I missed part of the story!

On Tue, May 31, 2005 at 05:43:45PM +0100, Patrick Welche wrote:
> It seems my NetBSD-3.99.3/i386 box isn't routing multicast across vlans..
> What have I missed?
> 
> options         MROUTING        # IP multicast routing
> #options        PIM             # Protocol Independent Multicast
> 
> I take it PIM is not necessary?
> 
> 
> ex0: flags=8a63<UP,BROADCAST,NOTRAILERS,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu 
> 1500
>         capabilities=3f00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM_Tx,UDP4CSUM
> _Rx,UDP4CSUM_Tx>
>         enabled=3f00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM_Tx,UDP4CSUM_Rx,U
> DP4CSUM_Tx>
>         address: 00:0a:5e:52:f6:69
>         media: Ethernet autoselect (100baseTX full-duplex)
>         status: active
> vlan1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
>         vlan: 1210 parent: ex0
>         address: 00:0a:5e:52:f6:69
>         inet 192.168.80.62 netmask 0xfffff000 broadcast 192.168.95.255
> vlan3: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
>         vlan: 1485 parent: ex0
>         address: 00:0a:5e:52:f6:69
>         inet 192.168.204.62 netmask 0xfffff000 broadcast 192.168.207.255
> 
> 
> So, ALLMULTI is defined on ex0, and the vlans seem happy with multicast.
> 
> mrouted is running with the default all commented out mrouted.conf file.
> I thought this meant it would forward multicast packets to all interfaces
> with MULTICAST, and indeed mrinfo says
> 
> 127.0.0.1 (localhost) [version 3.8,prune,genid,mtrace]:
>   192.168.80.62 -> 0.0.0.0 (local) [1/1/querier/leaf]
>   192.168.204.62 -> 0.0.0.0 (local) [1/1/querier/leaf]
> 
> 
> The aim is to multicast ghost images. The server is at 192.168.80.12, and
> a client at 192.168.204.227. I see the igmp requests come in on the
> interfaces, and indeed mrouted.cache contains:
> 
> Multicast Routing Cache Table (2 entries)
>  Origin             Mcast-group     CTmr  Age Ptmr IVif Forwvifs
>  192.168.192/20     224.77.0.0        2m   4m    -  3   
> >192.168.204.227
>  192.168.80/20      224.77.1.0        2m   4m    -  1   
> >192.168.80.12
> 
> However, no packets seem to be copied from one vlan to the other, which I
> think is why they each created a different group.

mrouted.dump doesn't mention the Mcast-groups listed in the table above:

vifs_with_neighbors = 0

Virtual Interface Table
Vif  Name  Local-Address                           M  Thr  Rate   Flags

 1  vlan1  192.168.80.62   subnet: 192.168.80/20   1   1      0   querier leaf
                         pkts in : 5
                         pkts out: 0

 3  vlan3  192.168.204.62  subnet: 192.168.192/20  1   1      0   querier leaf
                           groups: 239.255.255.253
                                   224.0.1.60     
                         pkts in : 10
                         pkts out: 0


> There are a few other vlans and the entire routing table is:
> 
> Multicast Routing Table (6 entries)
>  Origin-Subnet      From-Gateway    Metric Tmr In-Vif  Out-Vifs
>  192.168.240/20                        1   150   0    1* 2* 3* 4* 5*
>  192.168.192/20                        1   150   3    0* 1* 2* 4* 5*
>  192.168.80/20                         1   150   1    0* 2* 3* 4* 5*
>  192.168.32/20                         1   150   5    0* 1* 2* 3* 4*
>  192.168.16/20                         1   150   4    0* 1* 2* 3* 5*
>  192.168.0/20                          1   150   2    0* 1* 3* 4* 5*
> 
> and the ipf rules are
> 
> block in log quick all with short
> block in log quick all with ipopts
> pass in quick proto icmp all
> pass out quick proto icmp all
> pass out quick on vlan1 all
> pass in quick on vlan1 all
> pass out quick on vlan3 all
> pass in quick on vlan3 all
> pass in  quick on lo0 all
> pass out quick on lo0 all
> 
> so not exactly "restrictive".
> 
> 
> Cheers,
> 
> Patrick