November 11th, 2013

“Motivation for using Cassandra was: no single point of failure, easy setup and linear scalability.”

-Joshua Yuen, Senior Software Engineer at Polyvore

Joshua Yuen Senior Software Engineer at Polyvore


What does Polyvore do?

Polyvore is a new way to discover and shop for things you love. Visitors come to Polyvore to curate products, create sets, discover and buy lifestyle items.


How are you using Cassandra?

First for document oriented storage of image metadata and storage of messages/comments. We also make use of the TTL to store user stream information. We use Datastax Enterprise to store time series data and run MapReduce on Hadoop on top of Cassandra.


What was the motivation for using Cassandra and what other technologies was it evaluated against?

Motivation for using Cassandra was: no single point of failure, easy setup and linear scalability. The other technology we evaluated against was MongoDB. We tried to run MapReduce on MongoDB, it was not as easy as DSE.


Can you share some insight on what your deployment looks like?

We host in our own Datacenter. We have 2 main clusters, 12 nodes, SSD and around 8 TB of data.


What’s your favorite part about Apache Cassandra?

We have been using Cassandra since 0.6, it is fascinating to see how much it has been evolved.


What would you like to see out of Apache Cassandra in future versions?

I would like to see improvement on doing large range queries.


What’s your experience with the Apache Cassandra community?

It’s great that there is so much information, from administration, data modeling, different use cases, and more.