Monitoring Interfaces and Transceivers with NVUE
NVUE enables you to check the status of an interface, and view and clear interface counters. Interface counters provide information about an interface, such as the number of packets dropped, the number of inbound and outbound packets not transmitted because of errors, and so on.
Show Interface Configuration and Statistics
To check the configuration and statistics for an interface, run the nv show interface <interface-id> command:
cumulus@switch:~$ nv show interface swp1
operational applied pending
------------------------ ----------------- ------- -------
type swp swp
[acl]
evpn
multihoming
uplink off
ptp
enable off
router
adaptive-routing
enable off
ospf
enable on
area none
cost auto
mtu-ignore off
network-type broadcast
passive on
priority 1
authentication
enable off
bfd
enable off
timers
dead-interval 40
hello-interval 10
retransmit-interval 5
transmit-delay 1
ospf6
enable off
pbr
[map]
pim
enable off
synce
enable off
ip
igmp
enable off
ipv4
forward on
ipv6
enable on
forward on
neighbor-discovery
enable on
[dnssl]
home-agent
enable off
[prefix]
[rdnss]
router-advertisement
enable on
fast-retransmit on
hop-limit 64
interval 600000
interval-option off
lifetime 1800
managed-config off
other-config off
reachable-time 0
retransmit-time 0
router-preference medium
vrrp
enable off
vrf default
[gateway]
link
auto-negotiate disabled enabled
duplex full full
speed 1G auto
fec auto
mtu 9216 9216
[breakout]
physical-name swp1c1c2
state up up
stats
carrier-transitions 4
in-bytes 300 Bytes
in-drops 5
in-errors 0
in-pkts 5
out-bytes 9.73 MB
out-drops 0
out-errors 0
out-pkts 140188
mac 48:b0:2d:ef:52:b8
ifindex 3
Show Interface Counters
NVUE provides the following commands to show counters (statistics) for the interfaces on the switch.
NVUE Command |
Description |
|---|---|
nv show interface --view counters |
Shows all statistics for all the interfaces configured on the switch, such as the total number of received and transmitted packets, and the number of received and transmitted dropped packets and error packets. |
nv show interface <interface-id> counters |
Shows all statistics for a specific interface, such as the number of received and transmitted unicast, multicast and broadcast packets, the number of received and transmitted dropped packets and error packets, and the number of received and transmitted packets of a certain size. |
nv show interface <interface-id> counters errors |
Shows the number of error packets for a specific interface, such as the number of received and transmitted packet alignment, oversize, undersize, and jabber errors. |
nv show interface <interface-id> counters drops |
Shows the number of received and transmitted packet drops for a specific interface, such as ACL drops, buffer drops, queue drops, and non-queue drops. |
nv show interface <interface-id> counters pktdist |
Shows the number of received and transmitted packets of a certain size for a specific interface. |
nv show interface <interface-id> counters qos |
Shows QoS statistics for the specified interface. See Show Qos Counters. |
nv show interface <interface-id> counters ptp |
Shows PTP statistics for a specific interface. See Show PTP Counters. |
The following example shows all statistics for all the interfaces configured on the switch:
cumulus@switch$ nv show interface --view counters
Interface MTU RX_OK RX_ERR RX_DRP RX_OVR TX_OK TX_ERR TX_DRP TX_OVR Flg
-------------- ----- ----- ------ ------ ------ ----- ------ ------ ------ -----
BLUE 65575 1 0 0 0 0 0 4 0 OmRU
RED 65575 1 0 0 0 0 0 4 0 OmRU
bond1 9000 718 0 0 0 1091 0 0 0 BMmRU
bond2 9000 727 0 0 0 1088 0 0 0 BMmRU
bond3 9000 722 0 0 0 1089 0 0 0 BMmRU
br_default 9216 360 0 10 0 475 0 0 0 BMRU
eth0 1500 946 0 0 0 299 0 0 0 BMRU
lo 65536 651 0 0 0 651 0 0 0 LRU
mgmt 65575 283 0 0 0 0 0 4 0 OmRU
peerlink 9216 4972 0 0 0 5028 0 0 0 BMmRU
peerlink.4094 9216 3263 0 0 0 3224 0 0 0 BMRU
swp1 9000 721 0 0 0 1091 0 0 0 BMsRU
swp2 9000 730 0 0 0 1088 0 0 0 BMsRU
swp3 9000 725 0 0 0 1089 0 0 0 BMsRU
swp49 9216 2807 0 0 0 2691 0 0 0 BMsRU
swp50 9216 2165 0 0 0 2337 0 0 0 BMsRU
swp51 9216 685 0 0 0 690 0 0 0 BMRU
swp52 9216 703 0 0 0 722 0 0 0 BMRU
swp53 9216 738 0 0 0 710 0 0 0 BMRU
swp54 9216 682 0 0 0 730 0 0 0 BMRU
vlan10 9216 108 0 20 0 91 0 0 0 BMRU
vlan10-v0 9216 63 0 20 0 45 0 0 0 BMRU
vlan20 9216 104 0 20 0 88 0 0 0 BMRU
vlan20-v0 9216 58 0 20 0 44 0 0 0 BMRU
vlan30 9216 112 0 20 0 94 0 0 0 BMRU
vlan30-v0 9216 61 0 20 0 44 0 0 0 BMRU
vlan4024_l3 9216 1 0 0 0 82 0 0 0 BMRU
vlan4024_l3-v0 9216 0 0 0 0 36 0 0 0 BMRU
vlan4036_l3 9216 1 0 0 0 85 0 0 0 BMRU
vlan4036_l3-v0 9216 0 0 0 0 37 0 0 0 BMRU
vxlan48 9216 45 0 0 0 21 0 0 0 BMRU
The following example shows all statistics for swp1:
cumulus@switch$ nv show interface swp1 counters
operational applied
------------------- ----------- -------
carrier-transitions 4
Detailed Counters
====================
Counter Receive Transmit
----------------- ------- --------
Broadcast Packets 0 0
Multicast Packets 0 0
Total Octets 0 0
Total Packets 0 0
Unicast Packets 0 0
Drop Counters
================
Counter Receive Transmit
--------------- ------- --------
ACL Drops 0 n/a
Buffer Drops 0 n/a
Non-Queue Drops n/a 0
Queue Drops n/a 0
Total Drops 0 0
Error Counters
=================
Counter Receive Transmit
---------------- ------- --------
Alignment Errors 0 n/a
FCS Errors 0 n/a
Jabber Errors 0 n/a
Length Errors 0 n/a
Oversize Errors 0 n/a
Symbol Errors 0 n/a
Total Errors 0 0
Undersize Errors 0 n/a
Packet Size Statistics
=========================
Counter Receive Transmit
---------- ------- --------
64 0 0
65-127 0 0
128-255 0 0
256-511 0 0
512-1023 0 0
1024-1518 0 0
1519-2047 0 0
2048-4095 0 0
4096-16383 0 0
Ingress Buffer Statistics
============================
priority-group rx-frames rx-buffer-discards rx-shared-buffer-discards
-------------- --------- ------------------ -------------------------
0 0 0 Bytes 0 Bytes
1 0 0 Bytes 0 Bytes
2 0 0 Bytes 0 Bytes
3 0 0 Bytes 0 Bytes
4 0 0 Bytes 0 Bytes
5 0 0 Bytes 0 Bytes
6 0 0 Bytes 0 Bytes
7 0 0 Bytes 0 Bytes
...
The following example shows error counters for swp1:
cumulus@switch$ nv show interface swp1 counters errors
Counter Receive Transmit
---------------- ------- --------
Alignment Errors 0 n/a
FCS Errors 0 n/a
Jabber Errors 0 n/a
Length Errors 0 n/a
Oversize Errors 0 n/a
Symbol Errors 0 n/a
Total Errors 0 0
Undersize Errors 0 n/a
- NVUE does not show detailed statistics for logical interfaces, such as bonds, VLAN interfaces or subinterfaces. To see basic statistics for logical interfaces, run the
nv show interface <interface-id> link statscommand. - On NVIDIA Spectrum switches, Cumulus Linux updates physical counters to the kernel every two seconds and virtual interfaces (such as VLAN interfaces) every ten seconds. You cannot change these values. Because the update process takes a lower priority than other
switchdprocesses, the interval might be longer when the system is under a heavy load.
Interface Fault Detection
You can view interface status along with any detected local or remote fault with the nv show interface status command:
cumulus@switch:mgmt:~$ nv show interface status
Interface Admin Status Oper Status Protodown Protodown Reason Fault
--------- ------------ ----------- --------- ---------------- --------
...
swp45 up down disabled Local Fault
swp46 up down disabled Remote Fault
swp47 down down disabled No Fault
swp48 down down disabled No Fault
...
To view additional fault information for a specific interface, run the nv show interface <interface> link phy detail command:
cumulus@switch:mgmt:~$ nv show interface swp45 link phy detail
operational
-------------------------------- -------------------
...
linkdown-reason-code-local 23
linkdown-reason-status-local CABLE_WAS_UNPLUGGED
linkdown-reason-code-remote 1
linkdown-reason-status-remote UNKNOWN_REASON
...
- Interface fault detection is supported on NVIDIA Spectrum-4 and later platforms.
- Fault status reported by the
nv show interface statuscommand is only supported on physical switch ports and breakout interfaces. Fault state is not applicable to logical interfaces such as VLAN sub-interfaces, SVIs, loopback interfaces, VRF interfaces, and bonds. - Fault detection data is not retained across reboots or when the
switchdservice is restarted.
AmBER PHY Health Management
To show physical layer information, such as the error counters for each lane on a port and SNR information for media and host lanes (lane0 through lane7), run the nv show interface <interface-id> link phy health command.
This command highlights link integrity issues.
The nv show interface <interface-id> link phy health command shows SNR information for media and host lanes for OSFP 100G per lane capable transceivers on certain switches with the Spectrum-4 ASIC and later. The command does not show SNR information for SFP28 bonus ports.
- For lanes not in use for the optical cable, the command output shows 0.
- For copper ports, the command output shows n/a.
The effective-ber in the command output represents the uncorrectable bit error rate, which is the same as uncorrected FEC errors.
cumulus@switch$ nv show interface swp1 link phy health
operational
------------------------- -------------
time-since-last-clear-min 1
symbol-errors 0
symbol-ber 15E-255
raw-ber 15E-255
phy-received-bits 6214400000000
effective-errors 0
effective-ber 15E-255
Lane Stats
=============
Lane raw-ber snr-host snr-media phy-raw-errors
---- ------- --------- --------- --------------
0 15E-255 22.32 dB 23.14 dB 0
1 15E-255 25.24 dB 22.47 dB 0
2 15E-255 21.18 dB 22.30 dB 0
3 15E-255 25.12 dB 25.01 dB 0
4 15E-255 24.50 dB 24.62 dB 0
5 15E-255 0 dB 0 dB 0
6 15E-255 25 dB 24.62 dB 0
7 15E-255 22.18 dB 23.48 dB 0
To show a tabular summary of PHY health statistics for all switch port interfaces in a single command, such as the error counters for all ports, run the nv show interface --view ber command:
cumulus@switch$ nv show interface --view ber
Interface Eff-BER Raw-BER Sym-BER Sym-Errs Eff-Errs Last-Clr(min)
--------- ------- ------- ------- -------- -------- -------------
swp1 15E-255 15E-255 15E-255 0 0 1
swp2 15E-255 15E-255 15E-255 0 0 1
swp3 15E-255 15E-255 15E-255 0 0 1
swp49 15E-255 15E-255 15E-255 0 0 1
swp50 15E-255 15E-255 15E-255 0 0 1
swp51 15E-255 15E-255 15E-255 0 0 1
swp52 15E-255 15E-255 15E-255 0 0 1
swp53 15E-255 15E-255 15E-255 0 0 1
swp54 15E-255 15E-255 15E-255 0 0 1
- To filter the view for a range of interfaces, use the
--filteroption; for example, thenv show interface --view ber --filter "*>=swp49&*<=swp54"command shows PHY health statistics for swp49 through swp54. - To show the information in json format, use the
-o jsonoption; for example, thenv show interface --view ber --filter "*>=swp49&*<=swp54" -o jsoncommand shows PHY health statistics for swp49 through swp54 in json format.
To show physical layer diagnostic information for a port, run the nv show interface <interface-id> link phy detail command:
cumulus@switch$ nv show interface swp2 link phy detail
operational
-------------------------------- -----------
loopback-mode 0
fec-mode-request 4
module-info-ext 0
ethernet-compliance-code 128
ext-ethernet-compliance-code 24
memory-map-rev 7
module-fw-version 538837258
module-st 0
error-code-response 0
cable-proto-cap-ext 0
cable-rx-amp 0
cable-rx-pre-emphasis 0
cable-rx-post-emphasis 0
cable-tx-equalization 0
cable-attenuation-53g 0
cable-attenuation-25g 0
cable-attenuation-12g 0
cable-attenuation-7g 0
cable-attenuation-5g 0
pd-fsm-state 8
eth-an-fsm-state 6
phy-hst-fsm-state 0
psi-fsm-state 0
retran-mode-request 0
retran-mode-active 0
profile-fec-in-use 4
pd-link-enabled 16744453
phy-hst-link-enabled 0
eth-an-link-enabled 97452034
phy-manager-state 3
eth-proto-admin 0
ext-eth-proto-admin 0
eth-proto-capability 0
ext-eth-proto-capability 0
data-rate-oper 0
an-status 0
an-disable-admin 0
proto-mask 2
linear-direct-drive 0
cable-breakout 16
rx-output-valid 0
rx-power-type 1
active-set-host-compliance-code 0
active-set-media-compliance-code 0
mod-fw-fault 0
dp-fw-fault 0
rx-los-cap 0
tx-ad-eq-fault 0
rx-output-valid-change 0
tx-input-freq-sync 0
tx-cdr-state 15
rx-cdr-state 15
phy-manager-link-enabled 6130
core-to-phy-link-enabled 754
zero-hist 0
linkdown-reason-code-local -1
linkdown-reason-status-local N/A
linkdown-reason-code-remote -1
linkdown-reason-status-remote N/A
Datapath State
=================
Lane DP St Lanes
---- -----------
0 0
1 0
2 0
3 0
4 0
5 0
6 0
7 0
Switches with the Spectrum 1 ASIC do not support the nv show interface <interface-id> link phy health command or the nv show interface <interface-id> link phy detail command.
Clear Interface Counters
To clear counters (statistics) for all interfaces, run the nv action clear interface counters command.
cumulus@switch$ nv action clear interface counters
all interface counters cleared
Action succeeded
To clear the counters for an interface, run the nv action clear interface <interface-id> counters command:
cumulus@switch$ nv action clear interface swp1 counters
swp1 counters cleared
Action succeeded
The nv action clear interface <interface-id> counters command does not clear counters in the hardware.
Clear Interface Physical Layer Error Counters
To clear interface physical layer error counters, run the nv action clear interface <interface-id> link phy health command. The command clears the counters at the software level, but not the hardware level.
cumulus@switch:~$ nv action clear interface swp1 link phy health
Action executing ...
swp1 link phy health counters cleared.
Action succeeded
To clear physical layer error counters for a range of interfaces:
cumulus@switch:~$ nv action clear interface swp1-3,swp5,swp7-10 link phy health
Action executing ...
swp1 link phy health counters cleared.
Action executing ...
swp2 link phy health counters cleared.
swp3 link phy health counters cleared.
Action executing ...
swp5 link phy health counters cleared.
swp7 link phy health counters cleared.
Action executing ...
swp8 link phy health counters cleared.
swp9 link phy health counters cleared.
swp10 link phy health counters cleared.
If the specified interface is out of range; for example, if the switch supports up to 32 switch ports but you try to clear swp33, NVUE displays an error.
The nv action clear interface <interface-id> link phy health command does not clear counters in the hardware.
Reset a Transceiver
NVUE provides a command to reset a specific transceiver to its initial, stable state without having to be present physically in the data center to pull the transceiver.
The following example resets the transceiver in swp1:
cumulus@switch:~$ nv action reset platform transceiver swp1
Action executing ...
Resetting module swp1 ... OK
Action succeeded
The following example resets a range of transceivers:
cumulus@switch:~$ nv action reset platform transceiver swp1-3
Action executing ...
Resetting module swp1 ... OK
Action executing ...
Resetting module swp2 ... OK
Action executing ...
Resetting module swp3 ... OK
Action succeeded
The following example resets the transceiver in swp5 and swp7:
cumulus@switch:~$ nv action reset platform transceiver swp5,swp7
Action executing ...
Resetting module swp5 ... OK
Action executing ...
Resetting module swp7 ... OK
Action executing ...
Action succeeded
When the reset completes successfully, you see syslog messages similar to the following:
2024-12-06T07:12:37.996339+00:00 cumulus nvue-port-reset: The module reset was successfully completed on swp1
- To reset a transceiver on a split port, specify the parent port; for example, to reset swp1s0, run the
nv action reset platform transceiver swp1command. - If a cable is faulty, the
nv action reset platform transceiver <transceiver-idcommand completes successfully, but the details of the transceiver do not show until you resolve the issue or reboot the system if necessary,
Manage Transceiver Firmware
You can show and install firmware for a transceiver module directly through NVUE.
To show the current firmware version for a transceiver, upgrade status, and the firmware files available, run the nv show platform transceiver <transceiver-id> firmware command:
cumulus@switch:~$ nv show platform transceiver swp1 firmware
operational
------------------- ---------------------
actual-firmware 38.100.121
fw-upgrade-status ok
[files] fw-rel-38_100_121.bin
[files] fw-rel-38_100_122.bin
To show only the available firmware files and their paths, run the nv show platform transceiver <transceiver-id> firmware files command. Add a file name to show a single file:
cumulus@switch:~$ nv show platform transceiver swp1 firmware files
Available Firmware Files File Path
------------------------- --------------------------------------------------------------
fw-rel-38_100_121.bin /etc/fae_platform_firmware/transceiver/fw-rel-38_100_121.bin
fw-rel-38_100_122.bin /tmp/firmware_downloads/transceiver/fw-rel-38_100_122.bin
To install a firmware file on a transceiver module, run the nv action install platform transceiver <transceiver-id> firmware files <file> command:
cumulus@switch:~$ nv action install platform transceiver swp1 firmware files fw-rel-38_100_121.bin
Action executing ...
Burning fw-rel-38_100_121.bin on swp1 ...
Firmware burned on swp1 (mt4129_pciconf0_cable_0). Reset the module or power cycle the system to activate it
Action succeeded
Installing firmware on a transceiver does not activate it immediately. Reset the transceiver, or power cycle the switch, to bring the new firmware into use.
If the specified file is not staged for that transceiver, the action fails:
cumulus@switch:~$ nv action install platform transceiver swp1 firmware files fw-rel-38_100_121.bin
Action executing ...
Action failed
Firmware file not found: fw-rel-38_100_121.bin
Transceiver Thermal Control
To optimize transceiver thermal performance and maintain a cooler operating environment before reaching critical temperatures, you can set temperature thresholds for fan activation in the FAN algorithm, ensuring that they are below the module’s high warning threshold.
Lowering the thresholds leads to earlier and more frequent fan operation, increasing power usage and noise but protecting hardware performance and lifespan.
Setting the temperature thresholds does not change the module EEPROM-based alarm thresholds. The optical module continues reporting temperature alarms based on alarms or warning thresholds preprogrammed in the transceiver EEPROM.
Setting the temperature thresholds without proper guidance can result in transceiver damage, and might void your warranty and support agreements. This is an advanced configuration task; NVIDIA recommends you use the default transceiver temperature settings. Only modify this setting with approval from NVIDIA Technical Support.
- You can set temperature thresholds on the SN5610 switch only.
- Always use the parent transceiver name for configuration; for example, use swp1 not swp1s0.
To set the temperature threshold for a single port, run the nv set platform transceiver <transceiver-id> temperature setpoint command. You can set a value between 30 and 80. The temperature threshold for an interface must be below the module advertised high warning threshold. If you configure the setpoint to be above the module advertised high temperature warning threshold, the FAN algorithm uses the module advertised threshold.
cumulus@switch:~$ nv set platform transceiver swp2 temperature setpoint 60
cumulus@switch:~$ nv config apply
To set the temperature threshold for a group of ports:
cumulus@switch:~$ nv set platform transceiver swp2-10 temperature setpoint 60
cumulus@switch:~$ nv config apply
To set the temperature threshold for a group of comma-separated ports:
cumulus@switch:~$ nv set platform transceiver swp1,4,5,30 temperature setpoint 65
cumulus@switch:~$ nv config apply
- When you set the temperature threshold, the hardware management service (
hw-managament-tc.service) andswitchdrestart. - If you try to set the temperature threshold above the module advertised high temperature warning or if the module is not DOM capable, Cumulus Linux reports an invalid configuration warning.
To unset the temperature threshold and return to the default value, run the nv unset platform transceiver <transceiver-id> temperature setpoint command.
If you run the nv unset platform command, the switch clears out the temperature setpoints for all ports.
To verify the temperature threshold configuration for a port, run the nv show platform transceiver <transceiver-id> temperature command:
cumulus@switch:~$ nv show platform transceiver swp1 temperature
operational applied
-------- ----------- -------
setpoint 60 60
The nv show platform transceiver <transceiver-id> command also shows the temperature threshold for the specified port. See below.
Show Transceiver Information
To show the identifier, vendor name, part number, serial number, and revision for all modules, run the nv show platform transceiver command:
cumulus@switch:~$ nv show platform transceiver
Transceiver Identifier Vendor name Vendor PN Vendor SN Vendor revision
----------- ---------- ----------- ---------------- ------------- ---------------
swp1 QSFP28 Mellanox MCP1600-C001E30N MT2039VB01185 A3
swp10 QSFP28 Mellanox MCP1600-C001E30N MT2211VS01792 A3
swp11 QSFP28 Mellanox MCP1600-C001E30N MT2211VS01792 A3
swp12 QSFP28 Mellanox MCP1650-V00AE30 MT2122VB02220 A2
swp13 QSFP28 Mellanox MCP1650-V00AE30 MT2122VB02220 A2
swp14 QSFP-DD Mellanox MCP1660-W00AE30 MT2121VS01645 A3
swp15 QSFP-DD Mellanox MCP1660-W00AE30 MT2121VS01645 A3
swp18 QSFP28 Mellanox MCP1600-C001E30N MT2211VS01967 A3
swp20 QSFP28 Mellanox MFA1A00-C003 MT2108FT02204 B2
swp21 QSFP28 Mellanox MFA1A00-C003 MT2108FT02204 B2
swp22 QSFP28 Mellanox MFA1A00-C003 MT2108FT02194 B2
swp23 QSFP28 Mellanox MFA1A00-C003 MT2108FT02194 B2
swp31 QSFP28 Mellanox MCP1600-C001E30N MT2039VB01191 A3
To show a detailed view of module information for all ports that includes cable length, type, and diagnostics, current status and error status, run the nv show platform transceiver details command.
To show hardware capabilities and measurement information on the module in a particular port, run the nv show platform transceiver <transceiver-id> command:
cumulus@switch:~$ nv show platform transceiver swp2
cable-type : Active cable
cable-length : 3m
supported-cable-length : 0m om1, 0m om2, 0m om3, 3m om4, 0m om5
diagnostics-status : Diagnostic Data Available
status : plugged_enabled
error-status : N/A
vendor-date-code : 210215__
identifier : QSFP28
vendor-rev : B2
vendor-name : Mellanox
vendor-pn : MFA1A00-C003
vendor-sn : MT2108FT02204
temperature:
temperature : 42.56 C
high-alarm-threshold : 80.00 C
low-alarm-threshold : -10.00 C
high-warning-threshold: 70.00 C
low-warning-threshold : 0.00 C
temperature-setpoint : 60.00 C
alarm : Off
voltage:
voltage : 3.2862 V
high-alarm-threshold : 3.5000 V
low-alarm-threshold : 3.1000 V
high-warning-threshold: 3.4650 V
low-warning-threshold : 3.1350 V
alarm : Off
channel:
channel-1:
rx-power:
power : 0.8625 mW / -0.64 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -13.31 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -10.30 dBm
alarm : Off
tx-power:
power : 0.8988 mW / -0.46 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -11.40 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -8.40 dBm
alarm : Off
tx-bias-current:
current : 6.750 mA
high-alarm-threshold : 8.500 mA
low-alarm-threshold : 5.492 mA
high-warning-threshold: 8.000 mA
low-warning-threshold : 6.000 mA
alarm : Off
...
- The
nv show platform transceivercommands show information for front panel physical ports only (for example swp1). The commands do not show information for logical ports, such as SVIs, bonds, or eth0. - To show information for subinterfaces; run the
nv show interface <subinterface> transceivercommands.
You can also show transceiver data in a more condensed format with the nv show interface <interface-id> transceiver command:
cumulus@switch:~$ nv show interface swp1 transceiver
cable-type : Active cable
cable-length : 3m
supported-cable-length : 0m om1, 0m om2, 0m om3, 3m om4, 0m om5
diagnostics-status : Diagnostic Data Available
status : plugged_enabled
error-status : N/A
revision-compliance : SFF-8636 Rev 2.5/2.6/2.7
vendor-date-code : 210215__
identifier : QSFP28
vendor-rev : B2
vendor-oui : 00:02:c9
vendor-name : Mellanox
vendor-pn : MFA1A00-C003
vendor-sn : MT2108FT02204
temperature : 42.56 degrees C / 108.61 degrees F
voltage : 3.2888 V
ch-1-rx-power : 0.8625 mW / -0.64 dBm
ch-1-tx-power : 0.8988 mW / -0.46 dBm
ch-1-tx-bias-current : 6.750 mA
ch-2-rx-power : 0.8385 mW / -0.76 dBm
ch-2-tx-power : 0.9154 mW / -0.38 dBm
ch-2-tx-bias-current : 6.750 mA
ch-3-rx-power : 0.8556 mW / -0.68 dBm
ch-3-tx-power : 0.9537 mW / -0.21 dBm
ch-3-tx-bias-current : 6.750 mA
ch-4-rx-power : 0.8576 mW / -0.67 dBm
ch-4-tx-power : 0.9695 mW / -0.13 dBm
ch-4-tx-bias-current : 6.750 mA
To show channel information for the module in a particular port, run the nv show platform transceiver <transceiver-id> channel command. To show specific channel information for the module in a particular port, run the nv show platform transceiver <transceiver-id> channel <channel-id> command.
cumulus@switch:~$ nv show platform transceiver swp25 channel
channel:
channel-1:
rx-power:
power : 0.8625 mW / -0.64 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -13.31 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -10.30 dBm
alarm : Off
tx-power:
power : 0.8988 mW / -0.46 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -11.40 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -8.40 dBm
alarm : Off
tx-bias-current:
current : 6.750 mA
high-alarm-threshold : 8.500 mA
low-alarm-threshold : 5.492 mA
high-warning-threshold: 8.000 mA
low-warning-threshold : 6.000 mA
alarm : Off
channel-2:
rx-power:
power : 0.8385 mW / -0.76 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -13.31 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -10.30 dBm
alarm : Off
tx-power:
power : 0.9154 mW / -0.38 dBm
high-alarm-threshold : 5.40 dBm
low-alarm-threshold : -11.40 dBm
high-warning-threshold: 2.40 dBm
low-warning-threshold : -8.40 dBm
alarm : Off
...
To show the thresholds for the module for a specific interface, run the nv show interface <interface-id> transceiver thresholds command:
cumulus@switch:~$ nv show interface swp3 transceiver thresholds
Ch Value High Alarm High Warn Low Warn Low Alarm Alt Value
Threshold Threshold Threshold Threshold
------------------------------------------------------------------------------------------------------------------------
temperature - 42.74 C 80.00 C 70.00 C 0.00 C -10.00 C 108.94F
voltage - 3.2862 V 3.5000 V 3.4650 V 3.1350 V 3.1000 V
rx-power 1 -0.64 dBm 5.40 dBm 2.40 dBm -10.30 dBm -13.31 dBm 0.8625 mW
2 -0.70 dBm 5.40 dBm 2.40 dBm -10.30 dBm -13.31 dBm 0.8514 mW
3 -0.68 dBm 5.40 dBm 2.40 dBm -10.30 dBm -13.31 dBm 0.8556 mW
4 -0.60 dBm 5.40 dBm 2.40 dBm -10.30 dBm -13.31 dBm 0.8704 mW
tx-power 1 -0.48 dBm 5.40 dBm 2.40 dBm -8.40 dBm -11.40 dBm 0.8963 mW
2 -0.38 dBm 5.40 dBm 2.40 dBm -8.40 dBm -11.40 dBm 0.9154 mW
3 -0.19 dBm 5.40 dBm 2.40 dBm -8.40 dBm -11.40 dBm 0.9562 mW
4 -0.13 dBm 5.40 dBm 2.40 dBm -8.40 dBm -11.40 dBm 0.9695 mW
tx-bias-current 1 6.750 mA 8.500 mA 8.000 mA 6.000 mA 5.492 mA
2 6.750 mA 8.500 mA 8.000 mA 6.000 mA 5.492 mA
3 6.750 mA 8.500 mA 8.000 mA 6.000 mA 5.492 mA
4 6.750 mA 8.500 mA 8.000 mA 6.000 mA 5.492 mA
Show CPO Module and Laser Source Information
On a switch with co-packaged optics (CPO), the optics sit inside the switch ASIC package instead of in pluggable modules in the front panel ports, so optical data no longer maps to a single port. Cumulus Linux exposes this hardware through two new commands: nv show platform cpo, which holds the CPO modules, and nv show platform laser-source, which holds the external laser sources (ELS). Use these commands to inspect module state, per-channel optical power, and per-laser health when you suspect an optical fault.
A CPO module contains two optical engines (OE) and 32 lanes, which the CLI shows as channels. Each laser source is one ELS driving eight lasers. Cumulus Linux reads the mapping between ports, channels, and lasers from the switch firmware instead of assuming a fixed ratio, so the mapping can differ between platforms.
- The
nv show platform cpoandnv show platform laser-sourcecommands are available on CPO platforms only. On other platforms, use thenv show platform transceivercommands described earlier. - CPO and laser source data is read-only. There is nothing to configure.
- On a switch with more than one ASIC, these commands show the CPO modules for every ASIC.
- When a module is absent, the entry can still appear with whatever subcomponent state is available. Cumulus Linux does not report a missing value as valid data.
- When part of a module fails, the operational channels and lasers continue to report, and the failed components show an error status.
- When Cumulus Linux cannot read or parse a value, it shows
N/Aor0for that field and logs the error instead of returning a stale or synthetic value. Run the command again to retry the read. - Interface state does not affect these readings. Bringing an interface down, or changing its IP or VLAN configuration, does not change the underlying optical measurements.
Show CPO Modules
To show a summary of every CPO module on the switch, including its status, firmware version, the ports it carries, and the laser source and optical engines it uses, run the nv show platform cpo command:
cumulus@switch:~$ nv show platform cpo
CPO Status Error-status FW Version Ports Laser Source Optical Engines
---- -------- ------------------- ---------- --------- ------------ ---------------
cpo1 inserted fiber contaminated 42.40.15 swp1-swp4 els1 oe1, oe2
cpo2 inserted N/A 42.40.15 swp5-swp8 els2 oe3, oe4
To show all the data for one module, run the nv show platform cpo <cpo-id> command. The output covers all 32 channels of both optical engines, so it runs to between 100 and 200 lines:
cumulus@switch:~$ nv show platform cpo cpo1
status : inserted
error-status : fiber contaminated
identifier : CPO Virtual Module
fw-version : 42.40.15
ports : swp1,swp2,swp3,swp4
laser-source : els1
optical-engine:
oe1:
identifier : OE 16x
serial-number: MT2219FT03171
temperature : 36.8
oe2:
identifier : OE 16x
serial-number: MT2219FT031712
temperature : 36.8
threshold:
laser-source-input-power:
high-alarm : 2.0
low-alarm : -6.0
high-warning: 1.0
low-warning : -5.0
rx-power:
high-alarm : 2.0
low-alarm : -8.0
high-warning: 1.0
low-warning : -7.0
tx-power:
high-alarm : 2.0
low-alarm : -6.0
high-warning: 1.0
low-warning : -5.0
channel:
channel-1:
laser-source-input-power:
power : -1.0
alarm-status : off
alarm-severity: none
rx-power:
power : -1.0
alarm-status : off
alarm-severity: none
tx-power:
power : -0.6
alarm-status : off
alarm-severity: none
rx-los : false
tx-los : false
tx-fault : false
advanced-fault-opcode: none
dp-state : initialized
channel-2:
...
channel-32:
...
For each channel, laser-source-input-power is the power the laser source delivers into the optical engine, rx-power and tx-power are the received and transmitted optical power, and each of the three carries an alarm-status and an alarm-severity derived from the module thresholds. The rx-los and tx-los fields report loss of signal in each direction, tx-fault reports a transmit failure, and dp-state reports the state of the data path.
Because the full view is long, three child commands limit the output:
nv show platform cpo <cpo-id> channelshows thechannelblock only.nv show platform cpo <cpo-id> optical-engineshows theoptical-engineblock only.nv show platform cpo <cpo-id> thresholdshows thethresholdblock only.
cumulus@switch:~$ nv show platform cpo cpo1 optical-engine
optical-engine:
oe1:
identifier : OE 16x
serial-number: MT2219FT03171
temperature : 36.8
oe2:
identifier : OE 16x
serial-number: MT2219FT031712
temperature : 36.8
Show CPO Information for an Interface
To show only the part of a CPO module that carries a particular interface, run the nv show interface <interface-id> cpo command. The output has the same shape as the module view, adds a parent field naming the module, and restricts the channels and the optical engine to those the switch firmware assigns to that interface:
cumulus@switch:~$ nv show interface swp1 cpo
status : inserted
error-status : fiber contaminated
identifier : CPO Virtual Module
fw-version : 42.40.15
ports : swp1,swp2,swp3,swp4
laser-source : els1
parent : cpo1
optical-engine:
oe1:
identifier : OE 16x
serial-number: MT2219FT03171
temperature : 36.8
threshold:
...
channel:
channel-1:
...
channel-8:
...
Show Laser Sources
A laser source is not a transceiver, so it has its own commands. To list the laser sources on the switch with their vendor information and firmware version, run the nv show platform laser-source command:
cumulus@switch:~$ nv show platform laser-source
Laser-source Identifier Vendor name Vendor PN Vendor SN Vendor revision FW Version
------------ ---------- ----------- ------------- ------------- --------------- ----------
els1 ELS NVIDIA MCP7Y60-H01A MT2233VS02221 A4 42.40.15
els2 ELS NVIDIA MCP7Y60-H01B MT2233VS02222 A4 42.40.15
To show the detail for one laser source, including its temperature, power consumption, thresholds, and the state of all eight lasers, run the nv show platform laser-source <laser-source-id> command:
cumulus@switch:~$ nv show platform laser-source els1
status : inserted
error-status : laser_fiber_contaminated
vendor-date-code : 220505
identifier : ELS
vendor-name : NVIDIA
vendor-rev : A4
vendor-pn : MCP7Y60-H01A
vendor-sn : MT2443FT01035
fw-version : 42.40.15
parent : cpo1
temperature : 28.00
power-consumption: 3.52
icc-current : 122
threshold:
optical-power:
high-alarm : 1.0
low-alarm : -5.0
high-warning: 0.0
low-warning : -3.0
bias-current:
high-alarm : 14.0
low-alarm : 8.0
high-warning: 13.0
low-warning : 9.0
laser:
laser-1:
enabled : enabled
oper-status : off
error-status : laser_fiber_contaminated
power-restriction: off
health:
laser-age : 50
laser-health: 25
tec-health : 15
laser-optical-power:
target : 1.50
power : 32
alarm-status : off
alarm-severity: none
laser-bias-current:
current : 12.2
alarm-status : off
alarm-severity: none
frequency-error : 2.3
laser-mpd-current: 0.12
tec:
current : 55
voltage : 2.9
temperature: 23.5
laser-2:
...
laser-8:
...
The parent field names the CPO module the laser source feeds. For each laser, enabled is the administrative state and oper-status is the state the hardware reports, so a laser you enable that has not yet finished ramping shows enabled with an oper-status other than on. The health block reports the age of the laser and health values for the laser and its thermoelectric cooler (TEC), and the tec block reports the current, voltage, and temperature of that cooler.
As with the CPO commands, two child commands limit the output:
nv show platform laser-source <laser-source-id> lasershows thelaserblock only.nv show platform laser-source <laser-source-id> thresholdshows thethresholdblock only.
cumulus@switch:~$ nv show platform laser-source els1 threshold
threshold:
optical-power:
high-alarm : 1.0
low-alarm : -5.0
high-warning: 0.0
low-warning : -3.0
bias-current:
high-alarm : 14.0
low-alarm : 8.0
high-warning: 13.0
low-warning : 9.0
Upgrade Notes
When you upgrade a switch that has CPO hardware, Cumulus Linux detects the hardware and uses the CPO display automatically. You do not need to enable anything.
The scope of the data the transceiver commands return changes on CPO hardware: the switch reports CPO modules instead of front panel transceivers. Adjust any configuration file or script that parses nv show platform transceiver output on a CPO switch to use the nv show platform cpo and nv show platform laser-source commands instead. The output format does not change; it remains key-value pairs.
To export the same data to a monitoring system, see Platform Statistics.