Oracle Database Performing Slow | Windows Issue

Oracle Database Performing Slow | Windows Issue

This post is for the Oracle DBA professionals who usually face database slowness issue on daily basis but the scope of post is limited for the Oracle Database Server on Windows only. Since I have experienced the same, that's why this post is here.

While working with so many clients, database performing slow or application working slow due to database is quite common issue for me. The basic work around is to look for memory parameters in Oracle Database like SGA and PGA memory allocation and if you are using the ASMM (automatic shared memory management) feature then the parameters would be memory_target and memory_max_target. But every time your database goes slow, you don't keep on increasing the memory, do you?

Well, there are cases when you know that you have provided enough memory allocation to database and still it is performing very slow, even there is not that much of transaction activities going on. The other alternative that we often try is generating STATSPACK or AWR reports and then call the Oracle Support for help but that's not feasible for everyone (what if you don't have access to Oracle Support).

I am not saying that the above mentioned workarounds are useless but this is a case where above mentioned resolutions are not helping you and you need to fix things fast especially in the case of production environments where consistent performance is everything and you can't risk it. So here it is.

Shutdown your Database and Startup again...... That's it !!

Yes, this is all you have to do. Now this is not really a solution but can help you dramatically. A lot of Oracle Database Professionals would agree with me because they would have experienced the same. I don't know whether it's the incompatibility of Oracle Database with Windows as compared to UNIX/LINUX Environments or something like that but it works... surprisingly it really works.

So next time you face database slowness issue, make sure you take a database restart first to make the first check and if the problem still persists, then look for other resolutions.