Exadata Performance Degradation Identification

The most tuned way to do something in database is not to do it at all. But if you have to do something, then the second most tuned way is to reduce the data which is supposed to be traveling between you and the system. Less data moved means more throughput in lesser response time.

That is what actually Oracle Exadata database machine promises and delivers.

It smartly scans the query results from the storage, and gives them to the database. To make things more whimsical, the short query result is passed through high speed, vast capacity infiniband network to the database server. The scanning of query result is smart because everything is Oracle-aware at the storage and database level.

One of the major reason why even after deploying Exadata, users are experiencing performance degradation is that your queries might not be using the smart scanning of the result sets. Now how to identify that? Kerry Osborne gives a handy script here in this blog post to do just that.

Enjoy

Find Command with Interval in Linux

If you would like to just find files between two dates in a directory in Linux, do the following:


$ touch f1 -t 201001011130
$ touch f2 -t 201012311130

The above will create two files f1 of date 01 Jan 2010 and f2 of 31 Dec 2010.

Now to find files between 01 jan 2010 to 31 dec 2010:

find . -cnewer f1 -and ! -cnewer f2

Hope that helps.