I am able to extract time-series data using JSL and ODBC but JMP is extremely slow.
For example, for a dataset with 50k rows and 5 columns JMP takes 50x longer using the same ODBC driver but its native tool. We breakdown the queries not to have more than 1M rows per query, so the server can handle the load with ease.
I think JMP is prefetching or treating the query as normal SQL.
In the documentation (JMP 18) I found that there are multiple options, but I am not sure how to change them using JSL.
Any ideas?
SQL_connection_str = "DRIVER={...};HOST=servername;PORT=....;MAXROWS=1000000;CHARFLOAT=N;CHARTIME=N;CHARINT=N";
dbc = Create Database Connection( SQL_connection_str );
SQL_query = "SELECT ... WHERE..";
start=tick seconds();
dt_temp = Execute SQL( dbc, SQL_query );
end=tick seconds();
runtime=end-start;
show(runtime);
