Recordset memory considerations

From: Mark Carter <cartermark46_at_ukmail.com>
Date: 26 Mar 2004 07:54:59 -0800
Message-ID: <d3c9c04.0403260754.4c34b86e_at_posting.google.com>



I will need to access both an MS Access database and an Oracle database.

I want to look at historical values held on one of the databases. The problem is, the data might not exist for a given day, or might be rejected as an outlier; where what constitutes an outlier is a moving target depending on what data I find along the way. So although I know I want, say, 30 clean values, I don't know in advance how far back I'll have to look back to get them. The simplest solution is to select all the values prior to a certain date, for a subset that I am interested in, and then work backwards however long was necessary. The resulting recordset could be potentially be very large, and I wondered if there were any memory/resource implications of doing this. Or, should I consider a recordset to be somewhat like a pointer/cursor, which is actually quite a small object? Received on Fri Mar 26 2004 - 16:54:59 CET

Original text of this message