by Angela Guess
Hovhannes Avoyan is working on a series of blogs regarding some of the best options for NoSQL solutions. In his first post, Avoyan focuses on Apache Cassandra: “Apache Cassandra - like so many other NoSQL tools – is an open-source distributed database system. It was originally created at Facebook in 2008 – but with plenty of input from other sources, notably Google (BigTable) and Amazon (Dynamo). The result is that Apache Cassandra has an extremely scalable and fault-tolerant data infrastructure. Cassandra solves both real-time and analytical big data problems, from write-intensive workloads, to sub-millisecond caching layer reads, to analytical workloads involving petabytes of data using MapReduce.”
Avoyan continues, “Cassandra can integrate with Hadoop to provide a single solution for both analytics and real time needs. Hadoop MapReduce offers the ability to run massive analytical queries against terabytes of data. Cassandra offers caching on each of its nodes. Powered with Cassandra’s scalability characteristics, you can incrementally add nodes to the clusters to keep as much of your data in memory as you need. Thus, there’s no need for a separate caching layer. The Cassandra ring is composed on identical nodes and data is automatically replicated to multiple nodes. So, any node can be added/removed easily without manual migration of data. The result is since every node within the cluster is identical, there is no single point of failure or bottlenecks.”
Read more about Apache Cassandra here.
photo credit: Cassandra
