Penguin Computing Relion 1900e, Technical Manual

Page: 85 / 174

Relion 1900e/2900e Manual
Revision 1.0 70
ED1 – 0xA1
ED2 - CATERR type.
0: Unknown
1: CATERR
2: CPU Core Error (not supported on Intel
®
Server Systems supporting the Intel
®
Xeon
®
processor E5-2600
v3, v4product family)
3: MSID Mismatch
4: CATERR due to CPU 3-strike timeout
ED3 - CPU bitmap that causes the system CATERR.
[0]: CPU1
[1]: CPU2
[2]: CPU3
[3]: CPU4
When a CATERR Timeout event is determined to be a CPU 3-strike timeout, The BMC shall log the logical
FRU information (e.g. bus/dev/func for a PCIe* device, CPU, or DIMM) that identifies the FRU that caused the
error in the extended SEL data bytes. In this case, Ext-ED0 will be set to 0x70 and the remaining ED1-ED7 will
be set according to the device type and info available.
7.3.11.5 MSID Mismatch Sensor
The BMC supports a MSID Mismatch sensor for monitoring for the fault condition that will occur if there is a
power rating incompatibility between a baseboard and a processor. The sensor is rearmed on power-on (AC
or DC power on transitions).
7.3.12 Voltage Monitoring
The BMC provides voltage monitoring capability for voltage sources on the main board and processors such
that all major areas of the system are covered. This monitoring capability is instantiated in the form of IPMI
analog/threshold sensors.
7.3.12.1 Discrete Voltage Sensors
The discrete voltage sensor monitors multiple voltages from sensors around the baseboard and then asserts
a bit in the SEL event data for each sensor that is out of range. The sensor name for the asserted bit can be
retrieved via the Get Voltage Name IPMI function.
7.3.13 Fan Monitoring
BMC fan monitoring support includes monitoring of fan speed (RPM) and fan presence.
7.3.13.1 Fan Tach Sensors
Fan Tach sensors are used for fan failure detection. The reported sensor reading is proportional to the fan’s
RPM. This monitoring capability is instantiated in the form of IPMI analog/threshold sensors.
Most fan implementations provide for a variable speed fan, so the variations in fan speed can be large.
Therefore the threshold values must be set sufficiently low as to not result in inappropriate threshold
crossings.
Fan tach sensors are implemented as manual re-arm sensors because a lower-critical threshold crossing can
result in full boosting of the fans. This in turn may cause a failing fan’s speed to rise above the threshold and
can result in fan oscillations.