View

No Materialized Views are present in Databricks
When we query a view the underlying computation has to be performed again
The data is persisted and is accessible across sessions
They are tied to the database in which they are created

Temporary View

Temporary View deleted once we exist the current Spark Session
The table is not tied to any database

CREATE OR REPLACE TEMP VIEW events_temp_view
AS SELECT * FROM json.`${da.paths.working_dir}/raw/events-kafka/`;

INFO

Opening another Notebook, Install Python Package, Detaching from Cluster all creates a new session

Global Temporary View

Global Temp View is tied to the global_temp Database which is tied to the Spark Cluster
Users from different sessions that are using the same cluster can access the View
All long as the Cluster is running the view will continue to exist