This is a very minimal Funktor project that is handy for testing funktor during development.
It's the base project that you get by generating a new funktor app, with one addition. There's
a random job generator that you can run on a schedule to pump jobs into a test stack.
There's really no way to run this project locally as it's "cloud native" and is designed specifically to run in the AWS Lambda ecosystem. You'll want to deploy it to your AWS account to make use of it.
To deploy you'll need to have an AWS account, and will need to have credentials configured locally. If you've used AWS before you may already have everything in place.
If you have a file at ~/.aws/credentials then you may be good to go already.
TODO - Add some more details here about how to get set up. Probably a link to some resource or another.
This project uses the (Serverless framework)[https://www.serverless.com/] for deployment. Serverless is a node framework, so you'll need to have node installed.
TODO - Add more deets/links.
To deploy this project to your AWS account, first clone the repo, then cd into it.
Then run:
npm install
serverless deploy
Assuming everything goes smoothly you should now have a stack provisioned in CloudFormation and you
should see a funktor-testbed-dev-dashboard in your
CloudWatch Dashbaords list.
You'll notice that there is no activity happening on your dashboard yet. To help reduce the chances for unexpected expenses this project is setup so that you must manually start random jobs generating.
Which brings us to...
The function definition in config/function_definitions/random_job_generator.yml controls all aspects
of random job generation. The values you set there will be used by the generator code that lives in
lambda_event_handlers/random_job_generator.rb.
To get jobs started flowing you should change the value of reservedConcurrency to 1. (When you're
finished testing, don't forget to deactivate ranodom job generation!)
# config/function_definitions/random_job_generator.yml
handler: lambda_event_handlers/random_job_generator.RandomJobGenerator.call
timeout: 58
reservedConcurrency: 1
# snip ...After making that change, and saving the file, you can deploy again to start the action.
sls deploy
Then you should start seeing some activity in your dashboard.
Ther are a number of other options in that file, which are documented there, that you can use to affect the rate of new jobs coming in, how long they take to execute, and how many of them will throw a 'random' error (which is useful for testing retry logic).
Set reservedConcurrency back to 0, and then deploy to stop random job generation.
# config/function_definitions/random_job_generator.yml
handler: lambda_event_handlers/random_job_generator.RandomJobGenerator.call
timeout: 58
reservedConcurrency: 0
# snip ...After making that change, and saving the file, you can deploy again to stop the action.
sls deploy
A little while later you should see activity start to trail off, and then stop completely
after a while (15 minutes by default, which is controlled by MAX_JOB_DELAY_IN_SECONDS).