Keeping it in the database…

I’ve mentioned a couple of times recently that I’m in the process of installing and moving a whole bunch of servers from our current site to a new location. Well the first phase of this process was completed over the weekend. The move involved a change in platform from Alpha to Linux for a few projects, including some Oracle database servers.

I do pretty well all my Extraction Transformation Load (ETL) processes from within the database using SQL and PL/SQL, which by its nature is portable between Oracle supported platforms. As a result, everything just worked.

Some of the other non-Oracle projects use shell scripts and Perl (containing calls to OS commands), which can prove a little more problematic. I’ve just finished helping debug a Perl script that was failing due to the difference in output of the “ps -p <pid>” command on Alpha and RHEL. Imagine how big the impact would have been if the move had been between Windows and Linux!

This just reminds my how “keeping it in the database” has paid off for me again and again. When I treat Oracle as my platform life is easy. When I treat Oracle as a “bit bucket” and focus on external technologies I run into problems.

Just an observation by a biased PL/SQL junkie… Not flame bait… 🙂