Alon Goldshuv wrote:
I performed those measurement by executing *only the parsing logic* of the COPY pipeline. All data conversion (functioncall3(string...)) and tuple handling (form_heaptuple etc...) and insertion were manually disabled. So the only code measured is reading from disk and parsing to the attribute level.
Arguably this might exaggerate the effect quite significantly. Users will want to know the real time effect on a complete COPY. Depending on how much the pasing is in the total time your 20% improvement in parsing might only be a small fraction of 20% improvement in COPY.
Like you, I'm happy we have seen a 5 times improvement in parsing. Is it possible you can factor out something smallish from your patch that might make up the balance?
cheers andrew ---------------------------(end of broadcast)--------------------------- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match