I have a client consuming a chunked HTTP response.
HTTPoison.request(method, url, body, headers, stream_to: self(), timeout: 50_000, recv_timeout: 50_000)
However HTTPoison is able send messages to the process quicker than we can process them. The process mailbox is getting up to 30K messages and using all our memory. Is there any way to limit the rate at which HTTPoison fetches and forwards the response body to this process.
Happy to drop to the erlang/hackney interface if that exposes more options.