@mk_aem21 100% agreed, documentation is bit overwhelming and can misguide. Let me try to breakdown.
To start:
Adobe I/O - is the overall framework. Under aio umbrella, we have aio runtime action, web assets, events etc etc.. aio equivalent might be azure functions or aws lambdas. You can create microservices ( which is aio runtime actions) or microfrontends (which is aio web assts) using AIO Runtime. Once you create hello worlds and deploy, will give idea.
Next level, your runtime action (microservice) can be web-action or non-webaction. Meaning you can call the microservice from Postman OR create as cron job - non-webaction. For cron, aio uses OpenWhisk Alarm package.
Next level, your helloworld runtime action runs when you call from Postman. You want to integrate with other adobe products. For example, asset is published in AEM, and want to receive as event and send that data to 3rd party. For this, you ll start integrating your runtime with IO Events. The IO Events creates bridge to receive events, and then your runtime action (microservice) can process that event. This is the most useful part of aio.
Overall,
Adobe I/O is the framework equivalent to Azure functions.
Runtime is the coding space for developers to create micro-services / micro-frontends.
App Builder is the Cloud Infrastructure to run these stuff.
So you USE AIO framework, WRITE runtime actions/events, DEPLOY onto App Builder project.
To get started, just create a simple runtime action. Invoke using postman or CLI and validate.
To debug codes, agreed, its not straight. This is serverless cloud computation world. We dont have physical server that runs always. Instead only when microservice(runtime action) is invoked, a new server is created, code deployed, action executed and trashed immediately. All happens within seconds.
To debug, adobe uses Apache Openwhisk way like your link above. But agreed not easy to setup and get started. Easier way is to add log statements and when calling the action from postman, you can include
X-OW-EXTRA-LOGGING: on
headers. This will enable logging. Alternate is setup log forwarding from runtime into your org splunk and read splunk logs.
Overall this is a new beast. Coding on cloud world, thinking from microservices/microfrontend eyes. Hope this explanation helps to get started.
