Subject: Re: multicast daemon design
To: Jonathan Stone <jonathan@dsg.stanford.edu>
From: Bill Studenmund <wrstuden@netbsd.org>
List: tech-net
Date: 04/29/2005 12:46:59
--H5pgEA8DvTwLpheO
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri, Apr 29, 2005 at 11:39:24AM -0700, Jonathan Stone wrote:
>=20
> In message <rmir7gth44q.fsf@fnord.ir.bbn.com>,
> Greg Troxel writes:
>=20
> >Jonathan Stone <jonathan@dsg.stanford.edu> writes:
> >
> >> Bill: does that buy you anything, if other OSes have a similar limit?
> >
> >He's using v6, so no, because that already uses a linked list.
>=20
> Bill seems to be interested in portability, and having OpenSLP
> continue to work on other OSes, so I naturally assumed he'd have to
> not break OpenSLP over IPv4. OTOH I may be wrong.

I expect that any patches which break OpenSLP on other platforms will not=
=20
be well-received. :-)

Just to be clear, for IPv4, SLP uses one multicast address. So group=20
memberhsip limits aren't an issue.

For IPv6, it uses groups based on a hash of the address. I just looked at=
=20
the code more, and because of how one mask is set up, it will join up to=20
32768 groups (The lower two bytes of the hash are or'd with 0x1000 as part=
=20
of making the address).

Take care,

Bill

--H5pgEA8DvTwLpheO
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)

iD8DBQFCco8yWz+3JHUci9cRAh5yAJ9TxF5y/AHmTHTn7Vdo7wtkKFdqpQCfXO/j
Bps0wObDcrZInLR964fBc+g=
=ET3T
-----END PGP SIGNATURE-----

--H5pgEA8DvTwLpheO--