Symptom
The data displayed when executing the show controller phy | inc dBm is accurate.
sh controller Te0/6/0/0/6 phy | inc dBm
Total Tx power: 0.37 mW ( -4.29 dBm)
Lane 5 Tx power: 0.37 mW ( -4.29 dBm)
Total Rx power: 0.25 mW ( -5.95 dBm)
Lane 5 Rx power: 0.25 mW ( -5.95 dBm)
SNMP Poll for the Corresponding Optical Sensor OID displays inaccurate values
SNMPv2-SMI::mib-2.47.1.1.1.1.7.4423956 = STRING: "0/6-PORT-0-Tx Lane 6 Power"
SNMPv2-SMI::enterprises.9.9.91.1.1.1.1.4.4423956 = INTEGER: 0
SNMPv2-SMI::mib-2.47.1.1.1.1.7.4423968 = STRING: "0/6-PORT-0-Rx Lane 6 Power"
SNMPv2-SMI::enterprises.9.9.91.1.1.1.1.4.4423968 = INTEGER: 0
Conditions
Issue is seen in TenGigE interfaces using Lanes other then Lane 0.
This issue is triggered by the restart of the slice_mgr_proxy process in a linecard with slices in 10GE beakout mode.
Workaround
None
Recovery Process:
Process restart of slice_manager on the affected line card from the Sysadmin VM --> "process restart slice_manager loc " clears the issue.
Further Problem Description
The slice_manager will update the sensor data when one of the following events occur which will recover the issue for the affected interface:
- Process restart of slice_manager on the affected LC.
- LC Reload
- Optical Cable Pull/Insert
- Optics OIR
- Level change of > +/- 50 uW from previous reading. The slice_manager process checks for a change in the level every minute and will update if current value it is +/- 50 uW of the previous reading.
