Choosing the right amount of shards and replicas

In the beginning, when you started using Elasticsearch, you probably began by creating the index, importing your data to it and, after that, you started sending queries. We are pretty sure all worked well—at least in the beginning when the amount of data and the number of queries per second were not high. In the background, Elasticsearch created some shards and probably replicas as well (if you are using the default configuration, for example), and you didn't pay much attention to this part of the deployment.

When your application grows, you have to index more and more data and handle more and more queries per second. This is the point where everything changes. Problems start to appear (you can read about how we can handle the application's growth in Chapter 8, Improving Performance). It's now time to think about how you should plan your index and its configuration to rise with your application. In this chapter, we will...