As organizations start employing big data and NoSQL-based solutions across a number of projects, a data layer comprising of RDBMSs alone is no longer the best solution for all the use cases in a modern enterprise application. The following figure illustrates a situation that is rapidly disappearing across the industry:

A more typical scenario comprising of multiple types of data stores is shown in the following figure. Applications today use several types of data stores that represent the best fit for a given set of use cases. Using multiple data storage technologies, chosen based upon the way data is being used by applications, is called polyglot persistence. For example, Apache Spark is an excellent enabler of this and other similar persistence strategies in the cloud or on-premise deployments:

>

In the next section, we discuss the key concepts of batch and stream processing architectures.