Location not empty but not a Delta table

[Distributed_Com]

I need help or insight regarding the following errors. My instructors (Brooke Wenig with Conor Murphy) ran this code successfully on our course video, but I cannot replicate what she did.

Here is the code and below it is the outcome from my Coursera course, workspace section 4.5 notebook:

CREATE DATABASE IF NOT EXISTS Databricks;

USE Databricks;

DROP TABLE IF EXISTS fireCallsDelta;

CREATE TABLE fireCallsDelta

USING DELTA

PARTITIONED BY (City)

AS 

 SELECT * FROM fireCallsParquet

(3) Spark Jobs

• Job 3 View(Stages: 1/1)
• Job 4 View(Stages: 1/1, 1 skipped)
• Job 5 View(Stages: 1/1, 2 skipped)

Error in SQL statement: AnalysisException: Cannot create table ('`spark_catalog`.`databricks`.`fireCallsDelta`'). The associated location ('dbfs:/user/hive/warehouse/databricks.db/firecallsdelta') is not empty but it's not a Delta table

[dennyglee · Accepted solution]

The `fireCallsDelta` table already exists, and it appears it was created as a PARQUET table. A couple of options:

1. DELETE the original Parquet table as a separate statement (i.e., in a separate cell so it can complete itself before any other jobs executing)
2. Rename the table to something else if perhaps others are running the same demo.

[Distributed_Com]

Thanks. Renaming table as you suggested allowed the query to run successfully.

[Distributed_Com]

Done. Thanks.

[gilo12]

---

DELETE the original Parquet table as a separate statement

---

How can this be done? simple query "DROP TABLE .... " still failing with "cannot be found"