run time error forrtl: error (78): process killed (SIGTERM)

[pduarte]

Hi,

I am having the above run time error and I have seen already some posts about it but found no "easy" explanation for it. I am suspecting the model initial and boundary conditions files and I have a question about these. I am running a coupled physical-biogeochemical simulation and I added several variables to both the initial and boundary files. However, I noticed that these variables do not contain all the attributes of the other tracers in the netcdf file. I wonder if this may cause some problems messing with the initialization of something and thereafter causing the problems I am experiencing.

For example, the metadata corresponding to chlorophyll looks like this:
double chlorophyll(ocean_time, s_rho, eta_rho, xi_rho) ;
chlorophyll:_FillValue = 9.99999993381581e+36 ;

Whereas that corresponding to salt looks like this:
float salt(ocean_time, s_rho, eta_rho, xi_rho) ;
salt:long_name = "salinity" ;
salt:time = "ocean_time" ;
salt:coordinates = "lon_rho lat_rho s_rho ocean_time" ;
salt:field = "salinity, scalar, series" ;
salt:_FillValue = 1.e+37f ;

The error does not occur always at the same point so it seems to cause some random effect which suggests some variable not properly initialized. However, I have used all possible compile option and found nothing....

[kate]

If the attributes were a problem, I would expect a (slighty) more informative error message at the time of reading the file, not some random time later.

You're not running out of memory, are you?

[pduarte]

Hi Kate,

Thanks for your answers. No, I don't think there is a memory problem. It seems one of those difficult errors without a consistent pattern. I guess I will have to move further back in time in the repository I am using and see if I can be sure about when the error starts to show up...

Pedro