Escalating growth in data from web 2.0 applications.
Database 2.0 responds to the performance requirements in data storage and retrieval imposed by web 2.0 application. Web 2.0 applications are highly distributed in nature and have been characterized by the explosive growth in data that are generated. For instance the scale of data generated by social applications has all but excluded traditional RDBMS as viable alternatives.
RDBMS limitations due to high load.
The distributed nature of the internet can easily create situations in which high demand can overload an application RDBMS negatively impacting performance due to excessive read/write operations on disks.
RDBMS have shown limitations in scalability
For a web application introducing new services sometimes entail formulating complex SQL queries (join algorithm) which may not sit well with the database query optimizer.
Database scaling -up or -out does improve performance but only temporarily. Scaling up brings more expensive hardware , while scaling out will require a very complex modification and partitioning of the database tables.
Cassandra a fully open source and highly distributed database
Achieving high availability and recovery for data is done by using data replication across data center. The data model for Cassandra gives it the scalability: columns can be added dynamically under a hierarchical system with a user-defined super column on top. Under this scheme a virtually unlimited number of column can be added
Engineering solutions Cassandra
Very large data centers produce voluminous amount of data that need to be processed almost instantly. For example logs containing users sessions information in some cases not only require persistent storage as well as rapid transaction (write or read operation). RDBMS by the highly structured transactional integrity approach to storage and retrieval can not easily cope with high data throughput. Scalability allows for the addition of storage node to accommodate more data.
Cassandra is a very low cost innovative NoSQL approach to achieving persistent storage for data intensive applications.
The variety of use cases along with its Open Source nature allows for a wide variety of deployments options.
