[rabbitmq-discuss] Persister crashes Rabbit

Ben Hood 0x6e6562 at gmail.com
Sat Nov 15 17:29:46 GMT 2008


Ilya,

On Sat, Nov 15, 2008 at 4:48 PM, Ilya Grigorik <ilya at aiderss.com> wrote:
> So the messages don't get flushed out to disk at all? Right now, I'm running
> two publishers and no consumers.. I would expect RabbitMQ to keep a stack of
> messages in memory and overflow to disk - is this wrong?

Persistent messages are made persistent, this is part of the transfer
of responsibility. What you are observing is the fact the messages are
resident in memory as well.

There have been a number of discussions about disk overflow and paging
on this list, so it is feature we'd like to get around to addressing
at some stage.

Valentino has started a design document on this, and Tony has already
started knocking up some experimental code, but, as always, there is
no ETA on this feature.

> - That would also mean that the number of messages we can queue is
> essentially limited by the amount of available RAM?

ATM yes, hence why we have implemented producer flow control (see article).

> - I've seen a few mentions of RDBMS hooks for persistence, how far away are
> we from this?

I don't really recall anything concrete on this topic, apart from the
holy grail of pluggable queues - I would say that we're still a way
off this ATM.

HTH,

Ben




More information about the rabbitmq-discuss mailing list