[chbeck]

Hello,

I am performing LES simulations on a big cluster using ~1000 cores. In some cases during writing of the fields half of the processor dictionaries get written for time step n and the other half of the processor directories gets written for the time step n+1. This issue becomes apparent when trying to use the stored solutions for restart etc...

-Wall clock time per time steps is ~0.5s.
-Version 1.6 from the git repository from ~Dec 2009
-A lustre file system is used

Some modified settings:

minBufferSize=300000000

OptimisationSwitches
{
fileModificationSkew 10;
commsType nonBlocking; //scheduled; //blocking;
floatTransfer 0;
nProcsSimpleSum 0;
}

[mattijs]

The Mantis bug reporting should be up and running again. Could you please report any bugs there?

Thanks,

Mattijs