Business is on the move. As companies across all industries accelerate to stay ahead of their markets, they’re adding innovative digital solutions to meet customer needs and make their business more efficient and competitive. In turn, IT teams must keep pace to ensure their environment can easily adapt to the myriad of new technologies, applications and services.
But making ongoing updates and adapting to more-complex cloud environments is never what it seems on the surface, because you’re never just dealing with applications—you’re dealing with every bit of infrastructure (including other applications) that has some connection to them. Making extensive changes in the name of digital transformation or responding to unanticipated changes without knowing how it all comes together can produce a highly undesirable domino effect. But with IT information now stored across various systems and maintained by different teams, making the overall landscape siloed and inconsistent, such knowledge doesn’t always come easily. Mapping your application dependencies can reveal what’s lurking below and prevent business disruptions when you make changes—whether you’ve planned them or not.
Beyond Mere Asset Inventory: Application Dependencies
To effectively manage your systems, you need to know exactly what assets are in your IT environment and have accurate configuration data. IT staff must be aware of application-to-application and application-to-services dependencies as well as application-to-infrastructure dependencies. Often, just identifying the primary application isn’t enough, and a clear depiction of upstream and downstream relationships is necessary to fully prepare for changes and to inform personnel. For example, let’s say you’ve recovered your primary application in the public cloud, but you were unaware that an SaaS provider you used for a web service is part of your payment flow and required notification of an IP address change. Although such a dependency might easily go overlooked, it could prevent delivery of a critical business service. A clear picture of the overall impact of an operational issue, such as failed server, and how its effects might reverberate beyond the primary application will save you critical time and reduce the other consequences your organization would experience.
Some enterprises may find this information in a configuration-management database (CMDB). Others may use an automated application-discovery tool, particularly for service-impact analysis, asset management, compliance and configuration management. Autodiscovery, ITSM and asset-management solutions are designed to collect and consolidate all of this information. Many IT managers are enamored with the idea of an automated discovery tool. After all, with a name like autodiscovery, shouldn’t it be easy to produce a comprehensive picture?
The Limits of Autodiscovery
On the surface, it makes sense that an automatic sweep of your IT landscape should generate what you need. But more often than not, these tools produce excessive unnecessary data that ultimately fails to illustrate the interdependencies of applications for business-critical operations. We find that these tools may fail to provide complete contextual information, such as the business logic behind the use of a particular application, leaving you vulnerable when making decisions about what to move and when to move it.
Some other IT managers have an aversion to autodiscovery tools. If you have a unique IT setup—for example, a bespoke server with a highly unusual configuration running esoteric applications—how can you be confident an autodiscovery tool will get the information it needs? Can you be sure it’ll capture the legacy applications and all their dependencies buried in your IT landscape?
Mapping a Complete Strategy
Here are a few steps that you should take in preparing to execute or respond to change while maintaining full IT resilience:
- Ingest information describing all elements of your environment from a variety of sources, then normalize the data to filter out noise and build a single consistent source of truth. Although most would take it as a given that you must thoroughly map your environment, capturing everything can be much more challenging than many anticipate. IT environments are becoming increasingly complex, and the chances that critical application information will be siloed in various departments is only growing. The next point, therefore, involves bringing the tribal knowledge of your people into the mix.
- Validate the data and application dependencies with subject-matter experts to mitigate the risk of missing crucial relationships and dependencies, enabling teams to make better decisions. In fact, it’s critical in the discovery phase to engage the organization’s SMEs in short but meaningful interviews to obtain physical/virtual inventory, application inventory and dependencies.
- Once you’ve performed a thorough dependency analysis, you’re ready to create “move bundles” and events. Because you now have a complete view of how the various applications and other IT components depend on each other, you can decide which application groups are movable. This information can save tremendous time and resources in a data center transition, a move to the cloud or some other system improvement. Or it may enable you to respond more effectively and minimize the effects of an unforeseen disruption because you’re fully aware of how it might reverberate in your IT environment.
As you make decisions about your IT infrastructure, consider that you’re not just moving an application—you’re moving all the infrastructure connected to it. And in today’s dynamic environments, interdependencies between apps can be complex and extensive. Failure to account for those complexities and interdependencies across the IT landscape often leads to delays, breakdowns and disruption to critical business services, reducing company revenue and creating internal frustration.
About the Author
Eric Kraieski is vice president at Transitional Data Services (TDS), where he brings over 30 years of experience in emerging technology, enterprise software, product management and marketing leadership. Eric has worked with some of the world’s top companies to move data centers in complex environments. He has a B.S. in computer science from the Shippensburg University of Pennsylvania.