We have problems with the BFD between the ACX7509 and the T0 NSX-T.

The BFD adjacency is continuously down.

Oct 07 15:41:46 JTAC-re0 evo-pfemand[10323]: 0:dnx_algo_global_lif_allocation_verify: Error 'Entry exists' indicated, Asked for global lif 0x002ff07f WITH_ID, but it's already allocated

Oct 07 15:41:46 JTAC-re0 evo-pfemand[10323]: 0:dnx_algo_global_lif_allocation_allocate: Error indicated (Entry exists) on VERIFY

Oct 07 15:41:46 JTAC-re0 evo-pfemand[10323]: 0:bcm_dnx_tunnel_initiator_create: Error 'Entry exists' indicated

Oct 07 15:41:46 JTAC-re0 evo-pfemand[10323]: EVO_PFEMAND_ASIC_PROGRAM_FAILED: Failed to program the hardware with an error - entry exist in hardware

Oct 07 15:41:46 JTAC-re0 evo-pfemand[10323]: [Error] BrcmPlusOam: status_t BrcmPlusDnxOamBfdTx::bfdIpv4ExtendedSipTunnelCreate() Tunnel initiator create fail for src ip 0xa400b5e encapId 0x2feff1. BRCM ERROR : Entry exists

The session comes up with centralised mode using below CLI.

set routing-options ppm no-delegate-processing

But in distributed mode, we could see logs from symptoms.

PR1840565 was opened, suspecting interop issue & we identified the issue was due to PR1799467.

The fix is done via PR1799467.

PR1799467:

We create extended Sip Tunnel during the creation of BFD ipv4 endpoint. These tunnels will be cleared when the session is going down or during restart.

Tunnels are created in the respective ifd units, but during the deletion we are looping through all the units and deleting it.

Because of a limitation where the ifd container is deleted before tunnels are deleted & due to this, error is thrown where ever tunnel is not present.

Fixed in:

evo:23.2R2-S3-EVO evo:23.4R2-EVO evo:24.2R1-EVO evo:24.2R2-EVO evo:24.3R1-EVO evo:24.4R1-EVO evo:25.1R1-EVO