Why a Data-First Mainframe to Cloud Strategy is the Best Option for the Enterprise
Mainframe modernization projects continue to be hindered by unclear objectives and fragmented executive support.
Most modernization initiatives occur on long timelines, with uncertain impact that can harm business outcomes. As a result, IT leaders have taken a piecemeal approach to their modernization efforts due to competing objectives of reliability, stability, and security against business agility and flexibility.
The result is a trend towards a strategy of small, incremental, but always ongoing changes where the benefits of the changes are almost imperceptible. One approach that is gaining traction is migration of mainframe data to the cloud.
3 pitfalls executive leaders encounter when managing the risks of mainframe to cloud migrations
Unfortunately, this slow and steady approach comes with several pitfalls for IT leaders.
For starters, the sheer size and complexity of enterprise portfolios make prioritizing modernization tasks difficult, if not impossible. Many interconnections and dependencies that were not documented in the past are either poorly understood or lost entirely today.
Although IT leaders may strive to be comprehensive, the reality is portfolios are too big to perform an inventory of all applications in a cost-effective way. Complicating matters further is the divergent views of stakeholders on management goals, with IT managers usually taking an isolated ‘application’ view and business managers often taking a broader ‘capability’ view that can span applications, platforms, and ecosystems. This makes it difficult to build a clear roadmap that satisfies everyone.
Second, modernization projects are often viewed as a huge, monolithic undertaking by IT leaders, creating headwinds in the planning, approval, and implementation phases of the project. Because these projects are viewed as a type of Goliath that must be slayed, it is all too rare to find roadmaps that successfully prioritize modernization tasks into smaller, more manageable projects. Failure to compartmentalize project deliverables increases the probability of failure, which in turn builds a sense of reluctance around implementation.
Third, the combination of divergent priorities and unclear deliverables often results in unfinished projects that fail to make meaningful progress towards mainframe modernization, and culminate in an ongoing barrier between the mainframe and cloud ecosystems, leaving the mainframe silo intact as a standalone, non-participant. According to Gartner, IT leaders must pursue a cross-discipline cloud strategy if they wish to fully realize the benefits cloud computing offers. Therefore, those leaders should view a failure to connect their cloud and mainframe as a failure to realize the full potential business value of both.
What is a data-first approach to mainframe modernization?
Rather than embarking on application-centric mainframe modernization projects with indefinite timelines and little cost certainty, Infrastructure and Operations leaders looking to simplify and streamline their path to mainframe modernization should adopt a new strategy that begins by storing and managing mainframe data in cloud storage solutions.
By taking this ‘data-first’ approach to mainframe modernization projects, leaders can break through the complexity and uncertainty surrounding implementation roadmaps, and instead deliver value quickly while simultaneously building a runway for future improvements.
Model9 makes a data-first approach possible by using its new, patented technologies surrounding mainframe data storage, management, and transformation. Model9 allows companies to move mainframe formatted data into the cloud faster and more cost effectively than ever before via zIIP engines and TCP/IP, and our Cloud Data Manager for Mainframe allows companies to easily manage a bi-directional flow between their mainframe and cloud ecosystems. Once in the cloud, our transformation engine can transform mainframe data into open formats for use with cloud applications.
Infrastructure and Operations leaders can view a data-first approach as a mindset shift away from traditional mainframe modernization strategies. Instead of committing to replacing legacy storage systems, or implementing uncertain lift/optimize or replacement application strategies in the cloud, leaders can make smaller-scale changes that liberate siloed mainframe data for use in the cloud.
Below are three reasons why this data-first mainframe to cloud strategy is the best option for the enterprise:
#1: Use ‘data gravity’ to accelerate timelines
Data gravity is the concept that data attracts more data, and encourages the development of applications to use that data. In other words, wherever data is stored, more data will get stored there. And wherever more and more data is stored, applications will be developed to create more business value as the interrelationships between the collected data is analyzed and understood – customer buying behavior, competitive advantage, and productivity improvements can be harvested from aggregated data.
A data-first mainframe to cloud strategy leverages this concept to accelerate timelines. Instead of making the commitment to long and painful processes, companies can immediately start storing and managing their mainframe data in the cloud and let the inertia created move the project forward.
#2: Transform mainframe data into open formats in the cloud, without using mainframe cycles
Traditionally, the transformation of mainframe data into open formats used an ‘extract, transform, load’ process (ETL). This process transformed data into open formats in the mainframe, using mainframe cycles, before uploading it to the cloud. This is a costly process, and one that also diverts mainframe resources that could be used elsewhere. Remember that many mainframe billing algorithms are calculated based on general purpose engine cycles – specialty engine usage is excluded, so transferring workload from general purpose engines to specialty engines is advantageous.
Model9 replaces ETL with a new ‘extract, load, transform’ (ELT) process that extracts mainframe data, still in mainframe format and sends it via TCP/IP into the cloud first using specialty engines Once it is there in the cloud, Model9’s transformation engine transforms it into open formats.
In other words, Model9’s data-first approach means that as more data is migrated into the cloud, that data is in a place where it can be easily transformed into open formats with less friction than legacy transformation processes.
Because the transformation engine makes it so easy to transform data that is already in the cloud, companies are primed to actually start transforming that data at scale.
#3: Easily use mainframe data in your cloud ecosystem
As stated above, the mainframe holds some of the most mission-critical data an enterprise organization has. By driving data towards the cloud where it can be easily transformed, companies are accelerating the movement of that data into their open data cloud ecosystem, where it is made accessible for cloud AI/BI/ML tools to consume it and deliver huge amounts of business value.
In short, companies that can break down their mainframe data silos, liberate their mainframe data, and release it into their cloud ecosystem will have the capacity to monetize it, improve business operations, and more.
A data-first mainframe to cloud strategy is the best option for the enterprise
70% of Fortune 500 businesses have their core business operations powered by mainframe systems. Whether they are looking to perform mainframe modernization projects, or migrate off the mainframe entirely, using a data-first mainframe to cloud strategy to deliver a faster, safer, and easier way to tap into the wealth of data the mainframe holds is a strategic objective that cannot be ignored.
Data first mainframe to cloud migration strategies allow you to build ever-closer connections between the mainframe and the cloud without any application changes. Once this occurs, Infrastructure and Operations leaders can accelerate mainframe modernization timelines.
About the author
