[rabbitmq-discuss] Publisher confirms requires seperate call to get ack?

[Matthias Radestock]

On 07/01/14 19:11, Ryan Brown wrote:
> the Notes section confuses me a bit:
>
> "The broker loses persistent messages if it crashes before said messages
> are written to disk. "

Hopefully that bit isn't confusing/surprising.

> "Under certain conditions, this causes the broker to
> behave in surprising ways...."

The surprising bit, as the example illustrates, is that the broker may 
lose a persistent message *after it has sent it to a client*.

> I assume this indicates that an ack will not be recieved by the
> publisher thus, the publisher should attempt re-publishing. Correct?

That's exactly what it says at the end of the section you quoted...

"If the publisher's channel had been in confirm mode, the publisher 
would not have received an ack for the lost message (since the message 
hadn't been written to disk yet)."

Matthias.