lambda_deployment

Ruby gem for deploying lambda functions in AWS with Samson or another tool. This gem is used to update the function code only - the resources must already be provisioned using another tool (or manually).

This gem provides a binary called lambda_deploy which will deploy a zip or jar file to lambda, version it, and promote it to production (release). Usage:

lambda_deploy [-c path/to/configuration.yml] deploy|release

The deploy action will upload the function to S3, unless the SKIP_UPLOAD environment variable is set to true. It will then update the function to use the new code. If the environmental variable TAG is set it will also create a version and link that version to an alias named TAG.

The release action will link the production alias to the version referenced by the TAG alias. If TAG is not set this action is only required if the production alias is not already pointing to $LATEST.

The configuration file is intentionally small with the option to override environmental variables. The default action is to fetch values from the environment to ensure consistency within a given environment. These variables should be set within the deploy tool or using the configuration file.

Region in AWS where resources are located.

STS assume role to use if resources are in another account.

Name of S3 bucket to upload function to. Buckets can be provisioned per project or per environment.

Type of encryption to use with S3.

Version of the function being uploaded (e.g. git tag).

See the examples/lambda directory for a sample project layout.

Add the following line to your Gemfile and run bundle install:

gem 'lambda_deployment'

The only required fields are the project name (the function name in AWS), and the zip or jar file to upload (containing at least the script named after the lambda_function).

Here is an example:

file_name: lambda/foobar.zip  # required: zip or jar of lambda function and dependencies
project: my-dev-lambda-name   # required: name of lambda function
region: us-east-1             # optional: specify AWS region (will override $AWS_REGION)
s3_bucket: my-test-bucket     # optional: specify s3 bucket (will override $LAMBDA_S3_BUCKET)
s3_sse: AES256                # optional: set server side encryption on s3 objects (will override $LAMBDA_S3_SSE)
concurrency:                  # optional: set reserved concurrency limit (-1 to delete)
environment:
  FOO: bar                    # optional: set some env vars for your Lambda

• file_name is the file_name needs to be the path relative to the configuration file. It is also used, together with the TAG to generate the S3 key, not that the path, if any will not be part of the key. So 'dist/lambda.zip' will result in a S3 key such as 'lambda-v3.zip'.
• project must match the function name in the AWS console (the last value in the following example):

arn:aws:lambda:<region>:<account_id>:function:<project>

The lambda function and s3 buckets must already exist to use this gem. If you are creating actions for your lambda make sure they execute the production alias (see below). To do this append :production to the end of the ARN associated with your lambda function.

In the lambda console chose Create alias from the Actions drop down. The name should be production and version should point to $LATEST.