Hi!
We try to configure Multicast Service Reflection Feature on ME6524 (IOS: 12.2(33)SXI6, ADVIPSERV)
Documentation: http://www.cisco.com/en/US/docs/ios/ipmulti/configuration/guide/imc_serv_reflect_ps10591_TSD_Products_Configuration_Guide_Chapter.html
for rewriting Multicast Group and Source addresses (Multicast to Multicast)
We receive multicast streams from MCAST SP on Gi1/21 (groups: 233.26.198.195, 233.26.198.217) and using PIM-SM:
interface GigabitEthernet1/21
description === (TV feeding)
ip address 10.157.192.62 255.255.255.252
ip pim bsr-border
ip pim sparse-mode
!
! for rpf check pass:
ip route x.y.209.193 255.255.255.255 10.157.192.61
ip route x.y.140.122 255.255.255.255 10.157.192.61
!
Create Vif1:
!
interface Vif1
description === Multicast Service Reflection from X to Y (HDTV)
bandwidth 100000
ip address 172.31.255.81 255.255.255.248
ip pim bsr-border
ip pim sparse-mode
ip service reflect GigabitEthernet1/21 destination 233.26.198.195 to 239.191.254.125 mask-len 32 source 172.31.255.83
ip service reflect GigabitEthernet1/21 destination 233.26.198.217 to 239.191.254.124 mask-len 32 source 172.31.255.82
ip igmp static-group 233.26.198.217
ip igmp static-group 233.26.198.195
load-interval 30
!
interface Loopback0
ip address a.b.0.63 255.255.255.255
ip pim sparse-mode
!
ip pim rp-address a.b.0.63 PIM_IPTV_MCAST_PROVIDER_GROUPS override
ip pim rp-candidate Loopback0 group-list PIM_IPTV_GROUPS_HD priority 188
!
!
Always worked fine for 5-10 seconds (mroute created, traffic translated, and TV showing ).
!
I think this interval defined in hardware (mroute scan process and threshold between 'hardware-software' multicast shortcut installation (pps)):
!
Jul 11 10:04:26: MLSM: global scan in progress(1). set flag and exit
Jul 11 10:04:26: MLSM: (1)start another mroute scan
!
!and
mls ip multicast threshold 150
Then scan proccess is end in 'debug mls ip multicast all' output writes for all translated groups:
Jul 11 10:04:35: MLSM[0]: Build SC INS seq 83, size 34, fe->flags 0x80100020 rp:a.b.0.63 (172.31.255.82, 239.191.254.124, 0) oif: +1047(0x1) mfd
Jul 11 10:04:35: MLSM[0]: Build SC INS seq 165, size 34, fe->flags 0x100820 rp:a.b.0.63 (x.y.140.122, 233.26.198.217, 1048) oif: +0(0xF) sr_vif mfd
Jul 11 10:04:35: MLSM[0]: SC INS nack (172.31.255.82, 239.191.254.124, 0) seq_num 83 SC_ERR_GENERIC
Jul 11 10:04:35: MLSM[0]: chk seq (172.31.255.82, 239.191.254.124, 0) in_flags = 0x20 ,I
Jul 11 10:04:35: MLSM[0]: Dequeue pndQ-deq (172.31.255.82, 239.191.254.124, Vi1) 0 event flushed
Jul 11 10:04:35: MLSM[0]: proc FE_DELETE (0.0.0.0, 239.191.254.124, 0) !scmdb free
Jul 11 10:04:35: MLSM[0]: SC INS ack (x.y.140.122, 233.26.198.217, 1048) seq_num 165, olen 4, in_flags = 0x820
Jul 11 10:04:35: MLSM[0]: chk seq (x.y.140.122, 233.26.198.217, 1048) in_flags = 0x820 ,I olen 4, OIF: +0(0x0) oif-type 0xF I mfd
And after SC_ERR_GENERIC flow is stopped.
If all multicast flows going throught the software (mls ip multicast threshold 2000) always working fine, but cpu load is high.
In Release notes for ME6524 says that this feature supported in hardware from "providing forwarding in hardware in Cisco IOS Release 12.2(33)SXI4."
Would anyone help us?
