Optimal combination of extraction settings to increase performance (extraction type, segment size, fetch size, ...)

[shoerley]

Hi community,
I am reaching out Dbeaber enthusiasts regarding the parameters available when exporting some data. In my usecase, I have about 20 queries returning between 20k and 50k rows each, that I need to export to CSV files. Given the high amount of data, I'd like to set up export parameters so as to have the best possible performance (i.e. speed of execution) possible.

I could read somewhere that "multiple queries" enabled parallelization on server side, but I am pretty much unsure on this, as this parameter is not related to the number of threads, but to the segment size. Can someone share something about that ?

Also, can someone explain how segment size and fetch size work together ? I know their "raw" definitions, but lack of experience make me miss the understanding of their deep interaction.

Any help would be highly appreciated.

A screenshot of the extraction settings window :