It is the natural order of things that business and process functions don't die, but instead mutate into something new, better or at least more comprehensive.

Managing and monitoring business applications is a good example, for middleware grew into Enterprise Application Integration (EAI), which grew into the Enterprise Service Bus (ESB). Now, according to Gartner, ESB is set to meet its end, to be followed by......well, perhaps it is a case of `the ESB is dead, long live the ESB’.

History of ESB

The original justification for using ESBs may no longer be applicable, but the core reason for using them continues to exist. The growing trend towards businesses using hyperconverged platforms running microservices, with continuous delivery of updates to those microservices, makes the need for the next generation of `ESB’ arguably even more important.

In the `good old days’ the aim of the ESB was to connect applications inside the organization. These days, however, the need is to connect with partners and customers outside the organization in much more flexible, agile ways. Classic ESBs are too rigid for environments where businesses need to adapt to market and business practice changes that they cannot easily predict, but which can happen very quickly. Now businesses change their applications requirements every six months, and it is getting shorter.

What’s next?

So what is needed now may not be called ESB - API Service Gateway looks a likely candidate – but the underlying role for the monitoring and management of application collaboration will not have gone away. It will have become more important, so much so that it will need to be more automated, comprehensive, complex.....and faster.

It will need to allow businesses to interact with customers, partners, suppliers in a more direct way, in real time. Now businesses will need to expose more of their services, and that will need lightweight components that can easily manage their business processes and report on them. The model to bear in mind is what we all now use every day – smartphones. These provide lots of small, tailored apps where the objective is to provide a single service in a slim, slick, fast format, that is, above all other design goals, easy to use.

This is standard practice in the consumer world now and the same is already happening in the enterprise. CRM is already an external service rather than an internal application, and others will follow.

Business need for speed

The business need now is having the ability to provide whatever they need to build, monitor and manage the business processes and activities that they decide they need, at the time they are needed. That last part is important, for the lifecycle of business applications is getting shorter. It used to be years, and now it is months.

For some applications it is just weeks, including the implementation, monitoring and management components. The growing need is to be able to create short term, soup-to-nuts business `events' (for example an online, live marketing event) that is all done, from start to finish, in under a week.

With hyperconverged microservers running applications that are delivered in the `Dev/Ops’ model of install, update and remove using automated continuous delivery cycles, users need to rethink how they exploit the resources and capabilities coming available to them.

And having read that, I can imagine some starting to fear that they must prepare to rip-and-replace all their established business applications. After all, Gartner’s suggestion that the days of the ESB are numbered implies that very thought.

In practice this is not the case, of course. This is not an either/or situation, it is in practice a world where both ESBs and API Gateway Services will co-exist. And in such a world what is then needed is a bridge between them.

Why Integration Matters is extending monitoring capabilities

That is what Integration Matters can provide, for in that richer, more complex environment users will only be able to manage business processes that collaborate between the new cloud services and the established back-office business management tools if they understand what is going on. That means monitoring.

That is also why we are now expanding our scope beyond our start point of monitoring applications running in the TIBCO BusinessWorks environment to add capabilities dedicated to MuleSoft Anypoint and specific applications families, such as SAP. Our goal is to offer a real soup-to-nuts capability across everything the enterprise is likely to use.

And it is needed because the key now is our customers’ customer experience. They will not be interested in how our customers meet their expectations of delivery tomorrow afternoon, but if other companies can do it, our customers must as well. If it can be delivered tomorrow morning, so much the better. Their experience of the buying process is now the crucial measure for every part of the supply and delivery chains.

Monitoring the central applications in an enterprise can be great multiplier. If it has five ESBs, and that is quite common, it will need a tool that can look into all of them to create a single view of what is happening with the business processes, and where any problems can be found. That is what Integration Matters is doing now.

What we deliver is a precise measurement of the service level of a business based on what constitutes their service, what they think they are doing and what they would like to be doing. This applies equally whether it a business delivering one or two items to customers at a rate of hundreds of orders per second, or delivering one complex, multi-item package of goods and services a week, where the customer considers a five-day delivery cycle is good.

Our goal is to ensure that our customers can at least maintain those service levels, or improve on them, particularly in terms of the perception of their customers.

About the Author:  Hendrik Siegeln is co-founder and Managing Director of Integration Matters.