Mesos in Action
DESCRIPTION The modern data center is a complex arena, with physical and virtual servers, multiple OS environments, and complex networking that frequently spans multiple locations. The need to simplify has never been greater. Mesos, an innovative open-source cluster management platform, transforms the whole data center into a single pool of compute, memory, and storage resources that can be allocated, automated, and scaled as if working with a single super-computer. Mesos is an ideal environment for deploying containerized applications at scale, and it's generating a huge buzz in the big data world as a saner environment for running Spark and Hadoop. Mesos in Action introduces the Apache Mesos cluster manager and the concept of application-centric infrastructure. It guides readers from their first steps in deploying a highly-available Mesos cluster through deploying applications in production and writing native Mesos frameworks. It will show how to scale to thousands of nodes, while providing resource isolation between processes using Linux and Docker containers. It contains practical techniques for deploying applications using popular key frameworks, including Marathon, Chronos, and Aurora. Along the way, the book dives into Mesos internals, including fault tolerance, slave attributes, and resource scheduling and Mesos administration, including logging, monitoring, framework authorization, and slave recovery. KEY FEATURES Gets readers up and running quickly Covers need to know key concepts Combines community knowledge with real-world experience AUDIENCE Readers should be familiar with the core ideas of data center administration, including networking, virtualization, and application deployment on Linux systems. The Python-based code examples should be clear to readers using Mesos bindings for other popular languages, including C++, Go, and Scala. ABOUT THE TECHNOLOGY Apache Mesos is clustering software that allows IT professionals to abstract datacenter resources - CPU, memory, and storage - away from physical and virtual infrastructure and into a Mesos cluster. This allows software developers to write applications that use entire data centers as if they were a single, large computer.