search

acorn-io / mongodb-atlas

0 stars 2 forks source link

readme

Purpose

This folder defines an Acorn service which allows to create a Mongo Atlas cluster on the fly. To create the Mongo cluster, the Acorn interacts with Mongo Atlas APIs with user provided credentials.

In this very early version each cluster created by the service has the following characteristics, they are currently hardcoded but will soon become service's arguments:

Default values are:

If you set any of the following arguments, the service Acorn will not manage the resource.

See below for more details setting these values.

Notes:

Atlas Credentials

To use this service you need to have an Atlas Mongo account, to create an organization and a project within this one. Once deployped, you can click on the "cred-helper" link in the UI or from the CLI status. The link will open a browser page that instructs you how to get API keys for Atlas.

Once you save the credentials through the UI form, the cluster will be provisioned.

Usage

The examples folder contains a sample application using this Service. This app consists in a Python backend based on the FastAPI library, it displays a web page indicating the number of times the application was called, a counter is saved in the underlying MongoDB database and incremented with each request. The screenshot below shows the UI of the example application.

UI

To use the Mongo Service, we first define a service property in the Acornfile of the application:

services: db: {
  image: "ghcr.io/acorn-io/mongodb-atlas:v#.#-#"
}

Next we define the application container. This one can connect to the MongoDB service via environment variables which values are set based on the service's properties.

containers: {
  app: {
    build: {
   context: "."
   target:  "dev"
  }
    consumes: ["db"]
    ports: publish: "8000/http"
    env: {
      DB_HOST:  "@{service.db.address}"
      DB_NAME:  "@{service.db.data.dbName}"
      DB_PROTO: "@{service.db.data.proto}"
      DB_USER:  "@{service.db.secrets.user.username}"
      DB_PASS:  "@{service.db.secrets.user.password}"
    }
  }
}

This container is built using the Dockerfile in the examples folder. Once built, the container consumes the MongoDB Atlas service using the address and credentials provided through via the dedicated variables.

This example can be run with the following command (to be run from the examples folder)

acorn run -n app

After a few tens of seconds an http endpoint will be returned. Using this endpoint we can access the application and see the counter incremented on each reload of the page.

Running the app in Acorn Sandbox

Instead of managing your own Acorn installation, you can deploy this application in the Acorn Sandbox, the free SaaS offering provided by Acorn. Access to the sandbox requires only a GitHub account, which is used for authentication.

Run in Acorn

An application running in the Sandbox will automatically shut down after 2 hours, but you can use the Acorn Pro plan to remove the time limit and gain additional functionalities.

Consume an existing secret

This Acorn will need credentials to interact with the Atlas API. These credentials are stored in a secret named atlas-creds. The secret can be created ahead of time or you can run:

acorn login [APP_NAME]

This will prompt the user for the credentials needed to create the database on Atlas.

Note: this example uses an organization named Techwhale containing the project webhooks

Next create a public / private api key pair at the organization level

Organization api keys

Next get the project ID

Getting project ID

For this demo I set those 3 values in the following environment variables:

Next we need to create the secret atlas-creds providing the public and private keys as well as the Atlas project ID we want the MongoDB cluster to be created in.

Note: the following example uses environment variables already defined in the current shell

acorn secrets create \
  --type credential.cloud.mongodb.com/atlas \
  --data public_key=$MONGODB_ATLAS_PUBLIC_API_KEY \
  --data private_key=$MONGODB_ATLAS_PRIVATE_API_KEY \
  --data project_id=$MONGODB_ATLAS_PROJECT_ID \
  atlas-creds

Updating/Upgrading

When changing Tiers, you must follow the MongoDB atlas upgrade path. You can only upgrade an M0, M2, or M5 Cluster to an M10+ Cluster.

launching with clusterName, dbUser, dbAdminUser values

If you set the clusterName, the service Acorn will not manage it for you. It must already exist in Atlas. The reason for this, is if you create a cluster from a backup or some other source, Acorn will not take over the management.

If you set either of the users, you must pre-create the users in Atlas, create the user and/or admin secret in Acorn, and link them when launching the mongodb-atlas service Acorn.

Status

This service is still a work in progress. Feedback are welcome.