activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sridhar Komandur" <anon2...@komandur.com>
Subject Re: broker/discovery client help
Date Tue, 23 May 2006 18:51:15 GMT
Paul,

Reg your China/Chicago cluster examples.

If you plan on using multicast based discovery, note that
multicast packets don't travel natively over the Internet.

You need to work with your network/system admins - perhaps you
already have a tunnel between your sites with your own
private addressing.

cheers
- Sridhar

On 5/22/06, Paul Tomsic <ptomsic@yahoo.com> wrote:
>
> thanks.  the broker's would not be dynamic and
> changing, so once they're set up, it's a done deal.
> ideally, i'm just using discovery to enable ease in
> bringing up new brokers as is necessary.
>
>
> --- "Christopher G. Stach II" <cgs@ldsys.net> wrote:
>
> > Paul Tomsic wrote:
> > > thanks for the information.
> > > i think i'm confused, though, as to how
> > > consumers/producers would specify a specific
> > cluster
> > > and a machine?
> > >
> > > so say i've got consumers/producers in china that
> > need
> > > to connect to a cluster running in chicago named
> > > "foo".
> > > I've also got another cluster of brokers in
> > chicago
> > > running as "bar"
> > >
> > > how would the consumers/producers specify which
> > > cluster to connect to?
> > >
> > > would my consumers connect as
> > >
> > > "discovery:multicast://chicago.mycompany.com/foo"
> > >
> > > -and-
> > >
> > > "discovery:multicast://chicago.mycompany.com/bar"
> > >
> > > ??
> > >
> > > thanks again for the help..
> > >
> > > paul
> > >
> >
> > They subscribe to the same multicast group, your
> > multicast routers would
> > forward the packets, and as long as the TTLs are
> > high enough, they would
> > reach the other side.
> >
> >
> http://www.google.com/search?hl=en&q=ip+multicast+routing
> >
> > Since you said you have two "clusters", maybe you
> > should just set up
> > unicast TCP connectors and use those.  Unless you
> > have a bunch of
> > machines in each group, or they're very dynamic and
> > changing, maybe
> > discovery isn't even the ideal setup.
> >
> > --
> > Christopher G. Stach II
> >
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message