April 2018
Intermediate to advanced
468 pages
14h 34m
English
There are two broad categories of failure:
Pods might fail for many reasons. In a properly configured Kubernetes cluster (a cluster federation member or not), pods will be managed by services and ReplicaSets that can automatically handle pod failures. It shouldn't impact cross-cluster routing and load balancing done by the federated ingress. A whole cluster might fail due to problems with the data center or global connectivity. In this case, the federated services and federated ReplicaSets will ensure that the other clusters in the federation run enough pods to handle the workload, and the federated ingress will take care of routing client requests away from the failed ...
Read now
Unlock full access