Future of Application Mapping: Start Here, Start There, Start Anywhere
March 2, 2016 by Michelle Kerby
Application maps provide clear visibility into which parts of the IT infrastructure support a business service. An application map makes it possible to document and monitor business services. It enables IT to assess the impact of potential changes, check service continuity configurations, and other functions to ensure optimal business support.
But creating an application map is not a trivial exercise. Application owners have to manage complexity. There are a lot of moving parts tied up with a modern, distributed business application: network infrastructure, databases, storage, servers, and software. It’s common, for example, for the data center team to have an inventory of hardware and software assets. But that inventory doesn’t reveal how these assets connect to one another and how they support an application.
When data for application mapping is collected and entered manually—which is often the case—by different IT functions—the risk of inaccurate data rises. Plus, it takes so long, that by the time the data’s collected and consolidated, it’s likely out of date. So you end up with an inaccurate map, and that can cause big problems. Like taking a database offline thinking that it’s associated with a Tier 3 application, when it’s really supporting a Tier 1 application.
Auto-discovery and Application Mapping
So let’s move on. Manual application mapping is a dead end. What about auto-discovery tools? These tools collect data on network infrastructure, servers, software, and their relationships to build robust, tiered application maps. Auto-discovery also makes it feasible to easily update maps to reflect changes in the environment, which ensures timeliness and accuracy. Flexible reporting options accommodate a variety of documentation needs. By fostering collaboration between application owners and configuration managers, application maps advance IT-business alignment and help keep the environment responsive to business demands.
Where to Start Mapping
On their own, auto-discovery tools do not solve all the problems application owners face when trying to generate comprehensive and accurate maps. Where the mapping process starts can make a big difference in how accurate the map is.
Sometimes auto-discovery tools use a “top down” approach. That means the person creating the map has to specify the “top” entry-point for the application—often a URL that leads to a load-balanced service or to a web server. The choice of what to scan is also driven top-down. At each stage, the tool decides what to scan based upon the dependencies it detected at the previous stage.
But there’s not always just one entry-point. Application owners often think about where applications store data, which is akin to starting at the bottom. For off-the-shelf applications like Exchange, the natural place to start is with the Exchange software, rather than either the entry-points used by clients or the databases storing the data. Exchange servers are in the “middle,” with entry-point servers and load balancers above and databases below.
Start Anywhere Application Mapping
Start anywhere application mapping is the only effective way to accurately map diverse and complex applications. The start-anywhere approach gives application owners a better experience by allowing them to start with what they know, rather than forcing them to work out what constitutes the “top.”
BMC Discovery always looks everywhere, so it sees systems and their dependencies from all directions. It removes the key limit of the top-down approach, which is that often a dependency between two things can only be seen from one end. In those cases, the top-down approach only sees the dependency if it’s visible from the top.
Also start-anywhere mapping can start from multiple points simultaneously, which means you can cope much better with applications when parts or relationships are missing. Some dependencies will remain invisible to any tool, but with a rigid top-down approach, as soon as the tool meets something it can’t discover, that’s the end. With the start anywhere approach, you look from the top, bottom, and middle—pinpointing and filling in the gaps much more quickly.
Users employ a modelling editor to create the map. They can annotate and trim a map—making it easier to read. For example, a user might turn off the feature that shows software components associated with software instances. BMC can accomplish all this in about 15 minutes in most environments. And if the modeled environment changes, the map changes automatically.
Start Discovering Now
These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.
Share This Post