Neil Bothwick <neil@(protected):

> On Thu, 03 Apr 2008 06:51:28 +0200, Michael Schmarck wrote:
>
>> But nonetheless, there's still the risk that the KILL has
>> destroyed the application database (sort of - more correctly:
>> that the application and its database was in a "non consistent"
>> state when it received the signal).
>
> Yes, but in that case the application has already failed, otherwise it
> would have shut down on TERM.

Maybe it would have recoverd in "due time". The chances are pretty
slim, but they are >0.

> Emergency shutdowns aren't about
> eliminating any problems in the case of a serious system hang, they are
> about minimising such damage.

Absolutely correct! But Liviu asked, if there's a potential risk
to the system.

My answer is: Yes, there is! It is pretty low (for the reasons
you mentioned), but it is not 0.

Michael

--
gentoo-user@(protected)