July 2017
Beginner to intermediate
378 pages
10h 26m
English
Next, we will review the node types that make up a Hadoop cluster, what role they play, and how they integrate to ensure the cluster is a functioning unit. Servers in the same cluster communicate frequently with each other; they are very chatty.
Most cluster configurations keep the physical units close to each other (in networking terms) to minimize bandwidth bottlenecks. The managed cloud-based distributions of Hadoop handle this for you on the setup. But if you decide to go the route of manually creating your own cluster, you will need to keep networking considerations on top of mind when designing the architecture.
Getting to know Linux
Linux is ubiquitous in big data computing clusters. Since it is open source, ...