reverseproxy: Revert #4952 - don't ignore context cancellation in stream mode

i.e. Revert commit f5dce84a7028d1b116db7fead27ff8b2506baf78

Two years ago, the patch in #4952 was a seemingly necessary way to fix an issue (sort of an edge case), but it broke other more common use cases (see #6666).

Now, as of #6669, it seems like the original issue can no longer be replicated, so we are reverting that patch, because it was incorrect anyway.

If it turns out the original issue returns, a more proper patch may be in #6669 (even if used as a baseline for a future fix). A potential future fix could be an opt-in setting.

1 files changed, 1 insertions, 16 deletions

diff --git a/modules/caddyhttp/reverseproxy/reverseproxy.go b/modules/caddyhttp/reverseproxy/reverseproxy.go
index 08be40d62..ba9e5f3fc 100644
--- a/modules/caddyhttp/reverseproxy/reverseproxy.go
+++ b/modules/caddyhttp/reverseproxy/reverseproxy.go
@@ -106,11 +106,6 @@ type Handler struct {
 	// response is recognized as a streaming response, or if its
 	// content length is -1; for such responses, writes are flushed
 	// to the client immediately.
-	//
-	// Normally, a request will be canceled if the client disconnects
-	// before the response is received from the backend. If explicitly
-	// set to -1, client disconnection will be ignored and the request
-	// will be completed to help facilitate low-latency streaming.
 	FlushInterval caddy.Duration `json:"flush_interval,omitempty"`
 
 	// A list of IP ranges (supports CIDR notation) from which
@@ -773,17 +768,7 @@ func (h *Handler) reverseProxy(rw http.ResponseWriter, req *http.Request, origRe
 	}
 	req = req.WithContext(httptrace.WithClientTrace(req.Context(), trace))
 
-	// if FlushInterval is explicitly configured to -1 (i.e. flush continuously to achieve
-	// low-latency streaming), don't let the transport cancel the request if the client
-	// disconnects: user probably wants us to finish sending the data to the upstream
-	// regardless, and we should expect client disconnection in low-latency streaming
-	// scenarios (see issue #4922)
-	if h.FlushInterval == -1 {
-		req = req.WithContext(ignoreClientGoneContext{req.Context()})
-	}
-
-	// do the round-trip; emit debug log with values we know are
-	// safe, or if there is no error, emit fuller log entry
+	// do the round-trip
 	start := time.Now()
 	res, err := h.Transport.RoundTrip(req)
 	duration := time.Since(start)