[rabbitmq-discuss] Limiting the number of non-ack messages cause of subsystem

Wed Oct 26 17:32:26 BST 2011

Hi Johannes,

On Sun, Oct 23, 2011 at 02:59:25AM -0700, jhiemer wrote:
> yeah that's the other way I was thinking about. The problem is, that
> there might be different execution times for each thread. So imagine
> the following situation:
> 
> - 3 Threads runtime 10 seconds
> - 4 Threads runtime 25 seconds
> - 3 Threats runtime 200 seconds
> 
> Discarding the Consumer for the time till all the Threads are finally
> executed would result in an waiting time of 200 seconds. With the ACK
> approach I could automatically assign an new thread after 10 seconds.
> So from your side, that would be proper way? Are there any samples for
> it?

Erm, I'm afraid you've lost me. Does a new thread result in a new "slot"
immediately?

Matthew