Strony

środa, 12 marca 2008

asus p5b lm_sensors

Aquarius Server (my home file and vmware server) has waited reistallation with full raid configuration (many thanks to xoff for help and support :) ). Because it has many memory (6GB), some disks (about 1TB after raid) and rather good processor (C2D E6300) it made some noise. It has three 12cm coolers which blow through the case... and a cpu cooler made by Pentagram, but it make "any" sound, which is rather insufferable. I tried to remove the cpu cooler, and leave only the radiator (the case cooler is really near).
Asus P5B

All in all - the server is made to endure the big load (vmware with some testings enviroments based on windows 2k2, linux, and some java applications is enought to kill every processor, and make the system to swap :) ) so the cputemp without cooler should be monitored :)

The lm_sensors is great and enouht tool to monitor the system sensors. So this is my configuration:

/etc/sensors.conf
chip "w83627dhg-*"
# == VOLTAGES ==
# Vcore
label in0 "Vcore"
set in0_min 0.85
set in0_max 1.6

# +12V
label in1 "+12V"
compute in1 @*(66/10), @/(66/10)
set in1_min 10.2
set in1_max 13.8

# unused
ignore in2

# +3.3V
label in3 "+3.3V"
# this formula is stored in the ACPI, but it seems not to work
# compute in3 @*(68/34), @/(68/34)
set in3_min 2.97
set in3_max 3.63

# unused
ignore in4

# +5V
label in5 "5V"
compute in5 @*(32/10), @/(32/10)
set in5_min 4.5
set in5_max 5.5

# unused
ignore in6
ignore in7
ignore in8
ignore in9

# == TEMPERATURES ==
# Mainboard
label temp1 "Sys Temp"
set temp1_over 95
set temp1_hyst 60

# CPU
label temp2 "CPU Temp"
set temp2_over 95
set temp2_hyst 45

# unused
ignore temp3

# == FANS ==
# If you miss a certain fan set it to ignore
# Chassis 1
# ignore fan1
label fan1 "Cha1 Fan"
set fan1_min 800

# CPU
# ignore fan2
label fan2 "CPU Fan"
set fan2_min 600

# Powersupply
ignore fan3
#label fan3 "Pwr Fan"
#set fan3_min 800

# unused
ignore fan4

# Chassis 2
# ignore fan5
label fan5 "Cha2 Fan"
set fan5_min 800

# == CORE TEMPERATURES ==
# This requires the coretemp patch applied
# Maximums are autodetected by the driver depending on the CPU
# Core0
chip "coretemp-isa-0000"
label temp1 "Core0 Temp"

# Core1
chip "coretemp-isa-0001"
label temp1 "Core1 Temp"

You should also load two kernel modules (I have: 2.6.22 from PLD):
/etc/modules
w83627ehf
coretemp


now, after executing `sensors` you should get sth like this

w83627dhg-isa-0290
Adapter: ISA adapter
Vcore: +1.16 V (min = +0.00 V, max = +1.74 V)
+12V: +12.41 V (min = +1.37 V, max = +0.00 V) ALARM
+3.3V: +3.30 V (min = +2.46 V, max = +0.38 V) ALARM
5V: +5.07 V (min = +0.00 V, max = +0.20 V) ALARM
Cha1 Fan: 0 RPM (min = 1318 RPM, div = 128) ALARM
CPU Fan: 1268 RPM (min = 168750 RPM, div = 8) ALARM
Cha2 Fan: 0 RPM (min = 42187 RPM, div = 32) ALARM
Sys Temp: +34°C (high = +113°C, hyst = +0°C)
CPU Temp: +29.0°C (high = +80.0°C, hyst = +75.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Core0 Temp:
+40°C (high = +85°C)

coretemp-isa-0001
Adapter: ISA adapter
Core1 Temp:
+40°C (high = +85°C)


So i do not have to worry... this is rather good temp ;)

2 komentarze:

Bertl pisze...

pal, you got the w83627dhg system and cpu temp mixed up, your board has +29°C (measured inside the chip) and +34°C directly below the cpu (remote sensor)

Marcin Rybak pisze...

thanks for help :) U r right!