E-mail of the author of the article: [email protected] Address: Huizhou, Guangdong
▲ Purpose of this chapter
⚪ Understand the concept of Ganglia;
⚪ Understand the topology and execution process of Ganglia;
⚪ Master the installation and operation of Ganglia;
1. Introduction
1 Overview
1. Flume was originally developed by Cloudera and later contributed to Apache as a distributed, reliable mechanism for collecting, aggregating and moving log data.
2. In big data, more than 70% of the data in actual development comes from logs - logs are the cornerstone of big data.
3. Flume provides a very simple and flexible streaming mechanism for logs.
4. Version:
a. Flume0.X: also known as Flume-og. Relying on Zookeeper, the structure configuration is relatively complicated, and this version has been deactivated in the market now.
b. Flume1.X: also known as Flume-ng. It does not depend on Zookeeper, and its structure configuration is relatively simple. It is a commonly used version on the market.
2. Basic concepts
1. Event：
a. In Flume, each collected log is encapsulated into an Event object - in Flume, an Event corresponds to a log.
b. Event is essentially a json string, which contains two fixed partsÿ