Program seems slow

From: Bruce Bristol <bbristol_at_ix.netcom.com>
Date: 1997/08/09
Message-ID: <33ED6225.3D20@ix.netcom.com>#1/1

I have a COBOL program that simply reads a file and checks each record against our History Database to see if it's already been processed.

The problem is that the file contains 5.5 million records and it's taking over 8 hours to process! Also, it's the only thing running on a SparcCenter 2000 machine.

Are there any tuning parameters I can use to speed this up? I think it should be able to run much faster than it is.

There are 8 tables and I choose the appropriate table to select from based on a number.

Here's the SQL I'm using to check if a record has already been processed:

EXEC SQL
  SELECT COUNT(*)
    INTO :H-ROW-COUNT
    FROM DUAL
    WHERE EXISTS ( SELECT NULL

      FROM HIS_TABLE_A
      WHERE NUMBER_1 = :H-NUMBER-1 AND
            NUMBER_2 = :H-NUMBER-2 AND
            NUMBER_3 = :H-NUMBER-3 AND
            NUMBER_4 = :H-NUMBER-4 AND
            NUMBER_5 = :H-NUMBER-5 AND
            NUMBER_6 = :H-NUMBER-6 AND
            NUMBER_7 = :H-NUMBER-7 )

END-EXEC. Thank you!

-Bruce Received on Sat Aug 09 1997 - 00:00:00 CDT