Hi,

When creating a Spark view using SparkSQL ("CREATE VIEW AS SELCT ...") per default, this view is non-temporary - the view definition will survive the Spark session as well as the Spark cluster.

In PySpark I can use DataFrame.createOrReplaceTempView or DataFrame.createOrReplaceGlobalTempView to create a temporary view for a DataFrame.

Is there a way to create a non-temporary Spark View with PySpark for a DataFrame programatically?

spark.sql('CREATE VIEW AS SELCT ...') doesn't count 😉

I did not find a DataFrame method to do so...