By moving filtering earlier in the pipeline, we greatly reduce (in common cases) the number of records considered for further processing. This can result in dramatic speedups when the initial record selection is broad (large system, weeks of data, no node filtering).
By moving filtering earlier in the pipeline, we greatly reduce (in common cases) the number of records considered for further processing. This can result in dramatic speedups when the initial record selection is broad (large system, weeks of data, no node filtering).