lfg2025/archy
1
0
Fork 0
Code Issues 20 Pull Requests Actions Packages Projects Releases 61 Wiki Activity

fix(mesh): don't reset reconnect backoff for sessions that die young

Browse Source 
Extracted from beff5dd5 on archy-hwconfig (the rest of that commit is
flash-feature code staying on its branch): a device that connects then
drops within seconds — e.g. mid-boot-loop — kept resetting backoff to
5s forever, and every retry's open() toggles DTR/RTS which itself
resets ESP32-family boards. Backoff now only resets after a session
survives STABLE_SESSION_THRESHOLD (20s), so an unstable device gets
progressively longer quiet gaps to actually finish booting.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
archipelago 2026-07-26 07:13:27 -04:00
parent 729cee573a
commit 449ed7c1f7
1 changed files with 19 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
core/archipelago/src/mesh/listener/mod.rs
View File

@ -87,6 +87,18 @@ const RECONNECT_DELAY_INIT: Duration = Duration::from_secs(5);
/// Maximum reconnect delay (cap for exponential backoff). /// Maximum reconnect delay (cap for exponential backoff).
const RECONNECT_DELAY_MAX: Duration = Duration::from_secs(60); const RECONNECT_DELAY_MAX: Duration = Duration::from_secs(60);
/// Minimum time a session must run before we trust it enough to reset
/// backoff to the minimum. Without this gate, a device that connects then
/// fails again within a couple of seconds (e.g. mid-boot-loop) never backs
/// off — every retry immediately re-opens the port, which toggles DTR/RTS
/// (resets many ESP32 boards' MCU on native-USB and CP2102/CH340
/// auto-reset-circuit boards alike), turning a device that's merely
/// unstable into a self-sustaining boot loop that outlasts whatever
/// triggered the original instability. Confirmed live 2026-07-23: a Heltec
/// V3 stuck retrying every ~5-15s for 5+ minutes after a failed firmware
/// flash left it in a marginal state.
const STABLE_SESSION_THRESHOLD: Duration = Duration::from_secs(20);
/// Number of consecutive write failures before we consider the device dead /// Number of consecutive write failures before we consider the device dead
/// and trigger a reconnection cycle. /// and trigger a reconnection cycle.
const MAX_CONSECUTIVE_WRITE_FAILURES: u32 = 3; const MAX_CONSECUTIVE_WRITE_FAILURES: u32 = 3;
@ -560,6 +572,7 @@ pub fn spawn_mesh_listener(
return; return;
} }
let session_start = std::time::Instant::now();
match session::run_mesh_session( match session::run_mesh_session(
&state, &state,
&data_dir, &data_dir,
@ -582,13 +595,14 @@ pub fn spawn_mesh_listener(
{ {
Ok(()) => { Ok(()) => {
info!("Mesh session ended cleanly"); info!("Mesh session ended cleanly");
// Session was established before ending — reset backoff // Only trust a session that actually ran for a while —
reconnect_delay = RECONNECT_DELAY_INIT; // see STABLE_SESSION_THRESHOLD's doc comment.
if session_start.elapsed() >= STABLE_SESSION_THRESHOLD {
reconnect_delay = RECONNECT_DELAY_INIT;
}
} }
Err(e) => { Err(e) => {
// Check if session was ever connected (vs failed to open) if session_start.elapsed() >= STABLE_SESSION_THRESHOLD {
let was_connected = state.status.read().await.device_connected;
if was_connected {
reconnect_delay = RECONNECT_DELAY_INIT; reconnect_delay = RECONNECT_DELAY_INIT;
} }
error!("Mesh session error: {} (retry in {:?})", e, reconnect_delay); error!("Mesh session error: {} (retry in {:?})", e, reconnect_delay);