Airpak - System runs out of memory while post-processing a large size job
|
The case I am working on has a mesh size of about 1.5 M cells. I have completed a transient solution for a case involving species. The transient run was for 2hrs with 10 sec. time steps being saved at every 12 intervals (i.e. 120 sec). When I try to do post-processing on this job, I get an error message which says that the system is out of memory and acknowledging the error message results in the termination of Airpak. Useful information: Capacity of hard disk: 12 GB Size of current job: 6 GB RAM: 2 GB Operating system: Windows 2000 Space available on hard disk: 3.5 GB The scratch disk memory functionality was invoked to overcome the above described problem. When we assign this additional memory for post-processing, it is taken off the free space available on the local hard disk (up to the amount of virtual space that's configured for the machine) and is used as long as the job needs it. Subsequently it is given back to the system. Memory swapping can be invoked using the following procedure, - open a command prompt window (Start/Accessories/Command Prompt on Windows or terminal window on Linux) - switch to the directory which contains the airpak job (the model) directory - type the following at the prompt, For Windows: set ICEPAK_PP_MEMORY=-1 For Linux: setenv ICEPAK_PP_MEMORY -1 - hit Enter - start Airpak with text command airpak Post-processing will be slower than usual, but should be work fine without any errors. |
||
|