August 2017
Beginner to intermediate
278 pages
6h 40m
English
LinkedIn first created Kafka for its own log processing use case. As discussed, Kafka stores data for a configured period of time, and the default is seven days. The LinkedIn team felt the need to store data for any batch-reporting purpose or to use it later. Now, to store data in HDFS, which is a distributed storage file system, they started developing a tool that can use a distributed system capability to fetch data from Kafka. They developed Camus, a tool developed using map reduce API to copy data from Kafka to HDFS.
The following image gives a good idea of the Camus ...
Read now
Unlock full access