long running sim dies

From: Thomas C. Bishop (bishop_at_latech.edu)
Date: Mon Oct 13 2014 - 08:29:43 CDT

Dear NAMD,
we have been doing some benchmarks on my local computers lately and seems they are dieing
after running successfully from some time.
( NAMD 2.9 for Linux-x86_64-multicore, 32-way SMP opteron, 1 node, 1 physical node
Uname 3.11.10-21-desktop #1 SMP PREEMPT ... the machine has LOTS of ram and typically one user )

Nothing seems pathological w/ the system (all energies config etc.. are ok)
There is plenty of disk space for output..etc..

The only error I have was a note from one student
. I checked the output file and found a fatal error occured, it said that cannot find balancer (not sure, need to check it again).


Other runs died w/ no error message.
Could it possibly be that /usr/local/namd2, which is automounted, is disconnecting?
system logs do not indicate a hardware problem.

I presume if the namd2 executable is in same directory as the output that I/O will keep the automount active.
It this true or does namd open/close files between writes?


Just trying to figure this one out... guess I need to dust off the sys-admin manual.

Any ideas/comments appreciated.
Tom



-- 
*******************************
   Thomas C. Bishop
    Tel: 318-257-5209
    Fax: 318-257-3823
   www.latech.edu/~bishop
********************************

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:17 CST