On Friday, September 2, 2011, Alan Buxey <A.L.M.Buxey@lboro.ac.uk> wrote:
> I meant that if you use eg monit then you can have a service recovery rather than a corner case killing off your radius daemon in middle of night
>

Oh, definitely. We'd do that as a failsafe anyway. My main question is whether this is failing by design? If so, is there a best practice to avoid the crash (or something like monit to recover from a crash)? And, if not, is it a resolvable bug?

Thanks,

Norman