Practices for DB Scalability

From: C P <carlospena999_at_gmail.com>
Date: Fri, 1 Nov 2013 02:32:29 -0500
Message-ID: <CAPTPB12Ejduj46MqjLgXtBsvjEz2VEMCLgTfvjji7Kn-bv8Nxw_at_mail.gmail.com>

List,
I am looking for some of the most common and frequently used techniques to improve scalability of oracle database systems.

I see that minimizing the work like reducing latch contention, physical reads and buffer gets are important, I am trying to find out ways to do that, in both home grown (HG) and third party COTS applications where you cannot touch the SQL, more specifically in the third party COTS.

Here are some ways I know of to improve scalability:

• use bind variables (HG),
• use connection pooling (HG),
• careful about running big reports/long running queries that could clog IO and/or flood buffer cache; also could cause problems if a big table is involved and happens to be frequently updated table resulting in lot more IOs due to read consistency. (HG and COTS)
• use caching mechanisms in the database (HG and COTS) to improve response times.
• reduce concurrency where it can be done (HG, how to do that in cots) , but how?

Any more additions/modifications?

CP.

--
http://www.freelists.org/webpage/oracle-l

Received on Fri Nov 01 2013 - 08:32:29 CET