Chapter 10.  Strategies for Scaling

Up until now, we've generally covered topics that are relevant to any size of infrastructure, big or small. But when an infrastructure starts getting really big, there is an entire class of strategies that must be considered in order to handle the load. In this chapter, we'll cover the following topics:

• Building a hierarchy with syndics

• Using multiple masters

• Testing load with minionswarm.py

• Using external filesystems

• Managing the master using the wheel system

• Using reactors and Thorium to manage the master