Treehouse Software Customer Case Study:
Grede Foundries (Now Grede Holdings LLC)
This is the 13th installment in a continuing series of articles featuring tRelational and Data Propagation System (DPS), Treehouse Software's ADABAS-to-RDBMS product implementation, in several "real world" environments.
tRelational/DPS is a robust product set that provides modeling and data migration of legacy ADABAS data into new RDBMS-based platforms. The modeling and mapping facility (tRelational) auto-generates complete RDBMS schemata from existing ADABAS files and allows for easy mapping of ADABAS fields to already existing data warehouse or ERP schemata. After the mapping is completed, DPS can materialize (initially load) and propagate (subsequently keep synchronized) the ADABAS data into the RDBMS without requiring direct access to ADABAS.
tRelational/DPS is the answer for sites looking to completely replace a legacy ADABAS system, as well as for those wanting long-term data transfer and synchronization. A key differentiator of tRelational/DPS from the competition is its long-standing presence in the marketplace. For over 15 years, tRelational/DPS customers have been enjoying this mature, full-featured product.
Whether a site is completely replacing a legacy ADABAS system with an RDBMS-based system as was done at Grede Foundries (now Grede Holdings LLC), or if the need is a long-term data transfer and synchronization solution for data warehousing/Internet/Intranet/ERP applications, tRelational/DPS is the answer.
This article illustrates how Treehouse Software's tRelational/DPS product set and expert consultants helped Grede Foundries migrate its mainframe data to it's new home -- on time and under budget. Grede Foundries chose the Treehouse products for migration of their legacy ADABAS data as part of their plan to move from the mainframe by the end of June, 2010.
About Grede Foundries (Now Grede Holdings LLC)
Grede is a full-service manufacturer and supplier of innovative metal components to the transportation and industrial markets. Headquartered in Novi, Michigan, with sales and technical support throughout the United States and Europe, Grede delivers quality products to a global customer base.
Additionally, Grede's customers look to the company to manage logistics, packaging, and freight to ensure their products arrive on schedule. This requires leveraging global purchases, and working with suppliers to ensure they're meeting Grede's quality and delivery expectations.
A Time For Change
Until recently, Grede had relied on their mainframe database and file environment, which was running ADABAS v7.4.4 and NATURAL v3.1.5. The mainframe applications were to be retired, and the data had to be stored in a queryable format for archival purposes. The target RDBMS for data storage was SQL Server 2008 running on Windows Server 2008 64-bit.
A particularly challenging aspect of the migration was that many of the files were going to be active until a few days before the end of the project, and some files would be active the morning of the last day of the project.
Treehouse Was The Logical Choice
Treehouse had been a long-time provider of remote and on-site DBA services for Grede Foundries, and years ago, Grede was one of the very first tRelational/DPS users.
When the decision was made to retire their mainframe system, the Grede team contacted Treehouse and asked if it was possible to migrate all of their ADABAS data in a very tight time frame.
The migration plan needed to address the volume of files and records that had to be migrated in an extremely short period of time. The Grede mainframe was scheduled to be shut down for the final time the evening of June 30th. This meant that only 3 1/2 weeks were available from the start of the project to migrate 108 ADABAS files containing 63.7 million records. Additionally, 25 files were in use on June 29th, and 5 of those files were still in use the morning of June 30th. This meant that the final conversion of 25 files, containing 24.5 million records, had to be migrated from ADABAS to SQL Server during a 24-hour window.
The migration of the Grede's data was delivered by TSI consultants in four phases:
1. The Predict metadata for each ADABAS file was imported into tRelational for analysis and auto-generation of a relational schema. tRelational batch jobs were then run to build SQL DDL statements (later used to create the relational schema) as well as parameter files for DPS to be used during the data extraction process.
2. The ADABAS files were backed up to an Adasav, and then DPS jobs processed the Adasav using parameters created by tRelational in Phase 1. DPS generated record counts showing the amount of data processed, which were then compared against the Adarep. The DPS output was then FTPed to Windows 2008, split using the DPSSPLIT utility, and bulk loaded into SQL Server. During this process problems with the data, such as unprintable characters, duplicate MU/PE names across ADABAS files, etc. were discovered. Modifications were made to the tRelational data model, and pre-written column and field routines were added to the DPS parameters to handle the problems with the data.
3. The final extraction and load of the ADABAS data into SQL Server was conducted. Phase 3 began on static ADABAS files while Phase 2 continued on dynamic files. Then on the evening of June 29th, Treehouse consultants started the final extract and load of the last 25 ADABAS files.
4. The metadata stored in tRelational was extracted into several SQL Server tables and loaded into SQL Server. These tables show the ADABAS file and field names from Predict as well as the equivalent names in SQL Server.
In the end, the Grede team was very pleased that the migration went so smoothly and the new system was up and running on time, with no problems.
Treehouse Software is Ready to Help with Your ADABAS-to-RDBMS Project
Treehouse Software's tRelational/DPS product set has been the trusted best-of-breed ADABAS data migration solution by hundreds of customers since the mid-1990's, and is a proven solution for moving ADABAS data to relational databases, either for coexistence with, or full migration to new systems.