VisIt#

Introduction#

VisIt is a data analysis and visualisation application, developed to be able to analyse large datasets. We show here how to start the program on our facilities; however, please refer to the project documentation for details on how to use the application itself.

Although VisIt is flexible in the way it is deployed, the simplest method of invoking it is to run these commands on a login node:

$ module add visit
$ visit

This will provide the core functionality. More complicated methods are described below, offering different levels of performance. For example, the graphical interface can be run locally on a low-end workstation or laptop, while the rendering takes place in parallel on the cluster – taking advantage of the CPU, memory and storage resources available. Although not making use of the acceleration that graphics cards provide, it can allow larger datasets that cannot fit onto a graphics card to be analysed.

Running VisIt on the ARC Systems#

Please note that these methods depend on your workstation having an X server running and for you to have enabled X11 forwarding through your SSH client.

On a cluster login node (quick start)#

Recommended for very short or undemanding pieces of work. Login to the cluster and execute:

$ module add visit
$ visit

Please note that, as they are shared between all users, it is important not to tackle large pieces of work on the login nodes.

On the cluster compute nodes#

Recommended for larger pieces of work:

$ module add visit
$ qrsh -cwd -V -l h_rt=<hh:mm:ss> -l h_vmem=<vmem> visit

Or, where using more than one cpu is useful, to launch visit in parallel

$ module add visit
$ qrsh -cwd -V -l h_rt=<hh:mm:ss> -l h_vmem=<vmem> -pe ib <num> visit -np <num>

In the above commands, <hh:mm:ss> is the length of real time the program will be run for, <vmem> is the amount of memory required per core (e.g. “1G” for 1 GB RAM), and <num> is the number of cores required (please note it appears twice in the command).

Running VisIT on your workstation, offload storage and processing to the ARC systems#

Recommended where the graphical interface (menu items, etc.) feels sluggish, or where the best possible performance is required.

Select a version number from the list of installed versions of visit, then download and install a copy of that version appropriate to your machine from https://visit.llnl.gov. The list of installed versions is provided by executing the following command on the cluster:

$ module avail visit

If installing on a Windows platform, you will be asked to select the desired network configuration and given a choice of a number of institutions. Please select None .

Configure for remote execution#

Once installed, start the program and then exit it: this will create a number of directories.

Then, download the files corresponding to the cluster you wish to use:

host_arc3

host_arc4

On a Linux, Unix or Mac platform, please save these files to your .visit/hosts/ directory, found within your home directory.

On a Windows platform, please save these files to the application directory. Depending on whether VisIT was installed by an administrator for all users or just for your account, this may be found under Computer->C:->Profile Files/LLNL/VisIt <ver>/hosts or Computer->C:->Users->YourUsername->AppData->Local->Programs->LLNL->VisIT <ver>->hosts

Start VisIt once again and select the menu item Options->Host Profiles... This will open a window, which should contain an entry on the left hand side for each of the host files saved earlier. For each host in turn:

  1. Log into the cluster. Find the location of VisIT by loading the module for the version you want and examining a variable.

    module add visit echo $VISIT_HOME

  2. Highlight the hostname by clicking on it. On the right hand side,
    look for the configuration item Path to VisIt installation . If you cannot see it, click on the Machines and Host Settings tabs. Enter the location string found in step (1) above.

  3. Close the window with the Dismiss button

  4. Save settings with the Options->Save Settings menu item

Launch remote execution#

  1. With VisIt running on your workstation, select menu item File->Open file... to open the File open window.

  2. There will be a drop-down text box marked Host . Click on its arrow, which should show a list including localhost (your workstation) and each of the clusters you configured in the steps above. Select the cluster you wish to use.

  3. The File open window should now show files on the cluster. Select a file visit understands and press OK

  4. A Select options for <clustername> should appear. Select serial (login node)S for a single process on a login node (please note that login nodes are shared between all users of the cluster – it should only be used for short or undemanding pieces of work), or mpi (batch queue) to launch visit on the cluster’s compute nodes through the batch queue system. When launching through the batch queues, specify the number of cores required into Num procs , the length of time the resources are required into [Time limit] and any other batch queue options into Bank (the default value, -l h_vmem=1G requests 1G RAM per requested process).

Once VisIt has launched onto the login node or the batch queue has started running, use VisIt normally.

To double-check what resources are in use, select the menu item File->Compute Engines...

Developing plugins with VisIt#

When developing plugins to extend the software it is probably useful to use the same development environment used to build VisIt, in order to minimise compatibility problems between different C++ compilers, etc. If we have compiled a particular version of VisIt ourselves instead of using the version from the VisIt website, loading the visit module will set environment variable VISIT_BUILD_MODULES to contain the list of modules used build it. For example:

$ module add visit
$ echo $VISIT_BUILD_MODULES
gnu/native openmpi mkl qt