Storm and IMDB integration for dimensional data
IMDB stands for In-Memory Database. An IMDB is required to keep intermediate results while processing the streaming of events or to keep static information related to events, which is not provided in events, for example. Employee details can be stored in IMDB on the basis of employee IDs and events that are coming in and out of an office. In this case, an event does not contain complete information about employees to save the network costs and for better performance, Therefore, when Storm processes the event, it will take static information regarding the employee from the IMDB and persist it along with the event details in Cassandra or any other database for further analytics. There are numerous open source IMDB tools available on the market, but some famous ones are Hazelcast, Memcached, and Redis.
Let's see how to integrate Storm and Hazelcast. No special setup is required for Hazelcast. Perform the following steps:
- Add the dependencies:
<...
