What type of tables in Databricks are available across all clusters?

Global tables in Databricks are accessible from all clusters within the workspace. This means that no matter which cluster you are working with, you can query a global table without worrying about its scope being limited to a particular cluster. Global tables are designed for sharing data across multiple clusters, which is essential for collaborative data analysis and pipeline development.

In contrast, cluster-scoped tables are only available within the cluster where they were created, making them unsuitable for scenarios where you need data to be widely accessible. Notebook-scoped tables are tied to the session of a specific notebook and cannot be accessed by other notebooks or clusters, and temporary tables only exist for the duration of a session in a single cluster. These limitations make global tables the preferred choice for scenarios that require data sharing and cross-cluster access in Databricks.