harmony-service-lib

A library for Python-based Harmony services to parse incoming messages, fetch data, stage data, and call back to Harmony

Installing

Using pip

Install the latest version of the package from PyPI using pip:

$ pip install harmony-service-lib

Other methods:

The package is installable from source via

$ pip install git+https://github.com/harmony/harmony-service-lib-py.git#egg=harmony-service-lib

If using a local source tree, run the following in the source root directory instead:

$ pip install -e .

Usage

Services that want to work with Harmony can make use of this library to ease interop and upgrades. To work with Harmony, services must:

1. Receive incoming messages from Harmony. Currently the CLI is the only supported way to receive messages, though HTTP is likely to follow. harmony.cli provides helpers for setting up CLI parsing while being unobtrusive to non-Harmony CLIs that may also need to exist.
2. Extend harmony.BaseHarmonyAdapter and implement the #invoke to adapt the incoming Harmony message to a service call and adapt the service result to call to one of the adapter's #completed_with_* methods. The adapter class provides helper methods for retrieving data, staging results, and cleaning up temporary files, though these can be overridden or ignored if a service needs different behavior, e.g. if it operates on data in situ and does not want to download the remote file.

A full example of these two requirements with use of helpers can be found in example/example_service.py. Also see adapting-new-services for in depth guidance on service development using this library, especially the info on proper error handling.

Environment

The following environment variables can be used to control the behavior of the library and allow easier testing:

REQUIRED:

• STAGING_BUCKET: When using helpers to stage service output and pre-sign URLs, this indicates the S3 bucket where data will be staged

• STAGING_PATH: When using helpers to stage output, this indicates the path within STAGING_BUCKET under which data will be staged

• ENV: The name of the environment. If 'dev' or 'test', callbacks to Harmony are not made and data is not staged unless also using localstack

• OAUTH_UID, OAUTH_PASSWORD: Used to acquire a shared EDL token needed for downloading granules from EDL token-aware data sources. Services using data in S3 do not need to set this.

   NOTE: If `FALLBACK_AUTHN_ENABLED` is set to True (CAUTION!)
   these credentials will be used to download data *as* the EDL
   application user. This may cause problems with metrics and can
   result in users getting data for which they've not approved a
   EULA.
  

• OAUTH_CLIENT_ID: The Earthdata application client ID.

• OAUTH_HOST: Set to the correct Earthdata Login URL, depending on where the service is being deployed. This should be the same environment where the OAUTH_* credentials are valid. Defaults to UAT.

• OAUTH_REDIRECT_URI: A valid redirect URI for the EDL application.

• SHARED_SECRET_KEY: The 32-byte encryption key shared between Harmony and backend services. This is used to encrypt & decrypt the accessToken in the Harmony operation message. In a production environment, this should be injected into the container running the service Docker image. When running the service within Harmony, the Harmony infrastructure will ensure that this environment variable is set with the shared secret key, and the Harmony service library will read and use this key. Therefore, the service developer need not be aware of this variable or its value.

OPTIONAL:

• APP_NAME: Defaults to first argument on commandline. Appears in log records.
• AWS_DEFAULT_REGION: (Default: "us-west-2") The region in which S3 calls will be made
• USE_LOCALSTACK: (Development) If 'true' will perform S3 calls against localstack rather than AWS
• LOCALSTACK_HOST: (Development) If USE_LOCALSTACK true and this is set, will establish boto client connections for S3 & SQS operations using this hostname.
• TEXT_LOGGER: (Default: True) Setting this to true will cause all log messages to use a text string format. By default log messages will be formatted as JSON.
• HEALTH_CHECK_PATH: Set this to the path where the health check file should be stored. This file's mtime is set to the current time whenever a successful attempt is made to to read the message queue (whether or not a message is retrieved). This file can be used by a container's health check command. The container is considered unhealthy if the mtime of the file is old - where 'old' is configurable in the service container. If this variable is not set the path defaults to '/tmp/health.txt'.
• MAX_DOWNLOAD_RETRIES: Number of times to retry HTTP download calls that fail due to transient errors.

OPTIONAL -- Use with CAUTION:

• FALLBACK_AUTHN_ENABLED: Default: False. Enable the fallback authentication that uses the EDL application credentials. See CAUTION note above.
• EDL_USERNAME: The Earthdata Login username used for fallback authn.
• EDL_PASSWORD: The Earthdata Login password used for fallback authn.

Development Setup

Prerequisites:

• Python 3.9+, ideally installed via a virtual environment
• A local copy of the code

Install dependencies:

$ make install

Run linter against production code:

$ make lint

Run tests:

$ make test

Build & publish the package:

$ make publish

Releasing

GitHub release notes will automatically be generated based on pull request subjects. Pull request subject lines should therefore concisely emphasize library user-facing behavior and updates they should appear in the changelog. If more information is needed for release notes, note that in the PR content.