ANSYS Background Usage
This describes how to use ANSYS on the Sun workstations without sitting at the terminal . That is, in the background .
Since ANSYS uses a large amount of computer memory, disk space and CPU effort it is common to want to run an ANSYS job but not to wait for the results. For example, you may want to run a job or jobs overnight when there are fewer people competing for the limited amount of memory or disk space. This can be done easily using the background capabilities of UNIX. However, there are a few rules that we must follow in order to be fair and polite to other users of these machines.
- Do NOT run the jobs using the graphical interface and then lock the screen of the computer. This monopolizes the computer and is extremely rude.
- ANSYS batch runs should always be run using the unix nice command. This is because a user sitting at the terminal typing or using other software deserves quick responses from the computer. In a practice, this does not slow the execution of a batch job by a measureable amount!
- ANSYS jobs should be run one at a time and if at all possible, run between the hours of 10 pm and 8 am. At these times, the slower response of the computer will not interfere with users doing interactive work and will also take advantage of the lighter traffic load on our local engineering computer network.
In order to help you get started using the batch capabilities of Unix, a program that executes ANSYS (ie. a shell script ) called ansyshell has been created for you to copy and modify as you need. This shell script contains a number of features that you can use to
- Run jobs now but in the background
- Run jobs later at a specified time and day
- Run several jobs each with a different job name
If you have several jobs that you wish to run sequentially, an additional shell script can be created that runs them one at a time by executing the ANSYS shell script above repeatedly. Such a secondary script should reside in the directory where all these ANSYS runs will take place and might look like:
#!/bin/csh
#
ansyshell -i inputfile1 -j case1
ansyshell -i inputfile2 -j case2 -m 128 -d 64
ansyshell -i inputfile3
This will use the ansyshell script to run three jobs with different input files and different job names. The third job will use the default jobname file while the first two use case1 and case2 respectively. Also, the first and third job will use the memory and disk space limits set in the ansyshell script while the second job will be given the limits of 128 MBytes for memory and 64 MBytes for disk space. Most importantly, the second job will begin immediately after the first job is completed - however long that takes - and the third job will start after the second is finished. This way no more than 1 job runs at a time so the jobs finish as quickly as possible.
If you have any questions regarding the ANSYS script above or its use, please see Dr. Gordon or Dr. Reynolds.
Last update: March 6, 1998.
Return to ME Home Page
