Welcome to the 128th edition of Log Buffer, the weekly review of database blogs.
Let’s begin with some PostgreSQL blogs. Jignesh Shah shares his recipe for making a PostgreSQL 8.3 appliance based on OpenSolaris using VirtualBox. While we’re on appliances, Dave Page shows off PostgreSQL management on the iPhone with an application he himself wrote. Stealth DBA for the bus-rise home.
On Database Soup, Josh Berkus has been finding useless indexes. He begins, “I’d say, in general, that you can’t have really well-chosen indexes without the help of a performance expert. However, anyone can improve their use of indexes in PostgreSQL fairly easily using a few system tools … and a little improvement is all that a lot of users need.” And it’s what Josh offers.
Sometimes a DBA is asked to make something real purty-like, contrary to his or her nature though that may be. On the Postgres OnLine Journal, Leo Hsu and Regina Obe offer some help with the first of a series on Fusion charts and PostgreSQL. (”Fusion Charts . . . is a flash-based charting product that makes beautiful flash charts.”)
And now—hey what’s MySQL maven Baron Schwartz doing with a Postgres post on xaprb? He’s asking, what are your favorite PostgreSQL performance resources?
Maybe he’s considering crossing the floor out of weariness with all the contention in the MySQL world? Can’t say I blame him. Lately, the conversation in the MySQL ’sphere has been dominated by non-technical talk of the pluses and minuses of 5.1, of forking, community vs. enterprise, and so on. This week was no exception.
The week began with Jay Pipes’s advice to MySQL: “Drop the current roadmap . . . Forget Windows for now . . . Clean up the abysmal messiness of the code base . . .” It’s strong stuff and worth a read.
Speaking of which, the MySQL Performance Blog announced the Percona XtraDB Storage Engine: a drop-in replacement for standard InnoDB.
On High Availability MySQL, Mark Callaghan showed us how to make MySQL faster in one hour. Nice stuff. And real purty charts, too.
Let’s see what happened in SQL Server now. Kalen Delany opined that there is no such thing as a SQL Server, and she’s not the only one with an opinion on this (one would think) straight-forward matter.
Lots of thoughtful comment on that one, and a blog response from Ward Pond, who says that Linchi Shea makes an interesting point about hints, vis-a-vis the set-based and procedural paradigms.
The Data Management Journal looked into extracting numbers with SQL Server: “We all have perfectly normalized tables, with perfectly scrubbed data, right? I wish! Sometimes we are stuck with dirty data in legacy applications. What’s worse is that we are sometimes expected to do interesting things with dirty data. In this blog, I will show you how to extract a number from a varchar column that contains letters and numbers.”
Moving into things Oracle, Chen Shapira was thinking about a musical analogy for the DBA. Their not “rockstar programmers” or “jazz programmers”, says Chen. But I won’t give away her conclusion—click on.
Chet Justice, the Oracle Nerd, was pursuing ontology too in the second part of his Application Developers vs. Database Developers. (I wonder if it’s generally true that apps developers have such terrible manners.)
On An Expert’s Guide to Database Solutions, James Koopman suggested, maybe itÃ¢â‚¬â„¢s time to extend the DBAÃ¢â‚¬â„¢s realm of influence, using tools like Spotlight on Oracle.
Or perhaps with other tools, such as TOra? Here on the Pythian Blog, Brad Hudson posted his howto, Installing TOra with Oracle support on Ubuntu 8.04LTS (Hardy Heron).
Until next time, Happy Holidays to all our readers!