Skip to main content
Sign in

Workspaces & Environments

  • Updated

In Coalesce, you build and refine your data pipeline(s) in Workspaces. Once development is complete, you deploy your work to designated higher Environments, mirroring the development lifecycle defined by your organization, e.g. DEV -> QA -> UAT -> PROD. Workspaces and Environments are thus interrelated, but are very different in their usage contexts and behavior. A single Workspace may map to one (1) or more Environments, but an Environment will map to a single Workspace.

A comparison of the two objects is provided below:

Workspace: A sandbox-style integrated development environment in which the build of the data pipeline(s) is completed, refined, and preliminarily validated before it is merged into the comprehensive codebase and subsequently promoted to a higher environment. A "playground" environment without expectations of stability, persistence, etc. 
  • All Coalesce development is completed in a Workspace
  • Objects are created and data is loaded and refreshed manually on-demand, either for individual objects or for subsets of or the full data pipeline.
  • When a change is made to an object (e.g. adding or dropping a column, renaming a column, etc.), the object is dropped and recreated vs. ALTERed.
  • Deleted objects are not cleaned up (DROPPED) in the backend Snowflake instance
Environment:

A non-development environment (e.g. QA, UAT, PROD) to which the data pipelines built in a Coalesce Workspace(s) are deployed and refreshed. A stable and persistent environment, typically with governance standards in place.
  • No Coalesce development can be completed directly against an Environment
  • Objects are created in a steady state via a defined Git branch and commit
  • Data loading and refresh is kicked off on demand, via a schedule or designated trigger, and/or via an automated process, like a CI/CD pipeline
  • When a change is made to an object (e.g. adding or dropping a column, renaming a column, etc.), the object is ALTERed vs. dropped and recreated.
  • Deleted objects are cleaned up (DROPPED) in the backend Snowflake instance