By Manny Veiga
The somewhat tenuous relationship between cloud computing and mainframes was once again a big topic at our most recent event, SHARE in San Antonio. In fact, Frank DeGilio, Chief Architect for Cloud, IBM Systems Group, hosted a session this year all about the mainframe’s vital role in the future of cloud. Frank also wrote about this topic in a recent blog, and we found his take on the culture problems that hinder collaboration between mainframe and distributed systems professionals.
We asked him to elaborate in this Q&A, and he shared some great insights on how mainframe professionals can beat back misconceptions about their technology and put the mainframe front-and-center in future cloud conversations.
SHARE: You wrote that the mainframe is ideally suited to support the cloud, but that the big inhibitor to adoption is culture, not technology. What is it about the current culture or mindset today that blocks mainframes from being more widely adopted? And, what do mainframe professionals need to change on their end to start to transform their business culture?
Frank: The current culture or pervasive mindset of today’s applications developers (and lines of business for that matter) is based on a view that the mainframe is old, slow, and out of touch with modern techniques. The belief is that the mainframe can only do 3,270 connections and can’t leverage new technology. Developers expect mainframe application development to take months or longer and even worse they expect that the people involved – the mainframers – to say “no” to every new idea or approach. The development community needs proof that the mainframe can provide modern interfaces and support a development environment that is better than what can be done in the distributed world.
Likewise, mainframe professionals have to actively work against these stereotypes. They need to figure out how to provide something new and powerful to inspire other platform bigots that the mainframe has value. Application developers always take the Desire Path – the fastest path to production. Mainframe professionals need to figure out how to be that desire path.
SHARE: What's the most successful way for mainframe admins to make the case for mainframe to the rest of the business? Is the mainframe/cloud proposition about business benefits, technology benefits, or a mix of both?
Frank: It is definitely a mixture of both. We need to provide technically superior capabilities that provide those technical advantages at a price point that is at least competitive (if not better) with distributed solutions. This is actually easier than one might think and can be accomplished by taking advantage of the capabilities that are inherent in the platform like Sysplex and data sharing. Companies with proactive sysprogs are doing this today. Interestingly enough, companies with young technical professionals are gravitating toward forward-thinking mainframe professionals.
SHARE: What's stopping more collaboration between the mainframe and distributed sides of the business, and how can companies overcome it?
Frank: The biggest inhibitor to collaboration is politics. Generally, the mainframe and distributed platforms fall into different management chains where each is more focused on their own careers than what is right for the company. Even in companies that have distributed and mainframe teams working together, that collaboration all too often falls victim to politics from others who have an axe to grind against the mainframe. One company played the “it’s not strategic” card after finding out that the mainframe version of the tool was faster, more available, more reliable, more secure and one-fifth the cost of its distributed counterpart.
This is why mainframers have to be proactive in finding partners to make this work. One IT organization made mainframe adoption easy because they said, “What do you need? I am going to delight you by providing what you need quickly with a quality of service that you have never seen.”
SHARE: Are there any lessons mainframers can take from the way distributed folks work? On the flip side, how can distributed benefit from some of the tried and true mainframe processes and ways of working?
Frank: Mainframers can learn from distributed folks. Distributed folks are focusing on creating functionality quickly. Traditionally they have been a bit less focused on the kinds of RAS characteristics that drive mainframe proponents. They have been more focused on getting stuff out that takes advantage of the latest technologies. Mainframers obviously do not want to mimic the distributed model so much that they eliminate the RAS characteristics for which they are famous, rather they need to be able to adapt processes that enable them to move more agilely.
This is why building services is important. By building services, mainframers provide enterprise capability quickly with agility and without sacrificing the RAS characteristics inherent in mainframes. Distributed folks need to learn more about how to take advantage of the RAS characteristics available to them. This is the neat thing about building services, it brings the two communities closer together.
The companies who will win in this space are the ones that demonstrate that we are not creating distributed or mainframe solutions anymore. Instead, we are building cloud solutions that take advantage of both mainframe and distributed strengths. These applications can be developed quickly and agilely adapted to changing business needs.
SHARE: How does the cloud and mainframe connection engage and develop a new generation of mainframe professionals? What are some specific examples of ‘gateways’ through the cloud to mainframe that help new admins become more engaged in mainframe use?
Frank: There is a lot of misinformation out about mainframes. One of the things that both mainframers and distributed folks seem to believe is that the mainframe is hard to learn. It is true that the mainframe does things that can’t be done anywhere else and the way it attacks traditional IT problems is unique, but that doesn’t mean that it is hard or difficult to learn. The proliferation of services that are accessed off platform will increase interest in the capabilities inherent in the hardware and operating system. The services themselves will be the new gateways to understanding the mainframe.
People in our field are naturally curious and will, if given the proper incentive and freedom, learn about how to take advantage of a capability. Mainframers who take the opportunity to explain the merits of the system (without having to be too bits and bytes oriented up front) will find willing students. The key is to explain things at a level that is sufficient for a particular task to be done. People will learn more about the capabilities as the need arises. In the meantime, admins should be focusing on how to make automation available to newbies.
Watch Frank’s complete presentation, “The Mainframe Strikes Back! Reclaiming IT!” to learn more about this topic. You can find that presentation and others from recent SHARE events in the SHARE Live! content library.