What is application dependency mapping?

Application dependency mapping is an accurate map of an entire IT ecosystem, including infrastructure, applications, and their dependencies.

Modern business technology seldom exists in a vacuum. Today’s organizations incorporate a range of systems, applications, and hardware devices, all interconnected. And while this allows for powerful network infrastructure, it also creates certain problems. Namely, it creates a complex structure in which business critical applications and technologies depend on different servers and separate network devices. And when these dependencies go unrecognized, a simple change within an infrastructure can have far-reaching and disastrous effects on application.

A lack of application dependency mapping can lead to operational inefficiencies, fragmented technologies, poor or incomplete data quality, a lack of analytics in optimizing data, low or inadequate levels of automation, and communication issues.

The need to map applications dependencies is becoming widely recognized. Unfortunately, what isn’t nearly as universally understood is how to effectively perform this vitally important task. While there are many different techniques for application dependency mapping, each producing different results, there are four primary approaches that every business should be aware of.

Four methods of application dependency mapping

1. Sweep and poll

This is one of the oldest and lightest-weight methods. It sweeps and pings IP addresses, gathers the type of device that was pinged, and provides information about what is running on the server and the applications it is using. There are typically blueprints of the layout that it uses to search for information and identify the pieces that frame out the larger structure.

The advantages of the sweep-and-poll method is that it’s relatively easy to perform and allows users to sweep an entire network from a single location. The downside is that dynamic, complex environments can affect the sweep accuracy, and a single datacenter sweep can take a long time.

2. Network monitoring

Network traffic patterns are analyzed either at the packet-level using packet capture or using NetFlow at the flow level.

Network monitoring occurs in real time, making it easy to detect dependency changes as they happen. Also, because network monitoring does not refer to pre-built blueprints, this method is effective within less understood systems. Disadvantages of network monitoring include issues related to scaling and duplicate flow records, and problems with differentiating between different application-level dependencies.

3. Agent on server

Agents provide real-time monitoring of incoming and outgoing traffic. This allows them to find and understand the components while recognizing changes to any statuses as the topology changes.

Agents within the server provide the benefit of real-time monitoring and are capable of easily distinguishing between multiple applications running on the same IP address. However, an organization will need to place agents on every relevant, connected application, technology, and server to monitor it, potentially increasing the cost significantly.

4. Application dependency mapping

This leverages orchestration platforms. It deploys and maintains each underlying application component. Consequently, the orchestration understands, at all times, which individual components compose an application.

Effective dependency mapping creates a clear picture of how applications and other technologies work together. This provides several key benefits:

  • Get alerts on network issues and changes.
  • Drill into root causes of issues as they arise.
  • Reliably forecast and evaluate the impact of planned infrastructure changes on business and services.
  • Save time and costs associated with IT efforts.

Business mapping

Business mapping can provide a deep understanding of all servers and applications, especially with dependencies and communications. It helps organizations accurately envision their infrastructures, ensures that no systems are flying beneath the radar, and assists in retiring and consolidating assets.

Change management

Given the complexity and interconnectivity of modern business systems, even minimal changes can have far-reaching consequences. As organizations make changes in processes and technologies, it can have a trickle-down effect, impacting application performance. Whether changes are big or small, they must be monitored and addressed quickly to ensure that they don’t lead to downtime or other user issues. Application dependency mapping empowers IT to visualize individual changes, along with their potential infrastructure impact and the downstream and upstream application consequences they might have.

Root cause analysis

It’s crucial to reduce the time from incident to resolution, as poor performance or system failure can quickly lead to frustrated customers and increased churn. A comprehensive application dependency map can quickly detect problems within the organization, from delays and bottlenecks to failed connections and service issues. Fast identification allows for faster mitigation of the problem.

Proactive incident response

Accurate application dependency mapping clearly indicates which applications and systems would be impacted in the event of an attack or system outage, allowing for the creation of accurate simulations and plans.

It’s also easier to create a security plan ahead of time, like micro-perimeters and micro-segmentation, or transferring data to secure locations. By identifying at-risk areas, organizations are well-positioned for disaster recovery and backup solutions, limiting the impact in terms of governance or compliance.

IT operations management (ITOM) solutions are designed to help businesses take a more active approach to IT operations, using advanced automation and reliable, actionable insights. Service mapping methods provide the flexibility to choose the optimum option for any scenario.

These include the following:

Top-down

Top-down mapping creates a very-precise map of the applications and supporting infrastructure components that make up an application or technical service. It also identifies the relationships between these components. It is well suited for mapping mission-critical services. This includes cloud-native services—for instance, it can detect Lambda to Lambda calls and Lambda to RDS connections to build dynamic service maps. However, you will need to utilize no-code instructions (patterns) that tell ITOM Visibility how to discover non-standard services.

Tag-based

Tag-based mapping takes well-defined tagging policies and uses them to build service maps. For instance, it can create a service map containing all cloud resources tagged with a specific application service. This requires significantly less upfront effort than top-down mapping, but it only identifies the set of components that support the service—it does not identify the dependency relationships between these components. Tag-based mapping is well-suited for less mission-critical applications and for use cases that do not require dependency relationship information.

Intelligent traffic-based

Using machine learning to identify significant service-level relationships from traffic-flow data while filtering out distracting noise, intelligent traffic-based mapping is slightly less accurate than the top-down approach but is also much less labor intensive. It can be used to extend top-down maps or add relationships to tag-based maps.

Get started with IT Operations Management

Contact
Demo