[Pw_forum] restart problem

[Bhagawan Sahu]

Dear pwscfusers,

I am getting the following  error in a SCF restart run.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from seqopn : error #         1
     can't open a connected unit
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...

     Forces acting on atoms (Ry/au):


 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from seqopn : error #         1
     can't open a connected unit
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from seqopn : error #         1
     can't open a connected unit
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from seqopn : error #         1
     can't open a connected unit
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

This happens when I restart the ionic relaxation calculation from  a 
non-converged job 
(did not change anything in the input except  restart_mode= 'restart'). 

The non-converged job had

&control
    calculation='relax'
    restart_mode='from_scratch'

which was changed to

&control
    calculation='relax'
    restart_mode='restart',

for restarting from the previous run.

Is there a bug in the restart routine?  Pl. note the same code compiled on 
POWER5 (IBM) machine do not have such restart problem.

Sahu