Status Updates From Building Microservices: Des...

An event is a fact, a statement that something happened, along with some info about what happened.

A message is a thing we send over an asynchronous communication mechanism, like a message broker.

With event-driven collaboration, we want to broadcast that event, and a
typical way to implement that broadcast mechanism would be to put the
event into a message.

The message is the medium, the event is the payload.

This means in request-response, the requester has to have knowledge of what the downstream recipient can do, with event-driven, the
emitter doesn’t need to know what any microservices are able to do.

The intent behind an event is the opposite of a request. The
event emitter is leaving it up to the recipients to decide what to do.

With request-response, the microservice sending the request knows what should be done and is telling the other microservice what it thinks needs to happen next.

A Data Warehouse is a relational data store, highly structured, where large amounts of data are stored and they are used, usually for ETL pipelines.

A Data Lake is an object store. It is essentially a giant folder of files. It is not relational. There are no SQL tables. The Data Lake is just a dumb file system, but we can use Databricks on top of the data lake to query the data inside.

The communication through common data is suitable when for instance you need to share a large amount of data to another microservice. If you do it via API, network gets overloaded. If you do it via message brokers, it will be a terrible idea, as message brokers are intended for short frequent messages.

So the approach allows you to push the data into a data store from where microservices can read and use it.

Asynchronous communication through common data mean microservices define a location in which data is stored and at some point later another microservice picks up that data and uses it somehow.