[voq][pmon]xcvrd process show the backtrace in the syslog file on VOQ linecard

[mlok-nokia]

Description

[voq][pmon] xcvrd process show the backtrace in the syslog file on VOQ linecard.

Dec 11 01:16:45.075905 sonic ERR pmon#sfp: SFP index 37 out of range (1-36)
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd Traceback (most recent call last):
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/bin/xcvrd", line 8, in <module>
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     sys.exit(main())
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1620, in main
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     xcvrd.run()
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1539, in run
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     port_mapping_data, retry_eeprom_set = self.init()
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1499, in init
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     retry_eeprom_set = post_port_sfp_dom_info_to_db(is_warm_start, port_mapping_data, self.stop_event)
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 511, in post_port_sfp_dom_info_to_db
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     rc = post_port_sfp_info_to_db(logical_port_name, port_mapping, xcvr_table_helper.get_intf_tbl(asic_index), transceiver_dict, stop_event)
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 291, in post_port_sfp_info_to_db
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd     if not _wrapper_get_presence(physical_port):
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 127, in _wrapper_get_presence
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd     return platform_chassis.get_sfp(physical_port).get_presence()
Dec 11 01:16:45.079471 sonic INFO pmon#/supervisord: xcvrd AttributeError: 'NoneType' object has no attribute 'get_presence'

Steps to reproduce the issue:

1. boot the latest image on a VOQ line card which has internal ports
2. check the syslog file for the PMON module

Describe the results you received:

The below Backtrace in he syslog file

Dec 11 01:16:45.075905 sonic ERR pmon#sfp: SFP index 37 out of range (1-36)
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd Traceback (most recent call last):
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/bin/xcvrd", line 8, in <module>
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     sys.exit(main())
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1620, in main
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     xcvrd.run()
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1539, in run
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     port_mapping_data, retry_eeprom_set = self.init()
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 1499, in init
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     retry_eeprom_set = post_port_sfp_dom_info_to_db(is_warm_start, port_mapping_data, self.stop_event)
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 511, in post_port_sfp_dom_info_to_db
Dec 11 01:16:45.079362 sonic INFO pmon#/supervisord: xcvrd     rc = post_port_sfp_info_to_db(logical_port_name, port_mapping, xcvr_table_helper.get_intf_tbl(asic_index), transceiver_dict, stop_event)
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 291, in post_port_sfp_info_to_db
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd     if not _wrapper_get_presence(physical_port):
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd   File "/usr/local/lib/python3.7/dist-packages/xcvrd/xcvrd.py", line 127, in _wrapper_get_presence
Dec 11 01:16:45.079431 sonic INFO pmon#/supervisord: xcvrd     return platform_chassis.get_sfp(physical_port).get_presence()
Dec 11 01:16:45.079471 sonic INFO pmon#/supervisord: xcvrd AttributeError: 'NoneType' object has no attribute 'get_presence'

Describe the results you expected:

There should not such backtrace

Output of show version:

Latest code as Dec 01.

(paste your output here)

Output of show techsupport:

(paste your output here or download and attach the file here )

Additional information you deem important (e.g. issue happens only occasionally):

[prgeor]

Dec 11 01:16:45.075905 sonic ERR pmon#sfp: SFP index 37 out of range (1-36)

The index seems to be out of range due to which get_sfp() won't return the sfp object

[mlok-nokia]

Dec 11 01:16:45.075905 sonic ERR pmon#sfp: SFP index 37 out of range (1-36)

The index seems to be out of range due to which get_sfp() won't return the sfp object

The port list which is returned by the port_mapping includes the internal ports (either Backplane or Inband port). These internal ports don't have transceiver. They need to be skipped or removed from port_mapping. The old method sfputilbase.py and sfputilhelper.py under sonic-platform-common have this handling.

[prgeor]

Dec 11 01:16:45.075905 sonic ERR pmon#sfp: SFP index 37 out of range (1-36)
The index seems to be out of range due to which get_sfp() won't return the sfp object

The port list which is returned by the port_mapping includes the internal ports (either Backplane or Inband port). These internal ports don't have transceiver. They need to be skipped or removed from port_mapping. The old method sfputilbase.py and sfputilhelper.py under sonic-platform-common have this handling.

can you propose a fix similar to this sonic-net/sonic-platform-common#251

[mlok-nokia]

@prgeor Here is the fix: [voq][pmon]xcvrd process show backtrace in the syslog on VOQ linecard. sonic-net/sonic-platform-daemons#233

[prgeor]

this issue has now been fixed. closing.