Dataframe won't save as anything - table, global temp view or temp view

Question

A dataframe won't save as a table , be it permanent or temp - no error message is given.

I have tried :

df.createOrReplaceTempView("tmp_table")

# or

df.createGlobalTempView("tmp_table")

# or

df.writeTo("tmp_table")

I try to check the table using:

spark.sql("SHOW TABLES like 'tmp_table'")

# or

spark.sql("SHOW TABLES like 'global_temp.tmp_table'") # for global view

Nothing is displayed, there is not such view or table.

PySpark Version :3.1.2

Any ideas what could be at fault here?

Later Edit: this problem is happening when using spark streaming - it is not replicated in batch

Later Later Edit: Based on some logic change in the script, and because I need to process the json data, if I run this ( first time i use spark context in the script )

df_schema = spark.read.json(df.rdd.map(lambda x: x.data)).schema

the createOrReplaceTempView doesn't work.

If I don't try to do the schema logic and just run this first time with spark context

df = spark.read.json(another_df.rdd.map(lambda x: x.body))

The views are created properly. Could it be a spark context thing? Thanks