Failed Simulations due to ''Insufficient memory allocation for problem'' Answered

Hi, 

I am running a simulation regarding wave propagation in double layered tissue mimicking material. The thickness of the first layer changes from 500 um to 1.5 mm, the grid spacing for all three simulations was set to be 0.1428571E-04 m. The simulation was successful on the thickness of 1.0 mm and 1.5 mm, but failed several time on 500 um case. I download the console log to investigate the problem, it says " Insufficient memory allocation for problem " (see console log below).

Why the other two cases could be successfully solved except the first one ? How could I know the memory limitation for the console before I submit the job? 

 

console.log file

Memory allocation:

Numeric data- 60 million words ( 480 Mbyte)

Character data- 2500 thousand words ( 50 Mbyte)

**** PZFlex has detected an error****

****Subroutines tracked are: cormgr resdta sckttri2 plofld gphinp jobctl flex

****Error occurred in subroutine cormgr ****

****Error message****

Insufficient memory allocation for problem

sh: 1: cannot open /dev/nul: No such file

WARNING: There were 9 warnings generated during this run

 

1 comment

  • Official comment

    Hi there,

    When you are sweeping variables, ensure that the one requiring the largest memory is the starting value as this is what the estimate is based on and if the estimate doesn't estimate enough memory the job fails.

     

    A 500um layer will increase your timestep so you will need more memory, so set start value = 500um end value 1.5mm and this should work.

     

    Best Regards,

    Chloe.

    Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post