Johannesburg, 14 Apr 2023
Major financial institutions running complex, hybrid legacy mainframe environments are understandably reluctant to take a big bang approach to migrating their databases. However, they also need their data available in modern cloud environments for agility, analysis, new product development and to take advantage of modern application tools and packages.
The solution is to synchronise and replicate their databases to modern platforms in real-time, says Sam Selmer-Olsen, Managing Director at Bateleur Software, the South African distributor of Treehouse Software and tcVISION.
Selmer-Olsen says many large sites, particularly in the financial services sector, have developed large and complicated environments through decades of evolution, the acquisition of other companies and technology moving on. “They end up with a mix of databases including the more unusual, legacy databases such as IBM’s IMS DC/DB, DL1, or even Datacom/DB. It’s not a simple matter to move away from them, especially when there are potentially thousands of programmes written around them.”
He notes that some of the legacy systems still in use were developed 30 to 50 years ago and have been updated, changed and enhanced so much over the years that few, if any, team members are fully versed in the system design and dependencies. They typically have many thousands of application programs that access and update these old database structures.
“Moving to new environments in a big bang approach could mean multiple applications need to be rewritten because they all rely on the same integrated databases. In financial institutions, for example, home loans, vehicle finance, current accounts and business transactions all rely on and work on the same integrated database. Changing everything is a huge risk because the systems don’t just support their business – they are the business,” he says.
Selmer-Olsen says replication to a more modern infrastructure allows organisations to access the data using more modern application tools and packages, without having to move away from their legacy databases. “For example, we have recently seen interest in South Africa for IMS/DB real-time replication,” he says.
He notes that while tools exist for real-time replication, most do not support a mix of modern and legacy databases. “If you have a mix, which is very common, it becomes more complicated. The legacy databases tend not to be able to offer real-time data replication, and if they do it would be confined to their own technology platform and not to the myriad cloud and data streaming options that are available today. ETL batch replication to extract, transform and load the data on the other side is typically done only weekly, or even monthly, where massive volumes of data are involved. This means data for analytics and reporting are a week or more out of sync with transactional data. Modern organisations need to be able to extract from multiple sources and use one tool to replicate and synchronise it to distributed systems and cloud platforms in real-time,” he says.
“Real-time replication, particularly with bi-directional replication, solves the big bang problem and allows for a gradual transition to modern systems. tcVISION, a cross-system solution for real-time bidirectional data synchronisation and replication based on changed data, offers all the functions required for reliable and efficient synchronisation in heterogeneous environments. Treehouse, and developers BOS, have found significant new business with clients that have a mix of more modern databases like DB2, but also still have IMS/DB or DL1.”
tcVISION bi-directionally synchronises data in real-time between systems such as IBM DB2, Adabas, VSAM, IMS/DB, CA IDMS, CA DATACOM, or sequential files, and cloud and open systems targets such as AWS, Google Cloud, Microsoft Azure, Kafka and PostgreSQL.
Among its many benefits, tcVISION enables cost reduction through the relocation of data exchange processes from the mainframe to more cost-efficient platforms (eg, Unix) and reduces data transfer volumes through a focus on changed data. It also helps organisations overcome the challenge of scarce legacy mainframe skills and enables automatic processing of historically grown data structures, overcoming ‘data lock-in’.
Share