OpenShift single-DC HA design
In the previous chapter, we briefly covered HA and OpenShift HA in general, but we did'nt discuss how to practically design OpenShift in your data center environment.
Let's recall what main OpenShift components we have and how they provide redundancy:
- Openshift infrastructure nodes
- OpenShift master nodes
- OpenShift nodes
- Etcd key-value store
- Persistent storage
OpenShift infrastructure nodes
OpenShift infrastructure nodes are essential components that provide access from the outside of an OpenShift cluster. OpenShift infrastructure nodes scale horizontally, which means that we can add as many nodes as we need in order to add network throughput. If you recall from the previous chapter, we need to consider which VIP method to use. We have two main VIP methods:
- VIP using an external load balancer:
- IP failover using keepalived:
Both methods have their own pros and cons, but the one that provides better scalability and smooth migration to multi-DC design is the VIP using...
