Kubernetes: Difference between revisions

From Wiki RB4
Line 6: Line 6:


* '''Cluster''' is a group of machines running Kubernetes
* '''Cluster''' is a group of machines running Kubernetes
* '''Master System''' is the system which contrals a cluster in with which one interacts. The master manages the cluster like load balancing, backup, replication, scheduling, ...
* '''Master System''' is the system which contrals a cluster in with which one interacts. The master manages the cluster like load balancing, backup, replication, scheduling, ... The master system is the control pane consisting of the components:
** kube-apiserver: this is an API server that exposes the Kubernetes API
** etcd: key-value store of the cluster data
** kube-scheduler: monitors pods
** kube-controller-manager: runs controller processes
** cloud-controller-manager: embeds cloud-specific control logic
* '''Nodes''' are the machines in a cluster. Each node has a container runtime such as [[Docker|Docker]]. A Node can host multiple Pods.
* '''Nodes''' are the machines in a cluster. Each node has a container runtime such as [[Docker|Docker]]. A Node can host multiple Pods.
* '''Pods''' is a logical group of containers on a node.  
* '''Pods''' is a logical group of containers on a node.  
* Pods are wrapped as a Kubnetes '''Service'''.
* Pods are wrapped as a Kubnetes '''Service'''.
* '''Volumees''' are used to store data
* '''Volumees''' are used to store data

Revision as of 16:39, 6 February 2022

Concepts

  • Cluster is a group of machines running Kubernetes
  • Master System is the system which contrals a cluster in with which one interacts. The master manages the cluster like load balancing, backup, replication, scheduling, ... The master system is the control pane consisting of the components:
    • kube-apiserver: this is an API server that exposes the Kubernetes API
    • etcd: key-value store of the cluster data
    • kube-scheduler: monitors pods
    • kube-controller-manager: runs controller processes
    • cloud-controller-manager: embeds cloud-specific control logic
  • Nodes are the machines in a cluster. Each node has a container runtime such as Docker. A Node can host multiple Pods.
  • Pods is a logical group of containers on a node.
  • Pods are wrapped as a Kubnetes Service.
  • Volumees are used to store data