helmdt 0.6.1

A HELM Deployment toolkit

HELM Deployment Toolkit

A deployment toolkit that support:

• Build and push docker image. (Now supported container repositories: AzureCR)
• Apply helm packages and values.
• Link to aks. (Azure kubenetes services)

And you could use it with both command-line and guiding ways.

• Installation
  • via pip
  • via release
• Guiding Way
  • Manually specify the profile
    • -f | File
  • Actions
    • depcheck
    • build
    • apply
    • link
• Command Line Way
  • --yes | Yes
  • -n/--name | Name
  • -e/--env | Environment
  • --no-cache | Stop using cache
• Profile file structure
  • <root>
  • Kubernetes Context Profile
  • Docker Image Profile
  • HELM Package Profile
  • Helm Environment Profile
  • Helm Repository Profile
  • Variables
  • Example

Installation

via pip

pip install helmdt

via release

Download the latest stable release (.whl) from the release page.

pip install helmdt-x.x.x-py3-none-any.whl

Guiding Way

This way gives you a guiding flow to deploy your service by select options and input values.

helmdt [ACTION]

The helmdt tools will find helmdt.yaml, helmdt.yml or helmdt.json and load them as the profile. If those files are all not found, an error would displayed.

Manually specify the profile

If you want to manually specify the profile file, you could use -f argument. Just like:

helmdt -f profile.yaml [ACTION]
helmdt -f ../profile.yaml [ACTION]
helmdt -f /home/profile.yaml [ACTION]

These arguments must before the [ACTION] section.

-f | File

Set the profile file to use.

Actions

depcheck

Check the dependencies if they were installed.

build

Build the docker image and push it.

The tag of the image would be like: build-[SHORT GIT COMMIT HASH], for example: build-e7f38761.

apply

Apply the helm package and values. and some extra values would be passed to helm package:

• image.tag - The tag that the docker image used. It would first be checked if exists before applying.
• meta.* - The metadata from the profile would also passed to helm.

link

Link the kubenetes context and store it locally (into local kubenetes configuration).

Current supported service providers: Azure Kubenetes Services.

Command Line Way

By passing the arguments to helmdt, the helmdt tools would not run in guiding way. If some of the arguments are missing, it would still ask you on-demand.

helmdt link -n developers-internal
helmdt build -n stripe-worker -y --no-cache
helmdt apply -n stripe-worker -e dev -y

The follow arguments must be after the [ACTION] section.

--yes | Yes

Automatically skip all the prompts and answer yes for them. If some of the arguments are missing, and also the -y specified, an error would be displayed and the tools would be terminated.

-n/--name | Name

• For build, the name of the docker profile to build and push.
• For apply, the name of the helm package profile to apply.
• For link, the name of the kubenetes context.

-e/--env | Environment

The environment to apply. This argument only used in the apply action.

--no-cache | Stop using cache

bypass the --no-cache to docker build. This argument only used in the build action.

Profile file structure

<root>

• kubecontexts: (Kubenetes Context Profile) The contexts could be linked.
• docker: (Docker Image Profile) The profiles for docker build and push.
• helm: (Helm Package Profile) The profiles for helm apply.

Kubernetes Context Profile

• name: (string) The short name for the kubernetes context to be link, only used in the script.

• provider: (string) The name of the service provider. Now we only supported:

  • azure: Azure kubernetes Services.

    If you'd like to use Azure kubernetes Services, you should also specify the follow keys with the same level with provider.

    • subscription: (string) The subscription id in the azure portal.
    • resource_group: (string) The resource group that the kubernetes contained.
    • aks_name: (string) The name of the kubernetes service.

Docker Image Profile

• name: (string) The short name for the docker profile, just used in the script.
• image_name: (string) The image name placed in the local and pushed to the repository.
• repository: (string) The address of the repository.
• dockerfile: (string) The Dockerfile used for building.
• contextpath: (string) The path of building context. If not specified, the parent folder of dockerfile would be used by default.

HELM Package Profile

• name: (string) The short name for the helm profile, just used in the script.
• namespace: (string) The namespace to apply. If not exists, it would created automatically.
• dpname: (string) The name for the helm deployment, could be found in kubenetes/Deployment.
• package: (string) The path of the package.
• environments: (list of Helm Environment Profile) The environment entries.

Helm Environment Profile

• name: (string) The short name for the environment profile, just used in the script.
• values_file: (string) The values file.
• kubecontext: (string) The name of the kubenetes context where the helm package to be applied.

Helm Repository Profile

• name: (string) The name for the repository profile, will be added to helm.
• url: (string) The url of the repository.

Variables

• $(gitroot) - Would be replated to the root path of the current git repository.

Example

kubecontexts:
  - name: int
    provider: azure
    subscription: xxxxxxxx-xxxx-xxxx-xxxxxxxxxxxx
    resource_group: resource-group
    aks_name: service1-int
helm_repos:
  - name: bitnami
    url: https://charts.bitnami.com/bitnami
docker:
  - name: service1
    image_name: service1
    dockerfile: $(gitroot)/docker/service1/Dockerfile
    context: $(gitroot)/src
    repository: xxx.azurecr.io
helm:
  - name: service1
    dpname: service1
    namespace: ns1
    package: $(gitroot)/helm-packages
    environments:
      - name: dev
        values_file: $(gitroot)/helm-values/dev.yaml
        kubecontext: service1-dev

      - name: prod
        values_file: $(gitroot)/helm-values/prod.yaml
        kubecontext: service1-prod
  - name: service_from_repo
    dpname: service_from_repo
    namespace: ns1
    package: mariadb
    environments:
      - name: dev
        values_file: $(gitroot)/helm-values/dev.yaml
        kubecontext: service1-dev