HP-Health monitoring with Netdata

Monitors a local HPE ProLiant server using hp-health statistics a replacement for 'lm-sensors'.

Requirements

• A local installation of the "ProLiant Management Command Line Interface Utility (hpasmcli)" This is usually not part of the standard repos, get it directly from HPE.
• password-less sudo access to the used "show" commands (if not running netdata as root)

To allow sudo without password for that specific read-only command execute sudo visudo and add the line:

netdata ALL=(ALL:ALL) NOPASSWD:/usr/sbin/hpasmcli -s show temp;show fans;show powersupply

The plugin will check this configuration at startup and recommend a propper setting. To avoid cluttering your auth.log with netdata sudo messages, you can exclude these with:

Defaults:netdata !syslog, !pam_session

Charts

This module will produce following charts (if data is available):

• Temperature charts for each detected temperature sensor separated by sensor location
• Fan charts for each detected fan speed sensor separated by sensor location
• A fan partner overview listing the amount of fans per partner group (redundant fans in the same partner group can take over in the case of a fan failure)
• A power consumption chart per power supply
• A power supply chart counting the bays in ok condition

Alerts

This module will warn if:

• any temperature sensor comes close to 5°C of its predifened threshold value
• any fan partner group looses a fan
• a power supply changes to an unhealthy condition