Request for super-heroes workshop to be updated to use gRPC using the reactive approach

[dhoffer]

I am trying to understand now to use Quarkus's reactive stack for working with Kafka and GRPC. I am finding the learning curve to be steep as there are so many abstractions in the reactive approach that one cannot just follow code paths as they are hidden behind various annotations.

I just found this online self study workshop and it appears quite good and fairly current. However in the reactive approach it appears to use use web-sockets to update the UI.

https://quarkus.io/quarkus-workshops/super-heroes/

I would like to see this workshop updated to also show how to implement using gRPC.

Ultimately want to be able to understand how to do the following with gRPC & Kafka.

1. Create gRPC service where users can request batches of Kafka records per offsets and stream response back to user.
2. Create gRPC service where users can request that new Kafka records are streamed back to the user as they are available in Kafka, open server-to-client push.

Example needs to explain issues such as back-pressure, length of time server-to-client connection can be open/etc.

Ideally the workshop would explore the various options that are available using the reactive approach and pros/cons of each approach. Our use cause would be for extremely high data volume and need low latency from when data is published to Kafka and when it has reached consumers.