Python - pool.map_async pass params to the function

I'm getting data into a list of pandas.Dataframe. Then I need to send this data to a DB. It tool ~2min per iteration.

So I wanted to know of parallel processing is a good idea ? (no lock issue or something ?)

So I wanted to implement it from :

for df in df_list:
    # Send a DF's batch to the DB
    print('Sending DF\'s data to DB')
    df.to_sql('ga_canaux', engine, if_exists='append', index=False)
    db.check_data()

To something I find about multiprocessing:

with multiprocessing.Pool(processes=4) as pool:
    results = pool.map_async(df.to_sql(???), df_list)
    results.wait()

How can I pass the params I need in df.to_sql with map_async ?

EDIT :

I try something to pass N arguments like :

pool = multiprocessing.Pool()
    args = ((df, engine, db) for df in df_list)
    results = pool.map(multiproc, args)
    results.wait()

but I get the error TypeError: can't pickle _thread._local objects

EDIT2:

I changed a bit the way I'm doing mp and it's kinda work (179s vs 732s with the same sample dataset). But I'm facing an error when I try to read from the DB inside the pool.

# Connect to the remote DB
global DB
DB = Database()
global ENGINE
ENGINE = DB.connect()

pool = mp.Pool(mp.cpu_count() - 1)
pool.map(multiproc, df_list)
pool.close()
pool.join()

def multiproc(df):
    print('Sending DF\'s data to DB')
    df.to_sql('ga_canaux', ENGINE, if_exists='append', index=False)
    DB.check_data() // HERE

Error :

(psycopg2.OperationalError) SSL SYSCALL error: EOF detected
 [SQL: 'SELECT COUNT(*) FROM ga_canaux'] (Background on this error at: http://sqlalche.me/e/e3q8)

EDIT 3 When I try a on a bigger sample I got a DB timeout : psycopg2.DatabaseError: SSL SYSCALL error: Operation timed out