Welcome to the 18th Log Buffer, the weekly human-edited review of news from database blogs.
There’s no shortage of conferences and news about them, so let’s have a couple items on UK Oracle User Group Conference & Exhibition (UKOUG) to start. The titular Doug Burns of Doug’s Oracle Blog posts his personal agenda for the UKOUG. It’s interesting to me to see how busy this business of conference participation really is — three-and-a-half days crammed with information to take in! That’s overload. Doug, your last agenda item is very apt.
Incidentally, Pythian’s own Alex Gorbachev and Babette Turner-Underwood will be making presentations at UKOUG. If you’re there too, say hello, they’re really quite nice.
Oracle + Open Source has a list of some other Oracle-related events in the New York area.
Now for some more Oracle Open World aftermath.
Back on that Tooting blog, Arfur C. mentions the availability online (and the non-availability of some) of the presentations from Oracle Open World.
On ultramookie (currently tied with Radio Free Tooting in the Dave’s Favourite Blog Title showdown), Steve Kong has some points critical of the Oracle Linux thing. For instance: “Support Open Source. If you choose not to fork, then why not put some money, hardware, bandwidth and resources behind an good project? Try CentOS.”
TheOpenForce.com blog likewise takes a dim view of it: “Decisions are made by one guy at Oracle (much to the chagrin of many of his executives) and if Larry decides he wants to fork with Red Hat, then he tells his guys to get it done. So what do they do? They post CentOS for download on the Oracle web site and call it their own. And as part of that, Oracle is now distributing MySQL, which is ironic, but nonetheless appreciated.”
Ah, a segué! On the So What Co-operative, Jeff Hunter has a short piece about where he sees MySQL fitting in his enterprise: “In my opinion, I’d use MySQL for anything but the most mission critical applications. … (An) area where MySQL excels in serving database driven content directly on the webserver. … (When) you want an 80% solution, I think it’s the right choice.”
If that sounds to you like damning with faint praise, Joshua Drake’s PostgreSQL Blog takes it a step further. Alluding to Slashdot’s recent problems, Joshua suggests that their choice of MySQL as the DB backend could be improved on in a certain way: “Note that I am not bashing Slashdot. Instead I am noticing a place where they could provide better service to their readers. The recent migration to PostgreSQL 8.1 by Sourceforge … has proven a huge boon in performance.” Lookout! The gloves are off now! It’s almost as though there’s some kind of rivalry or competition between these two open source DBMSs.
Peter Zaitsev of the MySQL Performance Blog has a similar story, Undo area size restriction needed for InnoDB. “…MySQL Server was restarted … and (spent) hours to undo almost 300.000.000 of row operations being unavailable during all of this time. … It would be great…to add innodb_max_transaction_undo_rows or something similar which will protect from runaway transactions.”
The So What Co-op also has a recipe for a quick and dirty MySQL — with InnoDB — backup.
The mySQL DBA blog has a brief item on I/O schedulers in the Linux kernel, and how they affect MySQL/InnoDB performance: “Hands down the Deadline I/O scheduler is the best for INNODB traffic or RANDOM IO.”
If you aren’t already at MySQL Camp by now, you probably won’t make it, since it begins today. Nonetheless, here’s Jay Pipes’s last pre-camp pitch. If you are already at MySQL Camp… well… don’t forget to chew your food. And write your Mum!
On TheOpenForce.com, Zack Urlocker points to an article by MySQL AB’s CEO, Marten Mickos, outlining what he sees as the thirteen open source-hybrid business models. MySQL’s? “Software is free but if you embed it in closed source, you better pay a fee.”
Technical Notes and Articles of Interest offers a list of slogans for MySQL. To be honest, I’m not wild about any of them. I’m sure Log Buffer readers can come up with some better ones and leave them in the comments there. Maybe something like use the_fastest; select "MySQL" from DBMSs; ?
Bob Field of Bob’s World writes about editing a (MySQL) database with sed. Yes, that sed — good old unix-y sed. It’s so crazy it just might work… and it does work! I love little hacks like this.
Sean McCown tells a story of being virtually mixed-up, on Database Underground. He concludes that, “Virtual instances of DBs should be kept to dev and or test and QA. It’s also a good idea to use them in DR sites where you’d have to keep up and running during an outage, but not for long term processing.”
On his data manipulation for fun and profit, Adam Machanic announces the birth of SQLQueryStress, a simple query load tool for SQL Server: “It is not intended to replace tools such as Visual Studio Team System’s load tests, but rather to be a simple and easy-to-use tool in the DBA or database developer’s kit.”
On An Expert’s Guide to DB2 Technology, Chris Eaton has two posts on pivot queries: one on converting rows to columns; the other on, yes, converting columns to rows. It’s DB2-specific of course. Whenever I read examples like this, I always want to compare them with how the other DBMSs do them.
That’s all for this week. Look forward to Log Buffer #19, edited and published by Jay Pipes, fresh (or otherwise) from his adventures at MySQL Camp, on his Design, Develop, Discover, Define.
No comments