In the previous chapter, we successfully built our own cluster system with Proxmox on both the two-node and three-node models. We also know that a quorum disk is essential for a two-node cluster to operate. But what if we have a node failure? What will happen to the cluster and how will it behave?
In this chapter, we are going to test whether our newly built cluster is protected by HA or not. To understand the whole picture of different scenarios, the testing will be conducted in both two-node and three-node clusters. Before we can begin our test, we have to make some changes to our DRBD storage. The key sections that will be covered in this chapter are as follows:
Storage preparation for an LVM shared storage
Demonstration of live migration for OpenVZ and KVM
Building a VM with HA protection
Testing different scenarios, including network failure, network switch failure, and cluster node failure