With the advent of Big Data and the rapidly growing scale of web-applications, monolithic relational databases were replaced by scalable, partitioned, NoSQL databases and HDFS; individual queries to relational databases were replaced by the likes of Hive and Pig.
This growing scale and partitioned consumption model brought about by these systems, also put forth the need for smooth processing of “streams of events” at scale.
That’s when LinkedIn came up with Samza.
Apache Software Foundation’s incubation project since September 2013, Apache Samza is the distributed stream processing framework that incorporates Apache Kafka for messaging, and Apache Hadoop YARN to provide fault tolerance, processor isolation, security, and resource management.
Navina Ramesh of LinkedIn explains, “we recognized that some of our use cases couldn’t be implemented in Hadoop due to the large turn-around time that batch processing needed.”
“Unlike Hadoop, which is optimized for throughput, Kafka is optimized for low-latency messaging. We built a processing system on top of Kafka, allowing us to react to the messages — to join, filter, and count the messages. The new processing system, Apache Samza, solved our batch processing latency problem and has allowed us to process data in near real-time,” she explains in The New Stack article.
It essentially takes care of stream processing needs within LinkedIn, “to provide a lightweight framework for continuous data processing.” Originally open sourced, the framework enables building applications to process feeds of messages.
Read more here.
(Image credit: Jean and Fred, via Flickr)
