⚡️Serverless Framework Go Plugin

serverless-go-plugin is a Serverless Framework plugin that compiles Go functions on the fly. You don't need to do it manually before serverless deploy. Once the plugin is installed it will happen automatically. The plugin works with Serverless Framework version 1.52 and above.

dev.to: A better way of deploying Go services with Serverless Framework

Features

• Concurrent compilation happens across all CPU cores.
• Support for both serverless deploy and serverless deploy function commands.
• Support for serverless invoke local command.
• Additional command serverless go build.

Install

1. Install the plugin

   npm i --save-dev serverless-go-plugin

2. Add it to your serverless.yaml

   plugins:
     - serverless-go-plugin

3. Replace every Go function's handler with *.go file path or a package path. E.g.

   functions:
     example:
       runtime: go1.x
       handler: functions/example/main.go # or just functions/example

Configuration

Default values:

custom:
  go:
    baseDir: . # folder where go.mod file lives, if set `handler` property should be set relatively to that folder
    binDir: .bin # target folder for binary files
    cgo: 0 # CGO_ENABLED flag
    cmd: 'GOOS=linux go build -ldflags="-s -w"' # compile command
    monorepo: false # if enabled, builds function every directory (useful for monorepo where go.mod is managed by each function
    supportedRuntimes: ["go1.x"] # the plugin compiles a function only if runtime is declared here (either on function or provider level) 
    buildProvidedRuntimeAsBootstrap: false # if enabled, builds and archive function with only single "bootstrap" binary (useful for runtimes like provided.al2)

How does it work?

The plugin compiles every Go function defined in serverless.yaml into .bin directory. After that it internally changes handler so that the Serverless Framework will deploy the compiled file not the source file.

For every matched function it also overrides package parameter to

individually: true
exclude:
  - `./**`
include:
  - `<path to the compiled file and any files that you defined to be included>`

How to run Golang Lambda on ARM?

1. Add provided.al2 to supportedRuntimes and enable buildProvidedRuntimeAsBootstrap in plugin config
2. Append GOARCH=arm64 to your compile command (cmd line)
3. Change architecture and runtime in global config:

provider:
    architecture: arm64
    runtime: provided.al2

Warning! First deploy may result in small downtime (~few seconds) of lambda, use some deployment strategy like canary for safer rollout.