Hello,
I am from Juelich Supercomputing Centre, Germany, and we would like to create an animation of our climate data, using ParaView 5.5.2 on one node of our JURECA cluster,
The data has 720 time steps, stored in HDF5 files. We read the files via the XDMF reader.
Unfortunately, when we go from timestep to timestep, ParaView consumes about 1.5-1.9 GByte more memory each time. After several time stepes, all the memory of the node (1 TByte) is consumed and ParaView crashes. An animation of all 720 time steps can not be generated this way.
Is there any workaround to fix this behaviour?
Thanks in advance
Herwig