I’m no big fan of commercial database product. Oracle is a complicated beast, and I have never completely understood why big companies use such a complicated and expensive product to store small databases. A DBA friend of mine said that 90% of the databases he has seen in the last 10 years could be done in SQLite, 9% in MySQL, and only the remaining 1% would need Oracle. Maybe his assertion is a little bit exaggerated, well, he’s an Oracle DBA with more than 10 years experience, so he must be right in some way.
As UNIX engineer working in NMS (Network Monitoring System), I recently came to a project involving CA eHealth Performance Manager. It is as well expensive, but quite capable commercial product in the NMS field. Within eHealth, there is a database running on Oracle. CA says “don’t touch or mis-use the Oracle instance running within eHealth, it shall be a black box forever”.
Sadly, I have inherited an installation with eHealth 5.7 and the associated Oracle 9 database – and this database has been mis-used to death, or almost. I’m now faced with a reverse-engineering and re-factoring effort to externalize the custom tables that have been created within the EHEALTH instance.
This is the exact place where Oracle SQL Developer comes handy. It’s running on top of Java, so it works perfectly on OS X and Linux. Coupled with Vim and Sed, it’s the perfect reverse-engineering and fast-prototyping tool for the Oracle & SQL -impaired geek.
It can as well browse MySQL instances, but in a read-only fashion. If you need a read-write tool for MySQL, look at CocoaMySQL.
Post a Comment