Kraken will be decommissioned on April 30, 2014. For more information see Kraken Decommission FAQs
Kraken will be decommissioned on April 30, 2014. For more information see Kraken Decommission FAQs
The National Institute for Computational Sciences

Nautilus Software

VisIt

Category: Analysis-Visualization

Description

VisIt is an open source, turnkey application for visualization and analysis of large scale simulated and experimental data sets. It leverages several third-party libraries such as the Qt widget library for its user interface, the Python programming language for a command line interpreter, and the Visualization ToolKit (VTK) for its data model and visualization algorithms.

Interactive Use

The preferred way to use VisIt on Nautilus is to take advantage of its client/ server architecture and run a VisIt client locally while connecting to the server running on Nautilus.

Binary executables can be downloaded from the VisIt website. When installing VisIt locally, you can choose a particular site configuration. The NICS configuration will provide you with the host profiles you need to connect to Nautilus as a remote host.

To connect to Nautilus when running VisIt on your local machine, simply choose Nautilus from the 'Host' menu in the 'Open' dialog. You will be asked for your passcode. Now you should be able to navigate to your files on Nautilus. When you choose to open a file, you will be presented a dialog for launching VisIt's parallel engine. Here, you can choose the number of cores you wish to use. VisIt then launches its parallel engine on Nautilus behind the scenes and you can explore your data interactively on your local machine.

Running Scripts

Since VisIt can be scripted with Python, it can be run without its GUI or Viewer components, directly on Nautilus. You must first load VisIt with the module command:

> module load visit

To launch a parallel VisIt job from the login node, you must use the VisIt flags '-l' to specify the launch method and '-np' to specify number of cores. For example:

> visit -nowin -cli -s my_script.py -l qsub/mpirun -np 64

will automatically launch a job for 64 cores to run my_script.py using VisIt's parallel engine. To specify other PBS options with the VisIt command line, use these VisIt flags:

-p <part> Partition (queue) to run in
-b <bank> Bank (account) to charge
-t <time> Maximum walltime
-n <name> Name for job

For a complete list of VisIt options, do:

> visit -fullhelp

From within a PBS batch script, the '-l' flag is not needed--only specify the number of cores with '-np'. Here's an example script to use 64 cores:

#PBS -l ncpus=64,walltime=03:00:00
#PBS -A XXXYYY

module load visit
cd $PBS_O_WORKDIR
visit -nowin -cli -s my_script.py -np 64

Many examples of scripting VisIt with Python can be found on the visitusers wiki.

Support

This package has the following support level : Supported

Available Versions

Version Available Builds
intel pgi gnu Other
2.6.3
?
2.6.2
?
2.6.1
v
?
?
2.6.0
?
?
2.5.2
v
?
?
2.5.0
v
?
?
2.4.2
v
?
?
2.4.0_alt
?
2.4.0
v
?
?
2.3.2
v
?
?
2.3.1
u
?
?
2.3.0
u
?
?
2.2.2
v
?
?
2.2.1
v
?
?
2.2.0
v
?
?
2.1.2
v
?
?
2.1.1
v
?
?
2.1.0
v
?
2.0.2
?
?
?
2.0.1
v
?
?
2.0.0
v
1.12.2
v
1.11.2
?
v