FW: FLUKA Crashes

From: Harine, Katherine (Katherine.J.Harine@nasa.gov)
Date: Fri Oct 14 2005 - 22:17:08 CEST

Next message: Burge, F \(Frances\): "error message"

Previous message: Harine, Katherine: "FW: FLUKA Crashes"
Maybe in reply to: Harine, Katherine: "FW: FLUKA Crashes"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This is a multi-part message in MIME format.

------_=_NextPart_001_01C5D0FC.40B38E96
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

I reran one of these cases on a stand-alone 32-bit Red Hat Linux machine =
and it ran successfully. The machine that some runs crashed on was a =
64-bit SUSE node that was part of a cluster. I had "rsh n03" to get to =
a node and then ran the case which then crashed after several cycles in =
the original attempt that I wrote about earlier.

These crashes are similar to some I had a few weeks ago where runs would =
crash on the clustered nodes but run successfully on stand alone boxes. =
In the past even if the stand alone box used SUSE Linux and were 64-bit, =
they would run fine. For some reason the clustering causes the problem. =
In these recent cases it is interesting because the energies of the =
incident nuclei were much lower than the energies a few weeks ago when =
there were large hadronic cascades. Also the recent runs crashed at the =
same place that appears to be during the initialization of the run, =
rather than the transport itself.

- Katherine

-----Original Message-----
From: Harine, Katherine
Sent: Fri 10/14/2005 11:57 AM
To: giuseppe.battistoni@mi.infn.it; alfredo.ferrari@cern.ch; =
support@fluka.org; discuss@fluka.org
Subject: FLUKA Crashes
=20

I am getting crashes with the .out file ending:

.
.
.
.
   ***** Exp. pair production activated above 0.0000E+00 GeV *****
                                                                         =
      =20
   ***** Restricted bremsstrahlung energy loss added *****
   ***** Exp. bremsstrahlung activated above 3.0000E-04 GeV *****
                                                                         =
      =20
***** dp/dx tabulations in blank common start at location 1435141 =
*****
***** end at location 1450764 (I*4 addresses) *****
                                                                         =
      =20
  *** Range: P_max: 1.2300843 P_min 0.00398107152 ***
                                                                         =
      =20
***** Range tabulations in blank common start at location 1450767 =
*****
***** end at location 1450874 (I*4 addresses) *****
                                                                         =
      =20
                                                                         =
      =20
***** Sigtab : material number 26 "SILICON "

It always seems to be the same. It stops at this point, if it is going =
to stop. But some runs go for several cycles and then crash. Where =
should I look for problems? The .log file has no information.

- Katherine

------_=_NextPart_001_01C5D0FC.40B38E96
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7233.28">
<TITLE>FW: FLUKA Crashes</TITLE>
</HEAD>
<BODY>

I reran one of these cases on a stand-alone 32-bit Red =
Hat Linux machine and it ran successfully.  The machine that some =
runs crashed on was a 64-bit SUSE node that was part of a cluster.  =
I had "rsh n03" to get to a node and then ran the case which =
then crashed after several cycles in the original attempt that I wrote =
about earlier. 
 
These crashes are similar to some I had a few weeks ago where runs would =
crash on the clustered nodes but run successfully on stand alone =
boxes.  In the past even if the stand alone box used SUSE Linux and =
were 64-bit, they would run fine.  For some reason the clustering =
causes the problem.  In these recent cases it is interesting =
because the energies of the incident nuclei were much lower than the =
energies a few weeks ago when there were large hadronic cascades.  =
Also the recent runs crashed at the same place that appears to be during =
the initialization of the run, rather than the transport itself. 
 
- Katherine 
 
 
-----Original Message----- 
From: Harine, Katherine 
Sent: Fri 10/14/2005 11:57 AM 
To: giuseppe.battistoni@mi.infn.it; alfredo.ferrari@cern.ch; =
support@fluka.org; discuss@fluka.org 
Subject: FLUKA Crashes 
 
 
I am getting crashes with the .out file ending: 
 
. 
. 
. 
. 
   ***** Exp. pair production activated above  0.0000E+00 =
GeV ***** 
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
   ***** Restricted bremsstrahlung energy loss =
added          ***** 
   ***** Exp. bremsstrahlung activated above   =
3.0000E-04 GeV ***** 
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
 ***** dp/dx tabulations in blank common start at =
location   1435141 ***** 
 ***** end at location   1450764 (I*4 addresses) =
***** 
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
  *** Range: P_max:  1.2300843 P_min  0.00398107152 =
*** 
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
 ***** Range tabulations in blank common start at =
location   1450767 ***** 
 ***** end at location   1450874 (I*4 addresses) =
***** 
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
            &=
nbsp;           &n=
bsp;           &nb=
sp;           &nbs=
p;           &nbsp=
;            =
       
 *****  Sigtab : material number  26 "SILICON =
" 
 
 
 
 
It always seems to be the same.  It stops at this point, if it is =
going to stop.  But some runs go for several cycles and then =
crash.  Where should I look for problems?  The .log file has =
no information. 
 
- Katherine

</BODY>
</HTML>
------_=_NextPart_001_01C5D0FC.40B38E96--

Next message: Burge, F \(Frances\): "error message"

Previous message: Harine, Katherine: "FW: FLUKA Crashes"
Maybe in reply to: Harine, Katherine: "FW: FLUKA Crashes"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by hypermail 2.1.6 : Sat Oct 15 2005 - 13:00:37 CEST