Visualize Your System
Analyze and Understand Its Architecture

Caravel™ Insight

Caravel™ Insight provides comprehensive analysis information that allows you to quickly, accurately, and intuitively understand the assets of legacy systems such as JCL, CL, COBOL, and RPG running on mainframes (IBM z, Fujitsu MSP, NEC ACOS, etc.) and midrange systems (IBM i, AS/400). Insight is built with specifications that allow for flexible addition of functions, so you can implement support for resources (languages ​​and resource definition information) that Insight does not currently support, even if you are using a mechanism of your own specifications (for example, a JCL pre-conversion tool or COBOL source pre-conversion).

It clarifies all asset information and their dependencies, creating a 100% visual, interactive, and navigable system image. The analysis results of dependencies provide strong support for containerization and API of required functions by specifying entry points (data integration and operation starting points). The Java pseudocode generator generates reliable pseudocode to support the reengineering process.

Caravel™ Insightのフロー図

What you can check (partial excerpts)

  • About source and object consistency
  • Outputs objects that have no callee or reference destination
  • Refer to repository contents with search function
  • Output Metrics information for each object

4Features

01

Analysis of the current system

Details of individual sources and their dependencies are clearly visualized,
Capable of understanding the current status and analyzing the impact

(1)Analysis related to the entire system

  • Project statistics: System asset statistics, inventory
  • Network diagram: Specify the starting point you want to study, and investigate the correlation between program calls, files, and DBs

(2)Analysis related to individual resources

  • Object dictionary: programs (including JCL and CL), files (FILE, DB, screens, forms), index of unresolved objects
  • Individual program information: data structures in programs, program call relationships, file usage relationships, logic flowcharts, duplicate code
  • Individual file information: CRUD, data structure, and dependencies of programs used
  • Tips Report: Report on issues and potential problems in the transition.
02

Use to plan for improvement

JavaReporting tips on issues, duplicate code, and dead code for Java migration

  • Tips report: Reports on issues and problems that may arise during migration.
  • Duplicated code: Check duplicated code in each program and consider componentization
  • Dead code: Check unused code in each program and remove unnecessary logic
03

Utilize for Migration Planning

By functional analysis grouped by subsystem,
Visualization of common and isolated parts among groups

  • Functional Group Analysis (Clustering)
    Clustering processes by creating groups of jobs, subsystems, batches, online, etc., can clarify object dependencies.
    By creating an entry point group as a starting point and performing functional analysis, related objects are automatically identified and a repository of functional groups with all dependencies connected is created.
    By clustering multiple functional groups, it is possible to clearly determine which objects are commonly used among functional groups and which objects do not affect other functional groups at all.
    This feature can be used to plan conversion and testing of common parts (programs, files, and DBs) with priority, and to reduce duplication of work in migration planning.
  • Program metrics: Reports showing program quality and complexity, used to schedule testing from migration
  • Object dictionary and unresolved objects: Check dependencies and consider alternatives for programs without sources
04

Utilization of parts for restructuring and scratch development

Provides Java pseudo-code for COBOL and RPG conversion

  • If you choose to rebuild or scratch develop your COBOL or RPG system with major modifications, you can also reduce coding man-hours by diverting the generated pseudo-code.
    (If you choose to migrate using other methods without using our Converter or Express)
  • Individual program information: data structures in the program, program call relationships, file usage relationships, Java pseudo code