Skip to main content

The mainframe “cloud journey” is here

(Image credit: Shutterstock / issaro prakalung)

The phrase “where you sit is where you stand” is becoming increasingly prevalent in the modern tech space. It is a clever way of making the point that the position a person occupies professionally or in an organization, strongly influences their views on almost everything. Thus, cloud vendors and people that have committed to cloud or grown up with the cloud see only one future for IT, and it’s not on premises.

Similarly, mainframe experts have heard it all before. The rise of the minicomputer, growth of client-server computing and creation of rack servers were all supposed to be harbingers of the end – an end that never quite happened. Mainframes have continued to successfully power most large enterprises, but that doesn’t mean mainframe environments are or should be immutable.

The point is that an objective view must recognize the unique strength and staying power of the mainframe as well as the extensive capabilities available in the cloud. For example, Gartner, a leading technology research and advisory company, is predicting that by 2025 one-third of mainframe data storage will in fact reside in the cloud. Of course, predicting is one thing and achieving is another.

The fact of the matter is that traditional tools for data movement and transformation in the mainframe environment are wholly inadequate and, furthermore, organizations are being held back by not having a model for improvement. Either they are asked to take a futile “boil the ocean” approach of completely abandoning mainframe and moving lock, stock, and barrel to the cloud, or they are stymied by a sandbox “proof of concept” approach, which often leads to more dead ends, sterile experiments, and confusion rather than advancement.

Low-risk approach

A real-world tested approach built around the concept of a three part “journey” may be the answer. The journey begins by addressing an issue that is low-hanging-fruit, namely the cost and limitations of traditional mainframe archive storage. It invests in moving all or part of this category of storage to one of the many capable and secure cloud-based storage options. The journey then moves ahead based on wins and organizational goals. It is much more focused and results-oriented than a sandbox, proof of concept and it is the opposite of a bet-the-company, lift-and-shift approach.

The cloud journey is a low-risk approach because it starts by essentially integrating cloud into the mainframe environment. It focuses on retiring legacy tape storage systems and turning to the cloud to reduce data management costs and data center footprint.  Additionally, it offers an easy next step with immediate payback – a second step on the journey – in the form of affordable, agile analytics in the cloud which can help monetize data that is rarely accessible, let alone feasible to effectively use.

An example of taking this approach is a leading transportation sector company in the U.S. that wanted to take steps toward modernization. After weighing many options, the company chose to move archival data to an Amazon S3 cloud storage service with the help of designed-for-the-purpose software that put little or no burden on normal mainframe operations. The movement of data also helped to cut data center costs, particularly by eliminating the space, hardware, and management overhead of a proprietary tape storage system.

Once a company transfers data to the cloud, it can permanently avoid the costly, on-premise proprietary storage offerings that amount to simply running in place – without improvements in price-performance and with no increase in the ability to share data with applications and analytics. The key, though, is leveraging growing software capabilities, which can make data movement and transformation fast, easy, and affordable.

Software approach to data movement

Once this company‘s data was in the cloud, they took the second step on their journey by employing AWS Snowflake to begin the process of unlocking insights from within their store of operational data. This proved an equally worthwhile step on their modernization journey. The company found new power from the efficiency and scalability of cloud analytics, which strengthened their BI capabilities and opened doors to new, revenue-generating applications.

This leading transportation company is not alone in its desire and determination to better use their valuable mainframe data, but for many it has been a difficult path. While IBM has provided potent computing platforms to its customers, its offerings are often something companies simply can’t afford. So, it should come as no surprise that moving company data off of the mainframe has traditionally been difficult and financially impractical.

First and foremost, most methods rely heavily on mainframe processing to extract and transform data before loading it to a target. This is the traditional ETL process that has long been used for tasks such as populating data warehouses. Unfortunately, it can threaten critical operational processing and backup activities while burning up MSUs.

But, it’s no longer the only way. A software approach to data movement can offer radically different results – faster and more efficient data movement with little impact on the CPU and thus, little or no contribution to MSU billing. Instead of relying on the CPU, tasks can be assigned to mainframe zIIP engines. These System z Integrated Information Processors were originally conceived as helpers for DB2 processing loads but can also be put to work on other tasks, including data movement.

The other realization that makes the software approach so compelling is that the mainframe does not need to be responsible for transformation of the data at all -- just pulling it from storage and sending it on its way over TCP/IP. Transformation can then occur in the cloud when the time is right.

New “where you sit is where you stand” perspective

The “journey” model may end right there for many organizations. Simply reducing storage costs can be enough, but going from step one to step two on the journey and making data shareable with analytic tools and modern, agile applications is both achievable and exceptionally rewarding.

Once this data-centric step is accomplished, further refinements and even more substantive changes become possible. This is the third step, which involves developing a clear long-term IT strategy and putting applications – all or some – into the cloud as well. Although this step is possible, for most mainframe-oriented organizations, complete abandonment of the mainframe is unlikely. Still, having more data in the cloud offers wider options and makes possible a true mainframe-hybrid environment, which combines the very best of both cloud and on-premises.

In the not-so-distant future, based on growing experience with cloud data storage, mainframe professionals will have acquired a new “where you sit is where you stand” perspective that involves mastery of both the mainframe and the elegance, flexibility, and power of the cloud, even including multi-cloud. They will have recognized the cloud journey model and found their place and pace in it, bringing invaluable results.

Gil Peleg, Founder and CEO, Model9 (opens in new tab)

Gil Peleg - Founder and CEO of Model9, mainframe system programming and data management expert, and co-author of eight IBM Redbooks on z/OS Implementation. He holds a B.S.c in Computer Science and Mathematics.