Shulou(Shulou.com)06/03 Report--

1. Stand-alone mode

It is easy to install and runs the service on a single machine with almost no configuration, but only for debugging purposes. Without a distributed file system, read and write the file system of the local operating system directly.

2. Pseudo-distributed mode

Start namenode, datanode, jobtracker, tasktracker, secondary namenode and other processes on a single node at the same time to simulate each node running in a distributed way. The configuration is already very close to fully distributed.

3. Fully distributed mode

A normal Hadoop cluster consists of multiple nodes performing their respective functions.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.