[CIG-SHORT] Re: GeoFEST runs for strike-slip model (Jiangning L?)

[Jiangning Lu]

Hello, Charles

I got your files and tried to run them here, the result is the same:
the 1km case works well, but 500m and 250m cases still don't go.
I compared the 1km *.jpl files made by you and me, they are exactly the same.
It seems that the GeoFEST version I use just can't handle models with large
element numeber.

Probably GeoFEST at MIT need to be updated...
The GeoFEST version I use is:
GeoFEST 4.5
PYRAMID  DEMO Version 1.9,
The error info:

PYRAMID Unstructured AMR Library
DEMO Version 1.9,  9/ 6/2006 at 14 hrs 33 min 14 sec
Initialized:    32 processors using double precision reals
************************************************
Welcome to the Geophysical Finite Element Simulation Tool (GeoFEST 4.5)
(/home/jlu/bin/geofestmpi)
PYRAMID Current Time:   1.040000000016583E-004

p1_16053:  p4_error: net_recv read:  probable EOF on socket: 1
p3_8838:  p4_error: net_recv read:  probable EOF on socket: 1
p2_8816:  p4_error: net_recv read:  probable EOF on socket: 1
p5_5619:  p4_error: net_recv read:  probable EOF on socket: 1
.............

Jiangning

Quoting "Charles D. Norton" <Charles.D.Norton at jpl.nasa.gov>:

> Hello,
>
> The code seems to be running fine in parallel with my version of  
> GeoFEST-4.5 (no AMR) and the input
> files I've generated. I'm using the latest version of Pyramid, so 
> that  may be an issue.
>
> I'll transfer the input files, and the version of GFMeshParse that I  
> used as indicated below. If this still does
> not work then we can talk about how to upgrade your version of GeoFEST.