generator-axios-lambda

Yeoman generator to make developing Lambda Functions at Axios simpler and faster.

What's in here?

Files necessary for creating a Lambda function managed by Apex, a command line tool for building, managing, and deploying functions.

• generators
  • app a Yeoman app for generating a Lambda functions' subdirectories
  • node templates for a Node Lambda function
  • python template for a Python Lambda function

Setup

Within the generator-axios-lambda dir npm run setup

This will

• Install global dependencies: yo, yeoman-generator, gulp-cli
• Install local dependencies
• Link generators/app/ to yo

Generators

axios-lambda:node

yo axios-lambda:node function_name

Generate an AWS Lambda project with a Node.js function, named function_name

mkdir [project-name] && cd $_
yo axios-lambda:node [function-name]

Multiple functions

To generate multiple Lambda functions within a project, just re-use the generator

yo axios-lambda:node [function-name]

axios-lambda:python

yo axios-lambda:python function_name

Generate an AWS Lambda project with a Python function, named function_name

mkdir [project-name] && cd $_
yo axios-lambda:python [function-name]

Multiple functions

To generate multiple Lambda functions within a project, just re-use the generator

yo axios-lambda:python [function-name]

Developing

Utils

Decrypting secrets

def get_secret(secret_name, region_name):
    """
    A helper function for getting secrets from AWS Secrets Manager.

    Required: Function role permissions for Secrets Manager.
    Accepts: the name of the secret, the region the secrets is held in
    Returns: a dictionary of API secrets, binary data, or None if an error occurs
    """

Publishing a message via SNS

You can use SNS to create email alerts. Just create an SNS topic and subscribe to it.

def publish_to_sns(json_message, topic_arn, summary_string="SNS summary"):
    """
    A helper function for publishing a message to SNS.
    
    Required: Function role permissions for SNS:Publish.
    Accepts:
        * a JSON containing the message to publish
        * an SNS topic ARN string
        * an optional summary string describing this message
    Returns: None
    """

Writing to S3

def write_to_s3(data, filename, s3_bucket_name, output_type, s3_client=boto3.client('s3'), acl='private'):
    """
    A helper function for publishing a Python dictionary to S3 as a JSON file.

    Required: Function role permissions for S3:PutObject, S3:PutObjectAcl
    Accepts:
        * a Python dictionary or list of the data you wish to put on S3
        * the filename (and optionally its path) string to write to on S3
        * the S3 bucket's name as a string
        * the output filetype which is a string specifying CSV or JSON
    Returns: None
    """

Deploying

Connecting to AWS Lambda

We use a build tool apex. For more information, documentation, and examples visit apex.run.

If your lambda function utilizes an RDS database you'll need to follow these additional instructions to make sure the function has access to the database (for now this will need to be done in the console, ideally we can get this linked up to apex too):

1. Set up new Lambda functions in us-west-2 only, to keep them in the same region as production database.

2. When configuring the function, click on advanced settings and change the 'no VPC' setting to 'default VPC'. This will trigger two more options for subnet and security group.

3. Set the subnet group to the one labeled 'The Lambda Zone' and the security group to the one labeled 'Lambda Security Group'

Multiple Environments

Multiple environments are supported with the --env flag. By default project.json and function.json are used, however when --env is specified project.ENV.json and function.ENV.json will be used, falling back on function.json for cases when staging and production config is the same. For example your directory structure may look something like the following:

project.stage.json
project.prod.json
functions
├── bar
│   ├── function.stage.json
│   ├── function.prod.json
│   └── index.js
└── foo
    ├── function.stage.json
    ├── function.prod.json
    └── index.js

If you prefer your "dev" or "staging" environment to be the implied default then leave the files as project.json and function.json:

project.json
project.prod.json
functions
├── bar
│   ├── function.json
│   ├── function.prod.json
│   └── index.js
└── foo
    ├── function.json
    ├── function.prod.json
    └── index.js

Building a zip with dependencies for Lambda

Apex generates a zip file for you upon deploy, however sometimes it can be useful to see exactly what's included in this file for debugging purposes. The apex build command outputs the zip to STDOUT for this purpose.

Examples

Output zip to out.zip:

$ apex build foo > tmp/out.zip
$ apex deploy foo --zip tmp/foo.zip

OR

$ apex deploy foo -z tmp/foo.zip

Use environment variables

$ apex build foo > tmp/out.zip
$ apex deploy foo --env-file env.json --zip tmp/foo.zip

OR

$ apex deploy foo -E env.json -z tmp/foo.zip

Invoking

After you've deployed, you can run the function and check the logs.

apex invoke -L [function_name]

Scheduling

Scheduleding a Lambda function must be down within the AWS interface. Login to view the function apex deployed via console.aws.amazon.com

Testing

Select "Scheduled Event" from the dropdown to simulate cron trigger a lambda function

Triggering

Under the "Triggers" table, add a "CloudWatch Events - Schedule" trigger. If you have not set up a CloudWatch Event to trigger a lambda function, you may do so by creating a CloudWatch Event rule

Troubleshooting

apex --help

Check the logs

apex logs <function-name>

Want to watch the logs live?

apex logs <function-name> --follow

OR

apex logs <function-name> --f

Check the stats

apex metrics

Shows you:

• total cost ($USD)
• invocations (integer & $)
• duration (datetime & $)
• throttles (integer)
• errors (integer)
• memory (integer)

Rollback to previous version

apex rollback <function-name>