[CIG-LONG] Error: The side is moving in the wrong direction

[Walter Landry]

Mark Brandon <mark.brandon at yale.edu> wrote:
> I am having a problem with job that runs nearly to completion but then
> aborts with the following error:
> 
> The side is moving in the wrong direction.
> 750 747.272 750 739.772 750
> p0_30304:  p4_error: interrupt SIGx: 6
> 
> I included an expanded listing (below) of the log information where
> this error was generated. The full versions of the input.xml and log
> files are attched. As you can see from the log below, the program has
> already converged to a solution (after 1237 iterations), so it is
> unclear why it would abort at this point.

Gale has computed a solution to the first time step, but has not
finished advecting the grid.  In the course of advecting the grid, it
has found a point that it does not know how to interpolate to.  You
have set the right, left, bottom, front, and back sides to be static.
Since you did not set the top to be static, you probably need to set
LeftTop, RightTop, TopFront, TopBack, LeftTopFront, LeftTopBack,
RightTopFront, and RightTopBack to be static as well.

But I am a bit unsure about this, because I am not certain what you
are trying to accomplish with your boundary conditions.  Could you
explain what you are trying to simulate?

Cheers,
Walter Landry
walter at geodynamics.org