voter-lambda project

This project uses Quarkus, the Supersonic Subatomic Java Framework.

If you want to learn more about Quarkus, please visit its website: https://quarkus.io/ .

Local One Time Setup

Create an Execution Role

View the Getting Started Guide for deploying a lambda with AWS CLI. Specifically, make sure you have created an Execution Role. To use the provided scripts, you will either need to define a LAMBDA_ROLE_ARN environment variable in your profile or define it prior to executing the scripts like this:

export LAMBDA_ROLE_ARN="arn:aws:iam::1234567890:role/lambda-role"

Running the application in dev mode

You can run your application in dev mode that enables live coding using:

./gradlew :voter-lambda:quarkusDev

Packaging and running the application

The application is packageable using ./gradlew quarkusBuild. It produces the executable voter-lambda-1.0.0-SNAPSHOT-runner.jar file in build directory. Be aware that it’s not an über-jar as the dependencies are copied into the build/lib directory.

The application is now runnable using java -jar build/voter-lambda-1.0.0-SNAPSHOT-runner.jar.

If you want to build an über-jar, just add the --uber-jar option to the command line:

./gradlew quarkusBuild --uber-jar

Creating a native executable

(not working) You can create a native executable using: ./gradlew buildNative.

Or you can use Docker to build the native executable using: ./gradlew buildNative --docker-build=true.

You can then execute your binary: ./build/voter-lambda-1.0.0-SNAPSHOT-runner

If you want to learn more about building native executables, please consult https://quarkus.io/guides/gradle-tooling#building-a-native-executable .

Deploying the lambda to AWS

JDK8 Lambda

cd <root project>/voter-lambda
../gradlew quarkusBuild --uber-jar
sam local invoke --template sam.jvm.yaml --event payload.json
sh manage.sh create
Use AWS Lambda console to invoke it using the Test feature. Example test payload to pass: { "lastName": "shimono" }

Native Lambda

cd <root project>/voter-lambda
../gradlew buildNative --docker-build=true --enable-http-url-handler
zip -j build/function.zip src/main/resources/native/bootstrap build/voter-lambda-1.0-SNAPSHOT-runner
sam local invoke --template sam.native.yaml --event payload.json
sh manage.sh native create
Use AWS Lambda console to invoke it using the Test feature. Example test payload to pass: { "lastName": "shimono" }

AWS Lambda Performance Results

Runtime	Cold Start Latency*	Warm Latency	Resources Configured	Max Memory Used	Artifact Size
JDK 8	765 ms	.73 ms	256 MB	124 MB	38 MB
Custom (Native)	18 ms	.99 ms	256 MB	72 MB	11 MB

Conclusion

• Native results show JVM languages can be used for lambdas to have similar latency and resource utilization as other languages like Go, Python and Node
• Execution times were essentially the same when the lambda was warm
• Cold Start is a serious concern when the runtime was JDK 8 and latency is important

Note: Cold start latency fluctuated and were not consistent for both JDK 8 and Custom (Native).