Skip to content

Latest commit

 

History

History
61 lines (39 loc) · 2.13 KB

README.md

File metadata and controls

61 lines (39 loc) · 2.13 KB
Raw

CDP Public Cloud - Cloudera Machine Learning (CML)

Constructs a set of Cloudera Machine Learning (CML) workspaces within their own CDP Public Cloud Environment and Datalake. Generates via Ansible the AWS infrastructure and CDP artifacts, including SSH key, cross-account credentials, S3 buckets, etc.

Requirements

To run, you need:

  • Docker (or a Docker alternative)
  • AWS credentials (set via AWS_PROFILE)
  • CDP credentials (set via CDP_PROFILE)

Set Up

First, set up your ansible-navigator aka cdp-navigator environment -- follow the instructions in the NAVIGATOR document in cloudera-labs/cldr-runner.

Then, clone this project and change your working directory.

git clone https://github.com/cloudera-labs/cloudera-deploy.git; cd cloudera-deploy/public-cloud/aws/cml

Configure

Set the required environment variables:

export AWS_PROFILE=your-aws-profile
export CDP_PROFILE=your-cdp-profile

Tweak the definition.yml parameters to your liking. Notably, you should add and/or change:

name_prefix:    ex01      # Keep this short (4-7 characters)
admin_password: "Secret"  # 1 upper, 1 special, 1 number, 8-64 chars.
infra_region:   us-east-2

Note

You can override these parameters with any typical Ansible extra variables flags, i.e. -e admin_password=my_password. See the cldr-runner FAQ for details.

SSH Keys

This definition will create a new SSH keypair on the host in your ~/.ssh directory if you do not specify a SSH public key. If you wish to use an existing SSH key already loaded into AWS, set public_key_id to the key's label. If you wish to use an existing SSH key, but need to have it loaded into AWS, then set public_key_file to the key's path.

Execute

Then set up the CDP Public Cloud by running the playbook:

ansible-navigator run main.yml

Tear Down

Tear down the CDP Public Cloud by running the playbook:

ansible-navigator run teardown.yml