TheLanc3/rustdesk
1
0
Fork 0
mirror of https://github.com/rustdesk/rustdesk.git synced 2026-09-19 02:41:10 +03:00
Code Issues Packages Projects Releases Wiki Activity

kcp: make the congestion-control profile opt-in, not the default

Browse Source 
The branch had flipped KCP to nc=0 (built-in congestion window) for
every session. That is a transport-behavior change for all users made on
reasoning alone, and the reasoning does not decide it: which profile wins
depends on why packets are being lost.

nc=1 - what RustDesk has always shipped - never shrinks the send window,
so on a genuinely congested uplink it deepens the loss it is reacting to.
But nc=0's backoff is blunt: a fast retransmit halves the window while an
RTO sets cwnd = 1 outright (ikcp.c) and recovery slow-starts from one
packet, so on a link with random loss and no congestion - Wi-Fi
interference, a long-haul path - it reads loss as congestion and can
stall an interactive stream for seconds. That failure mode is also the
more visible one to a remote-desktop user.

No benchmark settles this either: a loopback A/B has no bottleneck queue,
hence no congestion to control, and would flatter nc=1 by construction.
Deciding it needs a shaped link or field data.

So keep the profile users already run and let the other one be asked for
("enable-kcp-congestion-control" = "Y"). Flipping the default later is a
one-line change once there is evidence. kcp-sys keeps its own test
covering the nc=0 path.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01ExUfAkYbq8UC9pQCiLy8TQ
This commit is contained in:
rustdesk
2026-08-07 15:55:28 +08:00
parent 79f7a72fb7
commit 654344dbd9
2 changed files with 23 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
src/kcp_stream.rs
View File

@@ -26,12 +26,11 @@ static KCP_RECV_ERR_LOG: hbb_common::log_throttle::LogThrottle =
hbb_common::log_throttle::LogThrottle::new(KCP_IO_ERR_LOG_INTERVAL);
impl KcpStream {
// Engage KCP's built-in congestion control (nc=0) unless disabled by option: pure turbo
// (nc=1) keeps blasting a full 1024-segment window through loss, which on constrained
// links amplifies brief loss into a spiral users experience as stalls or drops. This is
// sender-side only, so no wire negotiation is needed and either peer may run either
// profile. Requires kcp-sys from the `rustdesk-patches` branch, which wires the config
// factory into connection setup (on older revs the factory was stored but never consulted).
// Opt in to KCP's built-in congestion window (nc=0) instead of the pure turbo profile
// (nc=1) that has always shipped; see `get_kcp_cc_enabled` for why this is not the default.
// Sender-side only, so no wire negotiation is needed and either peer may run either profile.
// Requires kcp-sys from the `rustdesk-patches` branch, which wires the config factory into
// connection setup (on older revs the factory was stored but never consulted).
fn apply_kcp_config(endpoint: &mut KcpEndpoint) {
if crate::get_kcp_cc_enabled() {
endpoint.set_kcp_config_factory(Box::new(|conv| {