[rabbitmq-discuss] Maximum erlang processes & dead channels

[Alex Graul]

Hello,

I'm about to give some of our employees a framework for accessing and  
manipulating data from various resources in Ruby, some of which are  
accessed over AMQP/RabbitMQ 1.4.0. By the nature of what they're doing  
and how they're going to be using it, these scripts are likely to be  
crashed a heck of a lot of time, which usually wouldn't be a problem.  
However during heavy testing I ran into the errors consistent with the  
'too many processes' problem (see: http://www.nabble.com/Too-many-processes-td19706977.html 
  ). I think hard crashes mean the channel doesn't get closed, so  
eventually the max number of erlang processes is hit. Is there some  
mechanism by which these consumers/channels can be auto-timed out to  
stop this happening? Apologies if I'm missing something obvious.

Thanks for all the incredible work on RabbitMQ, it's a lovely system  
to work with.
-alex