[CP2K:388] Re: Block after " Adding QM/MM electrostatic....."

[Teodoro Laino]

brief summary:

Luca sent me his files yesterday and I had a look to them.. there  
were a couple of bug fixes (one for print level HIGH and another for  
the fist neighbor lists (this one just a memory leak that didn't show  
up previously))..

Anyway looks like the step where it's breaking is a little bit random  
(Luca rerun the system and he saw failures at step 95, 100 and 93)..
I'm still looking deeply with a parallel debugging into the run but I  
think that's some problem related to the machine (memory very  
probably or compiler??)

Teo

p.s.: will let you know..

On 5 Nov 2007, at 09:04, Axel wrote:

>
> hi guys,
>
> hard to tell anything without seeing the files, but have you checked,
> whether the box is still large enough to contain the QM subsystem
> properly?
> for the number of QM atoms, i would normally expect a somewhat larger
> box
> (unless your atoms comprise a compact sphere).
>
> please note the requirements of the tuckerman poisson solver (2x
> charge
> distribution sphere hast to fix into the box).
>
> ciao,
>    axel.
>
>
>
>
>
> >