[Swift-devel] Q about MolDyn

Veronika Nefedova nefedova at mcs.anl.gov
Mon Aug 6 11:56:58 CDT 2007


On Aug 6, 2007, at 11:51 AM, Ioan Raicu wrote:

> If there are no leftover directories, it could mean one of two things:
> 1) the job submission failed and the job never made it out to a  
> remote resource to create directories and execute
> 2) the job went through to the remote resource, created  
> directories, executed, finished with an exit code of 0, cleaned up  
> directories, ...
>
> In both instances, there are no directories at the end.  What we  
> are missing is the logs to determine which of the two cases from  
> above happened.  Note that there were absolutely no Exceptions  
> thrown in the Falkon logs, so I would be cautious to rule out (2)  
> from above.
>


There are no output files from any stages after the Stage 4 (in  
shared) for m050. So we can rule out any successful execution of  
Stage 5 and after...


> Just for a sanity check, can you run those two molecules manually,  
> just to double check that they work!  Could we do a 2 molecule run  
> through Swift with just the two suspect molecules to see if those 2  
> molecules complete OK?  Then, let's try another run with 244  
> molecules soon, as most of ANL/UC is free!
>
> Ioan
>
> Veronika Nefedova wrote:
>> m050 worked just fine last time we ran 244 molecules. And it  
>> doesn't look like an application failure but rather a job  
>> submission failure (no leftover directories).
>> m179 failed the last time we ran it but I thought it was fixed?  
>> Yuqing - could you please verify that I am using the right  
>> antechamber code on tg-uc.
>>
>> Nika
>>
>> On Aug 6, 2007, at 11:34 AM, Ben Clifford wrote:
>>
>>> have you tried running those molecules individually?
>>> -- 
>>>
>>
>>
>




More information about the Swift-devel mailing list