand extractions

# Trigger a BDI item extraction via REST
curl -X POST \
  "https://mfcs.oracleretail.com/bdi/extract/items" \
  -H "Authorization: Bearer ${ACCESS_TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{
    "extractType": "DELTA",
    "sinceDate": "2026-06-07",
    "outputFormat": "JSON",
    "destination": "cloud-storage://bucket/items/"
  }'

Custom Schema Provisioning

In the cloud, you cannot create your own database schemas. Instead, you request Oracle to provision a custom extension schema for you:

1. Submit Service Request (SR) to Oracle Support requesting a custom schema
2. Oracle provisions the schema with appropriate grants:
   • SELECT access on most base RMS views (not tables)
   • CREATE TABLE, CREATE VIEW, CREATE PACKAGE privileges in your custom schema
   • APEX workspace linked to your custom schema
3. You create custom objects in your schema:
   • CUST_ITEM_ATTR — custom item attributes table
   • CUST_LOC_ATTR — custom location attributes table
   • PKG_CUST_ITEM_PROCESSING — custom PL/SQL package
   • APEX applications for data entry and reporting

Operating in the Cloud

Daily Batch Monitoring

Instead of tailing log files on a server, cloud operations teams monitor batch execution through POM's web UI:

• Green = job completed successfully
• Yellow = job running
• Red = job failed — click for error details and stack trace

Performance Troubleshooting

Without AWR reports and direct execution plan access, performance troubleshooting follows a different workflow:

1. Identify slow-performing operations through POM execution times or user reports
2. Use APEX SQL Workshop to analyze your custom queries
3. For base table performance issues, open an Oracle Support SR with the query hash and execution time
4. Oracle's DBA team investigates and applies fixes (statistics gathering, plan management)

Upgrade Management

Oracle pushes updates automatically:

• Monthly patches: Bug fixes and minor enhancements (mandatory, automatic)
• Quarterly releases: Feature releases with new functionality (scheduled maintenance window)
• Annual major releases: Significant feature additions (coordinated with the retailer for testing)

You must maintain a regression test suite that runs after every Oracle patch to verify that your custom extension code still works.

Best Practices