This repo contains the package and dependencies to build the Arcimoto Global Dependencies Lambda Layer that is attached to each lambda during deployment. It also builds a dev and staging version of the layer. The layer automatically builds in the attached BitBucket pipeline.
- Cord Slatton - Repo Man (Authorizes changes to master branch)
- Keith Anderson
- Gary Malcolm
We use the Angular git commit style
Full (long) version:
<type>(<scope>): <subject>
<BLANK LINE>
<body>
<BLANK LINE>
<footer>
Short version
<type>(<scope>): <subject>
See https://github.com/angular/angular/blob/master/CONTRIBUTING.md#-commit-message-format more details
VS Code extension that helps you to make the commit msg, no need to remember your scope, etc.: https://marketplace.visualstudio.com/items?itemName=Jhecht.git-angular
We use the allowed types from https://github.com/angular/angular/blob/master/CONTRIBUTING.md#-commit-message-format
We use custom scopes per repository, see scopes below. If you need to add a scope that is allowed.
- args
- db
- dependencies
- exceptions
- note
- pipeline
- runtime
- tests
- user
- vehicle
The majority of the AWSLambda repo code is written in Python and meant to be executed by AWS Lambda in a Python 3 runtime environment. The currently used version of Python is 3.8, to maintain compatibility with the psycopg2 package. Many of the functions rely on the AWS SKD for Python, boto3.
Install boto3 using pip:
pip install boto3Additionally, in order to run Lambda functions which interact with AWS services, use the AWS Command Line Interface (CLI).
Install CLI using pip:
pip install awscliIn order to use the CLI, you will also need the proper AWS credentials, managed by the AWS Identity and Access Management (IAM) service. Have the Arcimoto AWS administrator create CLI credentials for you, or give you IAM permission to create them yourself. Then on the command line run:
aws configureand enter the Access Key ID and Secret Access Key that you just created.
To get the AWSLmbda Lambda functions running locally, first obtain a local copy by cloning the repository. From the directory that you want to install your local copy, use git on the command line:
git clone https://<username>@bitbucket.org/arcimotocode1/arcimoto-lambda-global-dependencies.gitWhere <username> is your bitbucket user name.
You must set up a python virtual env
python -m venv venv
and then install the package dependencies
pip install boto3 cerberus psycopg2 xmlrunner
at which point you should be able to run the tests
python -m unittest discover
Setup a bitbucket app password with read access to your account's repositories and substitute the username into the command:
pip install git+https://{{BITBUCKET_USERNAME}}@bitbucket.org/arcimotocode1/arcimoto-lambda-global-dependencies.gitUsed to identify resources and their configuration.
Global dependencies and corresponding layer configuration.
From the cerberus package extracted.
From the certifi package using the wheel file type, extracted.
The global_dependencies layer builds for a specific python3 minor (3.8) version due to compatibility with psycopg2 (postgres db interaction package).
The compatibility with psycopg2 requires a build for a specific runtime of python3 for our global_dependencies layer. Rather than compile psycopg2 for the AWS Lambda environment ourselves we can use awslambda-psycopg2 an open source repository that maintains compiled versions of psycopg2 for each version of python.
- Copy the
psycopg2-3.xdirectory from the awslambda-psycopg2 repository matching your intended minor version of python3 into your AWS Lambda projectdependenciesfolder and rename the folder topsycopg2before zipping it into a file calledpsycopg2.zip. This will then be used in any subsequent lambda creation or updates by the lambda-utility commands in this repository ONLY if they directly use thepsycopg2module as acommon_dependency. However, newer, upgraded lambdas rely on ourglobal_dependencieslayer. - Create a new layer version for the
global_dependencieswith a compatible runtime set to match the minor version of python3 you are upgrading to using the lambda-utility. After the layer is created via the lamda-utility it will return a version number: independencies.jsonset thelayers.global_dependencies.versionproperty to the new layer version number. This will cause subsequentcreateandupdatelambda-utility commands to use the new layer (and the upgraded version of python and psycopg2).
The BitBucket pipelines file [bitbucket-pipelines.yml] implements the Continuous Integration (CI)/Continuous Deployment (CD) for the AWSLambda repository.
The tests are run automatically on Pull Request (PR) creation/update from a feature branch beginning with the prefix TEL- to the dev branch.
These tests include:
- testing the
dependencies.jsonfile JSON validity - unit tests
Any failed tests will result in a failure of the pipeline, which will prevent merging the PR to dev until the PR is updated to fix the issue(s) and the pipeline is re-run successfully for the PR.
- Upon merge to
devfrom a feature branch prefixed withTEL-the pipeline creates a new global dependencies dev layer version.
- Upon merge to
stagingbranch the pipeline creates a new global dependencies staging layer version.
Upon merge to master branch the pipeline:
- publishes a new version using semantic versioning
- creates a new global dependencies production layer version
- creates a pull request in the dependent repository
awslambdafor the change in hash for this repo