Skip to content

Commit

Permalink
Minor code structure improvements
Browse files Browse the repository at this point in the history
  • Loading branch information
tigerblue77 committed Nov 17, 2024
1 parent ce7bb3f commit db64028
Show file tree
Hide file tree
Showing 3 changed files with 32 additions and 54 deletions.
55 changes: 19 additions & 36 deletions Dell_iDRAC_fan_controller.sh
Original file line number Diff line number Diff line change
Expand Up @@ -14,8 +14,7 @@ trap 'graceful_exit' SIGINT SIGQUIT SIGTERM
# readonly DELL_FRESH_AIR_COMPLIANCE=45

# Check if FAN_SPEED variable is in hexadecimal format. If not, convert it to hexadecimal
if [[ $FAN_SPEED == 0x* ]]
then
if [[ $FAN_SPEED == 0x* ]]; then
readonly DECIMAL_FAN_SPEED=$(printf '%d' $FAN_SPEED)
readonly HEXADECIMAL_FAN_SPEED=$FAN_SPEED
else
Expand All @@ -24,8 +23,7 @@ else
fi

# Check if the iDRAC host is set to 'local' or not then set the IDRAC_LOGIN_STRING accordingly
if [[ $IDRAC_HOST == "local" ]]
then
if [[ $IDRAC_HOST == "local" ]]; then
# Check that the Docker host IPMI device (the iDRAC) has been exposed to the Docker container
if [ ! -e "/dev/ipmi0" ] && [ ! -e "/dev/ipmi/0" ] && [ ! -e "/dev/ipmidev/0" ]; then
echo "/!\ Could not open device at /dev/ipmi0 or /dev/ipmi/0 or /dev/ipmidev/0, check that you added the device to your Docker container or stop using local mode. Exiting." >&2
Expand All @@ -40,15 +38,13 @@ fi

get_Dell_server_model

if [[ ! $SERVER_MANUFACTURER == "DELL" ]]
then
if [[ ! $SERVER_MANUFACTURER == "DELL" ]]; then
echo "/!\ Your server isn't a Dell product. Exiting." >&2
exit 1
fi

# If server model is Gen 14 (*40) or newer
if [[ $SERVER_MODEL =~ .*[RT][[:space:]]?[0-9][4-9]0.* ]]
then
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
Expand Down Expand Up @@ -78,19 +74,16 @@ IS_DELL_FAN_CONTROL_PROFILE_APPLIED=true
IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT=true
IS_CPU2_TEMPERATURE_SENSOR_PRESENT=true
retrieve_temperatures $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
if [ -z "$EXHAUST_TEMPERATURE" ]
then
if [ -z "$EXHAUST_TEMPERATURE" ]; then
echo "No exhaust temperature sensor detected."
IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT=false
fi
if [ -z "$CPU2_TEMPERATURE" ]
then
if [ -z "$CPU2_TEMPERATURE" ]; then
echo "No CPU2 temperature sensor detected."
IS_CPU2_TEMPERATURE_SENSOR_PRESENT=false
fi
# Output new line to beautify output if one of the previous conditions have echoed
if ! $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT || ! $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
then
if ! $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT || ! $IS_CPU2_TEMPERATURE_SENSOR_PRESENT; then
echo ""
fi

Expand All @@ -103,60 +96,51 @@ while true; do
retrieve_temperatures $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT $IS_CPU2_TEMPERATURE_SENSOR_PRESENT

# Define functions to check if CPU 1 and CPU 2 temperatures are above the threshold
function CPU1_OVERHEAT () { [ $CPU1_TEMPERATURE -gt $CPU_TEMPERATURE_THRESHOLD ]; }
if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
then
function CPU2_OVERHEAT () { [ $CPU2_TEMPERATURE -gt $CPU_TEMPERATURE_THRESHOLD ]; }
function CPU1_OVERHEAT() { [ $CPU1_TEMPERATURE -gt $CPU_TEMPERATURE_THRESHOLD ]; }
if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT; then
function CPU2_OVERHEAT() { [ $CPU2_TEMPERATURE -gt $CPU_TEMPERATURE_THRESHOLD ]; }
fi

# 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_OVERHEAT; then
apply_Dell_fan_control_profile

if ! $IS_DELL_FAN_CONTROL_PROFILE_APPLIED
then
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_OVERHEAT; 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_OVERHEAT; then
apply_Dell_fan_control_profile

if ! $IS_DELL_FAN_CONTROL_PROFILE_APPLIED
then
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
else
apply_user_fan_control_profile

# Check if user fan control profile is applied then apply it if not
if $IS_DELL_FAN_CONTROL_PROFILE_APPLIED
then
if $IS_DELL_FAN_CONTROL_PROFILE_APPLIED; then
IS_DELL_FAN_CONTROL_PROFILE_APPLIED=false
COMMENT="CPU temperature decreased and is now OK (<= $CPU_TEMPERATURE_THRESHOLD°C), user's fan control profile applied."
fi
fi

# If server model is Gen 14 (*40) or newer
if ! $DELL_POWEREDGE_GEN_14_OR_NEWER
then
if ! $DELL_POWEREDGE_GEN_14_OR_NEWER; then
# Enable or disable, depending on the user's choice, third-party PCIe card Dell default cooling response
# No comment will be displayed on the change of this parameter since it is not related to the temperature of any device (CPU, GPU, etc...) but only to the settings made by the user when launching this Docker container
if $DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE
then
if $DISABLE_THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE; then
disable_third_party_PCIe_card_Dell_default_cooling_response
THIRD_PARTY_PCIE_CARD_DELL_DEFAULT_COOLING_RESPONSE_STATUS="Disabled"
else
Expand All @@ -166,8 +150,7 @@ while true; do
fi

# Print temperatures, active fan control profile and comment if any change happened during last time interval
if [ $i -eq $TABLE_HEADER_PRINT_INTERVAL ]
then
if [ $i -eq $TABLE_HEADER_PRINT_INTERVAL ]; then
echo " ------- Temperatures -------"
echo " Date & time Inlet CPU 1 CPU 2 Exhaust Active fan speed profile Third-party PCIe card Dell default cooling response Comment"
i=0
Expand Down
28 changes: 12 additions & 16 deletions functions.sh
Original file line number Diff line number Diff line change
@@ -1,13 +1,13 @@
# Define global functions
# This function applies Dell's default dynamic fan control profile
function apply_Dell_fan_control_profile () {
function apply_Dell_fan_control_profile() {
# Use ipmitool to send the raw command to set fan control to Dell default
ipmitool -I $IDRAC_LOGIN_STRING raw 0x30 0x30 0x01 0x01 > /dev/null
CURRENT_FAN_CONTROL_PROFILE="Dell default dynamic fan control profile"
}

# This function applies a user-specified static fan control profile
function apply_user_fan_control_profile () {
function apply_user_fan_control_profile() {
# Use ipmitool to send the raw command to set fan control to user-specified value
ipmitool -I $IDRAC_LOGIN_STRING raw 0x30 0x30 0x01 0x00 > /dev/null
ipmitool -I $IDRAC_LOGIN_STRING raw 0x30 0x30 0x02 0xff $HEXADECIMAL_FAN_SPEED > /dev/null
Expand All @@ -17,17 +17,16 @@ function apply_user_fan_control_profile () {
# Convert first parameter given ($DECIMAL_NUMBER) to hexadecimal
# Usage : convert_decimal_value_to_hexadecimal $DECIMAL_NUMBER
# Returns : hexadecimal value of DECIMAL_NUMBER
function convert_decimal_value_to_hexadecimal () {
function convert_decimal_value_to_hexadecimal() {
local DECIMAL_NUMBER=$1
local HEXADECIMAL_NUMBER=$(printf '0x%02x' $DECIMAL_NUMBER)
echo $HEXADECIMAL_NUMBER
}

# Retrieve temperature sensors data using ipmitool
# Usage : retrieve_temperatures $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
function retrieve_temperatures () {
if (( $# != 2 ))
then
function retrieve_temperatures() {
if (( $# != 2 )); then
printf "Illegal number of parameters.\nUsage: retrieve_temperatures \$IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT \$IS_CPU2_TEMPERATURE_SENSOR_PRESENT" >&2
return 1
fi
Expand All @@ -39,8 +38,7 @@ function retrieve_temperatures () {
# Parse CPU data
local CPU_DATA=$(echo "$DATA" | grep "3\." | grep -Po '\d{2}')
CPU1_TEMPERATURE=$(echo $CPU_DATA | awk "{print \$$CPU1_TEMPERATURE_INDEX;}")
if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT
then
if $IS_CPU2_TEMPERATURE_SENSOR_PRESENT; then
CPU2_TEMPERATURE=$(echo $CPU_DATA | awk "{print \$$CPU2_TEMPERATURE_INDEX;}")
else
CPU2_TEMPERATURE="-"
Expand All @@ -50,22 +48,21 @@ function retrieve_temperatures () {
INLET_TEMPERATURE=$(echo "$DATA" | grep Inlet | grep -Po '\d{2}' | tail -1)

# If exhaust temperature sensor is present, parse its temperature data
if $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT
then
if $IS_EXHAUST_TEMPERATURE_SENSOR_PRESENT; then
EXHAUST_TEMPERATURE=$(echo "$DATA" | grep Exhaust | grep -Po '\d{2}' | tail -1)
else
EXHAUST_TEMPERATURE="-"
fi
}

# /!\ Use this function only for Gen 13 and older generation servers /!\
function enable_third_party_PCIe_card_Dell_default_cooling_response () {
function enable_third_party_PCIe_card_Dell_default_cooling_response() {
# We could check the current cooling response before applying but it's not very useful so let's skip the test and apply directly
ipmitool -I $IDRAC_LOGIN_STRING raw 0x30 0xce 0x00 0x16 0x05 0x00 0x00 0x00 0x05 0x00 0x00 0x00 0x00 > /dev/null
}

# /!\ Use this function only for Gen 13 and older generation servers /!\
function disable_third_party_PCIe_card_Dell_default_cooling_response () {
function disable_third_party_PCIe_card_Dell_default_cooling_response() {
# We could check the current cooling response before applying but it's not very useful so let's skip the test and apply directly
ipmitool -I $IDRAC_LOGIN_STRING raw 0x30 0xce 0x00 0x16 0x05 0x00 0x00 0x00 0x05 0x00 0x01 0x00 0x00 > /dev/null
}
Expand All @@ -88,12 +85,11 @@ function disable_third_party_PCIe_card_Dell_default_cooling_response () {
# }

# Prepare traps in case of container exit
function graceful_exit () {
function graceful_exit() {
apply_Dell_fan_control_profile

# Reset third-party PCIe card cooling response to Dell default depending on the user's choice at startup
if ! $KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT
then
if ! $KEEP_THIRD_PARTY_PCIE_CARD_COOLING_RESPONSE_STATE_ON_EXIT; then
enable_third_party_PCIe_card_Dell_default_cooling_response
fi

Expand All @@ -102,7 +98,7 @@ function graceful_exit () {
}

# Helps debugging when people are posting their output
function get_Dell_server_model () {
function get_Dell_server_model() {
IPMI_FRU_content=$(ipmitool -I $IDRAC_LOGIN_STRING fru 2>/dev/null) # FRU stands for "Field Replaceable Unit"

SERVER_MANUFACTURER=$(echo "$IPMI_FRU_content" | grep "Product Manufacturer" | awk -F ': ' '{print $2}')
Expand Down
3 changes: 1 addition & 2 deletions healthcheck.sh
Original file line number Diff line number Diff line change
Expand Up @@ -7,8 +7,7 @@
source functions.sh

# Check if the iDRAC host is set to 'local' or not then set the IDRAC_LOGIN_STRING accordingly
if [[ $IDRAC_HOST == "local" ]]
then
if [[ $IDRAC_HOST == "local" ]]; then
# Check that the Docker host IPMI device (the iDRAC) has been exposed to the Docker container
if [ ! -e "/dev/ipmi0" ] && [ ! -e "/dev/ipmi/0" ] && [ ! -e "/dev/ipmidev/0" ]; then
echo "/!\ Could not open device at /dev/ipmi0 or /dev/ipmi/0 or /dev/ipmidev/0, check that you added the device to your Docker container or stop using local mode. Exiting." >&2
Expand Down

0 comments on commit db64028

Please sign in to comment.