Mainframes have reliably served the heavy computing and security needs of large and medium-sized enterprises for decades. More recently, cloud computing has emerged promising scale, availability, simplicity, and lowered costs. But will it deliver?
Similarities and differences exist between mainframe and cloud computing, as well as various use cases for each. Companies will continue to have increasing needs for computing power, so the question many organizations face is: where will that power come from? Which type of computing will be best for what kind of situation? Are they really that different?
Mainframes and cloud systems both utilize a client/server model of computing. You don’t sit down at the cloud or at a mainframe. You sit down at your computer or device and access applications and data that are stored on mainframes or in the cloud. In both setups, the applications and data are housed separate from the typical user.
Mainframes and cloud computing both also have the capacity to scale elastically. Mainframes do this through dedicating more computing resources to a task or application. Cloud systems do it by calling upon additional servers and systems to handle load.
The mainframe is a single large host computer that the clients, and all their myriad terminals, can access. In cloud computing, the data, applications, and processing are distributed across many servers that can be clustered in one location or spread across many locations.
Security is another major difference. Historically, mainframes are very secure systems. Organizations control both the data and the access to it. Cloud computing, since it relies on an Internet connection, risks being hacked or accessed by intruders. Even private clouds fall short of what mainframes can do for enterprises seeking to secure their data and systems.
Finally, mainframes have proven themselves to be more reliable. Cloud applications become useless when an Internet connection goes down or if the server fails. On the other hand, mainframes have on-site IT staff to maintain them, have a history of reliability, and are often accessed through local networks.
Why the mainframe reigns in business
When people say ‘cloud computing,’ they often mean very different things. For consumers, cloud computing denotes things like Google Docs or Gmail, i.e. applications accessed through browsers. Consumers are quite familiar with cloud computing, but whether they know it or not, most have also interacted with a mainframe (ever used an ATM, booked a flight online, or got an insurance quote?). Many business processes, even when conducted over the phone or on the Internet, utilize a mainframe system.
In business, cloud computing can behave as a pay-for-play mainframe, but the data is never entirely under the enterprise’s control. It lives on a third party’s hardware. Additionally, one of the key selling points of cloud computing, the elastic scalability of the platform, is something that has been built into mainframe computing from early on. Mainframes are able to run thousands of virtual Linux servers concurrently. This is similar to having a data center, filled with racks of servers running multiple Linux instances. From a labor perspective, however, it takes far less effort to maintain and run one mainframe than the hundreds of servers it would take to operate at the same scale. In the enterprise, a mainframe system can work like a cloud system that is more secure and reliable.
The spread of the client/server model of personal computing during and after the 1980s was fueled more by the shrinking cost of personal computers than it was by superior technology. On those early distributed systems, the hardware and operating systems were usually designed separately and were optimized for cost savings rather than industrial-level utilization and strength. The sprawling data centers required many more system administrators to maintain than the centralized mainframes that would keep running without ever needing to reboot.
While cloud computing can replicate some of the advantages of mainframe systems, there is not a compelling case for companies to move mission critical systems into the cloud, and especially not onto servers they don’t directly control. The risk of having core business systems and data running on a cloud platform outside of an organization’s direct ownership and control far outweighs the benefits. The value of having full ownership and control over core business data cannot be overstated. Trusting core data and processes to third parties carries numerous risks. Regardless of how trustworthy third party cloud providers appear, using them will never be as secure as having the data locked down under your control. Rather than expose yourself to those risks, it is better to invest in solutions to make the most out of your mainframe.
Need help making the most out of your mainframe? Check out the mainframe software used by Global 500 companies for more than 25 years, and then request your free trial to see the difference ASPG software can make.