All things related to multicast

[Marshall Eubanks <>]

Hello;

On May 30, 2006, at 2:15 PM, Matthew Davy wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> On May 30, 2006, at 1:59 PM, Bill Owens wrote:
>
> > On Tue, May 30, 2006 at 01:11:55PM -0400, Matthew Davy wrote:
> > > So based on http://www.iana.org/assignments/multicast-addresses,
> > > which seems to be the authoritative source of multicast address
> > > assignment, the IANA reserved space is:
> > >
> > > 224.1.0.0 - 224.1.0.37  Reserved                            [IANA]
> > > 224.1.0.39 - 224.1.0.255 Reserved                           [IANA]
> > > 224.1.5.0 - 224.1.255.255 Reserved                          [IANA]
> > > 224.3.0.64 - 224.3.255.255     Reserved                         
> > > [IANA]
> > > 224.5.0.0 - 224.251.255.255  Reserved                           
> > > [IANA]
> > > 225.0.0.0 - 231.255.255.255 Reserved                 [IANA]
> > > 234.0.0.0 - 238.255.255.255 Reserved                     [IANA]
> >
> > A quick count, we have 2739 MSDP SAs in our NYC router just now,  
> > and about 690 of them would be blocked by this filter. I have no  
> > idea whether they are in the table because of unintentional  
> > leakage, or stupid application writers, or perhaps are actually  
> > being used. The largest single group in the reserved range is  
> > 234.21.81.1 has 271 sources, and is used by LimeWire. Of course,  
> > the largest group is 'legit' in that it has been approved by IANA,  
> > but probably shouldn't be getting out to the world; 224.1.0.38  
> > (that's Retrospect, the backup software, to save everyone from  
> > having to look). 313 sources. And our old friend 224.0.1.76, the  
> > IAPP group, is hanging in there with 111 access points chiming away.
>
> Thanks for the quick analysis !    This would clearly block a lot  
> of SAs that are being accepted today.   I wonder if users have  
> gotten used to just picking a group randomly out of the vast  
> reserved space and using it

Probably more application writers, but, yes, that is what I think is  
going on.

> instead of using GLOP or SAP or SSM   ?    If so, this is a  
> behavior that should probably change.   But the question is, what  
> we would we break if we implemented this filter ?

Here is what I see from here

There were  1468 SA-Cache Entries
There were     1 Duplicate S,G Entries
There were   354 SA-Cache Groups
There were  1152 SA-Cache Sources
There were   131 SA-Cache RPs
There were    55 SA-Cache ASs

The Most Active Group  is      224.1.0.38 with   296 members
The Most Active Source is   158.102.16.65 with    23 groups
The Most Active RP     is     194.81.46.1 with   321 entries
The Most Active AS     is             786 with   478 entries

There were   276 Groups with only one Sender

First Octet Histogram
Octet 224 had    160 entries or  45.20 %
Octet 225 had      6 entries or   1.69 %
Octet 226 had      1 entries or   0.28 %
Octet 227 had      2 entries or   0.56 %
Octet 228 had      5 entries or   1.41 %
Octet 229 had      1 entries or   0.28 %
Octet 230 had      5 entries or   1.41 %
Octet 231 had      1 entries or   0.28 %
Octet 233 had    159 entries or  44.92 %
Octet 234 had      9 entries or   2.54 %
Octet 236 had      2 entries or   0.56 %
Octet 237 had      1 entries or   0.28 %
Octet 238 had      1 entries or   0.28 %
Octet 239 had      1 entries or   0.28 %

I think that I am being filtered upstream...

Regards
Marshall



> > > IMO this is definitely the "right thing" to do (regardless of the
> > > benefits) and would go a long way to accomplishing all the potential
> > > goals that Bill Owens mentioned.
> >
> > I wasn't advocating for those goals; I was listing them as a  
> > question for the group, to try to ascertain goal or goals we're  
> > aiming at.
>
> Sorry, I didn't mean to indicate you were advocating those.   Just  
> noting that is would accomplish the goals of protecting various  
> things - if those were goals people were interested in.
>
> - - Matt
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.4 (Darwin)
>
> iD8DBQFEfIuplW/4XGQiy+sRAtyuAKCzgklt37FOXmRpu4E3PSYzzb58ngCgzFz2
> JTM1sdw59dlb+wbm8fhDH6A=
> =IiLY
> -----END PGP SIGNATURE-----