Kubernetes-Docker Cluster Management
Kubernetes Introduction
Kubernetes is an open source Docker container filing system, Kubernetes referred K8S.
Computing cluster node scheduling and dynamic management of the above operations
, the application will be grouped by the logical unit by using a concept of [Labels] and [PODS] of
1, K8S container for deploying applications, extensions and management
2, K8S provides a container scheduling, resource scheduling, resilient and elastic, deployment management, service discovery and a series of functions
3, Kubernetes goal is to enable the deployment of a container to be shipped Simple and efficient
4 official website: http://www.kubernetes.io
Kubernetes properties
1, self-healing
Restart the failed node failure containers, replacement and redeployment, the number of copies to ensure predictable; kill the health check fails container, and not before is not ready to handle client requests, to ensure that online services are not interrupted.
2, elastically stretchable
Use the command, UI or automatically based on CPU usage and rapid expansion of volume reduction application instances, high availability of business applications to ensure peak concurrent; resource recovery peak when traffic is low, at minimum cost to run the service
3, automatic deployment and rollback
K8S a rolling update the application, update a Pod, rather than delete all Pod, if a problem occurs during the update process, rolls back the changes to ensure that the upgrade is not affected business
4, service discovery and load balancing
K8S provides a unified access entry (UI internal IP address and a DNS address) of a plurality of containers, and containers of all loads associated balanced, so that the user need to consider the problem vessel IP
5, confidential and configuration management
Management of confidential data and application configuration. Without the need to expose sensitive data in the image, the sensitive data to improve security. And may be stored in some common K8S arranged in use to facilitate the application
6, store layout
Mount external storage systems, whether from local storage, public clouds (such as AWS), or network storage (NFS, GlusterFS) are part of a cluster resource use, greatly improving the storage flexibility of use
7, Batch
One-time task, the timing task; meet batch data processing and analysis of scenarios.
Kubernetes cluster architecture and components
master Components
1, cube-apiserver
Unified entrance Kubernetes API, clusters, each component coordinator, to provide RESTful API interface services, resource additions and deletions to all objects of investigation and monitoring operations to APIServer process for submission to the Etcd storage.
2、kube-controller-manager
Conventional processing background tasks cluster, a resource corresponding to a controller, and ControllerManager is responsible for the management of these controllers
3、kube-scheduler
Pod according to scheduling algorithm to select a new Node node is created, which could be deployed, can be deployed on a single node can be deployed on different nodes
4、etcd
Key distributed storage system for storing the cluster state data, such as object information Pod, Server and the like.
node assembly
1, omelets
kubelet is MAster Agent in the node Node, lifecycle management Run container, such container is created, the data volume Pod mounts, download Secret, node status, and obtaining container work. Pod kubelet converting into a group of containers.
2、bube-proxy
Pod achieve network node agent on Noed, maintenance, network planning and load balancing work four re =
3、docker或rocker
Vessel engine, run container