Data infrastructure optimization, availability & security software
Data integration & quality software
The Next Wave of technology & innovation

Why Mainframes Matter in the Age of Cloud Computing

Has cloud computing killed mainframes? You might think so. In fact, mainframes remain supremely important, even in the age of the cloud. Here’s why.

It’s easy enough to understand why one might think that mainframes are no longer relevant in the present age of “cloud-native” everything. Today, everyone and their mother is moving workloads to the cloud.

Public cloud computing platforms like AWS and Azure are generating record amounts of revenue, while private cloud frameworks, such as OpenStack, are more popular than ever.

You may think that, as a technology introduced more than a half-century ago, mainframes don’t have much to offer in a world dominated by much newer cloud computing technology.

Mainframes are the Original Cloud

But they do. The thing is, many of the qualities that make cloud computing so valuable to enterprises are also true of mainframes.

Think about it. Cloud computing has become popular because the cloud gives you:

  • On-demand access to compute and storage resources that can scale virtually without limit.
  • A more cost-efficient way to maintain infrastructure than relying on local commodity servers.
  • The flexibility to spin up different types of software environments. In the cloud, you can run any type of host operating system you want. You can use virtual servers, or you can use containers. You get a lot of flexibility.
  • A lower maintenance burden. When you move workloads to the cloud, you no longer have to maintain lots of individual on-premise servers.
Read our eBook: The Ultimate Guide to Mainframe Machine Data, mainframe predictions

To a large extent, mainframes offer precisely the same benefits. Consider the following points:

  • A mainframe provides a single infrastructure that delivers massive compute and storage resources whenever your applications require them.
  • If you use a mainframe, it’s probably one that you already own. You don’t have to purchase new infrastructure. In addition, a mainframe can last for decades, whereas a commodity server lasts usually just a few years. For both reasons, mainframes offer cost-efficiency.
  • Mainframes give you plenty of software flexibility. You can use a mainframe operating system like z/OS, or you can use Linux. You can even run Docker containers on your mainframe.
  • When you run applications on a mainframe, you only have one machine to maintain. This is simpler than maintaining dozens of on-premise commodity servers.
Why Mainframes Matter in the Age of Cloud Computing

Mainframes have long offered these benefits, but the cloud computing concept as we know it today has been around only for about a decade.

Essentially, then, mainframes were the original cloud. Going back decades, mainframes allowed companies to build cloud infrastructure avant la lettre (which is a fancy way of saying “before the term cloud infrastructure existed”).

Mainframes and Cloud Computing Today

Don’t get me wrong. I’m not anti-cloud. You should move workloads to the cloud when the cloud is the best fit for a particular type of workload.

But if you have a mainframe environment set up, there’s a good chance that your mainframe can provide many of the same benefits as a cloud platform. In that case, you’ll save time, effort and money by continuing to use your mainframe to provide the same cloud-like benefits that it has been delivering for years.

Expert Trevor Eddolls suggests making your mainframe the center of your cloud computing

In short, the cloud is not a replacement for your mainframe. It’s just a newer, flashier way of implementing the same types of functionality that mainframes offered starting decades ago.

Related: Expert Interview: Trevor Eddolls on Mainframe’s Role in Today’s IT Infrastructure (including the Cloud)

Of course, using a mainframe effectively today does require integrating your mainframe environment into the rest of your infrastructure — whether it involves the cloud or not.

Be sure to check out our eBook: The Ultimate Guide to Mainframe Machine Data

  • Brendan Coffey — September 5, 2017 at 10:33 am

    Excellent point of view Chris. The rush to use Clouds for traditional distributed workloads makes tons of sense,,,improve Time to Market, avoid capitlization of HW, gain some flexibility, and maybe save money. Not all of those translate to Mainframe environments when you look at Mainframe IaaS options like the one from IBM (shameless plug) and other providers in the market, you can gain all of those benefits without having to ‘cross the chasm’ and re-write workloads and re-platform away from z/OS.

  • Chris Probasco — September 8, 2017 at 10:48 am

    Great article, but lets not also forget your largest investment. The applications! Many client have huge investments in applications and IBM made a huge commitment years ago to our mainframe clients that if you develop an application to run on the mainframe you will never have to change change that application again for it to continue to execute. Think of the value this has brought and continues to bring to you and your clients. Think of the investments your clients have made in applications and application code over the years. Can anyone think of another platform cloud or not that can make that kind of statement?

Leave a Comment

Related Posts