Let the fan settle downward on a re-assert

A symmetric deadband catches the fan on the way up and holds it there:
on iz-pve0 under load, 17 of 20 samples ran at 38% while the curve
wanted 34-36%, and nothing would bring it down until the request fell a
full 8 points. Damping was turning into a permanently louder machine.

A re-assert may now settle the speed downward to the current request.
Upward moves still require a full deadband crossing, so the drift this
was built to prevent cannot come back - covered by a test.

Verified under a live migration: settled 38 -> 36 on the first
re-assert, then one change in ten minutes while the request wandered
36-42, CPU stable at 66-69c and disks a degree cooler.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

README.md

@@ -32,6 +32,13 @@ settle — measurably worse than stock. Measured on iz-pve0: stock iDRAC held on
 minutes off the same signal. A request for full speed is never held back, and the held
 value is re-pushed every `FAN_REASSERT_INTERVAL` in case the BMC forgets it.
 
+A deadband alone makes a machine permanently louder: it catches the fan on the way up and
+then nothing brings it down until the request falls a full deadband. Measured on `iz-pve0`
+under load, 17 of 20 samples ran at 38% while the curve wanted 34-36%. So a re-assert -
+and only a re-assert - is allowed to settle the speed **downward** to what is actually
+wanted. Upward moves still need a full deadband crossing, which is what stops this
+reintroducing the drift it was built to prevent.
+
 Note that iDRAC's own profile regulates to a target *RPM* (closed loop, rock steady),
 while manual control sets a *PWM* percentage (open loop, so measured RPM still wanders a
 percent or two at a fixed setpoint). That residual is inherent to manual control, not

fan_speed.sh

@@ -117,6 +117,14 @@ selftest() {
   # leak is what let the setpoint drift 32 -> 36 -> 31 on iz-pve0.
   check "re-assert re-pushes held value" 32 "$(next_fan_speed 34 32)"
 
+  echo "settling downward (damping must not make the machine permanently louder):"
+  check "holds above the want between re-asserts" 38 "$(settle_fan_speed 34 38 0)"
+  check "settles down to the want on re-assert"   34 "$(settle_fan_speed 34 38 1)"
+  check "small rise still held on re-assert"      38 "$(settle_fan_speed 42 38 1)"
+  check "real rise still adopted"                 46 "$(settle_fan_speed 46 38 1)"
+  # The old upward leak: a re-assert must never drift the setpoint up.
+  check "no upward drift on re-assert"            32 "$(settle_fan_speed 36 32 1)"
+
   echo "trend verdict (load is not a fault - only cooling that cannot keep up is):"
   TREND_SAMPLES=5
   HIGH_FAN_SPEED=50
@@ -293,12 +301,12 @@ one_pass() {
 
   # Hold the current speed unless the change is worth making - see the deadband
   # note in monitor.sh. Chasing every degree is what made this louder than stock.
-  local target now held=""
-  target=$(next_fan_speed "$DECIMAL_CURRENT_FAN_SPEED" "$APPLIED_FAN_SPEED")
-  [ "$target" = "$DECIMAL_CURRENT_FAN_SPEED" ] || held=" held, want $DECIMAL_CURRENT_FAN_SPEED"
+  local target now due=0 held=""
   now=$(date +%s)
-  if [ "$target" != "$APPLIED_FAN_SPEED" ] ||
-     [ $((now - APPLIED_FAN_SPEED_AT)) -ge "$FAN_REASSERT_INTERVAL" ]; then
+  [ $((now - APPLIED_FAN_SPEED_AT)) -ge "$FAN_REASSERT_INTERVAL" ] && due=1
+  target=$(settle_fan_speed "$DECIMAL_CURRENT_FAN_SPEED" "$APPLIED_FAN_SPEED" "$due")
+  [ "$target" = "$DECIMAL_CURRENT_FAN_SPEED" ] || held=" held, want $DECIMAL_CURRENT_FAN_SPEED"
+  if [ "$target" != "$APPLIED_FAN_SPEED" ] || [ "$due" = 1 ]; then
     apply_user_fan_control "$target"
     APPLIED_FAN_SPEED=$target
     APPLIED_FAN_SPEED_AT=$now

monitor.sh

@@ -97,6 +97,25 @@ next_fan_speed() {
   fi
 }
 
+# settle_fan_speed <wanted> <applied> <re-assert due, 0|1>
+#
+# A symmetric deadband catches the fan on the way up and then holds it there:
+# measured on iz-pve0 under load, 17 of 20 samples ran at 38% while the curve
+# wanted 34-36%, and nothing would bring it down until the request fell a full
+# deadband. That is how damping turns into a permanently louder machine.
+#
+# So on a re-assert - and only then - let the speed settle DOWNWARD to what is
+# actually wanted. Upward moves still require a full deadband crossing, so this
+# cannot bring back the leak that let the setpoint drift 32 -> 36 -> 31.
+settle_fan_speed() {
+  local wanted=$1 applied=$2 due=$3 target
+  target=$(next_fan_speed "$wanted" "$applied")
+  if [ "$due" = 1 ] && [ -n "$applied" ] && [ "$wanted" -lt "$target" ]; then
+    target=$wanted
+  fi
+  echo "$target"
+}
+
 ALERT_EMAIL=${ALERT_EMAIL:-Servers@ntfy1.izebra.xyz}
 ALERT_COOLDOWN=${ALERT_COOLDOWN:-3600}
 STATE_DIR=${STATE_DIR:-/root/fan_speed/state}