I am performing a long-running simulation in LIGGGHTS. It has approx 10,00,00,000 steps and dump files are written per 1000 time-steps. Working with such huge number of files, it becomes really difficult to perform lpp processing of all at a time. If I need to check the progress while the simulation is on the go everytime above command starts from beginning which is clearly a waste of time. Also, if any interruption happens or error comes then I have to perform lpp from beginning with the command lpp dump*.mixer.
I found some command over internet that may help in choosing the range of files for bla bla operation.
eg. ls dump[1][1-2][2-5][0-9][0-9].fcc results
dump11200.fcc dump11300.fcc dump11400.fcc dump11500.fcc dump12300.fcc and skip all other files. Similarly, using cp instead of ls only copies files listed above.
But, when I do lpp dump[1][1-2][2-5][0-9][0-9].fcc then it results
fcc11200_boundingBox.vtk fcc11300_boundingBox.vtk fcc11400_boundingBox.vtk fcc11500_boundingBox.vtk fcc11900_boundingBox.vtk
fcc11200.vtk fcc11300.vtk fcc11400.vtk fcc11500.vtk fcc11900.vtk
.
Why 12300 got excluded and why 11900 got included. I don't know whether it is bug or I am doing something wrong I don't know. Is there any other alternative to do so???
Please help!!
rahulsoni | Fri, 11/15/2013 - 13:06
Got it!!
lpp dump[1-4][1-2][2-5][0-9][0-9].fcc style command has worked. There was some silly problem. Checked with other systems, it worked.
PaulWinkler | Wed, 11/20/2013 - 12:57
Hi,
Hi,
LPP is not very well in handling large data. I had problems to convert dump files larger than 20GB. But you can use the dump vtk, since you have compiled LIGGGHTS with vtk support. There is an update macro for ParaView from Christian Richter, so you will get the latest computed steps in ParaView.
Regards,
Paul
rahulsoni | Wed, 11/20/2013 - 17:32
Okay, I will try to workout
Okay, I will try to workout your suggestions.
Thanks for same.