Last updated:
0 purchases
atlasinit 0.1.1
Atlas Init - A CLI for developing integrations with MongoDB Atlas
Currently, used with
https://github.com/mongodb/terraform-provider-mongodbatlas
https://github.com/mongodb/mongodb/mongodbatlas-cloudformation-resources
see atlas_init#repo_aliases for an up-to-date list
Requirements
Create an organization
Go to access_manager and click Create Api Key: https://cloud-dev.mongodb.com/v2#/org/{ORG_ID_IN_URL_FROM_1}/access/apiKeys
Tick all permissions
create directories and store a file in profiles/default/.env-manual (ATLAS_INIT_PROFILES_PATH/{profile_name}/.env-manual)
export AWS_PROFILE=REPLACE_ME # your AWS profile used to create resources or other env-vars supported by AWS TF provider
export MONGODB_ATLAS_ORG_ID=REPLACE_ME # ORG_ID_IN_URL_FROM_1
export MONGODB_ATLAS_PUBLIC_KEY=REPLACE_ME # with 2
export MONGODB_ATLAS_PRIVATE_KEY=REPLACE_ME # with 2
export ATLAS_INIT_PROJECT_NAME=YOUR_NAME # the name of the project
export MONGODB_ATLAS_BASE_URL=https://cloud-dev.mongodb.com/ # replace with https://cloud.mongodb.com/ if you are not a MongoDB Employe
# optional
TF_VAR_federated_settings_id=REPLACE_ME # will need to add organization: <https://cloud-dev.mongodb.com/v2#/federation/{FEDERATION_SETTINGS_ID}/organizations> (see internal testing wiki)
# if you want to use your locally built MongoDB atlas provider
# see appendix for details on the content
export TF_CLI_CONFIG_FILE=REPLACE_ME/dev.tfrc
# if you plan developing with cloudformation
export ATLAS_INIT_CFN_PROFILE=YOUR_NAME
export ATLAS_INIT_CFN_REGION=eu-south-2 # find a region with few other profiles
Two modes of running
1. pip install normal user
source .venv/bin/activate # ensure you are in your preferred python env
(uv) pip install atlas-init
# use export ATLAS_INIT_PROFILES_PATH=/somewhere/to/store/your/env-vars/and/tf/state
2. Local development, run from github repo
git clone https://github.com/EspenAlbert/atlas-init
cd atlas-init
brew install pre-commit uv hatch
# https://github.com/astral-sh/uv <-- python packaging lightning fast
# https://hatch.pypa.io/latest/ <-- uv compatible build system for python
pre-commit install
# check that everything works
pre-commit run --all-files
# configure your virtualenv
cd py
hatch test
export VENV_DIR=$(hatch env find hatch-test | grep py3.12) # hatch venv path for env=hatch-test
export VENV_PYTHON=$VENV_DIR/bin/python
$VENV_PYTHON # ensure you are in shell with python3.12 (cmd+d to exit)
cd .. # back to repo root
# open in your IDE with virtualenv enabled
code .
# select venv path from $VENV_PYTHON output as python interpreter
# to make it easy to invoke from any terminal
export pypath=$(pwd)/py # pwd=repo root(atlas-init)
echo "alias atlas_init='export PYTHONPATH=$pypath && \"$VENV_PYTHON\" -m atlas_init'" >> ~/.zprofile # replace with your shell profile
# test that it works
atlas_init # should show how to use the cli
3. pip install local wheel
will be used by the CI in other repos
atlasci_local_install
creates a local .venv builds the wheel from this repo and installs it
use export ATLAS_INIT_PROFILES_PATH=/somewhere/to/store/your/env-vars/and/tf/state
Commands
cd terraform/cfn/{YOUR_RESOURCE_PATH}
# if you used `pip install` replace `atlas_init` with `atlasci`
atlas_init # help info
atlas_init # initialize the terraform providers
atlas_init tf # help for tf specific commands
atlas_init cfn # help for cfn specific commands
atals_init apply # `terraform apply`
# use cmd+v if you plan on using other tools, e.g., cfn make commands
# see appendix on how to configure .vscode test env-vars
atals_init destroy # `terraform destroy`
Appendix
Configuring vscode to use your env vars
add to your settings.json
{
"go.testEnvFile": "/{SOME_PREFIX}/atlas-init/profiles/default/.env-vscode",
}
Content of dev.tfrc
usually, it will look something like this:
provider_installation {
dev_overrides {
"mongodb/mongodbatlas" = "REPO_PATH_TF_PROVIDER/bin"
}
direct {}
}
Re-generating the lock files
terraform providers lock \
-platform=darwin_amd64 \
-platform=linux_amd64 \
-platform=darwin_arm64 \
-platform=linux_arm64
# -platform=windows_amd64 \
For personal and professional use. You cannot resell or redistribute these repositories in their original state.
There are no reviews.