The "Hybrid" IT Environment — Mixing Legacy, Web and SaaS Systems — is the New Norm

IT executives are dealing with the real world issues presented by a sometimes chaotic mash-up of legacy, custom and new applications, infrastructure and tools developed or purchased over the last decade or two. This is very different from the homogenous modern enterprise depicted by analysts and marketers who describe a neat and glossy cloud-based environment that promises to alleviate the myriad operational challenges of “old IT.” According to research we did with senior IT executives, this “hybrid” IT environment appears to be a fact of life that is not going away any time soon:

“It used to be you just monitored the CPU and bandwidth. Now that healthy server could be tied to a web service that’s down. There are more moving parts that we don’t own within our applications. And with virtualization it’s even harder to tell where the problem is.”

“We run 400 applications, from Citrix to web-based to client-server to pure desktop, some are 25 years old.”

“We run a hybrid environment with multiple COTS packages. All on premises with plans to move to the cloud in 1-2 years.”

Managing "Hybrid IT" Increases Complexity — and Reactivity

Despite investments in tools and processes to deliver high service levels via proactive and preventative support, IT executives’ day-to-day mode can still be crisis-driven and reactive. Additional layers of anecdotal or incidental feedback as well as “optical” or subjective input contribute to the reactivity.

“Applications were failing and nobody knew. When they did know, 20 people sat in a room passing the buck pointing fingers. Just everyone looking at their own bit and saying ‘it looks OK to me.’”

“It’s more complicated now; there are more layers. First it was just desktop, then server, and then shared resources, now all these layers. The customer is saying ‘what’s going on?’ No one knew what was going on. As a customer it’s very scary.”

“In our business there are still too many ambulance drivers and not enough people investing in a plan.”

"Hybrid IT" Demands Application- and Platform-Agnostic Visibility and Control

Because the typical IT environment is comprised of hundreds of applications leveraging hundreds of platforms and networks, a new generation of Application Management software that is application- and platform-agnostic will become increasingly valuable. Here is what we believe are the five requirements of this new breed of software:

1. A Unique Design Leveraging the Application Process Component Layer

In order to view, manage and recover applications without customizing scripts for every scenario, platform, and application, a unique design approach is required. Where other tools tend to be built to support certain application types and technology, the software will instead leverage something common across all applications—what we at JumpSoft refer to as the Application Process Component layer (also referred to as the OS Process layer).

Your application can be broken down into the individual components (i.e. processes) that run and provide the services that make the application work: OS, database, application server, web server. You need to monitor each of these processes with an understanding of how they impact your application. Your application uses a web server: Is Apache or IIS actually running on your server? Are they actually serving up content? Same thing for the database: Is MySQL or MS MSQL actually running? Can you read and write to the database?

Application Process Components are the common building blocks by which applications are built and function, regardless of environment. This common framework provides the hooks that the Application Management software will leverage to achieve its cross-platform, application-agnostic functionality.

Through support of these universal technical building blocks, applications running on various, but common operating systems, databases, and application frameworks can be readily supported within the next generation of Application Management software.

These hooks exist in both legacy and next-gen applications, and allow application-agnostic management via this framework, versus API-level integration to support each specific application.

2. Stateful Awareness of the Application

Through a combination of built-in application-centric monitor types, the new generation of Application Management software will be aware of how an application is behaving, whether it is up, down, or partially impaired, including what parts of an application are impacted.

But just knowing the individual processes are working isn’t enough. There are lots of great monitors out there, but often they are very specific and require detailed knowledge to interpret. Classic APM tools can tell you down to the line of code how things are working, but what about the entire application? The trick is to consolidate all these monitors into a single view that represents the total health of the application and turn it into actionable information.

At a glance visibility is not about gathering and analyzing reams of performance data. In fact, when you’re trying to establish the health of the system, there is such a thing as “too much data.” The key is to boil that data down to “what to do about it to ensure services,” by leveraging multiple monitor types and points of view to derive true health and impact. And most importantly, what to do about it when things go awry.

3. Understanding of the Application’s Architecture, Components and Related Dependencies

Application Management software will associate application components with an understanding of the appropriate technical steps, in sequence, that are required to gracefully start or stop all, or parts of an application, in the context of its underlying IT infrastructure, in order to recover from, or work-around underlying issues impacting application operation.

4. Secure, Policy-Driven Action

The next generation of Application Management software will be managed by configurable business rules, in conjunction with operational context of the given state of an application. This ensures that the software will only drive the appropriate and prescribed steps in order to get an application to a defined “best-state” in the face of virtually any issue, or combination of issues. This ensures that an organization’s best practices will be uniformly and systematically applied to recover from application issues, negating the likelihood of human error.

5. Lightweight and User-Friendly — So It’s Practical and Cost-Effective to Deploy and Use

Just as importantly as the elegant engineering that goes into this application-centric Application Management tool is that it can be deployed quickly and made immediately available to all users without weeks of training, an army of consultants and hundreds of thousands or millions of dollars in customization. In other words, can I do it quickly, can I do it cheaply and will my staff actually be able to use it?

In essence, Application Management software should:

• Systematically enforce and/or automate an organizations’ best practices, in the context of the situation

• Incorporate an understanding of the technically correct approach

• Have full awareness on an app, or multiple apps

• Build in the model of “what would your people do” in this situation

These combined elements will enable Application Management software to intelligently identify and respond to any issue, or any series of complex events, resulting in rapid application recovery and service continuity. This includes items ranging from complete disaster recovery of an entire application, remediation of a failed application component or providing an automated notification of performance degradation within an application.