#44 Line interpolation for fan speed.

Dell_iDRAC_fan_controller.sh

@@ -15,11 +15,38 @@ trap 'graceful_exit' SIGINT SIGQUIT SIGTERM
 
 # Check if FAN_SPEED variable is in hexadecimal format. If not, convert it to hexadecimal
 if [[ $FAN_SPEED == 0x* ]]; then
-  readonly DECIMAL_FAN_SPEED=$(printf '%d' $FAN_SPEED)
-  readonly HEXADECIMAL_FAN_SPEED=$FAN_SPEED
+  readonly DECIMAL_LOW_FAN_SPEED_OBJECTIVE=$(convert_hexadecimal_value_to_decimal "$FAN_SPEED")
+  # Unused
+  # readonly HEXADECIMAL_FAN_SPEED=$FAN_SPEED
 else
-  readonly DECIMAL_FAN_SPEED=$FAN_SPEED
-  readonly HEXADECIMAL_FAN_SPEED=$(convert_decimal_value_to_hexadecimal $FAN_SPEED)
+  readonly DECIMAL_LOW_FAN_SPEED_OBJECTIVE=$FAN_SPEED
+  # Unused
+  # readonly HEXADECIMAL_FAN_SPEED=$(convert_decimal_value_to_hexadecimal "$FAN_SPEED")
+fi
+
+# Check if fan speed interpolation is enabled
+if [ -z "$HIGH_FAN_SPEED" ] || [ -z "$CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION" ] || [ "$CPU_TEMPERATURE_THRESHOLD" -eq "$CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION" ]; then
+  readonly FAN_SPEED_INTERPOLATION_ENABLED=false
+
+  # We define these variables to the same values than user fan control profile
+  readonly HIGH_FAN_SPEED=$FAN_SPEED
+  readonly CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION=$CPU_TEMPERATURE_THRESHOLD
+elif [[ "$FAN_SPEED" -gt "$HIGH_FAN_SPEED" ]]; then
+  echo 'Error : $FAN_SPEED have to be less or equal to $HIGH_FAN_SPEED. Exiting.'
+  exit 1
+else
+  readonly FAN_SPEED_INTERPOLATION_ENABLED=true
+fi
+
+# Check if HIGH_FAN_SPEED variable is in hexadecimal format. If not, convert it to hexadecimal
+if [[ $HIGH_FAN_SPEED == 0x* ]]; then
+  readonly DECIMAL_HIGH_FAN_SPEED_OBJECTIVE=$(convert_hexadecimal_value_to_decimal "$HIGH_FAN_SPEED")
+  # Unused
+  # readonly HEXADECIMAL_HIGH_FAN_SPEED=$HIGH_FAN_SPEED
+else
+  readonly DECIMAL_HIGH_FAN_SPEED_OBJECTIVE=$HIGH_FAN_SPEED
+  # Unused
+  # readonly HEXADECIMAL_HIGH_FAN_SPEED=$(convert_decimal_value_to_hexadecimal "$HIGH_FAN_SPEED")
 fi
 
 # Check if the iDRAC host is set to 'local' or not then set the IDRAC_LOGIN_STRING accordingly
@@ -45,23 +72,34 @@ fi
 
 # If server model is Gen 14 (*40) or newer
 if [[ $SERVER_MODEL =~ .*[RT][[:space:]]?[0-9][4-9]0.* ]]; then
-  DELL_POWEREDGE_GEN_14_OR_NEWER=true
-  CPU1_TEMPERATURE_INDEX=2
-  CPU2_TEMPERATURE_INDEX=4
+  readonly DELL_POWEREDGE_GEN_14_OR_NEWER=true
+  readonly CPU1_TEMPERATURE_INDEX=2
+  readonly CPU2_TEMPERATURE_INDEX=4
 else
-  DELL_POWEREDGE_GEN_14_OR_NEWER=false
-  CPU1_TEMPERATURE_INDEX=1
-  CPU2_TEMPERATURE_INDEX=2
+  readonly DELL_POWEREDGE_GEN_14_OR_NEWER=false
+  readonly CPU1_TEMPERATURE_INDEX=1
+  readonly CPU2_TEMPERATURE_INDEX=2
 fi
 
 # Log main informations
 echo "Server model: $SERVER_MANUFACTURER $SERVER_MODEL"
 echo "iDRAC/IPMI host: $IDRAC_HOST"
 
-# Log the fan speed objective, CPU temperature threshold and check interval
-echo "Fan speed objective: $DECIMAL_FAN_SPEED%"
-echo "CPU temperature threshold: $CPU_TEMPERATURE_THRESHOLD°C"
+# Log the check interval, fan speed objective and CPU temperature threshold
 echo "Check interval: ${CHECK_INTERVAL}s"
+echo "Fan speed interpolation enabled: $FAN_SPEED_INTERPOLATION_ENABLED"
+if $FAN_SPEED_INTERPOLATION_ENABLED; then
+  echo "Fan speed lower value: $DECIMAL_LOW_FAN_SPEED_OBJECTIVE%"
+  echo "Fan speed higher value: $DECIMAL_HIGH_FAN_SPEED_OBJECTIVE%"
+  echo "CPU lower temperature threshold: $CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION°C"
+  echo "CPU higher temperature threshold: $CPU_TEMPERATURE_THRESHOLD°C"
+  echo ""
+  # Print interpolated fan speeds for demonstration
+  print_interpolated_fan_speeds "$CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION" "$CPU_TEMPERATURE_THRESHOLD" "$DECIMAL_LOW_FAN_SPEED_OBJECTIVE" "$DECIMAL_HIGH_FAN_SPEED_OBJECTIVE"
+else
+  echo "Fan speed objective: $DECIMAL_LOW_FAN_SPEED_OBJECTIVE%"
+  echo "CPU temperature threshold: $CPU_TEMPERATURE_THRESHOLD°C"
+fi
 echo ""
 
 # Define the interval for printing
@@ -73,7 +111,7 @@ IS_DELL_FAN_CONTROL_PROFILE_APPLIED=true
 # Check present sensors
 IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT=true
 IS_CPU2_TEMPERATURE_SENSOR_PRESENT=true
-retrieve_temperatures $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
+retrieve_temperatures "$IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT" "$IS_CPU2_TEMPERATURE_SENSOR_PRESENT"
 if [ -z "$EXHAUST_TEMPERATURE" ]; then
   echo "No exhaust temperature sensor detected."
   IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT=false
@@ -93,35 +131,49 @@ while true; do
   sleep $CHECK_INTERVAL &
   SLEEP_PROCESS_PID=$!
 
-  retrieve_temperatures $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
+  retrieve_temperatures "$IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT" "$IS_CPU2_TEMPERATURE_SENSOR_PRESENT"
 
   # Initialize a variable to store the comments displayed when the fan control profile changed
   COMMENT=" -"
   # Check if CPU 1 is overheating then apply Dell default dynamic fan control profile if true
-  if CPU1_OVERHEAT; then
+  if CPU1_OVERHEATING; then
     apply_Dell_fan_control_profile
 
     if ! $IS_DELL_FAN_CONTROL_PROFILE_APPLIED; then
       IS_DELL_FAN_CONTROL_PROFILE_APPLIED=true
 
       # If CPU 2 temperature sensor is present, check if it is overheating too.
       # Do not apply Dell default dynamic fan control profile as it has already been applied before
-      if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT && CPU2_OVERHEAT; then
+      if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT && CPU2_OVERHEATING; then
         COMMENT="CPU 1 and CPU 2 temperatures are too high, Dell default dynamic fan control profile applied for safety"
       else
         COMMENT="CPU 1 temperature is too high, Dell default dynamic fan control profile applied for safety"
       fi
     fi
   # If CPU 2 temperature sensor is present, check if it is overheating then apply Dell default dynamic fan control profile if true
-  elif $IS_CPU2_TEMPERATURE_SENSOR_PRESENT && CPU2_OVERHEAT; then
+  elif $IS_CPU2_TEMPERATURE_SENSOR_PRESENT && CPU2_OVERHEATING; then
     apply_Dell_fan_control_profile
 
     if ! $IS_DELL_FAN_CONTROL_PROFILE_APPLIED; then
       IS_DELL_FAN_CONTROL_PROFILE_APPLIED=true
       COMMENT="CPU 2 temperature is too high, Dell default dynamic fan control profile applied for safety"
     fi
+  elif CPU1_HEATING || $IS_CPU2_TEMPERATURE_SENSOR_PRESENT && CPU2_HEATING; then
+    HIGHEST_CPU_TEMPERATURE=$CPU1_TEMPERATURE
+    if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT; then
+      HIGHEST_CPU_TEMPERATURE=$(max $CPU1_TEMPERATURE $CPU2_TEMPERATURE)
+    fi
+
+    # F1 - lower fan speed
+    # F2 - higher fan speed
+    # T_CPU - highest temperature of both CPUs (if only one exists that will be CPU1 temp value)
+    # T1 - lower temperature threshold
+    # T2 - higher temperature threshold
+    # Fan speed = F1 + ( ( F2 - F1 ) * ( T_CPU - T1 ) / ( T2 - T1 ) )
+    DECIMAL_FAN_SPEED_TO_APPLY=$((DECIMAL_LOW_FAN_SPEED_OBJECTIVE + ((DECIMAL_HIGH_FAN_SPEED_OBJECTIVE - DECIMAL_LOW_FAN_SPEED_OBJECTIVE) * ((HIGHEST_CPU_TEMPERATURE - CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION) / (CPU_TEMPERATURE_THRESHOLD - CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION))))
+    apply_user_fan_control_profile 2 $DECIMAL_FAN_SPEED_TO_APPLY
   else
-    apply_user_fan_control_profile
+    apply_user_fan_control_profile 1 $DECIMAL_LOW_FAN_SPEED_OBJECTIVE
 
     # Check if user fan control profile is applied then apply it if not
     if $IS_DELL_FAN_CONTROL_PROFILE_APPLIED; then

Dockerfile

@@ -22,7 +22,9 @@ ENV IDRAC_HOST local
 # ENV IDRAC_USERNAME root
 # ENV IDRAC_PASSWORD calvin
 ENV FAN_SPEED 5
-ENV CPU_TEMPERATURE_THRESHOLD 50
+ENV HIGH_FAN_SPEED 40
+ENV CPU_TEMPERATURE_THRESHOLD 60
+ENV CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION 50
 ENV CHECK_INTERVAL 60
 ENV DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE false
 ENV KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT false

README.md

@@ -84,6 +84,8 @@ docker run -d \
   -e CPU_TEMPERATURE_THRESHOLD=<decimal temperature threshold> \
   -e CHECK_INTERVAL=<seconds between each check> \
   -e DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE=<true or false> \
+  -e CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION=<decimal temperature lower threshold> \
+  -e HIGH_FAN_SPEED=<decimal or hexadecimal fan speed> \
   -e KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT=<true or false> \
   --device=/dev/ipmi0:/dev/ipmi0:rw \
   tigerblue77/dell_idrac_fan_controller:latest
@@ -102,6 +104,8 @@ docker run -d \
   -e CPU_TEMPERATURE_THRESHOLD=<decimal temperature threshold> \
   -e CHECK_INTERVAL=<seconds between each check> \
   -e DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE=<true or false> \
+  -e CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION=<decimal temperature lower threshold> \
+  -e HIGH_FAN_SPEED=<decimal or hexadecimal fan speed> \
   -e KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT=<true or false> \
   tigerblue77/dell_idrac_fan_controller:latest
 ```
@@ -124,6 +128,8 @@ services:
       - CPU_TEMPERATURE_THRESHOLD=<decimal temperature threshold>
       - CHECK_INTERVAL=<seconds between each check>
       - DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE=<true or false>
+      - CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION=<decimal temperature lower threshold>
+      - HIGH_FAN_SPEED=<decimal or hexadecimal fan speed>
       - KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT=<true or false>
     devices:
       - /dev/ipmi0:/dev/ipmi0:rw
@@ -147,6 +153,8 @@ services:
       - CPU_TEMPERATURE_THRESHOLD=<decimal temperature threshold>
       - CHECK_INTERVAL=<seconds between each check>
       - DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE=<true or false>
+      - CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION=<decimal temperature lower threshold>
+      - HIGH_FAN_SPEED=<decimal or hexadecimal fan speed when interpolation enabled>
       - KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT=<true or false>
 ```
 
@@ -161,19 +169,43 @@ All parameters are optional as they have default values (including default iDRAC
 - `IDRAC_USERNAME` parameter is only necessary if you're adressing a distant iDRAC. **Default** value is "root".
 - `IDRAC_PASSWORD` parameter is only necessary if you're adressing a distant iDRAC. **Default** value is "calvin".
 - `FAN_SPEED` parameter can be set as a decimal (from 0 to 100%) or hexadecimaladecimal value (from 0x00 to 0x64) you want to set the fans to. **Default** value is 5(%).
-- `CPU_TEMPERATURE_THRESHOLD` parameter is the T°junction (junction temperature) threshold beyond which the Dell fan mode defined in your BIOS will become active again (to protect the server hardware against overheat). **Default** value is 50(°C).
+- `CPU_TEMPERATURE_THRESHOLD` parameter is the T°junction (junction temperature) threshold beyond which the Dell fan mode defined in your BIOS will become active again (to protect the server hardware against overheat). **Default** value is 60(°C).
 - `CHECK_INTERVAL` parameter is the time (in seconds) between each temperature check and potential profile change. **Default** value is 60(s).
 - `DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE` parameter is a boolean that allows to disable third-party PCIe card Dell default cooling response. **Default** value is false.
+
+If you want to enable fan speed interpolation, add the following parameters :
+- `CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION` parameter enables fan speed interpolation once exceeded. Fan speed interpolation will increase your fan speed proportionally to **HIGH_FAN_SPEED** until **CPU_TEMPERATURE_THRESHOLD** is reached. This parameter must be less or equal to **CPU_TEMPERATURE_THRESHOLD**. **Default** value is 50(°C).
+- `HIGH_FAN_SPEED` parameter is the fan speed that will be set at `CPU_TEMPERATURE_THRESHOLD` when interpolation mode is enabled. In other words, it defines maximum fan speed before swiching back to the Dell default dynamic fan control profile (see `CPU_TEMPERATURE_THRESHOLD` parameter). **Default** value is 40(%).
+
+Example of how interpolation works:
+- `FAN_SPEED` = 10
+- `HIGH_FAN_SPEED` = 50
+- `CPU_TEMPERATURE_THRESHOLD_FOR_FAN_SPEED_INTERPOLATION` = 30
+- `CPU_TEMPERATURE_THRESHOLD` = 70
+
+| CPU temperature |                Fan speed                 |
+| --------------- | ---------------------------------------- |
+| 15 °C           | 10 %                                     |
+| 30 °C           | 10 %                                     |
+| 35 °C           | 15 %                                     |
+| 50 °C           | 30 %                                     |
+| 69 °C           | 49 %                                     |
+| 70 °C           | Dell default dynamic fan control profile |
+| 80 °C           | Dell default dynamic fan control profile |
+
+When using fan speed interpolation, we recommend decreasing **CHECK_INTERVAL**, for example "3" (seconds), to avoid the noise nuisance associated with a sudden increase in fan speed.
 - `KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT` parameter is a boolean that allows to keep the third-party PCIe card Dell default cooling response state upon exit. **Default** value is false, so that it resets the third-party PCIe card Dell default cooling response to Dell default.
 
 <p align="right">(<a href="#top">back to top</a>)</p>
 
 <!-- TROUBLESHOOTING -->
 ## Troubleshooting
 
-If your server frequently switches back to the default Dell fan mode:
+If your server frequently switches back to the default Dell default dynamic fan control profile:
 1. Check `Tcase` (case temperature) of your CPU on Intel Ark website and then set `CPU_TEMPERATURE_THRESHOLD` to a slightly lower value. Example with my CPUs ([Intel Xeon E5-2630L v2](https://www.intel.com/content/www/us/en/products/sku/75791/intel-xeon-processor-e52630l-v2-15m-cache-2-40-ghz/specifications.html)) : Tcase = 63°C, I set `CPU_TEMPERATURE_THRESHOLD` to 60(°C).
-2. If it's already good, adapt your `FAN_SPEED` value to increase the airflow and thus further decrease the temperature of your CPU(s)
+2. If it's already good, either :
+    - adapt your `FAN_SPEED` value to increase the airflow and thus further decrease the temperature of your CPU(s)
+    - enable and experiment fan speed interpolation mode
 3. If neither increasing the fan speed nor increasing the threshold solves your problem, then it may be time to replace your thermal paste
 
 <p align="right">(<a href="#top">back to top</a>)</p>