Restating failure

[Molcasito]

Dear Turbomole users and experts,

I have done a  calculation using CC2 and linear response to characterize the excited states energies.
for one case I have  increased the number of roots   from the previous calculation which finished fine
(but with less roots). When I try to re-start the calculation the Turbomole program is not doing anything
I have kept the restart.cc file and other files as well in the working directory.

The program just prints this message and does nothing
after 12 h 

Is any problem in the restarting code ? if not .......At what point I could restart and how  the CC2 calculation if the CC2
function was already converged, and the linear response vectors too.

   calculation of zero-order integrals is skipped because they were already
   available on disk

   restarted calculation begins with wave function model nb. 1 which is CC2


   **************************************************************************
   *                                                                        *
   *          OPTIMIZATION OF THE GROUND STATE CLUSTER AMPLITUDES           *
   *                                                                        *
   **************************************************************************
     restart CC2 from amplitudes found on the R0  file.


   Iter.   CC2 energy       Norm(Omega)    Norm(t1)  Norm(t2)     cpu    wall
   ...........................................................................

  ----------------------------------------------------------------------------
  hard_restart flag is disabled because the check sum for the
  css factor "$sccss" differs from the reference value:
  reference value:    0.525046681378430
  computed value :     1.00000000000000
  relative error :    0.904592554274703
  tolerance      :    0.100000000000000E-11
  the calculation will proceed...
  ----------------------------------------------------------------------------

Thanks in advanced

Oscar

[Arnim]

Hi Oscar,

it seems, that in this case there is an inconsistency with the informations in the restart.cc file.
If you remove restart.cc and run ricc2 again, then the converged cluster amplitudes on disk
will be read in. Some iterations are repeated in this case, which could be avoided if restart.cc
works, but maybe it is safer.

Cheers,
Arnim