Skip to content

elerch/lambda-zig

Repository files navigation

lambda-zig: A Custom Runtime for AWS Lambda

This is a custom runtime built in Zig (0.15). Simple projects will execute in <1ms, with a cold start init time of approximately 11ms.

Custom build steps are available for packaging and deploying Lambda functions:

  • zig build awslambda_package: Package the Lambda function into a zip file
  • zig build awslambda_iam: Create or verify IAM role for the Lambda function
  • zig build awslambda_deploy: Deploy the Lambda function to AWS
  • zig build awslambda_run: Invoke the deployed Lambda function

Build options:

  • function-name: Name of the AWS Lambda function
  • payload: JSON payload for function invocation (used with awslambda_run)
  • region: AWS region for deployment and invocation
  • profile: AWS profile to use for credentials
  • env-file: Path to environment variables file for the Lambda function
  • config-file: Path to lambda.json configuration file (overrides build.zig settings)

The Lambda function can be compiled for x86_64 or aarch64. The build system automatically configures the Lambda architecture based on the target.

A sample project using this runtime can be found at https://git.lerch.org/lobo/lambda-zig-sample

Lambda Configuration

Lambda functions can be configured via a lambda.json file or inline in build.zig. The configuration controls IAM roles, function settings, and deployment options.

Configuration File (lambda.json)

By default, the build system looks for an optional lambda.json file in your project root. If found, it will use these settings for deployment.

{
  "role_name": "my_lambda_role",
  "timeout": 30,
  "memory_size": 512,
  "description": "My Lambda function",
  "allow_principal": "alexa-appkit.amazon.com",
  "tags": [
    { "key": "Environment", "value": "production" },
    { "key": "Project", "value": "my-project" }
  ]
}

Available Configuration Options

Many of these configuration options are from the Lambda CreateFunction API call and more details are available there.

Option Type Default Description
role_name string "lambda_basic_execution" IAM role name for the function
timeout integer AWS default (3) Execution timeout in seconds (1-900)
memory_size integer AWS default (128) Memory allocation in MB (128-10240)
description string null Human-readable function description
allow_principal string null AWS service principal for invoke permission
kmskey_arn string null KMS key ARN for environment encryption
layers string[] null Lambda layer ARNs to attach
tags Tag[] null Resource tags (array of {key, value})
vpc_config object null VPC configuration (see below)
dead_letter_config object null Dead letter queue configuration
tracing_config object null X-Ray tracing configuration
ephemeral_storage object AWS default (512) Ephemeral storage configuration
logging_config object null CloudWatch logging configuration

VPC Configuration

{
  "vpc_config": {
    "subnet_ids": ["subnet-12345", "subnet-67890"],
    "security_group_ids": ["sg-12345"],
    "ipv6_allowed_for_dual_stack": false
  }
}

Tracing Configuration

{
  "tracing_config": {
    "mode": "Active"
  }
}

Mode must be "Active" or "PassThrough".

Logging Configuration

{
  "logging_config": {
    "log_format": "JSON",
    "application_log_level": "INFO",
    "system_log_level": "WARN",
    "log_group": "/aws/lambda/my-function"
  }
}

Log format must be "JSON" or "Text".

Ephemeral Storage

{
  "ephemeral_storage": {
    "size": 512
  }
}

Size must be between 512-10240 MB.

Dead Letter Configuration

{
  "dead_letter_config": {
    "target_arn": "arn:aws:sqs:us-east-1:123456789:my-dlq"
  }
}

Build Integration Options

You can also configure Lambda settings directly in build.zig:

// Use a specific config file (required - fails if missing)
_ = try lambda.configureBuild(b, dep, exe, .{
    .lambda_config = .{ .file = .{
        .path = b.path("deploy/lambda.json"),
        .required = true,
    }},
});

// Use inline configuration
_ = try lambda.configureBuild(b, dep, exe, .{
    .lambda_config = .{ .config = .{
        .role_name = "my_role",
        .timeout = 30,
        .memory_size = 512,
        .description = "My function",
    }},
});

// Disable config file lookup entirely
_ = try lambda.configureBuild(b, dep, exe, .{
    .lambda_config = .none,
});

Overriding Config at Build Time

The -Dconfig-file build option overrides the build.zig configuration:

# Use a different config file for staging
zig build awslambda_deploy -Dconfig-file=lambda-staging.json

# Use production config
zig build awslambda_deploy -Dconfig-file=deploy/lambda-prod.json

Environment Variables

Lambda functions can be configured with environment variables during deployment. This is useful for passing configuration, secrets, or credentials to your function.

Using the build system

Pass the -Denv-file option to specify a file containing environment variables:

zig build awslambda_deploy -Dfunction-name=my-function -Denv-file=.env

Using the CLI directly

The lambda-build CLI supports both --env flags and --env-file:

# Set individual variables
./lambda-build deploy --function-name my-fn --zip-file function.zip \
    --env DB_HOST=localhost --env DB_PORT=5432

# Load from file
./lambda-build deploy --function-name my-fn --zip-file function.zip \
    --env-file .env

# Combine both (--env values override --env-file)
./lambda-build deploy --function-name my-fn --zip-file function.zip \
    --env-file .env --env DEBUG=true

Environment file format

The environment file uses a simple KEY=VALUE format, one variable per line:

# Database configuration
DB_HOST=localhost
DB_PORT=5432

# API keys
API_KEY=secret123

Lines starting with # are treated as comments. Empty lines are ignored.

Service Permissions

Lambda functions can be configured to allow invocation by AWS service principals. This is required for services like Alexa Skills Kit, API Gateway, or S3 to trigger your Lambda function.

Using lambda.json (Recommended)

Add allow_principal to your configuration file:

{
  "allow_principal": "alexa-appkit.amazon.com"
}

Common service principals:

  • alexa-appkit.amazon.com - Alexa Skills Kit
  • apigateway.amazonaws.com - API Gateway
  • s3.amazonaws.com - S3 event notifications
  • events.amazonaws.com - EventBridge/CloudWatch Events

The permission is idempotent - if it already exists, the deployment will continue successfully.

Using the Zig Package Manager

To add Lambda package/deployment steps to another project:

  1. Fetch the dependency:
zig fetch --save git+https://git.lerch.org/lobo/lambda-zig
  1. Update your build.zig:
const std = @import("std");
const lambda_zig = @import("lambda_zig");

pub fn build(b: *std.Build) !void {
    const target = b.standardTargetOptions(.{});
    const optimize = b.standardOptimizeOption(.{});

    // Get lambda-zig dependency
    const lambda_zig_dep = b.dependency("lambda_zig", .{
        .target = target,
        .optimize = optimize,
    });

    const exe_module = b.createModule(.{
        .root_source_file = b.path("src/main.zig"),
        .target = target,
        .optimize = optimize,
    });

    // Add lambda runtime to your module
    exe_module.addImport("aws_lambda_runtime", lambda_zig_dep.module("lambda_runtime"));

    const exe = b.addExecutable(.{
        .name = "bootstrap",
        .root_module = exe_module,
    });

    b.installArtifact(exe);

    // Add Lambda build steps
    try lambda_zig.configureBuild(b, lambda_zig_dep, exe);
}

Note: The build function return type must be !void or catch/deal with errors to support the Lambda build integration.

Releases

No releases published

Packages

No packages published

Languages