Python & Cassandra - Best Friends
0 likes691 views
This document discusses the benefits of using Python and Cassandra together. It provides an overview of using virtual environments in Python to isolate projects. It also summarizes the Python driver for Cassandra, which allows connecting to Cassandra clusters and executing queries, and cqlengine, an object mapper that simplifies working with Cassandra from Python.
Python & Cassandra - Best Friends
- 1. Python & Cassandra, Best Friends Jon Haddad Technical Evangelist,Datastax @rustyrazorblade ©2013 DataStax Confidential. Do not distribute without consent. 1
- 2. Virtual Environments • Isolated environment per project • Build into Python 3.4, available • 3rd party lib in Python 2.7 • pip install virtualenvwrapper • Add to your shell startup file: • source /usr/local/bin/virtualenvwrapper.sh
- 3. Before we start • If you'd like to follow along locally, please: • have cassandra running • clone [email protected]:rustyrazorblade/python-presentation.git • create a virtualenv • mkvirtualenv tutorial • install requirements • pip install -r requirements.txt • Start the notebook • ipython notebook
- 4. Why Python? • Python is easy to read & write • Python is mature • It takes less time to develop an app in Python • Lots of existing libraries • Principle of least astonishment •You can be out the door and iterating while the other dev is still working • In most applications, most time is spent waiting on the database •Dev time is usually more costly than the extra server •Correct data model will save you more time than faster app code •Focus on architecture
- 5. Native Driver Overview • Provides means of connecting to a cluster and executing CQL3 queries • cassandra.cluster.Cluster keeps track of servers & schema • cassandra.cluster.Session manages connection pool • Cluster.metadata has keyspace, table information • Provides concurrency through cassandra.concurrency • session.execute()
- 6. cqlengine overview • Community project (but moving to DataStax) • Created by me & Blake Eggleston (also now at DataStax) • Object mapper on top of the native driver • Removes the need to write most CQL queries manually • Define models, work with regular python types
- 8. ©2013 DataStax Confidential. Do not distribute without consent. 8