Currently in our VertxHttpClient we do the following:
==> we take all the request's content into a byte array to give it to Vert.x HTTP client. By doing so, if the request's content is big, then it can lead to an out-of-memory.
We should consume the Entity through its rawContentInputStream and consume chunk-by-chunk the stream.
As we will use some blocking APIs (Inputstream#read then we should offload that operation of bytes transfers into a worker thread, in order not to block the event-loop. But we can debate this...