L2-TCAM_RM-3-TCAM_RM_RECVD_REPLY_NOMEM & L2-TCAM_RM-3-BLWM_FAILED_MSG_DUMP

[Rps-Cheers]

HI，

GSR12000

Ver 4.3.4

Does anyone know the reason for the following log? How should it be solved?

fib_mgr[158]: %L2-TCAM_RM-3-TCAM_RM_RECVD_REPLY_NOMEM : tcam_rm: TCAM could not satisfy allocation request, the node may not function as expected.  : pkg/bin/fib_mgr : (PID=98396) :  -Traceback= 78eca6fc 788ed4fc 788edb70 788ee2f0 78ecb898 78ecc7bc 78ecca98 78ec9db0 79b1c4d4 79b18908 79c83a58 79bffd6c 79bfff40 796a8948 7951a258 7946ce58

fib_mgr[158]: %L2-TCAM_RM-3-BLWM_FAILED_MSG_DUMP : TCAM RM:  ALLOC_VMR_SYM 36 REGN TYPE IPV6 REG-ID 148   vmrid 0x0 vmrid_p 0x5dd18b1c,  rc Not enough memory[12]   : pkg/bin/fib_mgr : (PID=98396) :  -Traceback= 78eca37c 78eca70c 788ed4fc 788edb70 788ee2f0 78ecb898 78ecc7bc 78ecca98 78ec9db0 79b1c4d4 79b18908 79c83a58 79bffd6c 79bfff40 796a8948 7951a258

This log is very frequent and fills the entire log cache. Will it affect the device?

[tkarnani]

Hi,

This error indicates the TCAM memory is full. we can verify with

show controllers pse tcam summary

if it is full, you will need to reduce the route scale. we can check the following output to see the current routes

show route ipv6 summary

show route ipv4 summary

Thanks

[Rps-Cheers]

Hi tkarnani

Thanks for your response.

RP/0/0/CPU0:router#show controllers pse tcam summary 
Tue Jan 5  
Cells Masks
============================= =============================
Node Dir Tcam Alloced Free Total %full Alloced Free Total %full
========= === ==== ============================= =============================
0/2/CPU0 Ing Fwd 950439 98137 1048576 90.6 950439 98137 1048576 90.6
0/2/CPU0 Ing Cls 2188 259956 262144 0.8 2188 259956 262144 0.8
0/2/CPU0 Ing Nf 0 262144 262144 0.0 0 262144 262144 0.0
0/2/CPU0 Egr Fwd 950136 98440 1048576 90.6 950136 98440 1048576 90.6
0/2/CPU0 Egr Cls 36 262108 262144 0.0 36 262108 262144 0.0
0/2/CPU0 Egr Nf 0 262144 262144 0.0 0 262144 262144 0.0

...

RP/0/0/CPU0:router#sho route ipv6 summary 
 Tue Jan 5 

Route Source Routes Backup Deleted Memory (bytes)
local 3 0 0 492 
connected 2 1 0 492 
bgp 17622 130401 18 0 21388716 
isis 100 388 2 0 74256 
Total 130794 21 0 21463956



RP/0/0/CPU0:router#sho route ipv4 summary 
 Tue Jan 5 

Route Source Routes Backup Deleted Memory (bytes)
connected 2 1 0 408 
local 3 0 0 408 
dagr 0 0 0 0 
bgp 17622 823212 8 0 111957920 
ospf 100 2201 0 0 394376 
Total 825418 9 0 112353112

From the above information, it is found that TCAM MEM has exceeded 90%. It seems that there are more BGP routes, right? Is there any other way besides reducing routing entries? This is a very difficult question!

[Rps-Cheers]

Can we try to clear the TCAM utilization? If so, what is the commands?