[rabbitmq-discuss] newbie with data feed application to Rabbit

[Jerry Kuch]

Hi, jdepp:

On Mon, Jan 14, 2013 at 10:43 AM, jdepp <j_depp_99 at yahoo.com> wrote:

> if you don't mind me asking a couple more just based on definitions - how
> is
> the routing key associated with a queue; is it just a sort of identifier to
> pass a msg to a specific queue?
> Sort of attaching a label (routing key) to a paper that needs to be filed
> and the queue is the folder that will hold the paper.
> Also what exactly is the virtual host for then within a broker. I am sorry
> about these questions, but I get a better understanding when explained in
> simple words.
>

The tutorials here should clarify some things:

http://www.rabbitmq.com/getstarted.html

Your understanding of routing keys is in the ballpark.  In brief:

A producer publishes to *exchanges*.  Queues are bound to exchanges and
these queues will receive messages published to the bound exchange
depending on:

   1. The type of the exchange
   2. The routing key used on message (if appropriate for exchange type)
   3. The binding key associated with the binding of the queue to the
   exchange

The tutorial examples show you the behaviors of a couple of different
exchange types and how they interact with routing and binding keys.  In
particular the "publish/subscribe," "routing," and "topics" examples show
how choice of exchange type and keys support fairly rich and flexible
routing of your messages.

A virtual host is just a way of partitioning up the resources of a RabbitMQ
broker.  Users can be given access to particular vhosts, and a given vhost
provides a sealed namespace fo the queues/exchanges/routings/etc. that
exist within it.  Thus queue "q_foo" in vhost "host_blah" and "q_foo" in
vhost "host_blarg" are different entities, with access to them governed
separately.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20130114/fc7c4926/attachment.htm>