viktor/infra
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fan-control: estimate + expose fan power (fan_watts_est)

Browse source 
The iDRAC reports only total DCMI watts + RPM (no per-fan power), so add a
cube-law fan-power estimate: fan_W ~= 0.0205*(RPM/1000)^3, calibrated to the
2026-06-05 sweep (fits within ~3W; ~2W floor -> ~99W full). The daemon reads
live RPM each loop and pushes pve_fan_control_fan_rpm + _fan_watts_est.
Surfaced in HA as sensor.r730_fan_power_est + a "Fan Power (est)" card on the
dashboard-it Server view, next to total power. 46 bash tests green; verified
live (9120rpm -> ~15W est).

[ci skip]

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Viktor Barzin 2026-06-05 11:10:27 +00:00
parent 324f2dc3bf
commit 51456a96f6
4 changed files with 41 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

17
docs/plans/2026-06-04-pve-fan-control-design.md
View file

@ -103,10 +103,19 @@ Manual fan mode bypasses the iDRAC's own protection, so it is backstopped:
## Observability
Pushes to the existing Pushgateway (`http://10.0.20.100:30091`, job
`fan_control`): `pve_fan_control_cpu_temp_celsius`, `_fan_percent`, `_mode`
(1 quiet / 2 cool / 0 fallback), `_ha_reachable`, `_fallback`. The existing CPU-
temp alert is unaffected.
Pushes to the Pushgateway (`http://10.0.20.100:30091`, job `fan_control`):
`pve_fan_control_cpu_temp_celsius`, `_fan_percent`, `_mode` (1 quiet / 2 cool /
3 manual / 0 fallback), `_ha_reachable`, `_fallback`, `_fan_rpm`, and
`_fan_watts_est`.
**Fan power is ESTIMATED** — the iDRAC exposes only total DCMI watts + RPM (no
per-fan power), so `_fan_watts_est` models it from RPM via the fan affinity law
(power ∝ RPM³), calibrated to the 2026-06-05 sweep: `fan_W ≈ 0.0205·(RPM/1000)³`
(≈2 W at the floor → ~99 W at full; fits the sweep within ~3 W). Surfaced in HA
as `sensor.r730_fan_power_est` + a "Fan Power (est)" card on the dashboard-it
Server view, next to total power (`sensor.r730_power_consumption`, redfish) — so
the fan tax of the control curve is visible. The existing CPU-temp alert is
unaffected.
## Testing

5
docs/runbooks/fan-control.md
View file

@ -24,7 +24,10 @@ can't run the fans wrong indefinitely. `CEILING` (83 °C) still overrides
everything → Dell auto. An HA change is applied within one daemon loop (~15 s).
Monitoring sensors on the same view: `sensor.r730_fan_speed` (redfish exporter),
`sensor.r730_fan_control_target` + `sensor.r730_fan_control_mode` (Pushgateway).
`sensor.r730_fan_control_target` + `sensor.r730_fan_control_mode` +
`sensor.r730_fan_power_est` (Pushgateway). `r730_fan_power_est` is an ESTIMATE of
total fan power (the iDRAC reports no per-fan power) — modelled from RPM via the
fan affinity law (∝ RPM³), calibrated to the power sweep (~2 W floor → ~99 W full).
The HA objects (helpers, the auto-revert automation, the REST sensors in
`rest_resources/{idrac_redfish_exporter,fan_control}.yaml`, and the dashboard

19
scripts/fan-control.sh
View file

@ -109,6 +109,12 @@ fc_pct_to_hex() { printf '0x%02x' "$1"; }
# fc_clamp <pct> -> 0..100
fc_clamp() { local p="$1"; (( p < 0 )) && p=0; (( p > 100 )) && p=100; echo "$p"; }
# fc_fan_watts <rpm> -> estimated TOTAL fan power (W). The iDRAC reports only
# total DCMI watts + RPM (no per-fan power), so this is a MODEL: fan power ∝ RPM³
# (fan affinity law), calibrated to the 2026-06-05 power sweep — fits within ~3W
# (~2W @4800rpm · ~17W @9360 · ~42W @12720 · ~99W @16920). Integer: 0.0205·(rpm/1e3)³.
fc_fan_watts() { echo $(( $1 * $1 * $1 * 205 / 10000000000000 )); }
# fc_resolve <ha_mode> <temp> <manual_pct> <presence> <current> <deadband> -> pct
# HA mode resolution (the hard ceiling is handled by the caller):
# manual -> clamp(manual_pct), no hysteresis
@ -145,6 +151,10 @@ read_cpu_temp() {
fc_parse_temp "$("$IPMITOOL" sdr type temperature 2>/dev/null | grep -E '^Temp ' | head -1)"
}
read_fan_rpm() { # Fan1 RPM — representative (all 6 fans are set together)
"$IPMITOOL" sdr type fan 2>/dev/null | awk -F'|' '/^Fan1/{gsub(/[^0-9]/,"",$5); print $5+0; exit}'
}
presence_cache="cool"; presence_ts=0
get_presence() {
local now; now="$(date +%s)"
@ -171,7 +181,7 @@ ha_entity_state() {
fc_json_str_field "$resp" state
}
push_metrics() { # <temp> <pct> <mode> <ha_ok> <fallback>
push_metrics() { # <temp> <pct> <mode> <ha_ok> <fallback> [fan_rpm] [fan_watts_est]
[[ -z "$PUSHGATEWAY_URL" ]] && return 0
local mode_num; case "$3" in quiet) mode_num=1;; cool) mode_num=2;; manual) mode_num=3;; *) mode_num=0;; esac
curl -fsS --max-time 5 --data-binary @- \
@ -186,6 +196,10 @@ pve_fan_control_mode $mode_num
pve_fan_control_ha_reachable $4
# TYPE pve_fan_control_fallback gauge
pve_fan_control_fallback $5
# TYPE pve_fan_control_fan_rpm gauge
pve_fan_control_fan_rpm ${6:-0}
# TYPE pve_fan_control_fan_watts_est gauge
pve_fan_control_fan_watts_est ${7:-0}
EOF
}
@ -239,7 +253,8 @@ main() {
if set_manual "$pct"; then log "temp=${temp}C ha_mode=${ha_mode} eff=${eff} fan=${pct}% (was ${current}%)"; current="$pct"
else log "WARN set_manual ${pct}% failed"; fi
fi
push_metrics "$temp" "$current" "$eff" "$ha_ok" 0
local rpm fan_w; rpm="$(read_fan_rpm)"; rpm="${rpm:-0}"; fan_w="$(fc_fan_watts "$rpm")"
push_metrics "$temp" "$current" "$eff" "$ha_ok" 0 "$rpm" "$fan_w"
(( RUN_ONCE == 1 )) && break || sleep "$LOOP_INTERVAL"
done
}

7
scripts/test-fan-control.sh
View file

@ -52,6 +52,13 @@ eq "resolve quiet=quiet curve" 73 "$(fc_resolve quiet 78 0 cool -1 3)"
eq "resolve auto+empty=cool" 51 "$(fc_resolve auto 60 0 cool -1 3)"
eq "resolve auto+present=quiet" 31 "$(fc_resolve auto 70 0 quiet -1 3)"
# --- fc_fan_watts: estimated fan power from RPM (cube-law, calibrated to the sweep) ---
eq "fan_watts 0" 0 "$(fc_fan_watts 0)"
eq "fan_watts 4800" 2 "$(fc_fan_watts 4800)"
eq "fan_watts 9360" 16 "$(fc_fan_watts 9360)"
eq "fan_watts 12720" 42 "$(fc_fan_watts 12720)"
eq "fan_watts 16920" 99 "$(fc_fan_watts 16920)"
# --- presence ---
now=1000000
eq "presence open -> quiet" quiet "$(fc_presence_mode Отворена 0 $now 900 Отворена)"