For chunked bodies, we overly aggressively consumed bytes from the client side reader buffer, causing us to drop a potentially future pipelined request after the chunked body. This fixes how we consume chunked body bytes so that a pipelined request is not dropped.
Making a draft for now while I figure out how to address the buffered request use case that the body_buffer test fails on.
For chunked bodies, we overly aggressively consumed bytes from the client side reader buffer, causing us to drop a potentially future pipelined request after the chunked body. This fixes how we consume chunked body bytes so that a pipelined request is not dropped.
Making a draft for now while I figure out how to address the buffered request use case that the
body_buffer
test fails on.