[gmx-users] tpbconv restart crashing on 3.3.1

[Alan Dodd]

Hello all,
A recent simulation has been running on a cluster for a couple of weeks.  5.something-ns in it crashed due to a hardware glitch.  All perfectly understandable so far.
The strange thing is, using tpbconv now results in a .tpr that crashes in the first step due to massive LINCS errors.  Even if you use the option -time to take the simulation way back, to say 0.5ns when it ought to still be fine, it still crashes.  Other simulations on the same hardware+software are running fine, other restarts from other jobs crashing at the same time have worked fine, so it doesn't look like a bug.  Any suggestions as to why LINCS errors could suddenly now prevent the simulation from running *any* of the steps that it previously ran just fine?  I just want a hint as to what we could have missed, really.

Alan Dodd
UoB


 
____________________________________________________________________________________
The fish are biting. 
Get more visitors on your site using Yahoo! Search Marketing.
http://searchmarketing.yahoo.com/arp/sponsoredsearch_v2.php