Hi, Do both environments use direct io and or async io? How much physical io is done by this query? You said you did a sql trace on the session and that you noticed the time spend on the "db file sequential read" event was higher. Do you mean that you see a higher elapsed time for this event then with the old environment? Are the occurrences of the "db file sequential read" event always taking the same time of do you see large fluctuations? You can use orasrp to generate a histogram of the events (http://www.oracledba.ru/orasrp) Does scattered reads also take longer on the new environment then on the old? Regards, Freek D'Hooge Uptime Oracle Database Administrator email: freek.dhooge@xxxxxxxxx tel +32(0)3 451 23 82 http://www.uptime.be disclaimer: www.uptime.be/disclaimer --- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Ram K Sent: maandag 22 augustus 2011 21:46 To: Blanchard, William G Cc: oracle-l Subject: Re: Is this hardware issue? I understand once the data is cached it can be retrieved quickly from the buffers. I am wondering why it would take 7/8 minutes to get some data from the IO layer. This behavior is NOT the case in the old hardware. It comes back much faster. The plans are same, once again. On Mon, Aug 22, 2011 at 2:20 PM, Blanchard, William G <William.Blanchard@xxxxxxxxxxxxx> wrote: This is expected behavior. The first time you run the query, it has to read everything into the buffer cache (db file sequential read). Every time after that (especially since nobody else is using the database) the information for the query is already cached and can be read right from the cache. WGB -- //www.freelists.org/webpage/oracle-l