April 2004 Archives

Are you already using Oracle compressed tables or thinking about the implications of adopting this technology?

This article by Sanjay Mishra goes into the pros and cons of this feature which has been available since Oracle 9 (Rel. 2). If you are running a Data Warehouse you should be considering the implications of using compressed tables in your system. But be wary it is not a solution to saving space all over your database!

If you already use Jalopy or you are thinking of using a code beautifier with Jdeveloper, then there is now an open source plugin available for JDeveloper 10g from SourceForge.

Steve Muench posted this article stating that the 10g Production release of Oracle's IDE "JDeveloper" was made available last week.
If you would like to download a copy it's available from OTN. There is also an online demo of JDeveloper in action available for viewing.

There is some discussion going on about the general quality of the training being offered by Oracle University on the 10g release. The question arising is "is it just a money grab"? This article by Damir Bersinic looks at why things aren't as good as they could be and what might work better.

This annotated 24x36 color Oracle data dictionary poster only displays the most important DBA dictionary views, the important data columns and the relationships between the DBA views. It's a great cheat-sheet for writing customized Oracle dictionary scripts.
Especially useful are the Automated Workload Repository (AWR) and Active Session History (ASH) DBA views which replace STATSPACK in Oracle10g. Here is an image of the new Oracle 10g reference poster.
UNISYS will be giving-away posters at IOUG Live! and OracleWorld, but if you cannot attend you can order a free copy of the poster here.

Row-level control which enables different users to be restricted in the data (rows) they are allowed to see, has been available to the Oracle world since version 8. On row-level access it is possible to define rules which validate to users identity and thus restrict the data (rows) which can be seen or changed. This is sometimes referred to as a Virtual Personal Database - VPD. This security article describes how that works.

The general thinking on indexes would seem to be 'Rebuild it regularly' if you want it to perform well, but that is not necessarily the correct way of thinking. By rebuilding you use-up a lot of system performance and down grade your systems performance during the rebuild as the index will not be available. This article is somewhat dated but neatly goes into the pros and cons of regular rebuilds.
We run a Data Warehouse application which has some very large indexes and we don’t regularly rebuild them, although we could probably find a recurring 8 hour time slot some time during the week.