9 Running ScyllaDB in production

 

This chapter covers

  • Configuring your ScyllaDB cluster
  • Sizing a production-ready database
  • Performing operations safely in production

Up to this point, you’ve been looking at ScyllaDB from the perspective of a user, examining how the database reads and writes data to fulfill an application’s requirements. In the previous chapter, you followed a top-down view of Scylla’s architecture, beginning at the high-level constructs, such as gossip and the hash ring, continuing down to how ScyllaDB stores data on-disk, and learned how these components interact to perform important cluster operations like repairs.

With that knowledge, you’re ready to become an Experienced and Knowledgeable ScyllaDB Operator and run a ScyllaDB cluster in production. In this chapter, you’ll learn the tools and techniques you need to operate ScyllaDB, beginning with configuring the database and continuing onward to performing increasingly complex cluster operations safely without affecting applications connecting to and using the database. Let’s first look at how you can build a production-level ScyllaDB cluster.

9.1 Building a production cluster

9.1.1 The config file

9.1.2 Seeds

9.1.3 Addresses

9.1.4 Authentication

9.1.5 Authorization

9.1.6 Snitches

9.2 Building your cluster

9.2.1 Designing your cluster topology

9.2.2 Computing your nodes

9.2.3 Testing the cluster

9.3 Managing the cluster

9.3.1 Repairing a node

9.3.2 Backing up your cluster

9.3.3 Compacting a node

9.3.4 Troubleshooting tables

9.4 Managing the node lifecycle

9.4.1 Stopping and starting a node

9.4.2 Replacing a node

9.4.3 Adding a node

9.4.4 Removing a node

Summary