[Xorp-users] Fwd: Problem in IPv6 SSM Multicast

Pavlin Radoslavov pavlin at icir.org
Tue Dec 4 02:05:08 PST 2007


> Thank you for the quick response. I intentionally did not add the mld
> configuration in Router 1 specifically because my understanding on the MLD
> as well as the IGMP protocol is that both are only intended for multicast
> routers to  receivers communication. Please don't hesitate to correct me if
> I'm wrong.

Yes, you are correct that IGMP and MLD are needed only on the router
with directly connected receivers.
I mentioned the missing mld just to make sure that you are aware of
it.

> After adding MLD to the router1 xorp configuration:
> 
> admin at demo_rtr1> show mld interface
> Interface    State    Querier         Timeout Version Groups
> sk0          UP       fe80::219:5bff:fe2f:146a    None       2      8
> vr0          DISABLED fe80::215:f2ff:fe3d:ac91    None       1      0

Looks good.

> Furthermore, I also happen to see these logs in stdout:
> 
> [ 2007/12/04 05:26:47 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:26:47 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:26:47 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found
> [ 2007/12/04 05:28:39 WARNING xorp_fea FEA ] proto_socket_read() failed: RX
> packet from ::1 to ::1 pif_index 4: no vif found

That's interesting. It looks like some IPv6 traffic is received on the
loopback interface. This might be harmless, but it could be a side
effect of some other problem. If you run tcpdump on the lo0
interface, can you figure out what kind of traffic is that.

> On Dec 4, 2007 5:00 PM, Pavlin Radoslavov <pavlin at icir.org> wrote:
> 
> > Router 1 is missing the "mld" configuration.
> >
> > After you add it, please send the output of the following commands
> > when running them on the router directly connected to the receiver:
> >
> > show mld interface
> 
> 
> Interface    State    Querier         Timeout Version Groups
> sk0          UP       fe80::219:5bff:fe85:cfc7    None       2      7
> vr0          DISABLED fe80::213:d4ff:fed8:6808    None       1      0

Looks good.

> >
> > show mld interface address
> 
> 
> Interface    PrimaryAddr     SecondaryAddr
> sk0          fe80::219:5bff:fe85:cfc7 2001:ec1:4001:10af::1
> vr0          fe80::213:d4ff:fed8:6808 2001:ec0:4000:beef::1

Looks good.

> >
> > show mld group
> 
> 
> Interface    Group           Source          LastReported Timeout V State
> sk0          ff02::2         ::              fe80::219:5bff:fe85:cfc7
> 243 1     E
> sk0          ff02::d         ::              fe80::219:5bff:fe85:cfc7
> 243 1     E
> sk0          ff02::16        ::              fe80::219:5bff:fe85:cfc7
> 243 1     E
> sk0          ff02::1:ff00:1  ::              fe80::219:5bff:fe85:cfc7
> 243 1     E
> sk0          ff02::1:ff2f:1468 ::              fe80::219:5bff:fe2f:1468
> 248 2     E
> sk0          ff02::1:ff85:cfc7 ::              fe80::219:5bff:fe85:cfc7
> 243 1     E
> sk0          ff02::2:15ba:6cf7 ::              fe80::219:5bff:fe85:cfc7
> 243 1     E

It doesn't seem that the client/receiver with IPv6 address
2001:ec2:4002:fa11:200:24ff:fec4:3235 has joined group ff3e::1234.
So your guess in your original email is correct: there is something
wrong with the receiver so it hasn't joined the multicast group.
Could you run tcpdump to capture all ICMPv6 (incl. MLD) traffic and
confirm that XORP is sending the periodic query messages, but the
receiver itself never sends MLDv2 Join messages.

BTW, what about MLDv1 Join messages? If you configure the client to
join some (*,G) multicast group does it initiate MLDv1 Join?

In any case, please make sure you don't have firewall rules that
stop ICMPv6 (the protocol number used by MLD).
You should check both the receiver and the router connected to it.

Regards,
Pavlin


> >
> >
> > Also, what do you mean by:
> > "there are no "Join" messages sent out from the client side"
> >
> > Do you mean that the client/receiver host didn't send MLDv2 Join, or
> > that the Router connected to it didn't send PIM-SSM Join toward the
> > sender?
> >
> 
> I was referring to the client/receiver host not sending an MLDv2 Join (upon
> entering the (S,G) parameters and pressing OK) to the upstream router
> connected to it.
> 
> >
> > Regards,
> > Pavlin
> >
> > Hansi < hantongs at gmail.com> wrote:
> >
> > > Hello All,
> > >
> > > I'm currently exploring ipv6 multicast specifically pim-ssm. The network
> > > topology setup I have is shown below:
> > >
> > >             +----------+                     +----------+
> > >             | Source/  +---------------------+          |
> > >             | Sender   |                sk0  | Router 1 |
> > >             +----------+                     +-----+----+
> > >                                                    | vr0
> > >                                                    |
> > >                                                    |
> > >                                                    |
> > >                                                    |
> > >                                                    |
> > >                                                    | vr0
> > >             +----------+                     +-----+----+
> > >             | Client/  +---------------------+          |
> > >             | Receiver |                sk0  | Router 2 |
> > >             +----------+                     +----------+
> > >
> > >
> > >
> > > Details:
> > >
> > > Source
> > > OS: FreeBSD 6.2
> > > Streaming Server: vlc-0.8.6_c
> > > sis1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> > >         options=8<VLAN_MTU>
> > >         inet6 fe80::200:24ff:fec4:3235%sis1 prefixlen 64 scopeid 0x2
> > >         inet 192.168.0.2 netmask 0xffffff00 broadcast 192.168.0.255
> > >         inet6 2001:ec2:4002:fa11:200:24ff:fec4:3235 prefixlen 64
> > autoconf
> > >         ether 00:00:24:c4:32:35
> > >         media: Ethernet autoselect (100baseTX <full-duplex>)
> > >
> > > Router 1
> > > OS: FreeBSD 6.2
> > > sk0: flags=8a43<UP,BROADCAST,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu
> > 1500
> > >         options=b<RXCSUM,TXCSUM,VLAN_MTU>
> > >         inet6 fe80::219:5bff:fe2f:146a%sk0 prefixlen 64 scopeid 0x1
> > >         inet 192.168.0.1 netmask 0xffffff00 broadcast 192.168.0.255
> > >         inet6 2001:ec2:4002:fa11::1 prefixlen 64
> > >         ether 00:19:5b:2f:14:6a
> > >         media: Ethernet autoselect (100baseTX <full-duplex,flag0,flag1>)
> >
> > >         status: active
> > > vr0: flags=8a43<UP,BROADCAST,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu
> > 1500
> > >         inet6 fe80::215:f2ff:fe3d:ac91%vr0 prefixlen 64 scopeid 0x2
> > >         inet 200.10.0.2 netmask 0xffffff00 broadcast 200.10.0.255
> > >         inet6 2001:ec0:4000:beef::2 prefixlen 64
> > >         ether 00:15:f2:3d:ac:91
> > >         media: Ethernet autoselect (100baseTX <full-duplex>)
> > >         status: active
> > >
> > > Router 2
> > > OS: FreeBSD 6.2
> > > sk0: flags=8a43<UP,BROADCAST,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu
> > 1500
> > >         options=b<RXCSUM,TXCSUM,VLAN_MTU>
> > >         inet6 fe80::219:5bff:fe85:cfc7%sk0 prefixlen 64 scopeid 0x1
> > >         inet 172.16.0.1 netmask 0xffff0000 broadcast 172.16.255.255
> > >         inet6 2001:ec1:4001:10af::1 prefixlen 64
> > >         ether 00:19:5b:85:cf:c7
> > >         media: Ethernet autoselect (1000baseTX
> > <full-duplex,flag0,flag1>)
> > >         status: active
> > > vr0: flags=8a43<UP,BROADCAST,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu
> > 1500
> > >         inet6 fe80::213:d4ff:fed8:6808%vr0 prefixlen 64 scopeid 0x2
> > >         inet 200.10.0.1 netmask 0xffffff00 broadcast 200.10.0.255
> > >         inet6 2001:ec0:4000:beef::1 prefixlen 64
> > >         ether 00:13:d4:d8:68:08
> > >         media: Ethernet autoselect (100baseTX <full-duplex>)
> > >         status: active
> > >
> > > Client
> > > OS: Linux Ubuntu Fiesty Fawn 7.04  2.6.20-16-generic
> > > Streaming Client: vlc-0.8.6_c
> > > eth0      Link encap:Ethernet  HWaddr 00:19:5B:2F:14:68
> > >           inet addr:172.16.0.2  Bcast:172.16.255.255  Mask:255.255.0.0
> > >           inet6 addr: 2001:ec1:4001:10af:219:5bff:fe2f:1468/64
> > Scope:Global
> > >           inet6 addr: fe80::219:5bff:fe2f:1468/64 Scope:Link
> > >           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
> > >           RX packets:71141 errors:0 dropped:0 overruns:0 frame:0
> > >           TX packets:103425 errors:0 dropped:0 overruns:0 carrier:0
> > >           collisions:0 txqueuelen:1000
> > >           RX bytes:6304908 ( 6.0 MiB)  TX bytes:137947821 (131.5 MiB)
> > >           Interrupt:17
> > >
> > >
> > > The problem is that I can't seem to recieve multicast streams from the
> > > streaming server. :( I haven't encountered issues w/ regards to ipv4 ssm
> >
> > > multicast. The syntax used for the vlc streaming server is: vlc -vvv
> > > video.xyz - -ipv6 - -sout udp:[ff3e::1234] - -ttl 12. I also tried
> > placing a
> > > backslash (\) character before each bracket but to no avail.  I
> > configured
> > > the client through vlc's graphical interface with
> > > udp://[2001:ec2:4002:fa11:200:24ff:fec4:3235]@[ff3e::1234].
> > > What I came to notice though is that there are no "Join" messages sent
> > out
> > > from the client side to indicate that it intends to join a channel, in
> > this
> > > case (2001:ec2:4002:fa11:200:24ff:fec4:3235, ff3e::1234). I suspect that
> > the
> > > it could be the client not initiating any "join" activity to a multicast
> >
> > > group although I can't also rule out that I overlooked something in my
> > XORP
> > > configuration file. I would appreciate if someone could lend me a hand
> > here.
> > >
> > > Attached is my xorp config file.
> > >
> > > Thanks.
> > > _______________________________________________
> > > Xorp-users mailing list
> > > Xorp-users at xorp.org
> > > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users
> >
> _______________________________________________
> Xorp-users mailing list
> Xorp-users at xorp.org
> http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users



More information about the Xorp-users mailing list