Integration approaches
When you use Code Pipeline as the source code manager (SCM), Code Pipeline itself acts as the back-end repository for mainframe applications. Git is not required, and developers work directly with Code Pipeline to manage application components. Using Code Pipeline as the SCM has the following benefits:
- System of record: Code Pipeline serves as the single repository of record for all mainframe assets.
- Triggers via webhooks: You can configure Code Pipeline events (such as Generate or Promote) to trigger downstream CI/CD orchestrators (for example, Jenkins, Azure DevOps, or GitLab) through webhooks.
- Simplified prerequisites: Because Git is not involved, the Git-related repository setup, branching strategies, and sync utilities (such as ispw-sync) do not apply. For more information, see Prerequisites.
- This approach is well suited for teams that want a streamlined, mainframe-centric CI/CD configuration in which Code Pipeline manages all application components and integrations directly, without the need for external Git repositories.
Comparison: Code Pipeline as SCM versus Git with Code Pipeline
| Aspect | Code Pipeline as SCM (Native) | Git with Code Pipeline (Hybrid) |
|---|---|---|
| Repository | Code Pipeline is the system of record. | Git is the primary repository. Code Pipeline syncs changes. |
| Preferability | Best for teams implementing mainframe-centric, full CI/CD pipelines Code Pipeline as the system of record, with simplified setup and direct integration to CI/CD orchestrators. | Teams adopting enterprise DevOps standards that unify mainframe and distributed development through Git-based workflows, pull requests, and end-to-end traceability. |
| Integration | Direct webhooks from Code Pipeline via an orchestrator (Jenkins, Azure DevOps, GitLab, or Bitbucket) | Git commits or PRs trigger orchestrators by using ispw-sync to push changes into Code Pipeline |
| Developer workflow | Developers work directly in Code Pipeline with Eclipse plug-in, VSCode Extension, ISPF, or a web UI. | Developers code in Git and sync to ensure mainframe lifecycle in Code Pipeline. |
| Branching | Code Pipeline levels (DEV → QA → PROD) control workflow, allowing for parallel development via a fixed number of parallel paths | Git branching strategy (main, develop, feature) maps to Code Pipeline levels, using Code Pipeline sandboxes for an unlimited number of parallel paths |
| Toolchain requirements | Only CES (REST API) and orchestrator | Git hosting (GitHub, GitLab, Bitbucket, Azure respositories), CES, and orchestrator |
| Testing | Trigger BMC AMI DevX Total Test via webhook pipelines | SonarQube can be integrated into a standard pipeline with or without Git. |
| Pros | Simpler, fewer moving parts, and faster adoption | Full DevOps alignment with enterprise pipelines, and better collaboration |
| Cons | Limited to mainframe ecosystem and no Git-based collaboration | Slightly more setup and scripting (Git, sync, and orchestration) |
Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*