[rabbitmq-discuss] Protection against slow consumers

[Matthias Radestock]

Jiri,

On 01/11/10 16:33, jiri at krutil.com wrote:
> I see what you mean, but what if the subscriber uses a large prefetch
> buffer and consumes without ever acking?

Currently unacked messages are held in memory. Changing that is on our 
todo list.

> Also, even if long queues get persisted to disk, this consumes resources
> and the queues can't grow forever.
>
> Is there no way how can we limit the queue size or message time to live?

Given that your queues are exclusive and thus get deleted on consumer 
disconnect, presumably your concern here is that while consumers are 
connected queues could still grow to the point where disk space is 
exhausted. Right?

If so then per-queue message TTLs might be an answer, which is a new 
feature we are introducing in the next release.


Regards,

Matthias.