Restarting MD run using OT does not use saved wave functions

[Christopher O'Brien]

I have been using CP2K to make very long DFT/MD runs that must be stopped 
and restarted. I am restarting the calculations using the '-read_restart' 
command-line option of CP2K. When restarting runs using OT for the SCF 
calculations, I notice that the first SCF-cycle after restarting does not 
converge, even after 100 steps.

This indicates that CP2K is not reading the WFN file upon restarting. Thnd 
restarts proceed smoothly. 

How can OT be made to read the WFN file?

My observations:

   - Manually telling CP2K to use the WFN file results in errors. It says 
   that the number of MOs do not match what is in the WFN file.
   - Using diagonalization algorithms for SCF, this behavior does not occur 
   and restarts proceed smoothly. 
   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150731/06f31439/attachment.htm>