BerkeleyDB locking subsystem

[Peter Corlett]

On 31 Jul 2006, at 18:19, Thomas Busch wrote:
[...]
> that's exactly what I did but perl doesn't allow to catch the kill -9
> signal (probably with a reason). My problem occurs when a users
> abruptly ends a CGI script.

It's not a Perl thing. *No* Unix program can trap SIGKILL. Yes, there  
is a very good reason, which will become obvious once you write a  
runaway script that's trapped a load of signals.

I suspect you're actually getting SIGPIPE from your CGI script when  
you're trying to print stuff to a browser that's gone away. Are you  
trapping that signal?