Export events fails for large datasets

Idea created by kimo on Nov 29, 2010
    • n314
    • jwielki
    • BruceLang
    • kimo
    • vtn
    • dknudsen
    After creating an XY event layer it is essential to export it to a real featureclass for performance and stability.

    When the event layer is created it appears instantly, which is great but ominous. When you export to a featureclass, the CPU runs hot, with no I/O. One would expect that the in memory features would just be streamed, but something is created in memory that takes a very long time, say an hour. Finally I/O starts but if the feature count is large (eg 1 million or so) it crashes. Demo sized sets work just fine, but it does not scale.