Skip to content

Latest commit

 

History

History
39 lines (30 loc) · 1.64 KB

Lambdas.md

File metadata and controls

39 lines (30 loc) · 1.64 KB

mesh-deco on AWS λ

You can host meshing worker nodes on AWS Lambda in order to scale the worker pool "infinitely". The notation for this is identical to the treatment of conventional servers, and the config file might look like this:

{
    "workers": [
        {
            "url": "https://[MY UUID].execute-api.us-east-1.amazonaws.com/[MY_URL]/"
        }
    ]
}

It is recommended that you use zappa to deploy worker nodes to Lambda endpoints. For more instructions, see Deploying with Zappa

Deploying with Zappa

It is recommended that you use zappa to deploy worker nodes to Lambda endpoints. Zappa converts Flask servers into deployed services (public) using AWS Lambda.

Here is an example zappa.config.json:

{
    "dev": {
        "app_function": "worker_mesh_server.App",
        "aws_region": "us-east-1",
        "profile_name": "[YOUR PROFILE]",
        "project_name": "meshdeco",
        "runtime": "python3.6",
        "s3_bucket": "mesh-worker-[YOUR RANDOM ID]",
        "slim_handler": true
    }
}

You must replay YOUR PROFILE with the name of your AWS profile, if you have multiple in your AWS auth file. (If you only have one profile in this file, you may omit this key-value pair.)

Note that slim_handler should be set to true to enable the installation of C libraries (required) on the Lambda node. As a result, you must specify a s3_bucket, for which you may provide an arbitrary string. The bucket will be created for you at runtime.