[f-nsp] Multicast is being switched by LP CPU on MLXe?

[Alexander Shikoff]

Hi!


Well, I'd like to bring this thread up again hoping to catch 
someone who has also hit this issue.

Today I upgraded software to 05.9.00be, and situation is still
the same: with enabled Multicast Traffic Reduction,
multicast traffic is being switched by LP CPU.

Current test VLAN configuration is:
!
vlan 450 name ITCons2DS_test 
 tagged ethe 7/1 to 7/2 ethe 9/5 ethe 11/2 ethe 12/8 ethe 13/8 
 multicast passive
 multicast pimsm-snooping
!

telnet at lsr1-gdr.ki#show vlan 450

PORT-VLAN 450, Name ITCons2DS_test, Priority Level 0, Priority Force 0, Creation Type STATIC
Topo HW idx    : 65535    Topo SW idx: 257    Topo next vlan: 0
L2 protocols   : NONE
Statically tagged Ports    : ethe 7/1 to 7/2 ethe 9/5 ethe 11/2 ethe 12/8 ethe 13/8 
Associated Virtual Interface Id: NONE
----------------------------------------------------------
Port  Type      Tag-Mode  Protocol  State     
7/1   TRUNK     TAGGED    NONE      FORWARDING
7/2   TRUNK     TAGGED    NONE      FORWARDING
9/5   TRUNK     TAGGED    NONE      FORWARDING
11/2  TRUNK     TAGGED    NONE      FORWARDING
12/8  TRUNK     TAGGED    NONE      FORWARDING
13/8  TRUNK     TAGGED    NONE      FORWARDING
Arp Inspection: 0
DHCP Snooping: 0
IPv4 Multicast Snooping: Enabled - Passive
IPv6 Multicast Snooping: Disabled

No Virtual Interfaces configured for this vlan


IGMP snooping works, I'm able to see In/Out interfaces and current
active querier:

telnet at lsr1-gdr.ki#show ip multicast vlan 450
----------+-----+---------+---------------+-----+-----+------
VLAN       State Mode      Active          Time (*, G)(S, G) 
                           Querier         Query Count Count 
----------+-----+---------+---------------+-----+-----+------
450        Ena   Passive   192.168.210.1   119   1     1     
----------+-----+---------+---------------+-----+-----+------

Router ports: 12/8 (11s)

Flags-  R: Router Port,  V2|V3: IGMP Receiver,  P_G|P_SG: PIM Join

  1    (*, 239.32.4.130) 00:34:48       NumOIF: 1       profile: none
          Outgoing Interfaces:
               e9/5 vlan 450 ( V2) 00:34:48/40s

  1    (91.238.195.1, 239.32.4.130) in e11/2 vlan 450 00:34:48          NumOIF: 1  profile: none
          Outgoing Interfaces:
               TR(e9/5,e7/1) vlan 450 ( V2) 00:34:48/0s
          FID: 0xa0a9     MVID: None     


Right after multicast stream start flooding from Eth11/2 out of TR(e9/5,e7/1),
the CPU load on LP 11 increases:

telnet at lsr1-gdr.ki#show cpu-utilization lp 11

17:25:10 GMT+02 Tue Dec 13 2016

SLOT  #:               LP CPU UTILIZATION in  %:
             in 1 second:  in 5 seconds:  in 60 seconds: in 300 seconds:
    11:        6             6              6               6  



And I see these packets processed by LP CPU:

LP-11#debug packet capture include vlan-id 450
[...]
91.238.195.1 -> 239.32.4.130 UDP [2000 -> 2000] 
**********************************************************************
[ppcr_tx_packet] ACTION: Forward packet using fid 0xa0a9
[xpp10ge_cpu_forward_debug]: Forward LP packet
Time stamp : 00 day(s) 11h 32m 49s:,
TM Header: [ 1022 00a9 a0a9 ]
Type: Multicast(0x00000000) Size: 34 Mcast ID: 0x9a0 Src Port: 2
Drp Pri: 2 Snp: 2 Exclude Src: 0 Cls: 0x00000001
**********************************************************************
00: a0a9 0403 5e50 41c2-7840 0abc 4400 0000  FID     = 0xa0a9
10: 0100 5e20 0482 f4cc-55e5 4600 0800 4588  Offset  = 0x10
20: 0540 08df 0000 3d11-5cb4 5bee c301 ef20  VLAN    = 450(0x01c2)
30: 0482 07d0 07d0 052c-0000 4701 e11a 8534  CAM     = 0x00055e 
40: 5a95 fb85 94ee 0b69-9938 967a c827 f571  SFLOW   = 0
50: 73cc 8e72 98cc 82e0-436e 30f1 4414 f400  DBL TAG = 0
60: 11fd 7b2b c8be d9ca-d0fa 44d0 45b5 53e5
70: a386 ac24 cc0b 9698-c0a2 ff65 9f32 6b14
Pri CPU MON SRC   PType US BRD DAV SAV DPV SV ER TXA SAS Tag MVID
4   0   0   11/2  3     0  1   0   1   1   1  1  0   0   1   0

91.238.195.1 -> 239.32.4.130 UDP [2000 -> 2000] 
**********************************************************************
[ppcr_rx_packet]: Packet received
Time stamp : 00 day(s) 11h 32m 49s:,
TM Header: [ 0564 8a23 0040 ]
Type: Fabric Unicast(0x00000000) Size: 1380 Class: 4 Src sys port: 2595
Dest Port: 0  Drop Prec: 1 Ing Q Sig: 0 Out mirr dis: 0x0 Excl src: 0 Sys mc: 0
**********************************************************************
Packet size: 1374, XPP reason code: 0x00045286
00: 05f0 0403 5c50 41c2-7841 fffe 4400 0000  FID     = 0x05f0
10: 0100 5e20 0482 f4cc-55e5 4600 0800 4588  Offset  = 0x10
20: 0540 08e0 0000 3d11-5cb3 5bee c301 ef20  VLAN    = 450(0x01c2)
30: 0482 07d0 07d0 052c-0000 4701 e11f c052  CAM     = 0x00ffff(R)
40: e9df e2fb 1f9d 0c1d-354a 7df5 f0df edab  SFLOW   = 0
50: 1145 566c 4c59 2557-f7cf c708 a75e 5a29  DBL TAG = 0
60: 1704 9f8b 151c b66b-957a 51eb ac99 772d
70: 07e7 23d7 f84a 50ac-5864 452d 7f70 0495
Pri CPU MON SRC   PType US BRD D
4   0   0   11/2  3     0  1   0   1   1   1  0  0   0   1   0


I have no ideas why it happens. "Multicast Guide" clearly tells
that these packets should be processed in hardware.
Please advice!

Thanks!

-- 
MINO-RIPE