Changes between Version 30 and Version 31 of ParaView


Ignore:
Timestamp:
05/03/17 12:01:13 (7 years ago)
Author:
Jens Henrik Goebbert
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • ParaView

    v30 v31  
    2424There are other more advanced configurations that can be used for Remote Visualization or separation of services (Data, Rendering, and User Interface). \\
    2525Each mode separates the three main components (User Interface, Computation, Rendering) in different ways.\\
    26 More details on possible !ParaView scenarios can be found [wiki:ParaView/Jureca here].
     26More details on possible !ParaView scenarios can be found [wiki:ParaView/Jureca here] or [https://daac.hpc.mil/software/ParaView/ParaView_on_Multiple_Processors.html here].
    2727
    2828----
    … …  
    4242
    4343=== Parallel !ParaView
    44 * [http://www.paraview.org/Wiki/ParaView/ParaView_Readers_and_Parallel_Data_Distribution ParaView Readers]
     44If your dataset is large you might be limited by the size of the memory or the cpu performance of a single node. \\
     45Executing filters on your data in !ParaView might be slow and takes too long or even fails because of memory limitations. \\
     46One possible solution can be to run !ParaView im parallel across multiple nodes.
    4547
    46 ==== Parallel Data Management
    47 Data must be distributed across parallel processes to take advantage of resources
     48==== How to start !ParaView in parallel
     49First of all, __parallel__ !ParaView is only available on vis-compute nodes (not on vis-login nodes).\\
     50We recommend the following:
     51* Start a VNC session on multiple vis-compute nodes using the login tool 'Strudel' [wiki:vnc3d (more details)].
     52* Start !ParaViews 'pvserver' with MPI multiple times on the nodes.
     53  * click the icon 'start PVServers' on the VNC desktop (profile vis).
     54* Start !ParaView and connect to the pvservers (localhost:11111).
     55
     56==== How to read data to !ParaView in parallel
     57Just starting !ParaView in parallel __does not__ result necessarily in using the compute resources (cpu and memory) of more than one node or using them on the allocated nodes equaly.
     58
     59Two main issues must be considered:
     60* Parallel Data Management - the data must be distributed across parallel processes to take advantage of resources
     61* Parallel Work Managemnet - the work must be distributed across parallel processes to take advantage of
     62
     63===== Parallel Data Management
     64Data must be distributed across parallel processes to take advantage of resources.\\
     65
    4866 
    4967 * Some !ParaView readers import in parallel