Handle case of 0 connect retries better

With a default of 0 connect_retries, confusing log messages show up.
The logs suggest that a retry is going to be carried out, but it isn't.
Also, a sleep is carried out, even if no retries are to happen. This
check prevents this from happening.

CHANGELOG.md

@@ -6,6 +6,12 @@
 ### Under the hood
 - Add changelog, issue templates ([#119](https://github.com/fishtown-analytics/dbt-spark/pull/119), [#120](https://github.com/fishtown-analytics/dbt-spark/pull/120))
 
+### Fixes
+- Handle case of 0 retries better for HTTP Spark Connections ([#132](https://github.com/fishtown-analytics/dbt-spark/pull/132))
+
+### Contributors
+- [@danielvdende](https://github.com/danielvdende) ([#132](https://github.com/fishtown-analytics/dbt-spark/pull/132))
+
 ## dbt-spark 0.18.1.1 (November 13, 2020)
 
 ### Fixes

dbt/adapters/spark/connections.py

@@ -401,7 +401,7 @@ def open(cls, connection):
                         msg += ', is your token valid?'
                     raise dbt.exceptions.FailedToConnectException(msg) from e
                 retryable_message = _is_retryable_error(e)
-                if retryable_message:
+                if retryable_message and creds.connect_retries > 0:
                     msg = (
                         f"Warning: {retryable_message}\n\tRetrying in "
                         f"{creds.connect_timeout} seconds "