[VoQ chassis] Error when creating Portchannels resulting in SAI error

[judyjoseph]

Description

Portchannel interface creation fails in Nokia chassis with the error, resulting in orchagent/swss docker exit

SAI_API_LAG:brcm_sai_dnx_create_lag:138 SYSTEM_PORT_AGGREGATE_ID >= 128

Mar 16 05:50:50.631409 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- addLag: Create an empty LAG PortChannel1025 lid:2000000000d5e
Mar 16 05:50:50.631806 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- updatePortOperStatus: Port PortChannel1025 oper state set from unknown to down
Mar 16 05:50:50.631819 STG01-0101-0200-01T2-lc04 WARNING swss0#orchagent: :- flushFDBEntries: Couldn't flush FDB. Bridge port OID: 0x0 bvid:0,
Mar 16 05:50:50.651452 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- addLag: Create an empty LAG PortChannel1026 lid:2000000000d5f
Mar 16 05:50:50.651744 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- updatePortOperStatus: Port PortChannel1026 oper state set from unknown to down
Mar 16 05:50:50.651744 STG01-0101-0200-01T2-lc04 WARNING swss0#orchagent: :- flushFDBEntries: Couldn't flush FDB. Bridge port OID: 0x0 bvid:0,
Mar 16 05:50:50.672296 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- addLag: Create an empty LAG PortChannel1027 lid:2000000000d60
Mar 16 05:50:50.672572 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- updatePortOperStatus: Port PortChannel1027 oper state set from unknown to down
Mar 16 05:50:50.672572 STG01-0101-0200-01T2-lc04 WARNING swss0#orchagent: :- flushFDBEntries: Couldn't flush FDB. Bridge port OID: 0x0 bvid:0,
Mar 16 05:50:50.673062 STG01-0101-0200-01T2-lc04 ERR syncd0#syncd: [06:00.0] SAI_API_LAG:brcm_sai_dnx_create_lag:138 SYSTEM_PORT_AGGREGATE_ID >= 128
Mar 16 05:50:50.673062 STG01-0101-0200-01T2-lc04 ERR syncd0#syncd: :- sendApiResponse: api SAI_COMMON_API_CREATE failed in syncd mode: SAI_STATUS_INVALID_PARAMETER
Mar 16 05:50:50.673109 STG01-0101-0200-01T2-lc04 ERR syncd0#syncd: :- processQuadEvent: attr: SAI_LAG_ATTR_SYSTEM_PORT_AGGREGATE_ID: 128
Mar 16 05:50:50.673230 STG01-0101-0200-01T2-lc04 ERR swss0#orchagent: :- create: create status: SAI_STATUS_INVALID_PARAMETER
Mar 16 05:50:50.673240 STG01-0101-0200-01T2-lc04 ERR swss0#orchagent: :- addLag: Failed to create LAG PortChannel1028 lid:2000000000d61
Mar 16 05:50:50.673240 STG01-0101-0200-01T2-lc04 ERR swss0#orchagent: :- handleSaiCreateStatus: Encountered failure in create operation, exiting orchagent, SAI API: SAI_API_LAG, status: SAI_STATUS_INVALID_PARAMETER
Mar 16 05:50:50.673254 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- notifySyncd: sending syncd: SYNCD_INVOKE_DUMP
Mar 16 05:50:50.673391 STG01-0101-0200-01T2-lc04 NOTICE syncd0#syncd: :- processNotifySyncd: Invoking SAI failure dump
Mar 16 05:50:50.676645 STG01-0101-0200-01T2-lc04 NOTICE swss0#orchagent: :- sai_redis_notify_syncd: invoked DUMP succeeded
Mar 16 05:50:51.035837 STG01-0101-0200-01T2-lc04 INFO swss0#supervisord 2023-03-16 05:50:51,035 INFO exited: orchagent (terminated by SIGABRT (core dumped); not expected)
Mar 16 05:50:52.041014 STG01-0101-0200-01T2-lc04 INFO swss0#supervisor-proc-exit-listener: Process 'orchagent' exited unexpectedly. Terminating supervisor 'swss'

Steps to reproduce the issue:

1. Create more lags in the device, in various linecards.

Describe the results you received:

Could not create a particular PortChannel1028, and orchagent/swss docker exit

Describe the results you expected:

Output of show version:

202205 latest

(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):

[judyjoseph]

Looks like in SUP we have the mac lag id as 512 : https://github.com/sonic-net/sonic-buildimage/blob/master/device/nokia/x86_64-nokia_ixr7250e_sup-r0/chassisdb.conf

But Broadcom J2c+ supports max 128, for these code

[judyjoseph]

Broadcom CSP raised : Case CS00012287482 increase the limit to 1024 in SAI/SDK. Waiting for this to be merged.

Changes in SONiC:

1. All chassis platforms to update the LAG limits to 1024 in chassis_db.conf
2. Add the soc property in the config bcm file for supporting 16-member system wide LAGs
   trunk_group_max_members=16

[rlhui]

@judyjoseph please update if fixes are all there, thanks

[judyjoseph]

Updated the platforms with 1k LAG interfaces, Brcm SAI already in 202205 branch