Manage blueprint repositories
A blueprint repository is a remote repository that contains templates and source code for blueprint functionality. Each time you run the XL CLI xl blueprint
command, it fetches files from the blueprint repository.
Repository types
You can define one or more of the following blueprint repository types:
- Local server
- HTTP
- GitHub online repository
- Bitbucket Cloud
- Bitbucket Server (on-premise)
- GitLab (Cloud and on-premise)
Define blueprint repositories
Defined blueprint repositories are stored in the ~/.xebialabs/config.yaml
file. This file is created automatically when you run any XL CLI command after installing the XL CLI. When you run the xl blueprint
command, this file's presence on your system enables you to select one of the available blueprints stored in a repository.
- On initial installation, the
config.yaml
file is configured to access the Deploy/Release public blueprint repository provided in the Deploy/Release public software distribution site. - You can also configure your own HTTP blueprint repository and update the
config.yaml
file to point to it. - You can define multiple blueprint repositories in your
config.yaml
file.
HTTP repository configuration fields
Here are the configuration fields for an HTTP repository in the config.yaml
file:
Field | Expected value | Default value | Required | Description |
---|---|---|---|---|
name | — | — | Yes | Repository configuration name |
type | http | — | Yes | Repository type |
url | — | — | Yes | HTTP repository URL, including protocol |
username | — | No | Basic authentication username | |
password | — | No | Basic authentication password |
Note: Only basic authentication is supported at the moment for remote HTTP repositories.
Local repository configuration
The type: local
repository is mainly intended to be used for local development and tests. Any local path can be used as a blueprint repository with this type.
Field | Expected value | Default value | Required | Description |
---|---|---|---|---|
name | - | - | Yes | Repository configuration name |
type | local | - | Yes | Repository type |
path | - | - | Yes | Full local path where blueprint definitions can be stored. ~ can be used for stating the current user's home directory under Unix systems. |
ignored-dirs | - | - | No | List of comma-separated directories, to be ignored while traversing the local path. Example: .git , some-other-dir |
ignored-files | - | - | No | List of comma-separated files, to be ignored while traversing the local path. Example: .DS_Store , .gitignore |
Notes
- In the case of local repositories, if the path is set too generically - such as
~
- the traversal path will be big and may result in the blueprint command running very slowly. - in development you can use the
-l
flag to use a local repository directly without defining it in configuration. For example, to execute a blueprint in a local directory~/mySpace/myBlueprint
you can runxl blueprint -l ~/mySpace -b myBlueprint
.
Define multiple repositories
You can specify multiple blueprint repositories in your config.yaml
file.
Important notes:
- Only one of the listed repositories will be active at a given time.
- The active blueprint repository should be stated using the
current-repository
field in the configuration file. - If the defined blueprint repository cannot be reached, an error displays.
- If the
current-repository
field is not stated, an error displays.
Here is the format for the blueprint section of the config.yaml
file that points to a GitHub repository, the public Digital.ai HTTP repository, a local repository that you create, and Bitbucket Cloud, Bitbucket Server, and GitLab repositories:
blueprint:
current-repository: XL Blueprints
repositories:
- name: xebialabs-github
type: github
repo-name: blueprints
owner: xebialabs
token: my-github-token
branch: master
- name: xebialabs-dist
type: http
url: http://dist.xebialabs.com/public/blueprints
- name: test
type: local
path: /path/to/local/test/blueprints/
ignored-dirs: .git, .vscode
ignored-files: .DS_Store, .gitignore
- name: Bitbucket Cloud
type: bitbucket
owner: xebialabs
repo-name: blueprints
branch: master
token: bitbucket-token
- name: Bitbucket server
type: bitbucketserver
user: xebialabs
url: http://localhost:7990
project-key: XEB
repo-name: blueprints
branch: master
token: bitbucket-token
- name: Gitlab
type: gitlab
owner: xebialabs
url: http://localhost
repo-name: blueprints
branch: master
token: gitlab-token
xl-deploy:
authmethod: basic
password: admin
url: http://localhost:4516
username: admin
xl-release:
authmethod: basic
password: admin
url: http://localhost:5516
username: admin
Note that the xebialabs-github
repository is declared as the default in this example.
GitHub repository configuration fields
You can maintain blueprints in one or more GitHub repositories and specify these details in your config.yaml
file.
Here are the configuration fields for a GitHub repository in the config.yaml
file:
Field | Expected value | Default value | Required? | Details |
---|---|---|---|---|
name | — | — | Yes | Repository configuration name |
type | github | — | Yes | Repository type |
repo-name | — | — | Yes | GitHub remote repository name |
owner | — | — | Yes | GitHub remote repository owner Can be different than the user accessing it |
branch | — | master | No | GitHub remote repository branch to use |
token | — | No | GitHub user token, please refer to GitHub documentation for generating one. Repo read permission is required when generating the token for the XL CLI |
Note: When the token
field is not specified, the GitHub API will be accessed in unauthenticated mode and the rate limit will be much less than the authenticated mode. According to the GitHub API documentation, the unauthenticated rate limit per hour and per IP address is 60, whereas the authenticated rate limit per hour and per user is 5000. You should set the token
field in your configuration so as not to receive any GitHub API related rate limit errors.
Define a single GitHub repository
Here is an example of the blueprint section of a config.yaml
file that is configured to access a GitHub repository:
blueprint:
current-repository: my-github
repositories:
- name: my-github
type: github
repo-name: blueprints
owner: mycompany
branch: master
token: my-github-token
Define multiple GitHub and HTTP repositories
You can specify multiple GitHub and/or HTTP blueprint repositories in your config.yaml
file.
Important notes:
- Only one of the listed repositories will be active at a given time.
- The active blueprint repository should be stated using the
current-repository
field in the configuration file. - If the defined blueprint repository, or
current-repository
field is not stated, the XL CLI will auto-update theconfig.yaml
file with the default Deploy/Release Blueprint repository.
Here is the format for the blueprint section of the config.yaml
file that points to the public XebiaLabs HTTP repository and a second GitHub repository you create:
blueprint:
current-repository: my-github
repositories:
- name: xl-dist
type: http
url: https://dist.xebialabs.com/public/blueprints/
- name: my-github
type: github
repo-name: blueprints
owner: mycompany
branch: master
token: GITHUB_TOKEN
Manually specify a blueprint using the blueprint
command
You can choose to explicitly specify a local or remote folder path to a blueprint when running the blueprint
command. Supported options depend on the version of the DevOps Platform software and XL CLI you are running. See the xl blueprint
command in the XL CLI command reference for details.
Repository structure example
To better understand the file and folder structure of a blueprint repository, review the public Deploy/Release Blueprint repository.
For example, you can drill down from the root of this repository to see how the Microservice Application on Amazon EKS blueprint is structured:
blueprints
├── index.json
└── aws/
└── microservice-ecommerce/
├── blueprint.yaml
├── xebialabs.yaml
├── cloudformation/
│ ├── template1.yaml.tmpl
│ └── template2.yaml
│
└── xebialabs/
├── xld-environment.yaml.tmpl
├── xld-infrastructure.yaml.tmpl
├── xlr-pipeline.yaml.tmpl
└── README.md.tmpl
The repository structure consists of:
-
index.json
file: Theindex.json
file at the root level of an HTTP blueprint repository provides an index listing off the blueprints stored in the repository, enabling you to select one of these blueprints using the XL CLI.For example, the
index.json
file in the Deploy/Release public repository defines the available blueprints:[
"aws/monolith",
"aws/microservice-ecommerce",
"aws/datalake",
"docker/simple-demo-app"
]Notes:
- The
index.json
file is not needed for a GitHub type repository. - If you choose to set up a new HTTP repository, you must update the JSON file to reflect your new repository.
- To automatically generate an
index.json
file on your release pipeline, you can refer to the samplegenerate_index.py
python script in the official Deploy/Release Blueprint GitHub repository.
- The
-
Blueprint template files: All files with
tmpl
extension are templates for the blueprint. These template files will be passed through generator to create "ready-to-use" YAML files. -
Regular files and folders: All files and directories will be copied directly.
File details
Here are the file details for the Microservice Application on Amazon EKS blueprint example.
microservice-ecommerce/
├── blueprint.yaml
├── xebialabs.yaml
├── cloudformation/
│ ├── template1.yaml.tmpl
│ └── template2.yaml
│
└── xebialabs/
├── xld-environment.yaml.tmpl
├── xld-infrastructure.yaml.tmpl
├── xlr-pipeline.yaml.tmpl
└── README.md.tmpl
blueprint.yaml
file: Each application must have ablueprint.yaml
in which you specify the required user prompts and files used for the blueprint.- See the Blueprint YAML format for a description of this file structure.
- For a working example, open the XebiaLabs Microservices e-commerce blueprint.yaml file to review the metadata, parameters, variables and files defined for this blueprint.
xebialabs.yaml
file: This file is an entry point forxl apply
command. For your convenience, this file combines all Deploy and Release YAML templates as anImport
kind, enabling you to apply a blueprint with a single command.cloudformation
folder: This folder is specific to AWS, containing CloudFormation templates used to provision the AWS infrastructure from Deploy. Other blueprint types will include folders and files specific to the type of application.xebialabs
folder: You place your Deploy/Release YAML templates in this folder. This folder will also include any generated files, including.gitignore
,values.xlvals
andsecrets.xlvals
files.