Awesome

Running AWS Lambda Custom Runtime in Knative

In November 2018, AWS announced support for Lambda custom runtime using a straightforward AWS lambda runtime API.

In this repository you find a function invoker implemented in Go, which provides the AWS Lambda runtime API. You also find a Knative build template. Using this build template you can run AWS Lambda custom runtimes directly in your Kubernetes cluster using Knative.

The AWS Lambdas execution environment is replicated using the Docker image amazonlinux and some environment variables.

AWS custom runtime walkthrough

This repository contains an example lambda function written in bash with a AWS custom runtime described in this AWS tutorial. To run this function use our tm client to talk to the knative API.

Install AWS custom runtime:

tm deploy task -f https://raw.githubusercontent.com/triggermesh/aws-custom-runtime/main/runtime.yaml

Deploy function:

tm deploy service lambda-bash -f https://github.com/triggermesh/aws-custom-runtime --runtime aws-custom-runtime --build-argument DIRECTORY=example --wait

In output you'll see URL that you can use to access example/function.sh function

AWS Lambda RUST example

RUST is also verified to be compatible with this runtime. Though official readme has build instructions, it is more convenient to use docker.

Clone repository:

git clone https://github.com/awslabs/aws-lambda-rust-runtime
cd aws-lambda-rust-runtime

Build binary and rename it to bootstrap:

docker run --rm --user "$(id -u)":"$(id -g)" -v "$PWD":/usr/src/myapp -w /usr/src/myapp rust:1.31.0 cargo build -p lambda_runtime --example basic --release
mv target/release/examples/basic target/release/examples/bootstrap

Deploy runtime using tm CLI:

tm deploy runtime -f https://raw.githubusercontent.com/triggermesh/aws-custom-runtime/main/runtime.yaml
tm deploy service lambda-rust -f target/release/examples/ --runtime aws-custom-runtime

Use your RUST AWS Lambda function on knative:

curl lambda-rust.default.k.triggermesh.io --data '{"firstName": "Foo"}'
{"message":"Hello, Foo!"}

AWS Lambda C++ example

Build custom runtime:

cd /tmp
git clone https://github.com/awslabs/aws-lambda-cpp.git
cd aws-lambda-cpp
mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release -DBUILD_SHARED_LIBS=OFF -DCMAKE_INSTALL_PREFIX=/tmp/out
make && make install

Prepare example function:

mkdir /tmp/hello-cpp-world
cd /tmp/hello-cpp-world


cat > main.cpp <<EOF
// main.cpp
#include <aws/lambda-runtime/runtime.h>

using namespace aws::lambda_runtime;

invocation_response my_handler(invocation_request const& request)
{
   return invocation_response::success("Hello, World!", "application/json");
}

int main()
{
   run_handler(my_handler);
   return 0;
}
EOF


cat > CMakeLists.txt <<EOF
cmake_minimum_required(VERSION 3.5)
set(CMAKE_CXX_STANDARD 11)
project(bootstrap LANGUAGES CXX)

find_package(aws-lambda-runtime REQUIRED)
add_executable(\${PROJECT_NAME} "main.cpp")
target_link_libraries(\${PROJECT_NAME} PUBLIC AWS::aws-lambda-runtime)
aws_lambda_package_target(\${PROJECT_NAME})
EOF

Build function:

mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=/tmp/out
make

Deploy with tm CLI:

tm deploy task -f https://raw.githubusercontent.com/triggermesh/aws-custom-runtime/main/runtime.yaml
tm deploy service lambda-cpp -f . --runtime aws-custom-runtime

C++ Lambda function is running on knative platform:

curl lambda-cpp.default.k.triggermesh.io --data '{"payload": "foobar"}'
Hello, World!

Events wrapping

Triggermesh AWS custom runtime supports events wrapping for better interoperability of functions and data originated from or targeted at the different platforms. Currently, there are two events wrapper available besides the default "passthrough" one:

API Gateway wrapper ensures that HTTP requests are digestible by the AWS Lambda functions and decodes their responses to the simple readable format
CloudEvents wrapper converts function responses into CloudEvents event objects.

Events wrapper can be enabled by setting function's environment variables and may have different set of configurable parameters. Let's take a look at CloudEvens example:

Generate sample Go function using tm CLI
```
tm generate go
```

Open go/serverless.yaml deployment manifest and add events wrapper env variables:

...
environment:
   RESPONSE_WRAPPER: CLOUDEVENTS
   CE_TYPE: go-klr-cloudevent
...

Deploy function:
```
tm deploy -f go --wait
```

After CLI report that deployment is succeeded, send a request to the function endpoint:

curl -d '{"Name":"Joe"}' https://go-demo-service-go-function.default.dev.munu.io

The result will be encoded into CloudEvents format:

Validation: valid
Context Attributes,
  specversion: 1.0
  type: go-klr-cloudevent
  source: go-demo-service-go-function
  subject: klr-response
  id: 7fa17c84-56f0-488a-b7de-e45d7fb3b7b1
  datacontenttype: application/json
Data (binary),
  "Hello Joe!"

Support

We would love your feedback on this tool so don't hesitate to let us know what is wrong and how we could improve it, just file an issue

Code of Conduct

This plugin is by no means part of CNCF but we abide by its code of conduct