[rabbitmq-discuss] What is a good tradeoff for number of connections and channels/connection for a client ?

[Alexandru Scvortov]

Hello Arun,

> We are currently using rabbitmq 1.7.2 with the erlang client.
> We have some code that sits on top of the erlang client library that abstract away the need for the client to actually know anything about the rabbitmq connections or channels.
> Basically, we create a pool of connections, and try to multiplex multiple channels across these pooled connections.

Why are you creating a pool of connections?

Why are you multiplexing channels across them?  It sounds as if you've got
a number of channels and are selecting one at random to execute some
command.  Is that it?


Unless you have some strange requirements, one connection should be
enough.

Channels have a logical role.  For instance, some AMQP methods raise
exceptions that close the channel they were exectued in.

Otherwise, there shouldn't be any reason to have multiple channels.
Specifically, there shouldn't be any performance gains.

Cheers,
Alex