The mainframe is not dead. Most banking, insurance and government institutions still run legacy code on mainframe systems, and they don't plan to replace those anytime soon. Yet such systems aren't immune to the changes brought forth by the cloud revolution. The relationships between cloud infrastructure and mainframes are critical.
Mainframes in the cloud context
Mainframes are, in one aspect, similar to the cloud. They provide computing platforms form many different business systems, and through operating systems such as zLinux, they can be highly scalable. The comparisons also fall short: mainframes cannot provide the latest technological features such as big data, artificial intelligence, autoscaling and advanced monitoring.
Yet mainframes remain popular because they provide advanced performance-tuning capabilities. This quality is why mainframes remain the computing platform of choice for high-end transactions systems.
In some cases, as IT estates migrate their other infrastructure to the cloud and distributed computing, the mainframe is the last 'on-premises' server. Thus, companies prioritise improving the speed, bandwidth and security by which mainframes and cloud environments connect. Those capabilities are very mature, prompting the question: how can we further improve such integrations?
The truth is that, even though the implementations have advanced, the concepts still harken to the early days of cloud. Most mainframe-based enterprises rely on old interfaces created when distributed computing had just entered the market, utilising integration designs available at the time. Though the market has improved integration security, bandwidth and speed, they still rely on yesterday's methodologies.
Modernising mainframes with the cloud
There are new business applications on mainframes and richer ways to connect mainframe and cloud environments. The way to modernise symbiotic mainframe-cloud ecosystems is to enable more mainframe business applications by exposing their functionality via application programming interfaces (APIs).
Using software from mainframe vendors such as Software AG and IBM, we can identify the critical business capabilities of mainframe applications and convert them to services, accessible via APIs through modern TCP/IP application protocols.
Yet, while it's relatively simple to identify such opportunities, creating those service APIs can be much more complicated when you use the wrong approaches. Though the market offers several tools to create such service APIs, there are only two fundamental ways to construct WSDL or REST format APIs, namely program call and background terminal sessions.
- Program calls are popular because they are traditional and easily understood by programmers. Modernisation tools discover COBOL or NATURAL/ADABAS code, converting the copybooks and parameters into a CORBA-like interface definition language (IDL). These are easily converted into Java, DOT NET, Web Services, IBM MQ, and other enterprise components. Such distributed computing environment components request data and services from the mainframe applications, thus preventing the mainframe from consuming the cloud's enterprise services and data. The changes can be minimal to construct user-friendly interfaces with cohesive data objects.
- Background terminal sessions work in one direction – towards the users – and the mainframe applications act as servers. The user's interactive sessions with the terminal screens are executed in the background. Meanwhile, specified screen data is scraped and updated through the Web service interface the user engages (for more on this approach, please read mWtech's release, The legacy dilemma: Transformation without impact). Though it requires more imagination to program the screens' flow, integrating mainframe applications and cloud-based users this way requires almost no changes to the legacy code.
Not all integration houses understand these nuanced yet effective ways to blend mainframe and cloud ecosystems. We humbly position mWtech (middleware technologies) as an expert in the field. We provide these specific integration services to our customers who do have all flavours of enterprise applications in their distributed computing environment.
mWtech helps mainframe owners identify services from mainframe applications and consume these services, whether using the cloud or on-premises choices, via Web APIs. Our extensive experiences, built from years of working with mainframe and cloud integrations, prove that enabling legacy systems with the right APIs can cut modernisation costs more than 10 times.
You don't need to bury your mainframe ecosystem just because the cloud has come knocking. To demonstrate these results and our faith that mainframes and cloud systems can elevate each other, mWtech provides free pre-sales hours to interested companies, demonstrating how easy it is to bring legacy applications to the Web era.
Please contact us for details and queries and allow mWtech to show how our integration methodologies can enhance your environment and cut modernisation costs. You will undoubtedly realise that digital transformation is much more tangible and achievable than you might think, without kicking your mainframe to the curb.
Share