Getting started

Dependencies

You will need following things which you have to install on your own depending on your system.

• Homebrew (for Mac OS) /Linuxbrew (for Linux)
• Java 8
• Docker, Docker Compose >= 1.13.1, 1.22.0

After having all mentioned above dependencies just use the following commands:

1. Install aws, aws-sam clojure and babashka>=0.3.7

   brew tap aws/tap && brew install awscli aws-sam-cli clojure/tools/clojure  borkdude/brew/babashka
   

2. Configure default AWS profile via aws-cli. This is necessary for making a bucket and deploying an application. If you just want to test holy-lambda on local then this step is not necessary, but you will be able to use only a limited set of commands.

   aws configure
   

First project

1. With clj-new set up, scaffold a new project

   clojure -M:new -m clj-new.create holy-lambda basic.example 
   

   You should see following project structure when cd to the project directory:

    .
    ├── README.md
    ├── bb.edn
    ├── deps.edn
    ├── envs.json
    ├── resources
    │   └── native-agents-payloads
    │       └── 1.edn
    ├── src
    │   └── basic
    │       └── example
    │           └── core.cljc
    └── template.yml
   
    5 directories, 7 files
   

2. Try to sync dependencies in project:

   cd basic.example && bb stack:sync
   

   The first sync is not always successful. If you see a message of not successful sync, then remove the .holy-lambda folder and run bb stack:sync once again. The purpose of the sync command is to gather all dependencies from bb.edn, deps.edn for both Clojure, Native and Babashka runtime. Syncs checks as well whether any additional layers for runtime should be published. If :self-manage-layers flag is set to false, then holy-lambda will automatically publish all necessary layers and output ARN of each.

3. At this point you should have .holy-lambda directory in your project. If not then go to troubleshooting. Now you can choose one of three runtimes:

   • :babashka
   • :native
   • :java

   The template is adjusted in the way that all of the runtimes should work flawlessly and all you need to change is a :runtime value in bb.edn.

   1. :babashka runtime Babashka runtime is probably the best one to start the journey with holy-lambda.

      Run bb stack:invoke:

      ❯ bb stack:invoke
      [holy-lambda] Command <stack:invoke>
      Invoking basic.example.core.ExampleLambda (provided)
      arn:aws:lambda:eu-central-1:443526418261:layer:holy-lambda-babashka-runtime:20 is already cached. Skipping download
      Skip pulling image and use local one: samcli/lambda:provided-66f9fa34aefd240f454416a79.
      
      Mounting /home/fierycod/Workspace/Work/holy-lambda/examples/bb/babashka/basic.example/src as /var/task:ro,delegated inside runtime container
      START RequestId: 78eca6ee-0389-442c-ad81-9a0e3d64d3fb Version: $LATEST
      END RequestId: 78eca6ee-0389-442c-ad81-9a0e3d64d3fb
      REPORT RequestId: 78eca6ee-0389-442c-ad81-9a0e3d64d3fb  Init Duration: 0.04 ms  Duration: 198.68 ms     Billed Duration: 200 ms Memory Size: 2000 MB    Max Memory Used: 2000 MB
      {"statusCode":200,"headers":{"Content-Type":"text/plain; charset=utf-8"},"body":"Hello world. Babashka is sweet friend of mine! Babashka version: 0.3.7"}
      

      After some time you should see above output. First invocation is rather slow locally since AWS SAM has to download runtime image for babashka.

Troubleshooting

1. Running bb stack:sync results in:

   Project did not sync properly. Remove .holy-lambda directory and run
   

   Solution:

   bb stack:purge && bb stack:sync
   

2. Commands are failing:

   Solution:

   bb stack:purge && bb stack:doctor
   

   Fix all errors reported by the tool. If you still experience any issue please report it at Github.

3. GraalVM native-image compilation fails due to not enough RAM memory on MacOS

   Solution*: Increate the RAM limit in Docker UI preferences.