Python MySQLdb: connection.close() VS. cursor.close()

If I use MySQLdb to connect to MySQL-Server through Python. I create a connection and a cursor like this:

connection = MySQLdb.connect(...)
cursor = connection.cursor()
# process

When the MySQL-processing is done one should close the connection. Now I was wondering: Is it sufficient to close the connection by doing:

connection.close()

or do I have to close the cursor first and then the connection? Like this:

cursor.close()
connection.close()

I will re-iterate the best practice at everyone who comes across the sql connection using MySQLdb or any other package to connect python2/3 needs to know this

(Following mock run assumes that you have a table named tablename in your sql database. It has got 4 columns/fields with names field1,field2,field3,field4). If your connection is local (same machine) then it is 127.0.0.1 also known as "localhost".

The process is to be simple 7 steps

1. Create connection
2. Create cursor
3. Create Query string
4. Execute the query
5. Commit to the query
6. Close the cursor
7. Close the connection

Here is a simple step by stem mock run

mydb = MySQLdb.connect(host=host, user=user, passwd=passwd, db=database, charset="utf8")
cursor = mydb.cursor()
query = "INSERT INTO tablename (text_for_field1, text_for_field2, text_for_field3, text_for_field4) VALUES (%s, %s, %s, %s)"
cursor.execute(query, (field1, field2, field3, field4))
mydb.commit()
cursor.close()
mydb.close()

Connection and cursor are different. connection is at the SQL level while cursor can be considered as a data element. You can have multiple cursors on the same data within single connection. It is an unusual occurrence to have multiple connections to same data from the same computer.

More has been described here "The cursor paradigm is not specific to Python but are a frequent data structure in databases themselves.

Depending on the underlying implementation it may be possible to generate several cursors sharing the same connection to a database. Closing the cursor should free resources associated to the query, including any results never fetched from the DB (or fetched but not used) but would not eliminate the connection to the database itself so you would be able to get a new cursor on the same database without the need to authenticate again."

Closing the cursor as soon as you are done with it is probably the best bet, since you have no use for it anymore. However, I haven't seen anything where it's harmful to close it after the db connection. But since you can set it as:

cursor = conn.cursor()

I recommend closing it before, in case you accidentally assign it again and the DB connection is closed as this would throw an error. So you may want to close it first in order to prevent an accidental reassignment with a closed connection.

(Some don't even close it at all though as it gets collected by the garbage collector (see:https://stackoverflow.com/questions/2330344/in-python-with-sqlite-is-it-necessary-to-close-a-cursor))

References: https://stackoverflow.com/questions/5669878/python-mysqldb-when-to-close-cursors

https://stackoverflow.com/questions/2330344/in-python-with-sqlite-is-it-necessary-to-close-a-cursor

Closing a connection should be good enough here in this particular context. If you are working with multiple cursors etc. you need to care about proper resource management.

Use with, this tool allows you to create a temporary cursor that will be closed once you return to your previous indentation level.

from contextlib import closing
with closing( connection.cursor() ) as cursor:
    (indented) use the cursor

(non-indented) cursor is closed.
connection.close()