Remote Jupyter kernel default for new DataSpell notebooks

Remote Jupyter kernel default for new DataSpell notebooks

DataSpell integrates Jupyter Notebook execution with remote kernel connectivity, allowing code to run on external servers through SSH tunnels or Jupyter server URLs configured in the Managed Servers settings. For projects that consistently target the same remote compute environment, establishing a persistent default kernel would reduce repetitive manual selection.

Assuming a current DataSpell release, the unresolved configuration behavior concerns how the IDE determines the initial kernel for a newly created .ipynb file. While the IDE permits switching between local interpreters and remote server connections on a per-notebook basis, it is not clear whether a project-level or IDE-wide preference can override the default kernel assignment. The interaction between project interpreter settings, managed server configurations, and notebook metadata remains ambiguous.

Does DataSpell support enforcing a specific remote Jupyter server kernel as the default for all new notebooks within a project? If a default is configured, how does it apply to existing notebooks that already reference a local kernel? Are there documented limitations on kernel persistence when the remote server connection is interrupted or when switching between projects?