Scalable/Elastic Datacenter
The Challenge
Enterprises need to manage, scale, and store their data, which is exploding in volume, type, and velocity. Information creation is so prolific it defines a new category: big data. Big Data offers tremendous potential for deep, real-time insights that provide greater efficiencies, ROI and opportunity for business transformation.
Businesses worldwide are not only focused on reducing costs and improving efficiency, but are also addressing business growth needs including business analytics and real-time access to their data. Enterprises are looking for greater efficiencies beyond individual department and the enterprise. Combined with rapid growth of Internet scale data and the availability of new technologies to harness this big data, businesses are easily able to respond to opportunities leading to greater competitive advantage.
The Solution
An ideal scalable system does not require reconfiguration and supports incremental addition of hardware. Not many systems can pull off scalability without the need to reconfigure and rebooting. The Apache Cassandra distributed database is a system that effectively provides ideal horizontal scalability.
Cassandra is a distributed database composed of a cluster of nodes that load balance storage and request processing. With Cassandra you can add new nodes to the cluster any time with out interruption in service. As your application grows and meets its throughput capacity, Cassandra allows gracefully adding commodity hardware and incrementally scaling on demand. New resources in the cluster are automatically detected and efficiently used.
Cassandra employs many sophisticated techniques described in Amazon’s Dynamo’s paper and Google’s BigTable paper. Some of these techniques include consistent hashing, reconciliation of reads, anti-entropy, gossip based protocols and membership and failure detection.
