Boost Deployment Efficiency with Structured Environment Management

Goal

Improve the efficiency of the deployment stage in version control by clearly distinguishing environment configurations.

Environment List

Development Environment (Classification: Test, Role: Development) – Rapidly changing code development and self‑testing environment.

Integration Environment (Classification: Test, Role: Development) – SIT environment mainly used for integration testing.

Test/UAT Environment (Classification: Test, Role: Test) – Functional testing and user acceptance testing environment.

Pre‑Production Environment (Classification: Production, Role: Operations) – Part of production used for specific user feature validation.

Production Environment (Classification: Production, Role: Operations) – Main production environment accessed by most users.

Principles

Maintain distinct development, integration, UAT, pre‑production, and production environments.

Test and production environments must be physically isolated.

Test and production environments should be created as similarly as possible.

Each environment must have clear responsibilities without over‑reach or misuse.

Configuration should be separate from code and fully version‑controlled.

Environment resources should be provisioned automatically, without manual configuration.

Use a unified infrastructure service baseline to create all environments.

Standards

Environment management must follow IT and security department standards.

Deployments must not include unrelated content.

Production data must never be restored directly in non‑production environments without sanitization.

Test/UAT environments are a single set managed by QA, with development assistance; QA starts work from UAT and hands over to users for acceptance testing.

Every deployment to test/UAT must pass a smoke test; failures require fix or rollback.

Test environments must not be linked to production environments (e.g., A‑system UAT cannot connect to B‑system pre‑production).

OS and middleware versions in integration, UAT, and pre‑production must match production.

Recommendation: New version deployments should first pass a smoke test.

Recommendation: Systems should interconnect environments of the same classification (e.g., A‑system UAT with B‑system UAT).

Recommendation: Test environment data should be version‑tracked and correspond to application versions.

Recommendation: Non‑production machine configurations, network topology, and counts may differ from production.

Recommendation: Provide scripts or backups for test seed data to facilitate testing.

Recommended Practices

System Environment Management

Automate one‑click creation of development, integration, UAT, and production environments according to standards.

Strictly prohibit manual changes in production; any change requires detailed plan and approval.

Maintain a unified infrastructure baseline that is regularly updated and hardened, offering development interfaces for downstream use.

Ensure system environments comply with security hardening standards.

Adopt Infrastructure‑as‑Code for managing infrastructure environments.

Common Operations

Environment Request and Preparation At project start, request required environments via ITSM tickets; follow host‑naming standards. Record machine‑project mappings in the DevOps platform and assist in writing CI/CD scripts. Deploy a simple "Hello World" to verify the CI/CD pipeline.

Environment Deployment After project initialization in DevOps, trigger deployment directly from the platform.

Temporary Environments When standard environments are insufficient (e.g., for stress testing), request machines and resources via ITSM, ensure cluster and host names reflect purpose, and coordinate with DevOps to set up temporary integration and deployment pipelines.

Environment Recycling Submit a recycle request to decommission resources.

Development to Test Flow Daily front‑end and back‑end integration, as well as upstream/downstream system integration, primarily occurs in Development and SIT environments. Once functionality is stable and passes smoke testing, deploy to UAT for QA acceptance and regression testing. Any release that fails the smoke test must be rolled back or fixed immediately to avoid impacting QA or users.

Metrics

Deployment success rate for each environment.

Resource cost utilization for each environment instance.