J’ai réussi la Swisscom TV fonctionne

sans coupure sur mon Cisco 887VA-W

=> enfin!!!

Le secret ou plutot les secrets?

L’IP RP (Rendezvous Point) chez Swisscom est bel et bien 1.1.1.1!!!
sans IP RP le flux s’arrete après environ 10 secondes (flux unicast avant multicast)
ip igmp query-interval 125 ou 60 ou 30 ne fonctionne pas (set igmp query-intvl 125)
ip igmp query-interval 12 fonctionne – default Cisco (60) et Motorola (125)
sans une valeur aussi basse le flux s’arrete avant 60 seconds!
Les timeout chez Motorola sont en deci-secondes tandis et chez Cisco en milli-secondes!

> je n’ai pas bcp de temps pour analyser ces données,

> mais au premier coup d’oeil je dirai que ce n’est pas possible que lP Rp soit 1.1.1.1!

Oui, pour moi je suis persuadé que IP RP Rendezvous Point est bien 1.1.1.1

chez Swisscom. Ce qui est un choix discutable vu que c’est un block routable

qui a été enlevé de la boggus list par IANA..

Au niveau du Cisco, j’ai obtenu ceci:

%SEC-6-IPACCESSLOGSP: list INTERNET-TO-WAN denied igmp 1.1.1.1 -> 224.0.0.1 (0), 1 packet

Au niveau du Motorola Netopia 3397GP et après avoir activé le mode bridge/router

et le mode log igmp… bingo encore 1.1.1.1!!!

Netopia-3000/31499700> config

Config Mode v1.3
Netopia-3000/31499700 (top)>> set igmp log-enable on
Netopia-3000/31499700 (top)>> save
WARNING: Concurrent-bridging/routing has been turned ON!

Configuration data saved.
Netopia-3000/31499700 (top)>> top
Netopia-3000/31499700 (top)>> cd..
Error: Command (‘cd’) not found

Netopia-3000/31499700 (top)>> exit

Netopia-3000/31499700> show log

Message Log:
2/19/12 09:10:21 AM L3 TR-069: Closing connection on auth fail
2/19/12 09:10:21 AM L3 SSL: Closing Connection: rms.bluewin.ch
2/19/12 09:10:25 AM L3 IGMP: [ENWAN1] Received General Query host: 1.1.1.1 max-reponse-time 2 s
2/19/12 09:10:25 AM L3 IGMP: [en0] Received General Query host: 1.1.1.1 max-reponse-time 2 s
2/19/12 09:10:26 AM L3 IGMP: [ENWAN1] Sending v2-Report group: 239.255.255.250
2/19/12 09:10:26 AM L3 IGMP: [ENWAN1] Sending v2-Report group: 224.0.0.251
2/19/12 09:10:26 AM L3 IGMP: [ENWAN1] Sending v2-Report group: 239.186.63.6
2/19/12 09:10:26 AM L3 IGMP: [en0] Received v2-Report group: 239.186.63.6 host: 172.16.78.128
2/19/12 09:10:26 AM L3 IGMP: [en0] Received v2-Report group: 239.255.255.250 host: 172.16.78.128
2/19/12 09:10:27 AM L3 IGMP: [en0] Received v2-Report group: 224.0.0.251 host: 172.16.78.2
2/19/12 09:10:36 AM L3 IGMP: [ENWAN1] Received General Query host: 1.1.1.1 max-reponse-time 2 s
2/19/12 09:10:36 AM L3 IGMP: [en0] Received General Query host: 1.1.1.1 max-reponse-time 2 s
2/19/12 09:10:36 AM L3 IGMP: [ENWAN1] Sending v2-Report group: 239.186.63.6

Netopia-3000/31499700>

Meme si 1.0.0.0 a été assigné à APNIC il semble que Swisscom l’utilise en interne.

Tu peux d’ailleur pinger 1.1.1.1 depuis ta connexion Swisscom!

ping -c 1 1.1.1.1
PING 1.1.1.1 (1.1.1.1): 56 data bytes
64 bytes from 1.1.1.1: icmp_seq=0 ttl=126 time=19.913 ms

--- 1.1.1.1 ping statistics —
1 packets transmitted, 1 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 19.913/19.913/19.913/0.000 ms

1.1.1.1 n’arrive clairement pas chez au USA avec 19ms!

traceroute 1.1.1.1
traceroute to 1.1.1.1 (1.1.1.1), 64 hops max, 52 byte packets
1 * 172.16.78.1 (172.16.78.1) 0.591 ms 0.601 ms
2 1.1.1.1 (1.1.1.1) 19.706 ms 19.553 ms 19.902 ms

Voir aussi cet article 🙂

http://packetlife.net/blog/2010/feb/5/ripe-plays-with-1-0-0-0-network-apnic-allocation/

inetnum: 1.1.1.0 - 1.1.1.255
netname: Debogon-prefix
descr: APNIC Debogon Project
descr: APNIC Pty Ltd
country: AU
admin-c: AR302-AP
tech-c: AR302-AP
mnt-by: APNIC-HM
mnt-routes: MAINT-AU-APNIC-GM85-AP
mnt-irt: IRT-APNICRANDNET-AU
status: ASSIGNED PORTABLE
changed: hm-changed@apnic.net 20110922
source: APNIC

> le cmd sont normalmement

> set igmp snooping on

> set igmp robustness 2

> set igmp query-intvl 125

> set igmp query-response-intvl 100

> set igmp version 3

> set igmp fast-leave on

> set igmp log-enable off

> set igmp default-forwarding-allow on

> set ip igmp-forwarding on

Voici la configuration qui fonctionne avec la Swisscom TV et Swisscom VDSL2

!
service internal
!
memory-size iomem 20
!
controller VDSL 0
operating mode vdsl2
modem co5
!
ip pim rp-address 1.1.1.1
!
ip multicast-routing
!
ip igmp snooping
!
interface Ethernet0
ip dhcp client broadcast-flag clear
ip address dhcp
no ip virtual-reassembly in
no ip route-cache

no ip mfib cef input
no ip mfib cef output

ip pim sparse-mode

ip igmp query-interval 10
ip igmp querier-timeout 60
ip igmp query-max-response-time 8

ip igmp last-member-query-interval 1000
ip igmp last-member-query-count 2
ip igmp version 3
ip igmp proxy-service
load-interval 30
auto discovery qos
service-policy output QoS-Ethernet0
!
interface Vlan1
ip address 172.16.78.1 255.255.255.0
ip nat inside
no ip virtual-reassembly in
ip tcp adjust-mss 1452
ip pim sparse-mode

ip igmp query-interval 10
ip igmp querier-timeout 60
ip igmp query-max-response-time 8

ip igmp last-member-query-interval 1000
ip igmp last-member-query-count 2
ip igmp version 3
ip igmp proxy-service
ip igmp helper-address 1.1.1.1
ip igmp explicit-tracking
load-interval 30
auto discovery qos
service-policy output QoS-Vlan1
!
class-map match-any QoS-Video
match protocol rtp video
match ip dscp af41
!
policy-map QoS-Ethernet0
class class-default
fair-queue
!
policy-map QoS-Vlan1
class QoS-Video
set dscp af41
class class-default
set dscp default
!
ip forward-protocol nd
!
scheduler allocate 20000 1000
!
end

> Faut pas que tu n’en dormes pas du week-end

> Peut-être d’autres membres plus spécialisés que moi pourraient t’aider?

Je pense que c’est bon maintenant.

Merci encore de ton aide!

Faut dire que j’ai été dégouté pendant trois jours de repasser sur mon vieux

Motorola Netopia de 2008… pour que ma ptite famille regarde la Swisscom TV…

C’était étrange quand même qu’en connectant uniquement la Swisscom TV

sur le Cisco 887VA, le fux Swisscom TV se fige… C’est pas la QoS qui allait aider

car avec 22Mbps/5Mps c’est suffisant pour un flux de 12Mbps (HD) ou 5Mbps (SD)

DETAILS TECHNIQUES

> set igmp query-intvl 125
>
> Query Interval– the amount of time in seconds between IGMP General Query messages sent by the
> querier gateway. The default query interval is 125 seconds

C’ETAIT L’ORGINE DU STOP DE L’IMAGE APRES MOINS DE 60 SECONDES
(LE STOP APRES 10 SEC A ETE FIXE AVEC L’ADRESSE RendezvousPoint => 1.1.1.1)

CONFIG
ip igmp query-interval 10

CISCO
ip igmp query-interval?
<1-18000> Query interval in seconds

INFO
Autant 125 (Motorola/RFC default) et 60 (Cisco default) et meme 30 ne fonctionne pas bien avec Swisscom TV.
Il faut ip igmp query-interval 12!!!

<< Note: We recommend that you use the default IGMP query interval and timeout period.
The Cisco IOS software uses a default IGMP query interval of 60 seconds,
which is different from the RFC standard default of 125 seconds.
Using a lower default IGMP query interval of 60 seconds allows routers
to stop forwarding traffic faster when a member crashes without sending
leaves (in IGMPv2 or IGMPv3 environment) >>

> set igmp last-member-query-intvl 10
>
> Sets the last member quer y inter val: the amount of time in tenths of a second that the IGMP gateway
> waits to receive a response to a Group-Speciﬁc Quer y message. The last member quer y inter val is also
> the amount of time in seconds between successive Group-Speciﬁc Quer y messages.
> The default is 1 second (10 deci-seconds)

CONFIG

ip igmp last-member-query-interval 1000

CISCO
ip igmp last-member-query-interval?
<100-25500> Last member query interval in milliseconds

INFO
Il faut faire très attention les valeurs Motorola sont en deci-seconds tandis que Cisco en millo-seconds!!!

> set igmp last-member-query-count 2

CONFIG
ip igmp last-member-query-count 2

CISCO
ip igmp last-member-query-count?

<1-7> Last member query count

> set igmp query-response-intvl 100
>
> Sets the query-response interval range: from 5 deci-seconds (tenths of a second) – 255 deci-seconds.
> The default is 100 deci-seconds.
>
> Last Member Query Interval – the amount of time in tenths of a second that the IGMP gateway
> waits to receive a response to a Group-Speciﬁc Query message. The last member query interval is
> also the amount of time in seconds between successive Group-Speciﬁc Query messages. The
> default last member query interval is 1 second (10 deci-seconds).

INFO
pas info j’ai cru que c’était ip igmp last-member-query-interval??

> set igmp version 3

CONFIG

ip igmp version 3

> set igmp snooping on

CONFIG
ip igmp snooping

INFO
et pas ip igmp snooping vlan 1

> set igmp default-forwarding-allow on

ip multicast-routing

> set igmp fast-leave on
>
> Fast Leave – set to off by default, fast leave enables a non-standard expedited leave mechanism.
> The querier keeps track of which client is requesting which channel by IP address. When a leave
> message is received, the querier can check its internal table to see if there are any more clients on
> this group. If there are none, it immediately sends an IGMP leave message to the upstream querier.

INFO
“ip igmp snooping fast-leave” pas dispo sur mon IOS

> set igmp robustness 2
>
> Robustness – a way of indicating how sensitive to lost packets the network is.
> IGMP can recover from robustness minus 1 lost IGMP packet. The default value is 2.

INFO
Pas dispo sur ma version IOS

> set igmp unsol-report-intvl 10
>
> Sets the unsolicited report interval: the amount of time in seconds between repetitions of a particular
> computer’s initial report of membership in a group. The default is 10 seconds

INFO
Pas info sur une commande similaire sur IOS

> set igmp wireless-m2u on
>
> This command allows you enable or disable wireless multicast-to-unicast if igmp snooping is set to on.
>
> Wireless Multicast to Unicast conversion – Only available if IGMP Snooping is enabled. If set to
> on, the Gateway replaces the multicast MAC-address with the physical MAC-address of the wireless
> client. If there is more than one wireless client interested in the same multicast group, the Gateway
> will rever t to multicasting the stream immediately. When one or more wireless clients leave a group,
> and the Gateway determines that only a single wireless client is interested in the stream, it will once

> again unicast the stream.

INFO
Pas info pour IOS

REFERENCES
Page 158 http://broadband.motorola.com/consumers/products/7107-65-10NA/downloads/AdminHndbkV784-sku29_34.pdf
http://www.cisco.com/en/US/docs/ios/ipmulti/command/reference/imc_02.html

http://www.sadikhov.com/forum/index.php?/topic/186519-cisco-88xva-%26-8211%3B-part-5-vdsl-advanced-configuration-%26-8211%3B-iptv-via-multicast/

http://www.ditc.eu/downloads/Multicast_Cisco_T-Home_IPTV.pdf