On 8-6-2009 16:25 Aleksandar Ivanisevic wrote:
Arjen van der Meijden
writes:
As soon as those second-queue-consumers exit however, they don't
consume much messages. Most of the time it seems that only one is
actually doing any work and the other three are idle, even if the
queue has more than
Arjen van der Meijden
writes:
> As soon as those second-queue-consumers exit however, they don't
> consume much messages. Most of the time it seems that only one is
> actually doing any work and the other three are idle, even if the
> queue has more than 100 messages. The same holds for the
> thi
Well, I seem to have fixed it... I completely overlooked that the
default for the prefetch is actually larger than 1. Setting it to 1
fixes the problem. Still it seems to be a bit of a waste that several
newly connected consumers can be idle while because a "old" one that is
processing, is actu
Hello list,
In a somewhat complicated set of consumption-processing, I have three
queues. One to indicate a new job, which is consumed by a single
long-running consumer that sends messages to a second queue.
That second queue is consumed by four consumers that kill themselves
after having proc
