Monitoring Capabilities

In this chapter

This chapter describes the monitoring capabilities of the SmartNICs. The parameters described in the chapter can be monitored from the application through the API or by using the monitoring tool (see DN-0449).

Monitoring parameters can be divided into optical module parameters and SmartNIC parameters. They consist of both sensor parameters and state parameters.

Note: The sensor parameter values are updated every 5 seconds. This means that the values received at a user application can be up to 5 seconds old, and therefore might not reflect the current situation.

Optical module parameters

This table shows the sensor parameters available for monitoring for each optical module type. The mandatory parameters are available in all the relevant optical modules, whereas the optional parameters are module-dependent.

Note: QSFP28 modules, CFP4 modules and QSFP+ modules (except for bidirectional modules) have four sensors for some of the parameters, one for each laser. These sensors measure TX bias current, TX power and RX power. Therefore, four values are available for each of these parameters.

Note: Bidirectional QSFP+ modules are not required to have sensors.

Module	SmartNICs	Mandatory Parameters	Optional Parameters
QSFP28	NT200A02-2×100/40, NT200A01-2×100/40, NT200A01-2×100	QSFP28 temp.	Supply voltage TX bias current (x4) TX power (x4) RX power (x4)
CFP4	NT200C01-2×100, NT100E3-1-PTP	CFP4 temp.	Supply voltage TX bias current (x4) TX power (x4) RX power (x4)
QSFP+	NT200A02-2×100/40, NT200A02-2×40/8×10, NT200A02-8×10/2×40, NT200A01-2×100/40, NT200A01-2×40, NT200A01-2×25/10/2×40, NT200A01-8×10/2×40, NT80E3-2-PTP, NT80E3-2-PTP-8×10/2×40, Intel® PAC A10 GX	QSFP+ temp.	Supply voltage TX bias current (x4) TX power (x4) RX power (x4)
SFP28	NT200A01-2×25/10/2×40 (with a pluggable module adapter (PMA – see DN-0515) enabling SFP28 and SFP+ modules to be plugged into a QSFP28 port)	SFP28 temp. Supply voltage TX bias current TX power RX power	Laser temp. TEC current
SFP+	NT200A02-2×40/8×10, NT200A02-8×10/2×40, NT200A01-2×25/10/2×40 (with a pluggable module adapter (PMA – see DN-0515) enabling SFP28 and SFP+ modules to be plugged into a QSFP28 port), NT200A01-8×10/2×40, NT80E3-2-PTP-8×10/2×40, NT40A01-4×10/1-SLB, NT40E3-4-PTP, NT20E3-2-PTP		SFP+ temp. Supply voltage TX bias current TX power RX power
SFP	NT40A01-4×10/1-SLB, NT40A01-4×1, NT40E3-4-PTP, NT20E3-2-PTP		SFP temp. Supply voltage TX bias current TX power RX power

For each optical module parameter the current value as well as the lowest and highest registered values can be retrieved. For some parameters also high and low alarm levels can be shown as well as the alarm state.

Note: One QSFP28/CFP4/QSFP+/SFP28/SFP+/SFP temperature level is significant for the NT SmartNICs: At 75 °C a message is created in the driver log, and a high-temperature event is written to the hardware log. The event can be monitored from the application through the API.

Monitoring parameters for NT SmartNICs

This table shows the most important SmartNIC parameters of which the NT SmartNICs provide monitoring.

Parameter	Mode of Operation
Link status	Ethernet link status per port
PCB temperature	PCB temperature level and alarm
FPGA temperature	FPGA temperature level and alarm Automatic over-temperature shutdown Note: Three FPGA temperature levels are significant: 90 °C – a message is created in the driver log, and a high-temperature event is written to the hardware log. The event can be monitored from the application through the API. 95 °C – the driver triggers the FPGA to shut down. A message is created in the driver log, the event is written to the hardware log, and the system LED flashes red. The SmartNIC will remain in this state until the server is powered off then on again (a hard reset). Any applications that were running must be restarted. The event can be monitored from the application through the API when the server is powered on again. 125 °C – for safety an automatic fail-safe FPGA shutdown is conducted by the FPGA (built in). The event is written to the hardware log, and can be monitored from the application through the API. Note: When the LogToSystem parameter in the ntservice.ini file (see DN-0449) is set to `1`, the warning messages are also logged in the system log.
Time synchronization status	Status on SmartNIC time synchronization and loss of synchronization
Fan speed	Speed of the cooling fan Note: An out-of-range fan speed event is written to the hardware log if the fan rotates at a speed outside a specific range. The event can be monitored from the application through the API. For NT200A02-SCC and NT200A01-SCC the low limit is 5264 RPM and the high limit is 7450 RPM. For NT200C01-SCC, NT100E3-1-PTP, NT80E3-2-PTP and NT80E3-2-PTP-8×10/2×40 the low limit is 4738 RPM and the high limit is 6566 RPM. For NT40A01-4×10/1-SLB, NT40A01-4×1, NT40E3-4-PTP and NT20E3-2-PTP the low limit is 4000 RPM and the high limit is 6000 RPM.
Critical temperature	Temperature of critical components
Power	Total power consumption

Monitoring parameters for Intel® PAC with Intel® Arria® 10 GX FPGA

The Intel® PAC with Intel® Arria® 10 GX FPGA provides monitoring of FPGA temperature level with alarm and automatic shutdown as shown in this table.

Parameter	Mode of Operation
FPGA temperature	FPGA temperature level and alarm Automatic over-temperature shutdown Note: Two FPGA temperature levels are significant: 90 °C – a message is created in the driver log. The event can be monitored from the application through the API. 95 °C – the driver triggers the FPGA to shut down. A message is created in the driver log, and the system LED flashes red. Note: When the LogToSystem parameter in the ntservice.ini file (see DN-0449) is set to `1`, the warning messages are also logged in the system log.

Parameter

Mode of Operation

FPGA temperature

FPGA temperature level and alarm

Automatic over-temperature shutdown

Note: Two FPGA temperature levels are significant:

90 °C – a message is created in the driver log. The event can be monitored from the application through the API.
95 °C – the driver triggers the FPGA to shut down. A message is created in the driver log, and the system LED flashes red.

Note: When the LogToSystem parameter in the ntservice.ini file (see DN-0449) is set to 1, the warning messages are also logged in the system log.