Hi All,
I am facing a strange problem. We are using postgres on linux. Our
application code uses ODBC API (Driver: unixODBC)
We observed that process size shoots up when a SQLGetData is done from
application code on a table(s) with more than a 100,000 rows. Upon
experimentation we found that if we restrict the SQLGetData to work on
< 20,000 rows at a time then the size didn't grow. But since we
couldn't hard code the number of rows in application, we want to know
if there is a way to call SQLGetData() (or other equivalent ODBC API)
where in a fixed pool of memory can be specified by the application.
With this the driver will return the SQLGetData call once the buffer
size max is reached. This will be a generic solution for us.
Could some one let us know if some kind of solution is possible?
PS: In case some one knows of any other news group please let me know
so that I can post it there too.
thanks in advance,
--kv