Update GOS_misc.md

2025-01-01 21:39:04 +00:00
parent 405a517466
commit 3b24dc9cfc
1 changed files with 127 additions and 1 deletions
@@ -90,3 +90,129 @@ We recommend to do following tuning before deploying server to production:
 ## Other sources

 * https://github.com/leandromoreira/linux-network-performance-parameters
+
+
+# IPMI
+
+## IPMI setup
+
+- Verify if your server has IPMI.
+
+```
+modprobe ipmi_devintf
+modprobe ipmi_si
+```
+If second command fails, you might not have IPMI.
+Check if `ipmitool` is installed, if not, install it.
+Then verify if it works:
+```
+ipmitool sdr
+```
+
+### IPMI network setup
+
+First verify current settings:
+```
+ipmitool lan print
+```
+
+Assign IP to IPMI interface:
+```
+ipmitool lan set 1 ipsrc static 1.2.3.4/24
+ipmitool lan set 1 defgw ipaddr 1.2.3.1
+ipmitool lan set 1 access on
+```
+
+Set username and password. First you need to find channel with users.
+Try:
+```
+ipmitool user list
+ipmitool user list 1
+ipmitool user list 2
+etc to 8
+```
+When it will show you list of users, you can set username and password for this channel:
+
+```
+ipmitool user set name 1 admin
+ipmitool user set password 1 password
+```
+TODO: Some boards doesnt allow to change username, and some have limits on password length.
+
+
+
+# Monitoring more than basic metrics
+
+## Temperatures
+
+### IPMI
+
+If IPMI present, you can use ipmitool to get temperatures:
+```
+ipmitool sensors
+```
+
+If you don't have ipmitool please contact to obtain latest version of GlobalOS.
+
+### lm-sensors
+
+```
+modprobe i2c-dev
+modprobe coretemp
+modprobe k10temp
+modprobe pinctrl_lewisburg
+```
+
+In `/sys/class/hwmon/` you will find entries for your sensors, search for one with name "coretemp" or "k10temp". for example
+
+```
+grep "" /sys/class/hwmon/hwmon*/name
+/sys/class/hwmon/hwmon1/name:coretemp
+/sys/class/hwmon/hwmon2/name:pch_lewisburg
+
+grep "" /sys/class/hwmon/hwmon1/temp*
+/sys/class/hwmon/hwmon1/temp10_crit:98000
+/sys/class/hwmon/hwmon1/temp10_crit_alarm:0
+/sys/class/hwmon/hwmon1/temp10_input:30000
+/sys/class/hwmon/hwmon1/temp10_label:Core 8
+/sys/class/hwmon/hwmon1/temp10_max:88000
+/sys/class/hwmon/hwmon1/temp11_crit:98000
+....
+
+grep "" /sys/class/hwmon/hwmon2/temp*
+/sys/class/hwmon/hwmon2/device/temp:37000
+
+```
+
+`pch_lewisburg` is chipset temperature
+
+
+### Power consumption monitoring
+
+- Power meter (ACPI interface)
+```
+Check in /sys/class/hwmon/ entries where device/name contains "power" or "energy".
+```
+
+- Power meter (IPMI)
+```
+ipmitool sensor
+```
+
+### Disk temperature
+
+Note: Smartctl is available in latest GlobalOS versions.
+
+```
+smartctl -a /dev/sda | grep Temperature
+```
+
+## SFP/SFP+ modules
+
+```
+ethtool -m ethX
+```
+
+
+
+