Field Replaceable Units and Sensor Data Records

Field replaceable units (FRUs) are major modules in the chassis that contain active electronic circuitry. FRUs can store information-such as board serial number, part number, name, and asset tag-that can be read using the System Setup Utility (page 49). The BMC stores FRU information for the baseboard in a nonvolatile storage component on the board.

The BMC uses Sensor Data Records (SDRs) to identify the sensors in the system for monitoring. SDRs provide a list of the sensors, their characteristics, location, type, and type-specific information, such as default threshold values, factors for converting a sensor reading into the appropriate units (mV, rpm, degrees Celsius), and information on the types of events that a sensor can generate. The BMC stores SDR information in a nonvolatile storage component on the baseboard.

You can use the FRU/SDR Load Utility (page 57) to initialize or update the FRU and SDR information. Intel server boards are shipped from the factory with some sensors disabled, because the actual configuration of the chassis is only determined when the user completes the system configuration. Chassis-specific information in the FRU, such as chassis part number, is also absent. For these reasons, it is important to run the FRU/SDR Load Utility as part of the system setup process. You should also run the FRU/SDR Load Utility whenever you change the number of fans, processors, or power supplies in the server.

System Event Log

The BMC manages a system event log (SEL), where it records significant or critical system events. Such events include temperatures and voltages out of range, fan failures, and other sensor-related events. The BIOS, software, and other devices can also log events by sending messages to the BMC. The SEL is stored in nonvolatile storage.

You can view the current contents of the SEL by using the System Setup Utility (page 48).

Platform Event Management

Events can trigger alerts and other actions by the BMC. The server is configured with the following set of standard events:

•Temperature sensor out of range

•Voltage sensor out of range

•Fan failure

•Chassis intrusion

•Power supply fault

•BIOS uncorrectable ECC error

•BIOS POST error

•Processor fault resilient booting (FRB) failure

•Fatal nonmaskable interrupt (NMI) from a source other than the front panel switch

•Watchdog timer reset, power down, or power cycle

•System restart (reboot)

Description