This project has moved. For the latest updates, please go here.

mc.exe timeout/crash for long-running process

Coordinator
Jun 1, 2011 at 7:20 PM
Edited Jun 1, 2011 at 7:23 PM

A summary of email conversation with Rob:

Rob started 10M photons it at ~5:15pm yesterday, expecting the simulation to take 25 hours. Here is a snapshot:

http://i52.tinypic.com/2iapevr.jpg

Coordinator
Jun 1, 2011 at 8:33 PM
Edited Jun 1, 2011 at 8:43 PM

An update from Carole:

After scanning the input file, I see one potential problem. The input file specifies the creation of two databases and those file are probably getting *very* large for 10M.

I have edited your 10K input file and made that option nil and attached it (Note to self, change default infile to not specify the creation of databases).

Please update N and give this a try for 1M and 10M.

Carole

Aside: not sure how to deal with attachments and images in this discussion tab - maybe we should adopt a different online service.

In the meantime, I recommend using codepaste.net and tinypic.com for code and screenshots, respectvely.

Here's Carole's XML code: http://codepaste.net/j51na1

Jun 1, 2011 at 10:16 PM

I started a new launch of 1M photons at 2:15pm (with Carole's new .xml file) and it seems on track to finish around 4:45pm (it just announced 40% completion).  I'll post back soon when/if it finishes.

Jun 1, 2011 at 10:19 PM

Also, in response to Carole... the two databases left behind after the 10M launch are 95MB and 78MB.

I'm not sure if that's what you guys call "*very* large"... I think that's a few albums worth of .mp3s though.

Jun 2, 2011 at 1:54 AM

The simulation completed successfully and I started a 10M photon launch.  In the meantime, I attempted to utilize David's vts_parser (v2) and you can see the results here.  I'm not sure what I'm looking at... does this look correct?

 

Also, I recommend imgur.com for uploading pictures (tinypic.com wouldn't upload my .jpgs for some reason) :D

Coordinator
Jun 2, 2011 at 3:37 AM
Thanks for the tip!

You're looking at 2D maps of R(rho,omega). Vertical axis is the first variable, rho; horizontal axis is the second, omega. The amplitude map is plotted
in log(A), phase is in radians. You can look at all the results generated by typing 'whos' in Matlab. Note the 'results' variable.
Typing 'results' will show you all outputs (detectors) generated by your simulation run, including the full 2D arrays. The midpoints of the
bins are pre-calculated for you, so that you can access values at a constant rho or omega, for example,

'figure; plot(results.ROfRhoAndOmega.Omega_Midpoints, results.ROfRhoAndOmega.Amplitude(10, :));'

will plot the amplitude at the 10th rho bin, which would be at

'results.ROfRhoAndOmega.Rho_Midpoints(10)'

I typed that from my iPhone, so caveat emptor...let me know how it goes.
Coordinator
Jun 2, 2011 at 4:23 PM
This discussion has been copied to a work item. Click here to go to the work item and continue the discussion.