[rabbitmq-discuss] Question about initial cluster connect & client-side cluster awareness

Holger Hoffstätte holger at wizards.de
Thu Sep 18 17:10:00 BST 2008

I've been experimenting with a cluster of rabbits (keeping the current
limitations in mind) and was wondering about the client side of things.
How should the problem of the initial connection be approached? As far as
I can tell there are no built-in mechanisms in place to have a client

- dynamically discover a broker or a cluster of nodes on a network

- try to connect to a number of hosts in order to get an initial connection

- be aware of the available cluster nodes via the known_hosts field (as
described in the docs) so that when that node dies, the client can
automatically reconnect to a live member

Would this be an accurate description or am I missing something? I was
happy to see the known_hosts returned (a great first step!) but I am for
now more concerned about initial connection bootstrap in a loosely coupled
environment where the hosts may come and go.

Would love to hear any thoughts on this.


PS: maybe soemwhat OT but as far as autodiscovery is concerned, Erlang/OTP
 itself does not seem to have any built-in mechanisms for this, right?
Like a node-aware service registry or something. DNS-SD aka zeroconf would
work and be cross-platform/standards-compliant etc. but is not really
sufficient, IMHO.

More information about the rabbitmq-discuss mailing list