Basic concepts and terminology of Kubernetes
The Node, Pod, Replication Controller, Service, etc. in K8s can be regarded as a kind of "resource object". The resource object can be added, deleted, modified, and checked through the kubectl tool provided by K8s and stored in etcd for persistent化Storage. From this perspective, K8s is a highly automated resource control system that realizes automatic control and automatic error correction by tracking and comparing the difference between the "resource expected state" saved in the etcd library and the "actual resource state" in the current environment. Advanced Features.
1. Master
The Master node is responsible for the management and control of the cluster. Basically, all control commands of K8s are run on the Master node. The Master node usually occupies an independent server (3 servers are recommended for high-availability deployment).
The following key processes are running on the Master node:
- Kubernetes API Server (kube-apiserver): Provides the key service process of the HTTP Rest interface. It is the only entry for adding, deleting, modifying, and checking all resources in Kubernetes, and it is also the entry process for cluster control.
- Kubernetes Controller Manager (kube-controller-manager): The automation control center for all resource objects in Kubernetes.
- Kubernetes Scheduler (kube-scheduler): The process responsible for resource scheduling (Pod scheduling).
The Master node also needs to start an etcd service, because the data of all resource objects in K8s are all stored in etcd.