[rabbitmq-discuss] Silent crash causes persistent durable message loss

[Tim Watson]

On 21/05/2012 22:12, Matthias Radestock wrote:
> On 21/05/12 22:08, Tim Watson wrote:
>> Even with an OOM failure, we should at least be able to see an
>> erl_crash.dump file somewhere.
>
> I said 'oom *killer*'. http://lwn.net/Articles/317814/
>
> Matthias.

Owch, that's gotta hurt. What SIGNAL does OOM killer use? Can it provoke 
a regular core dump?