Python utilities for cassandra

At this point in time the only module provided. It provides a near drop in replacement for cassandra.execute_concurrent which is able to execute statements concurrently with a memory footprint in O(concurrency) instead of O(statements). This yields orders of magnitude less memory usage and throughput with large numbers of statements.

This implementation has been offered for integration in the python driver for Cassandra from Datastax: [https://datastax-oss.atlassian.net/browse/PYTHON-274].

The examples below execute with a very limited and constant memory footprint, while the original implementation consumes > 200 megabytes to start with and which quickly grows as statements are being executed.

Insert values from a generator:

insert = session.prepare('insert into tbl (key, value) values (?, ?)')
execute_concurrent(session, ((insert, ('key', i)) for i in xrange(1000 * 1000)))

Calculate the sum where the select statements are yielded from a generator:

select = session.prepare('select value from tbl where key = ?')
print sum(r[0].value for _, r in execute_concurrent(session, ((select, (str(i),)) for i in xrange(1000 *1000))))