[CIG-LONG] Tracers

[Leonardo Cruz]

It seems that the problem was over partitioning. We were using two nodes with 8 cores each. Increasing the resolution or decreasing the number of cores solved the problem. 
On a related issue, is there a maximum number of tracer particles that you can add to the input file? I seem to be hitting a wall after 40 particles.
Thanks!
Leonardo

----- Original Message -----
From: "Walter Landry" <wlandry at caltech.edu>
To: leocruz at stanford.edu
Cc: cig-long at geodynamics.org
Sent: Thursday, September 12, 2013 1:29:56 PM
Subject: Re: [CIG-LONG] Tracers

Back to the list.

Leonardo  Cruz <leocruz at stanford.edu> wrote:
> We have issues with the our cluster. We can run all the cookbook
> files using only 1 node but only three files (extension3D.json,
> non_newtonian.json, and thermal_file.json) if we run 2+ nodes. Do
> you know if there's a chance that some parameter or variable in the
> input file could be determining how many cores it can run on?

All of the cookbooks should run fine with multiple nodes.  You might
run into a problem with a large number of nodes for some of the small
examples, but 2 nodes should work for all of the examples.

Cheers,
Walter Landry