Airflow - MSSQL connection works on UI but fails in DAG run - airflow

Question

On Airflow 2.2.4 (Postgres and Celery).
I have a connection created for Microsoft SQL Server. When I try to test the connection, UI asked me to enter API Auth user/password (Basic Auth) and showed up a green flash "Connection successfully tested"
But,
When I use the same connection ID in an Operator definition to run some SQL queries, I am getting an error as below.
pymssql._mssql.MSSQLDatabaseException: (20009, b'DB-Lib error message 20009, severity 9:\nUnable to connect: Adaptive Server is unavailable or does not exist (<IP>)\nNet-Lib error during Connection timed out (110)\n
DB-Lib error message 20009, severity 9:\nUnable to connect: Adaptive Server is unavailable or does not exist (<IP>)\n
Net-Lib error during Connection timed out (110)\n')
Here is what my code from my custom operator looks like,
def get_update_num(self):
conn = None
try:
query = f"""
UPDATE SOME_TABLE
SET COL_A = COL_A+1
OUTPUT INSERTED.COL_A
WHERE COL_ID='12345'
self.log.info(f"SQL = {query}")
conn_id = self.conn_id # This is equal to MSSQL_CONNECTION
hook = MsSqlHook(mssql_conn_id=conn_id)
conn = hook.get_conn()
hook.set_autocommit(conn, True)
cursor = conn.cursor()
cursor.execute(query)
row = cursor.fetchone()
self.log.info(f"row = {row}")
return row[0]
except Exception as e:
message = "Error: Could not run SQL"
raise AirflowException(message)
finally:
if not conn:
conn.close()
Any help would be much appreceated.