Skip to content

nasa/cumulus

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

Cumulus Framework

npm version Coverage Status

About Cumulus

Cumulus is an open source cloud-based data ingest, archive, distribution, and management framework developed for NASA's future Earth Science data streams. This repo supports the development, deployment, and testing of Cumulus and supplies useful tips on configuration, workflow management, and operations. To learn more about Cumulus and NASA's Earth Observing System Data and Information System (EOSDIS) cloud initiatives go to More Information.


πŸš€ Getting Started

Below is in-depth guidance to help get you started with your Cumulus development. To get a quick start on Cumulus deployment go to our Getting Started section.

Contents


πŸ“– Documentation

πŸ”¨ Development

The Cumulus core repo is a monorepo managed by Lerna. Lerna is responsible for installing the dependencies of the packages and tasks that belong in this repo. In general, Cumulus's npm packages can be found in the packages directory, and workflow tasks can be found in the tasks directory.

To help cut down on the time and disk space required to install the dependencies of the packages in this monorepo, all devDependencies are defined in the top-level package.json. The Node module resolution algorithm allows all of the packages and tasks to find their dev dependencies in that top-level node_modules directory.

TL;DR - If you need to add a devDependency to a package, add it to the top-level package.json file, not the package.json associated with an individual package.

Installation

This is for installation for Cumulus development. See the Cumulus deployment section for instructions on deploying the released Cumulus packages.

Prerequisites

  • NVM and node version 16.19.0
  • AWS CLI
  • Bash
  • Docker (only required for testing)
  • docker-compose (only required for testing pip install docker-compose)
  • Python 3.10
  • pipenv

You may use brew to install the prerequisites. Visit Homebrew documentation for guidance.

Install the correct node version:

nvm install
nvm use

Install Lerna

We use Lerna to manage multiple Cumulus packages in the same repo. You need to install Lerna as a global module first:

npm install -g lerna

Install Local Dependencies

We use npm for local package management. Run the following to get your dependencies set up.

npm install
npm run bootstrap

Build all packages:

npm run build

Build and watch packages:

npm run watch

To add new packages go to Adding New Packages for guidance.

Running the Cumulus APIs locally

Start the API:

npm run serve

Or start the distribution API:

npm run serve-dist

See the API package documentation for more options.

πŸ“ Tests

Unit Tests

LocalStack

LocalStack provides local versions of most AWS services for testing.

The LocalStack repository has installation instructions.

Localstack is included in the docker-compose file. You only need to run the docker-compose command in the next section in order to use it with your tests.

Docker containers

Turn on the docker containers first:

npm run start-unit-test-stack

Stop localstack/unit test services:

npm run stop-unit-test-stack

Run database migrations

npm run db:local:migrate

Using an AWS-hosted Elasticsearch server

The tests can be run against an Elasticsearch server running in AWS. This is useful if you are using an ARM-equipped Mac and are unable to run the old Intel version of Elasticsearch in Docker. These instructions assume that you have a deployment of Cumulus available, and the deployment name is "EXAMPLE".

Pre-Reqs
  • The AWS CLI is installed
  • The Session Manager plugin for the AWS CLI is installed
  • jq is installed
  • Your Cumulus deployment specified a key_name in cumulus-tf/terraform.tfvars that will grant you access to the EC2 instances that are part of that deployment
  • You are able to SSH into one of your EC2 instances (you are connected to a NASA VPN if required)
Configure ssh

Add the following to your ~/.ssh/config file

Host i-*
  User ec2-user
  ProxyCommand sh -c "aws ssm start-session --target %h --document-name AWS-StartSSHSession --parameters 'portNumber=%p'"
  StrictHostKeyChecking no
  UserKnownHostsFile /dev/null
Start the ssh tunnel to Elasticsearch

Open an SSH tunnel to Elasticsearch with the following command.

./bin/es-tunnel.sh EXAMPLE

At this point you can send requests to https://localhost:8443 and get responses from your Elasticsearch domain running in AWS. Note that, because you're tunneling TLS-encrypted traffic, the certificates are not going to match. The test code handles this already but, if you're using curl, make sure to use the -k option to disable strict certificate checks.

$ curl -k https://localhost:8443
{
  "name" : "ABC123",
  "cluster_name" : "123:abc-es-vpc",
  "cluster_uuid" : "abc-Ti6N3IA2ULvpBQ",
  "version" : {
    "number" : "5.3.2",
    "build_hash" : "6bc5aba",
    "build_date" : "2022-09-02T09:03:07.611Z",
    "build_snapshot" : false,
    "lucene_version" : "6.4.2"
  },
  "tagline" : "You Know, for Search"
}
Run the tests

With the tunnel configured, you can now run the tests with the following command:

env \
  LOCAL_ES_HOST_PORT=8443 \
  LOCAL_ES_HOST_PROTOCOL=https \
  LOCAL_ES_HOST=localhost \
  LOCALSTACK_HOST=127.0.0.1 \
npm test

Run tests

Run the test commands next

export LOCAL_ES_HOST=127.0.0.1
export LOCALSTACK_HOST=127.0.0.1
npm test

Coverage tests

If tests are working, run coverage tests

export LOCAL_ES_HOST=127.0.0.1
export LOCALSTACK_HOST=127.0.0.1
npm run test:coverage

These tests will fail if coverage drops below certain thresholds or if unit tests fail.

an environment variable can be set to only measure and not threshold

export FAIL_ON_COVERAGE=false
npm run test:coverage

Additionally, you can facilitate updating coverage values with the included coverage script

npm run coverage -- --update

Integration Tests

For more information please read this.

Running tests via VS Code debugger

Copy the .vscode.example directory to .vscode to create your debugger launch configuration. Refer to the VS Code documentation on how to use the debugger.

πŸ”¦ Code Coverage and Quality

For more information please read this.

πŸ“¦ Adding New Packages

Create a new folder under packages if it is a common library or create folder under cumulus/tasks if it is a lambda task. cd to the folder and run npm init.

Make sure to name the package as @cumulus/package-name.

Running command in all package folders

lerna exec -- rm -rf ./package-lock.json

Cleaning Up all the repos

npm run clean

Contribution

Please refer to: https://github.com/nasa/cumulus/blob/master/CONTRIBUTING.md for more information.

πŸ›’ Release

To release a new version of cumulus read this.


More Information

For more information about this project or more about NASA's Earth Observing System Data and Information System (EOSDIS) and its cloud work, please contact Katie Baynes or visit us at https://earthdata.nasa.gov.