Slide 25
Slide 25 text
How Many Clusters?
OpenShift Sizing
25
▸ A cluster is a distinct/discrete OpenShift environment with its own set of masters and infrastructure hosts
▸ A customer determines whether or not they wish to have discrete/separated OpenShift clusters
▸ Dev, Stage, Prod
▸ High Availability and Disaster Recovery
▸ Cluster per “tenant” or LOB vs. many tenants sharing the same cluster
▸ What are you optimizing for: Utilization, Speed/Flexibility, Ability to change quickly
▸ Compliance and regulatory reasons
▸ Other reasons
▸ Each discrete OpenShift environment = 1 Cluster
▸ For highly-available clusters, 3 Masters are required.
▸ For highly-available infrastructure (Registry, Router, Logging, Metrics, etc), at least 3 Infrastructure Nodes are required.
Example:
Customer will run development, staging, and production applications in OpenShift inside a single HA OpenShift cluster
(1 cluster - 3 masters, 3 infrastructure).