Writing a Pandas DataFrame object as MongoDB collection using a user defined function

Question

I am using pymongo to insert a pandas dataframe (here it is coll) as collection to MongoDB database using the following code:

mongo_url = "mongodb://%s:%s" % (host, port) # not disclosed
client = pm.MongoClient(mongo_url)
db = client[dbname]
db.coll.insert_many(coll.to_dict('records'))

This works fine. But whenever I am trying to perform the same operation using a user defined function, though I am not getting any Python error, but the database is not updated. Below is my function:

def testWrite(host, port, dbname, coll):
    mongo_url = "mongodb://%s:%s" % (host, port) # not disclosed
    client = pm.MongoClient(mongo_url)
    db = client[dbname]
    db.coll.insert_many(coll.to_dict('records'))
    return 0

Here the insert_many() method is not working (i.e. not updating the database) without throwing any kind of error. Please suggest something so that I can pass the dataframe name in a function and push it to the database as a collection.