Cutoff radius question in hse06 calculation

[jade panda]

Dear CP2K users and developers,

I am trying to perform an scf-calculation with the HSE06-functional.
but I get the following warning:

WARNING in hfx_types::: Periodic Hartree Fock calculation requested with use of a truncated or shortrange potential. The cutoff radius is larger than half the minimal cell dimension. This may lead to unphysical total energies. Reduce the cutoff radius in order to avoid possible problems.       

In this paper: J. Chem. Theory Comput. 2009, 5, 3010–3021, I find the reason is that exchange energy calculation will include "a large contribution for an electron and its 'periodic image' if Rc is large enough". For only Γ k-point simulation, the Rc is need to be smaller than L/2 written in the hfx_types.F. If I increase the OMEGA parameter, the cutoff radius decreases and warning disappears. But how large for OMEGA (how small for cutoff radius) is reasonable?

In addition, performing hse06 calculation in VASP before, I usually use default parameter (omega = 0.2 1/Å) and do not consider the supercell size and k-points numbers. Is it wrong, or right?  I don't find more details about this in VASP website or lecture? Can you tell me some advices or information? Thanks very much.

Best regards!

Aolei Wang