Log Buffer #162: A Carnival of the Vanities for DBAs

Sep 18, 2009 / By David Edwards

Tags: , , , , , , , ,

Welcome to the 162nd edition of Log Buffer, the weekly review of database blogs.

Oracle

The big news this week came was Oracle’s unveiling the OLTP Oracle Database Machine & Exadata v2, as reported by Alex Gorbachev.

Kevin Closson covered it, of course: Oracle Drops Exadata In Favor of Sun FlashFire Based OLTP Database Machine?, and he and his readers kick it around in a diverting way.

The dbaStreet blog offered their very thorough HOWTO, 11gR2 rac installation on 64 bit Linux step by step.

Iggy Fernandez was questioning Method R, in an excellent interview with Cary Millsap.

If that’s not enough “R” for you, there’s another item in the “Q” — Cary himself on the importance of diagnosing before resolving.

The ORACLE-BASE Blog observed that the RELIES ON clause is no more. To which Tom Kyte responded, every day I learn something new, which also links to Tom’s podcast on 11g Release 2 “things”.

SQL Server

The Rusano blog shared their thoughts on SQL Server’s boolean operator short-circuit. “Many developers that come from an imperative language background like C are relying on boolean short-circuit to occur when SQL queries are executed.  . . . I’ve stayed close enough to Microsoft CSS front lines for 6 months to see actual cases pouring in from developers bitten by the short-circuit assumption.”

Perhaps these developers could use a good book. On Dave does Data appeared a review of MS SQL Server 2008 T-SQL Fundamentals — according to Dave, ” . . . a must have for most people.”

Michelle Ufford shared some T-SQL of her own, sharing a script that creates a monitoring process for performance counters.

John Paul Cook offered his script to create all primary keys, ” . . . a script that generates a script to drop or create all of the existing primary keys in a database.”

Ben Nevarez was into the statistics, publishing the second part of his series, How the Query Optimizer Uses Statistics.

Paul S. Randal was also in a statistical frame of mind this week, as reflects his latest survey: how do you create and maintain statistics? (And are they better than lies and damn lies?)

MySQL

The MySQL Performance Blog announced a new statistical bauble: statistics of InnoDB tables and indexes available in xtrabackup.

Porting a web site from MS SQL to MySQL is the second part of a series from Tom Hotchkiss, a.k.a. Mr Stapler.

Shlomi Noach shared some of his thoughts on ranking without self join in SQL, which he introduces thus: “The common way of solving the classic SQL problem of ranking, involves a self join. I wish to present a different solution, which only iterates the table once, and provides the same output.”

Roland Bouman responded with another ranking trick: “Shlomi’s trick reminds me somewhat of the trick I came across little over a year ago to caclulate percentiles. At that time, several people pointed out to me too that using user-defined variables in this way can be unreliable.”

Those MySQL@Facebook people say, “Most of the tests in the MySQL regression test suite are hardwired to use MyISAM or InnoDB. This has made it difficult for new storage engines to run these tests. PBXT has spent too much time modifying these tests and maintaining a huge patch for that modification. Guess what? Now we all need the storage-engine independent test suite.” (Google cache.)

Ronald Bradford responded with his item on engine-agnostic MySQL test cases: “I could not agree more.  . . .  There is however a way to do it with the current mysql-test syntax.”

That’s all for now. Please share your favourite DB blogs from this week in the comments. See you in a week!

Leave a Reply

  • (will not be published)

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>