On the left is my personal, if you need further communication, please add. On the right is my public number "OpenStack private cloud", if interested, please pay attention.
in the previous post, " Kubernetes Getting Started learning mentioned in the Katacoda experimental environment can be further familiar with the use of k8s , the specific URL is:
https://www.katacoda.com/courses/kubernetes/
Recently spent some time to complete these experiments, a total of 17 experimental scenes, each scene about 10 minutes , after these experiments, the k8s have a general understanding. These scenarios contain these things:
Learn how to start a single-node Minikube cluster with DNS and Kube UI
Booting kubernetes clusters with Kubeadm
Learn how to use Kubectl to start a container and make it accessible
Learn how to use YAML definitions to deploy containers
This scenario teaches you how to deploy the Guestbook sample using Kubernetes
Learn about the different network methods available
Learn how to define host-and path-based ingress routing
Learn how to use liveness and readiness probes to ensure container health
Learn how to deploy a Cri-o-based KUBEADM cluster
Learn how to run a stateful service on kubernetes
Learn how kubernetes helps ensure confidentiality
Learn how to deploy an existing Docker compose definition using Kompose
How to get the running service from the source code in Kubernetes
Use Kubernetes's Helm Package Manager to deploy Redis
Use Kubernetes's Helm (Classic) Package Manager to deploy Redis
Learn how to run weave Scope on kubernetes
Learn how to deploy server-less functionality to kubernetes using kubeless
17 experiments covered the installation configuration and use of the k8s, in addition, for the practical use of the helm can achieve standardized modular functional components (such as Web Services, MARIADB, Redis, etc.) automated deployment, Includes automated image downloads and a YAML deployment profile that makes its own images, uploads images, and automatically configures standardized k8s based on Dockerfile. There is also a k8s visualization component weave Scope, which can realize the management of the Web GUI of the whole k8s cluster, more intuitive and convenient management and maintenance k8s cluster.
After the experiment, personal feeling k8s really achieved the upper application and the infrastructure, the original need to consider the platform differences, infrastructure, dual-machine redundancy, capacity expansion, software version upgrades and so on a series of very human and time things, in the k8s inside all is very very simple to be solved.
K8s is a container-managed Docker,node node that is similar to OpenStack-based compute nodes that use Docker to carry out the hosting of the container, and master like the OpenStack Controller node completes the management of node and the scheduling of pods, Automated health monitoring, node resource monitoring, and other automated pod copy deployment or migration.
K8s's main management command is KUBECTL, can accomplish most of the k8s management tasks, conceptually, if you want to learn k8s, you must first familiar with the use of Docker, KUBECTL related Management command of pod structure and Docker very similar.
Developers using k8s or some complex, if need to smooth and efficient use of k8s, should still need to use Helm,helm to achieve most of the application of the standard modularity, such as Web applications, various middleware applications, all kinds of open source data applications, but need to solve the problem of network speed, Because all applications are implemented by the container, the general container file is larger, so the subsequent use needs to consider preparing the local mirror in advance.
Usually k8s itself monitoring and maintenance, can be done by deploying weave Scope.
For the k8s application scenario, is very suitable for the application complex, the software application version changes frequently, the resource use change big Internet enterprise, for small and medium scale IT system, the individual thought still has some small material big use. In addition, you can also consider the direct use of Red Hat k8s commercial version of OpenShift, you do not have to deploy K8S platform, directly use.
Personally, for k8s, should be seen as a revolutionary product, because it really does the upper layer for the underlying infrastructure decoupling, if a company uses k8s architecture, the maintenance of the application is a container maintenance and flow, can be placed on the local physical machine k8s platform, can also be directly to the public cloud, and the degree of automation is high enough to greatly liberate the maintenance personnel, so that the maintenance staff will be more focused on platform architecture aspects.
Kubernetes (k8s) Experiment Summary