Picture it: A customer purchasing a concert ticket enters their order online. This person is using a mobile application (or a browser site) that sends their transaction through the provider’s network, its front-end portal or web services, all the way into its supporting infrastructure. That back-end could be made up of some mix of traditional application servers or mainframes, which work together to process the transaction and fulfill the order.
That’s how a cross-platform business service is supposed to work—touching distributed systems, mainframes, or both, depending on what makes the most sense for that service. Modern applications are increasingly being developed this way because companies want to make sure that, no matter what, customer-facing applications are able to offer the best possible experience.
The problem is that, from an organizational perspective, many enterprises aren’t set up to manage and monitor cross-platform environments. For the most part, mainframers still keep an eye on their mainframe responsibilities. Distributed teams do the same. There’s separation even within each team—an IMS specialist might never need to worry about CICS. There’s very little cross-team collaboration or monitoring, so when something goes wrong with the customer application, the customer may be the first one who actually notices.
This must change in order to support cross-platform business services.
The Old Way of Working
The historical model has mainframes as the “system of record” storing mission-critical data and resources, and distributed environments as the “systems of engagement” that power less-critical siloed applications like email or webservers. However, such a model no longer works, because the typical use cases for each environment have changed dramatically.
At a recent SHARE event CA Technologies Product Manager Lowell Higley presented on the topic ‘Removing Platform Silos - Why Change? Why Now?’ The discussion focused on how IT organizations are unifying ops across platforms in order to keep their mainframe systems of engagement working together in harmony with their hybrid systems of engagement.
Twenty years ago, corporate email running on a distributed system would not be considered mission critical, and losing access might not have been a huge deal. Today, it could cripple a company’s productivity, demonstrating how some “systems of engagement” have become just as mission-critical as mainframes. Similarly, the use cases for mainframes have evolved. While still often the “system of record,” companies also want to leverage the processing heft of the mainframe to power a range of use cases—mobile, data and the cloud—that might previously have been earmarked for distributed systems.
A Mountain of Monitoring Tools
Despite thinner layers of separation between mainframe and distributed, there remains a rather large and inconvenient chasm separating the teams that manage these environments, said Higley. One particularly frustrating problem is the number of tools required to monitor and manage each environment. A single organization could have half a dozen or more different pieces of software to keep an eye on different subsystems within both mainframe and distributed environments.
A single IT generalist—your front-line customer service tech—has to monitor all of these tools to make sure things are running smoothly. What happens if all the monitoring systems report back that they’re working fine, but a customer says they still can’t access your mobile app? Is that a problem for the mainframe team, a distributed team, both, or neither? How long will it take to fix the issue with such little clarity?
Cross-Platform Monitoring
In a new era of application development, enterprises need better tools in order to monitor and manage cross-enterprise business services. A “single pane of glass” for cross-enterprise monitoring could save companies time and money by simplifying problem identification and resolution, from the IT generalist down through to the individual technology line expert.
But achieving these benefits won’t happen overnight. Enterprises will need to work toward a common goal, Higley said, by getting key players from IT, mainframe, and distributed lines of business together to create a unified vision for the entire company. You can start with baby steps—a collaborative project here or there, even a new seating plan to put mainframers right next to the Linux team, for example—but the idea is to work toward a fully collaborative environment.
When an app problem occurs, the IT support team should know exactly who to call and how to escalate it for the fastest possible solution. That’s the ticket for smarter IT and happier customers.