{"id":357,"date":"2013-04-26T12:50:05","date_gmt":"2013-04-26T12:50:05","guid":{"rendered":"http:\/\/ri.itservices.manchester.ac.uk\/csf-apps\/?page_id=357"},"modified":"2016-11-25T12:49:41","modified_gmt":"2016-11-25T12:49:41","slug":"remotex","status":"publish","type":"page","link":"https:\/\/ri.itservices.manchester.ac.uk\/csf-apps\/software\/applications\/paraview\/remotex\/","title":{"rendered":"Running ParaView Entirely on the CSF (remote X)"},"content":{"rendered":"

Overview<\/h2>\n

This page describes how to run ParaView entirely on the CSF over a remote X connection. This is not<\/strong> the preferred method of running ParaView. However, if you are unable to install the ParaView GUI locally on your own computer then this method will work.<\/p>\n

You will need to have an X-server installed on your local computer. See further information about how to start X-Windows and GUI applications on the CSF<\/a>.<\/p>\n

The two methods of running ParaView described here are:<\/p>\n

    \n
  1. Running only the ParaView GUI interactively – for small datasets only<\/li>\n
  2. Running the ParaView GUI interactively and multiple pvserver<\/code> processes to process and render your datsets in parallel – for large datasets<\/li>\n<\/ol>\n

    NOTE:<\/strong> In both cases the paraview<\/code> GUI must not<\/strong> be run on the login node. You must<\/strong> run it interactively using qrsh<\/code> as described below.<\/p>\n

    Running paraview GUI Interactively (small datasets)<\/h2>\n

    After loading the ParaView modulefile submit an interactive job to the short<\/code> resource. Note that interactive jobs will not run in the standard 7 day resource.<\/p>\n

    \r\nmodule load apps\/gcc\/paraview\/3.14.1\r\nqrsh -V -cwd -b y -l inter -l short paraview\r\n<\/pre>\n
    The ParaView GUI will start once the system has scheduled your interactive session. <\/p>\n
    Full details of interactive job limits are given on the qrsh page<\/a>.<\/p>\n
    Running MPI pvserver batch processes and paraview GUI Interactively (large datasets)<\/h2>\n
    The paraview<\/code> GUI should not<\/strong> be running at this point.<\/p>\n
    After loading the ParaView modulefile use a jobscript to launch the pvserver<\/code> MPI processes. The choice of Parallel Environment (PE) depends on whether you want to use InfiniBand-connected nodes for multi-node larger jobs or a single compute node for a smaller jobs. InfiniBand-connected nodes require that the number of cores used is a multiple of 12 or 24 (and a minimum number of cores must be used). Single-node (SMP) jobs can use a smaller number of cores (up to 24).<\/p>\n
    pvserver in SMP (single-node smaller jobs)<\/h3>\n
    Setup paraview by loading the following modulefile on the login node:<\/p>\n
    module load apps\/gcc\/paraview\/3.14.1<\/pre>\n
    (note that the 4.0.1 version cannot be used – the paraview<\/code> GUI executable is not installed on the CSF in this version. It only has the off screen pvserver<\/code> and pvbatch<\/code> executables).<\/p>\n
    Create the following jobscript (e.g., naming it pvserver.sh):<\/p>\n
    \r\n#!\/bin\/bash\r\n#$ -S \/bin\/bash\r\n#$ -V\r\n#$ -cwd\r\n\r\n#$ -pe smp.pe 4       # Change number of cores (max 24)\r\n\r\n#$ -l short\r\n     #\r\n     # Optional (and max 12 cores must be used above)\r\n\r\n# Ensure the paraview modulefile is loaded\r\n\r\nmpirun -n $NSLOTS pvserver --use-offscreen-rendering\r\n\r\n<\/pre>\n
    Then submit the above jobscript using<\/p>\n
    \r\nqsub jobscript<\/em>\r\n<\/pre>\n
    where jobscript<\/em> is the name of your file (e.g., pvserver.sh<\/code>)<\/p>\n
    pvserver over InfiniBand (larger multi-node jobs)<\/h3>\n
    Set up paraview by loading the following modulefile on the login node:<\/p>\n
    module load apps\/gcc\/paraview\/3.14.1-ib<\/pre>\n
    (note that the 4.0.1 version cannot be used – the paraview<\/code> GUI executable is not installed on the CSF in this version. It only has the off screen pvserver<\/code> and pvbatch<\/code> executables).<\/p>\n
    Create the following jobscript (e.g., naming it pvserver-ib.sh):<\/p>\n
    \r\n#$ -S \/bin\/bash\r\n#$ -V\r\n#$ -cwd\r\n#$ -pe orte-24-ib.pe 48\r\n     #\r\n     # Change number of cores, MUST be multiple of 24.\r\n\r\n# Ensure the paraview modulefile is loaded\r\n\r\nmpirun -n $NSLOTS pvserver --use-offscreen-rendering\r\n<\/pre>\n
    Then submit the above jobscript using<\/p>\n
    \r\nqsub jobscript<\/em>\r\n<\/pre>\n
    where jobscript<\/em> is the name of your file (e.g., pvserver-ib.sh<\/code>)<\/p>\n
    Find where the Batch Job is Running<\/h3>\n
    You must now wait<\/strong> until the pvserver jobs are running. Use qstat<\/code> to check the job queue. Once the state<\/code> reported by qstat<\/code> has changed from qw<\/code> to r<\/code> you will see something like:<\/p>\n
    \r\njob-ID  prior   name     user     state submit\/start at     queue                          slots \r\n------------------------------------------------------------------------------------------------\r\n5724  0.05534 pvserver.q xxxxxxxx r     08\/28\/2012 17:06:57 R410-short.q@int01<\/strong>.prv.csf       4\r\n<\/pre>\n
    Make a note of the queue<\/strong> value, specifically the hostname<\/strong> after the @<\/strong>. In this example we see int01<\/code><\/strong> (you don’t need the .prv.csf….). <\/p>\n
    We now run the paraview<\/code> GUI client and connect it to the rank 0 MPI process running on the backend nodes. To run paraview<\/code> use:<\/p>\n
    \r\nqrsh -V -cwd -b y -l inter -l short paraview --server-url=cs:\/\/int01<\/strong>\r\n   #\r\n   # Change int01 to match the hostname used by your pvserver batch job.\r\n   # You MUST specify -l short for interactive (-l inter) jobs.\r\n<\/pre>\n
    Once the qrsh<\/code> command has been scheduled to run, the ParaView GUI will appear. The Pipeline Browser<\/em> area should show an icon representing the backend node running the pvserver<\/code> rank 0 process, as show below. This indicates you have successfully connected the paraview<\/code> GUI to the backend pvserver<\/code> processes. Data processing and rendering will be performed in parallel by the backend processes where possible, and only lightweight command-and-control actions will be performed by the GUI process.<\/p>\n