Establish and Document the Application's Upstream and Downstream Dependencies

This chapter explains how to establish and document what an Application depends on, and what depends on it. It addresses the dependency chain identified during design; the full baseline of dependency types, applied across the Application’s life, is addressed in the chapter “Establish and Document the Application’s Dependencies.”

Upstream and Downstream Dependencies

Upstream dependencies are the Applications, Systems, services, technologies, vendors, and other resources the Application relies on in order to work. If an upstream dependency fails or changes, the Application may be affected.

Downstream dependencies are the Applications, Systems, business processes, and users that rely on the Application. If the Application fails or changes, they may be affected.

Dependencies overlap with, but are broader than, APIs and Integrations. An Application may depend on a shared identity service, a scheduler, a certificate authority, a reference data feed, or a vendor’s support, none of which necessarily appears as a data Integration.

Establish the Dependencies

Identify the Application’s upstream and downstream dependencies, starting from its APIs, Integrations, and hosting infrastructure, and extending to shared platforms, technologies, vendors, and the business processes and users that rely on it. For each, establish the nature of the dependency and the impact if it fails.

Document and Share the Dependencies

Record upstream and downstream dependencies, with their impact, in the Application Documentation, preferably with a dependency diagram, and link each one to its own record where it has one.

Remember to review dependencies with every Release.

Best Practice: Establish and Document the Application’s Upstream and Downstream Dependencies

It is a best practice to both establish and document the solution for an Application’s dependency chain: identify what it relies on and what relies on it, with the impact of each failing, and keep that record current.

Benefits: Knowing upstream dependencies tells support teams where to look when an Application fails for reasons outside itself. Knowing downstream dependencies tells everyone who must be warned before a change or outage, and lets the enterprise judge the true business impact of each Application.

Advance Maturity Deliberately: Crawl, Walk, Run

• Crawl. Dependencies are listed manually from the knowledge of the Application team.

• Walk. Dependencies are recorded consistently, linked to the records of the dependent and depended-upon Applications and technologies, and reviewed with every Release.

• Run. Dependency records are validated against automatically discovered connections and enterprise models, and impact analysis across all Applications is available on demand.

Connections to Related IF4IT Practices and Inventories

The IF4IT Enterprise Model and Modeling Best Practices describes how relationships between Applications and other enterprise entities are modeled. The Software Technologies Inventory and Attributes and the Vendors Inventory and Attributes describe two of the most common kinds of upstream dependency.