pyodbc is very slow on my machine but not on other machines

Question

I am accessing a db2 (9.7) server running on the same network as my computer using python 2.7 and pyodbc 3.0.6.

The following is a very stripped down example of something similar to what i am trying to do.

import pyodbc as db

def dbcur(connectname):
    con = db.connect(connectname)
    cur = con.cursor()
    return cur,con


def main():

    print 'a'
    cur,con = dbcur('DSN=MYODBCCON')
    print 'b'


    # X.name is a short varchar
    qry = """
        select
        X.name as N
        from schema.table as X
        """

    print 'c'

    cur.execute(qry)

    print 'd'

    c = 0
    for row in cur:
        c+=1
        if not c%100:
            print c, row

    return

if __name__ == '__main__': main()

This code runs very slow on my windows 8 computer (100 rows in 10 seconds). However, on my colleagues computer with windows 7 this is lightning fast. His computer is simmilarly specced to mine, and he has the same versions of pyodbc and python, so i do not belive this to be an issue.

The problem persists when i use both IBM DB2 ODBC DRIVER and IBM DATA SERVER DRIVER FOR ODBC.

Any suggsetions as to why this is so slow?

I am aware of things such as cur.fetchmany(), but i still expect this code to be much faster without using this.

UPDATE:

It turns out that ODBC was slow because tracing was turned on. I disabled tracing and the performance was back to what I expected it to be. (Tracing can be found under ODBC Data Sources in Windows 7 and 8.)

Answer 1

It turns out that ODBC was slow because tracing was turned on. I disabled tracing and the performance was back to what I expected it to be. (Tracing can be found under ODBC Data Sources in Windows 7 and 8.)