Introduction DRAFT

Imagine a towering skyscraper, built decades ago. It’s still functional, housing critical operations, but the blueprints are faded, the wiring is a tangled mess, and only a handful of people truly understand how it all works. That’s a legacy system in a nutshell. Just like understanding the structural integrity of that skyscraper is crucial before any renovations, understanding the architecture of your legacy COBOL system is paramount before embarking on any modernization journey.

This chapter is your architectural survey, equipping you with the tools and techniques to ‘deconstruct’ the monolith, identify its hidden complexities, and prepare for a successful transformation. We’ll explore three critical aspects of legacy system architecture:

1. Documentation Archaeology and Knowledge Extraction - How to recover lost insights from scattered documentation and subject matter experts
2. Business Rule Identification and Isolation - Techniques to separate valuable business logic from implementation code
3. System Interdependency Mapping - Methods to visualize the complex web of connections in your legacy landscape

By the end of this chapter, you’ll be ready to ‘read’ your legacy system like a seasoned architect, laying the groundwork for effective modernization planning.

Documentation Archaeology and Knowledge Extraction: Recovering Lost Insights

Legacy COBOL documentation often resembles a neglected library: incomplete and poorly understood. This section guides you in becoming a ‘documentation archaeologist’ to uncover hidden system knowledge. We’ll cover recovering documents, interviewing Subject Matter Experts (SMEs), and reverse engineering code to build a knowledge base essential for modernization.

Our journey will cover:

• Recovering Lost Documentation: Finding and interpreting incomplete or outdated documentation.
• Interviewing Subject Matter Experts (SMEs): Effective questioning for knowledge capture.
• Reverse Engineering COBOL Code: Understanding code via dependency and control flow analysis. See Chapter 2 for COBOL basics.
• Building a Knowledge Base: Consolidating extracted information.

Begin by locating all existing documentation. Valuable information often resides in unexpected places.

Unearthing Existing Documentation

The first step is to find any existing documentation. Think of it as a treasure hunt; you need to know where to dig.

Here are some places to look:

• Physical Documents: Look for physical documents stored in filing cabinets, archives, or offices. Even outdated documents can provide valuable clues.
• Online Repositories: Shared drives, version control systems (Git, Azure DevOps (TFS/VSTS), GitLab, Bitbucket), wikis. Search for system-related documents (.doc, .docx, .pdf, .txt, .xls, .xlsx, .html, .md, .csv, .json, .yaml, .sql). Leverage version control history to uncover older documentation versions and commit messages that might contain valuable information.