Shulou(Shulou.com)06/03 Report--

First, the basic overview of Helm 1, the purpose of using Helm?

From the previous application deployment process, it is known that when deploying containerized applications on kubernetes systems, resource configuration manifests need to be manually written in advance to define resource objects, and each configuration definition is basically hard-coded, which basically cannot be reused. For large-scale application scenarios, application configuration, distribution, version control, lookup, rollback, and even viewing will be a user's nightmare. Helm can greatly simplify the difficulty of application management.

2What is the "Gimme Helm"?

To put it simply, Helm is the application package manager of kubernetes, similar to yum or apt-get on Linux systems, which can be used to help users find, share and use kubernetes applications. The current version is maintained by CNCF (Microsoft,Google,Bitnami and Helm community). Its core packaging function component, called chart, can help users create, install and upgrade complex applications.

Helm packages kubernetes resources (Deployment,service, configmap, etc.) into a charts, and each charts produced and tested is saved to the charts repository for storage and distribution. In addition, Helm implements configurable release, which supports version management of application configuration, and simplifies the version control, packaging, release, deletion and update operations of kubernetes deployment applications. The Helm architecture components are shown in the following figure:

3What are the advantages of Helm? Managing complex applications: charts can describe even the most complex program structure, providing a definition of reusable application installation. Easy to upgrade: use in-place upgrades and custom hooks to solve update challenges. Simple sharing: charts is easy to versioning, sharing and hosting through public or private services. Rollback: a quick rollback can be easily achieved using the "helm rollback" command. The core terminology of 4Gen Helm

For Heml, it has the following key concepts:

Charts: a Helm package that contains images, dependencies, resource definitions, etc., needed to run a kubernetes application, and, if necessary, a definition of service; it is similar to APT's dpkg file or yum's rpm file. Repository:Charts repository for centralized storage and distribution of Charts, similar to Perl's CPAN, or python's pyPI. Config: the configuration information used by the application to instantiate the installation run. Release: the application instantiates the configuration and runs a Charts instance in the kubernetes cluster; on the same cluster, a charts can be installed multiple times using different Config, and a new Release is created for each installation. 5Gen Helm architecture

Helm is mainly composed of Helm client, Tiller server and Charts repository (Repository). The communication diagram among Helm members is as follows:

Heml client: Helm client is a command line client tool written in GE language and interacts with Tiller server based on gRPC protocol. It mainly accomplishes the following tasks:

Local charts development. Manage Charts warehouse. Interact with the Tiller server (send Charts to install, query information about release, and upgrade or uninstall existing Release).

Tiller server:Tiller server is a containerized service application that runs in a kubernetes cluster. It receives requests from Helm clients and interacts with kubernetes APi server if necessary. It mainly accomplishes the following tasks:

Listen for requests from Helm clients. Merge charts and configuration to build a Release. Install Charts to the kubernetes reporter and track the appropriate Release. Upgrade and uninstall Charts.

Charts repository: only when there is a need for distribution, Charts files for the same application should be packaged and submitted to a specific charts repository in a compressed archive format. The warehouse can be run either as a public trust guan platform or as a user-built server for specific organizations and individuals.

Second, deploy Helm1 and install Helm Client

There are two ways to install Helm client: precompiled binaries and source code compilation and installation. The precompiled binary program is installed in this paper.

1) download the binary package and install:

Download address of binary installation package: https://github.com/helm/helm/releases, you can choose different versions, such as installing version 2.14.3:

[root@master helm] # wget https://get.helm.sh/helm-v2.14.3-linux-amd64.tar.gz[root@master helm] # tar zxf helm-v2.14.3-linux-amd64.tar.gz [root@master helm] # ls linux-amd64/helm LICENSE README.md tiller# copy or move its binary command (helm) to the directory pointed to by the system PATH environment variable [root@master helm] # cp linux-amd64/helm / Usr/local/bin/# View helm version [root@master helm] # helm versionClient: & version.Version {SemVer: "v2.14.3" GitCommit: "0e7f3b6637f7af8fcfddb3d2941fcc7cbebb0085", GitTreeState: "clean"} Error: could not find tiller// executes the helm version command and finds that the helm client version is v2.14.3 Prompt server tiller is not installed yet.

2) Command completion

Helm has many subcommands and parameters. In order to improve the efficiency of using the command line, it is usually recommended to install helm's bash command completion script as follows:

[root@master helm] # echo "source

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.