[rabbitmq-discuss] Is the C client a show-stopper for rabbit?

Alexis Richardson alexis at rabbitmq.com
Tue Feb 15 16:41:11 GMT 2011


Well, at this point I should suggest writing a test harness ;-)


On Tue, Feb 15, 2011 at 3:51 PM, Jiri Krutil <jiri at krutil.com> wrote:
>>>> What's the typical fanout ratio going to be?
>>>
>>> Between 1 : 1000 and 1 : 5000.
>>
>> So if the ingress is, eg, 3,000 mps, the egress might be 3,000,000?
>> That seems high.
>
> Yeah, that's really a lot, but it probably won't be that much. It's hard to
> estimate these things, because the inbound messages that appear very often
> usually have lower fanout ratio. It is not easy to estimate this precisely
> right now.
>
> If the egress would be too high, we could:
> - limit the ingress to some extent (reduce price update rate)
> - reduce parallelism of the calculation, resulting in fewer consumers and
> lower fanout ratio
>
>
>
>
>


More information about the rabbitmq-discuss mailing list