Case Study Commerzbank

Country:     Germany
Industry:     Banking
Solution:     BCV5




The merger with Dresdner Bank left Commerzbank supporting computing services for more than 15 million clients using two separate IT environments. The need to bring all of this Db2 data under one common umbrella in an efficient and cost effective manner became the most critical task for the bank. BCV5 proved to be the right tool for the job.

Mission: With over 62,000 employees and a balance sheet in excess of 844 Billion Euros, Commerzbank ranks as one of Germany’s most important financial institutions. With the Dresdner Bank acquisition it now boasts that one out of every five German’s is an account holder. Founded in 1870 the enterprise, located in Frankfurt/Main, now provides financial products to over 15 million clients.

Commerzbank offers private and corporate customers a broad and attractive line of financial products on a worldwide scale. In 2009 the Dresdner Bank was the third largest bank in Germany and employed about 21,000 professionals. Since, the merger Commerzbank has maintained a positive course in becoming one of the European community’s leading banks.

Mergers have wreaked operational havoc for many well known enterprises. Bringing together two cultures along with their data and procedural orientations is a monumental task. Commerzbank has taken a proactive approach in dealing with the worldwide financial crisis while still implementing a very successful Growing Together Integration Project. The goal is to end with an organization with one culture, one data reservoir and one IT operation.

The schedule calls for the Dresdner Bank data center to go offline in mid-2011. Commerzbank’s planning process initiated in 2009 resulted in a tight schedule, a strategy for merging data and systems, and created a lean budget with which to accomplish the feat. Few if any of the bank’s employees are unaffected by this ambitious project.

The challenge for the IT staff of these two institutions is far from simple. The banks address different lines of financial business needs. The database engine that stands behind these applications, Db2 for z/OS, became a very important part in that process. The database professionals, the application developers and also quality assurance needed to develop a very close cooperation to create new data models and change or enhance existing models for a total amount of data in the area of 40 TB. Testing demands dominate a good deal of this projects timeline and cost. Refreshing the Db2 test data is repeated many times and needs to be done quickly, with minimal resource usage and as automated as possible.

The integration of clients from both of these companies required merging some very large databases. The need to extract live data from production without impacting operational windows was critical. The original estimates revealed that at least 16 TB of Db2 data would need to be extracted repeatedly for testing purposes. The problem revealed itself when it proved impossible to accomplish this task in a timely manner using the UNLOAD/LOAD utility or other popular 3rd party vendor tools.

After an extensive search and evaluation process they opted for UBS Hainer’s product BCV5. The decision was easy because this product offered the best functionality, performance, reliability and ROI in the marketplace. During a trial period,Commerzbank was impressed by the ability to bring down their average copy time from 16 days to 2 days.

Solution: BCV5 was specifically designed to copy large amounts of data quickly and efficiently within a Db2 group or from sysplex to sysplex. It is commonly used to copy test data from a production environment to environments that serve quality assurance, acceptance testing and throughput capacity loading. BCV5 excels at refreshing preproduction area databases. Its high efficiency rating and low CPU time consumption allow it to be used both in ad-hoc daily work and on a nightly basis, without impacting critical operational windows.

BCV5’s multithreaded copy facility boosted performance without needing any third party hardware assists. Compared to common copy methods BCV5 saves up to 90% in replication time, resource consumption and staff effort. Days of copy time are cut to hours, and hours are reduced to minutes. BCV5 met all of Commerzbank’s requirements. It has made it possible to consistently provide fresh test data in a timely manner over night. Weekend overtime processing was all but eliminated, and much of

it is now executed on autopilot during weekday nightly runs. Their staff have confirmed that tasks that previously took one to two weeks to complete are now relegated to scheduler tasks, and that considerable specialists have been freed up to perform other work.

Another challenge was the “living” target environment. The new applications and the different data models of Dresdner Bank required constant modifications to the data structures, such as adding new columns to existing tables, or changing data types. BCV5’s integrated structure compatibility check along with BCV5’s structure compare component (SC) also solved this problem. BCV5 has an intelligent mechanism for detecting the most efficient way to move data while observing constraints that are caused by the different structures of the source and target environment.

A large number of databases, combined with different source and target environments, required the creation of a vast amount of copy tasks. Although BCV5 comes with an easy to use ISPF interface that allows the creation and management of these copy tasks, the sheer number of them called for a more elegant solution, namely the batch interface (BI) component. A database manager of the project praised this component’s capabilities, stating that the creation of such an enormous number of copy processes could not be accomplished in the given time without it. Many different sets of objects and databases required special attention and thousands of copy variants had to be setup.

There was also a requirement to freeze the data and structures captured in a first step for processing at a later time, and this backup had to be repeatedly restored. In contrast to BCV5’s straight copy method, the Ice Box option (IB) of BCV5 makes it possible to save both data and structure for multiple use in one or different target environments. It can store monthly or yearly breakpoints that can be restored as a functional environment or moved to a different operational environment and still work. This results in restoring the environment as it existed at a certain point in time along with the previously saved data. Commerzbank now uses Ice Box to build a shadowplex. Data from production is copied to the shadowplex for later processing. This allows them to get fresh data at any time from this archive without impacting the production environment. It provides a stable source of test data that can be used to refresh many different testbeds.

Conclusion: Commerzbank needed a tool that met their requirements, in BCV5 they found a product that exceeded all expectations and had value well beyond the life of the project for which it was acquired. It has become one of the data centers most used products in day to day operations involving the copying of Db2 data. As a result of this success, UBS Hainer has been engaged by Commerzbank to work on an ongoing basis to further optimize and automate copy procedures. [/slider]