Database Decoupling

Zero-Loss DB2 to AWS Migration

Transforming Legacy IBM i Data to Amazon Aurora

Stop letting flat, non-relational DDS architectures dictate your integration capabilities. Our automated IBM i database migration pipeline securely translates legacy DB2/400 physical and logical files into highly scalable, fully relational Amazon Aurora PostgreSQL schemas — with zero manual rewriting.

The Challenge — Breaking Free from Legacy DB2 Constraints

In a traditional AS/400 environment, the database is tightly coupled to the operating system and applications. Our WMSPOCv3 case study highlighted the core bottlenecks:

Non-Relational Structures

Classic DDS Physical Files (PF) often lack enforced foreign keys or referential integrity at the database level.

Legacy Data Types

Date columns are frequently stored as packed DECIMAL(8,0) fields rather than standard SQL DATE or TIMESTAMP formats.

Proprietary Concepts

The IBM i "Multi-Member file" concept has no direct equivalent in standard RDBMS systems.

Cryptic Naming

10-character field limits (e.g., ITITMCD, ITQOH) inherited from rigid RPG conventions limit data readability and BI tooling.

The Solution — Automated Schema Conversion

Using the AWS Blu Age (Gapwalk) toolchain, we automatically generate fully PostgreSQL-compatible DDL from your original DDS source files.

How We Map Your AS/400 Data to AWS

Physical Files (PF)

Converted to CREATE TABLE with native PostgreSQL partitioning.

Packed DECIMAL Dates

Preserved as DECIMAL(8,0) PLUS automatic _LASTUPDATE_TIMESTAMP and _CREATION_DATE audit columns added to every table.

Multi-Member Files

Converted to PostgreSQL PARTITION BY RANGE — preserving the IBM i concept in a cloud-native way.

DDS Logical File Access Paths

Converted to explicit PostgreSQL Composite Indexes (CREATE INDEX) — every access pattern preserved.

No FK Constraints? Fixed.

Full referential integrity (FOREIGN KEY constraints) added automatically — something the original IBM i schema never had.

WMSPOCv3 — Real Case Study Results

3 core business tables were fully migrated with zero business logic changes. 15+ JICS runtime tables (IBM i CICS equivalent) also auto-migrated — with full referential integrity added.

DB2 Table	Business Purpose	Key Fields
WMSITEMPF	Item Master	Item Code, UOM, Bin Location, Qty On Hand, Qty Allocated
WMSWHSTPF	Warehouse Master	Warehouse ID, Address, Zones, Aisles, Bays, Capacity (cu ft)
WMSLEDGPF	Transaction Ledger	Tran#, Type, Status, Created By, Date

Legacy AS/400 (DDS)

     A          R WMSREC
     A            ITITMCD       10A         COLHDG('ITEM CODE')
     A            ITDESC        50A         COLHDG('DESCRIPTION')
     A            ITUOM          2A         COLHDG('UOM')
     A            ITQOH         11P 3       COLHDG('QTY ON HAND')
     A            ITLSTDT        8S 0       COLHDG('LAST REC DATE')
     A          K ITITMCD

Target AWS (Aurora PostgreSQL)

CREATE TABLE schema_wms.wmsitempf (
    ititmcd VARCHAR(10) NOT NULL,
    itdesc  VARCHAR(50),
    ituom   VARCHAR(2),
    itqoh   NUMERIC(11,3),
    itlstdt NUMERIC(8,0),
    _creation_date TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    _lastupdate_timestamp TIMESTAMP,
    PRIMARY KEY (ititmcd)
) PARTITION BY RANGE (ititmcd);

Notice the automated injection of standard audit timestamps and native PostgreSQL partitioning logic without losing the core schema layout.

The Target — Amazon Aurora PostgreSQL

Fully Managed

No DB2/400 DBA overhead — AWS handles patching, backups, and failover automatically.

Auto-Scaling Storage

Up to 128 TB without any manual DBA intervention.

Multi-AZ Replication

Built-in high availability — 6-way replication across 3 Availability Zones.

5x Throughput

5x the throughput of standard PostgreSQL at comparable cost.

Serverless Option

Aurora Serverless v2 — pay-per-use for dev/test environments.

Native REST Ready

Java Spring Boot services connect directly via JDBC/JPA — no middleware layer required.

See How Your DB2 Schema Maps to Aurora

Get a free automated schema assessment for your DB2 environment.