grads-2.0.a1 segmentation fault

[Arlindo da Silva]

On Thu, May 1, 2008 at 10:37 AM, Matt Alonso <matt.alonso at gmail.com> wrote:

> Was anyone ever able to find a solution to this issue?  I think I am
> experiencing a similar problem using gradsnc v1.9b4.  I have a 13MB netCDF
> file (samples can be found here,
> http://www.cdc.noaa.gov/cgi-bin/db_search/DBListFiles.pl?did=34&tid=20349&vid=1246)
> which contains Reanalysis rainfall rate data.
>
> My script - based off of a script I found in the list archives - attempts
> to loop over all points in the file for all timesteps (366) and at each
> point (192x94) it will write ASCII output to a file.  If I run it for less
> than 4 timesteps it works like a charm but if I set it to use 4 or more
> timesteps it will always fail on the 4th one at point 161,69 running the
> command to display the value.  My first instinct was that the data at that
> particular point/time was somehow corrupted but subsequent tests showed that
> it doesn't matter what timestep I start at it will always fail the 4th time
> through at that exact X,Y.
>
> Anyone have any ideas about what the issue might be?  Thanks in advance.
>


Have you monitored the memory usage as your script progresses? There is a
memory leak in the sdf part of the code that may cause you to use up all the
available memory. One alternative is to create a ctl for your netcdf file:

   http://grads.iges.org/grads/gadoc/SDFdescriptorfile.html

which exercises a very different part of the GrADS code and have no memory
leak as far as I know. See if this works.

  Arlindo.







--
Arlindo da Silva
dasilva at alum.mit.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20080501/096d5e93/attachment.html