ENG-2640: MySQL periodic integration test (#1179)

Sets up MySQL data integration tests to run periodically.

Co-authored-by: Andre Giron <[email protected]>

.github/workflows/periodic-integration-tests.yml

@@ -103,6 +103,23 @@ jobs:
     timeout-minutes: 60
     name: SDK Integration Tests against Data Connectors
     services:
+      mysql:
+        image: mysql:8.0
+        env:
+          MYSQL_USER: aqueduct
+          MYSQL_PASSWORD: Password123!
+          MYSQL_DATABASE: aqueducttest
+          MYSQL_ROOT_PASSWORD: Password123!
+        ports:
+          # Maps tcp port 3306 on service container to the host
+          - 3306:3306
+        options: >-
+          --name=mysql
+          --health-cmd="mysqladmin ping"
+          --health-interval=10s
+          --health-timeout=5s
+          --health-retries=3
+
       postgres:
         image: postgres:15
         env:

integration_tests/sdk/data_integration_tests/conftest.py

@@ -15,6 +15,7 @@
         ServiceType.SQLITE,
         ServiceType.SNOWFLAKE,
         ServiceType.MARIADB,
+        ServiceType.MYSQL,
     ],
     "s3_test": [ServiceType.S3],
     "mongo_db_test": [ServiceType.MONGO_DB],

integration_tests/sdk/setup_integration.py

@@ -164,6 +164,10 @@ def _setup_postgres_db():
     _execute_command(["aqueduct", "install", "postgres"])
 
 
+def _setup_mysql_db():
+    _execute_command(["aqueduct", "install", "mysql"])
+
+
 def _setup_relational_data(client: Client, db: RelationalDBIntegration) -> None:
     # Find all the tables that already exist.
     existing_table_names = set(db.list_tables()["tablename"])
@@ -219,6 +223,9 @@ def setup_data_integrations(client: Client, filter_to: Optional[str] = None) ->
             if integration_config["type"] == ServiceType.POSTGRES:
                 _setup_postgres_db()
 
+            if integration_config["type"] == ServiceType.MYSQL:
+                _setup_mysql_db()
+
             client.connect_integration(
                 integration_name,
                 integration_config["type"],

sdk/aqueduct/integrations/sql_integration.py

@@ -26,7 +26,7 @@
 
 LIST_TABLES_QUERY_PG = "SELECT tablename, tableowner FROM pg_catalog.pg_tables WHERE schemaname != 'pg_catalog' AND schemaname != 'information_schema';"
 LIST_TABLES_QUERY_SNOWFLAKE = "SELECT table_name AS \"tablename\", table_owner AS \"tableowner\" FROM information_schema.tables WHERE table_schema != 'INFORMATION_SCHEMA' AND table_type = 'BASE TABLE';"
-LIST_TABLES_QUERY_MYSQL = "SELECT table_name FROM INFORMATION_SCHEMA.TABLES WHERE table_type = 'BASE TABLE' AND table_schema NOT IN ('mysql', 'sys', 'performance_schema');"
+LIST_TABLES_QUERY_MYSQL = "SELECT table_name AS tablename FROM INFORMATION_SCHEMA.TABLES WHERE table_type = 'BASE TABLE' AND table_schema NOT IN ('mysql', 'sys', 'performance_schema');"
 LIST_TABLES_QUERY_MARIADB = "SELECT table_name AS \"tablename\" FROM INFORMATION_SCHEMA.TABLES WHERE table_type = 'BASE TABLE' AND table_schema NOT IN ('mysql', 'sys', 'performance_schema');"
 
 LIST_TABLES_QUERY_SQLSERVER = (
@@ -53,6 +53,7 @@ def list_tables(self) -> pd.DataFrame:
         Returns:
             pd.DataFrame of available tables.
         """
+
         if self.type() in [ServiceType.BIGQUERY, ServiceType.SNOWFLAKE]:
             # Use the list integration objects endpoint instead of
             # providing a hardcoded SQL query to execute