Showing current/actives engines of SQLAlchemy

I have a function to run queries in a MySQL database and stores the results in a DataFrame, like that:

def getData(connect_string, query, echo=False):

    sql_engine = sql.create_engine(connect_string, echo=echo)

    df = pd.read_sql_query(query, sql_engine)

    return df

This function is called multiple times in multiple files, sometimes passing the same connect_string.

At this moment, I'm concern about memory usage and/or have too many connections on database. My first thought was kill Engine after use, but the official docs says:

The Engine is not synonymous to the DBAPI connect function, which represents just one connection resource - the Engine is most efficient when created just once at the module level of an application, not per-object or per-function call.

So now I'm looking for a way to check if I have an active connection with same connect_string and use they instead of create a new one, but I cannot find any reference/documentation explaining how to do this.

Right now, I'm considering create a dict to store created connections, like that:

engines = {'mysql+pymysql://scott:tiger@localhost/test1': <Engine Object1>,
           'mysql+pymysql://scott:tiger@localhost/test2': <Engine Object2>}

And then modify my function to this:

def getData(connect_string, query, echo=False):
    if connect_string not in engines.keys():
        sql_engine = sql.create_engine(connect_string, echo=echo)
        engines[connect_string] = sql_engine
    else:
        sql_engine = engines[connect_string]
    df = pd.read_sql_query(query, sql_engine)

    return df

But I don't think that is the best approach to do this.

Any thoughts?