Apache Flume is a distributed service designed for efficiently collecting, aggregating, and moving large volumes of log data, primarily aimed at delivering data to Apache Hadoop's HDFS. Its architecture is simple and flexible, focusing on streaming data flows while ensuring robustness and fault tolerance through various failover and recovery mechanisms. This resource addresses issues related to HDFS and streaming data/logs, demonstrating how Flume can effectively resolve these challenges. The book begins with an architectural overview of Flume, detailing each component and guiding readers through the installation and compilation processes. It covers the use of channels and channel selectors, providing in-depth explanations of architectural components such as Sources, Channels, Sinks, Channel Processors, and Sink Groups, along with their configuration options. This allows for customization of Flume to meet specific needs. Additionally, it offers insights into writing custom implementations, enhancing your understanding and ability to apply them. By the end of the book, readers will be equipped to construct a series of Flume agents that transport streaming data and logs from their systems into Hadoop in near real time.
Subas D'Souza Reihenfolge der Bücher (Chronologisch)
