In Chapter 1, Building Your Own Kubernetes Cluster, we set up our Kubernetes cluster with several different provisioning tools. Based on your Kubernetes provisioning tool, there is an easy way to set up EFK stack. Note that Elasticsearch and Kibana are heavy-duty Java applications. They require at least 2 GB of memory each.

Therefore, if you use minikube, your machine should have at least 8 GB of memory (16 GB is recommended). If you use kubespray or kops to set up Kubernetes cluster, Kubernetes node should have at least 4 core CPUs and 16 GB of memory in total (in other words, if you have 2 nodes, each node should have a minimum of 2 core CPUs and 8GB of memory).

In addition, in order to demonstrate how to gather the application logs efficiently, we create one additional namespace. It will help you to search your application log easily:

$ kubectl create namespace chap9
namespace "chap9" created