Based on this AWS doc both layer 4 and layer 7 ALBS terminate the client connection and establish a new TCP connection with the back-end instances (connection are reused on both ends).
- Is there any info on how these connections are managed? (i.e 1 server connection per client connection?)
- How is TCP backpressure handled? If the server stops reading TCP packets, will the client be made aware of it so it can backpressure?
