Log Buffer #136: A Carnival of the Vanities for DBAs
Feb 20, 2009 / By David Edwards
This is the 136th edition of Log Buffer, the weekly review of database blogs. Welcome.
Let’s start with the bad news first. The relational database is doomed. Or is it? That is the line of inquiry taken by the item by Tony Bain on ReadWriteWeb. “Recently, a lot of new non-relational databases have cropped up both inside and outside the cloud. One key message this sends is, ‘if you want vast, on-demand scalability, you need a non-relational database’. . . . Is this a sign that relational databases have had their day and will decline over time?”
Whether or not you believe that your beloved relational DB’s days are numbered, you will be glad to hear that it ain’t necessarily so that DBA work is doomed to get boring. So explains Chen Shapira, illustrating her thesis with a sophisticated graph.
Eddie Awad’s got the same idea. He’s got Stephane Faroult’s Oracle Database Architecture in Less than 10 Minutes video. Maybe the point & click DBAs lamented by Jeff Hunter of the So What Co-operative should watch, to begin to get to grips with, “the fundamental basics,” as Doug Burns says.
Cary Millsap responds to what Doug said in his own blog on the matter of throughput versus response time. Here’s a preview: “ . . . you can’t know whether a system is optimal without knowing whether its tasks are efficient, and you can’t know whether a given task is efficient without profiling it.”
Christian Antognini is always Striving for Optimal Performance, and this week he found that virtual column-based partitioning in 11g might lead to wrong results.
The OptimalDBA (by name, Daniel Fink) proffers some criteria in answer to the question, when is a SQL statement too long?
Anders Karlsson delves into the mystery of what really is a valid date. “Most RDBMS systems aren’t as forgiving as MySQL, so trying to insert a date in a DATE column like this ’2009-02-30′ will cause an error with, say Oracle. . . . [Is] 2009-02-30 a valid date or not? Look in your calendar, and you see that it’s not. On the other hand, ask someone who work with [financial] instruments, and the answer will be maybe!”
On his MySQL in Communications blog, Henrik Ingo responds to Anders with his own post on date-handling in MySQL, which he begins, “With dates, just as in many other cases, MySQL is very flexible in how you can input data. . . . MySQL does *not* by default allow you to enter a non-existant date, e.g. ’31st of February 2009′. [This] can be enabled with an SQL_MODE setting . . . ”
Matt Yonkovit, a Big DBA Head, has a quick one on MySQL Memory allocation & TMPDIR. He writes, “I wanted to remind folks to not forget about allocating memory to a tmpfs for the tmpdir. Certainly this is not going to help everyone, but those who create lots of temporary table will find the performance boost most welcome.”
Shlomi Noach offers his HOWTO on manually installing multiple MySQL instances on Linux. He says, “Installing a single MySQL instance on a linux machine is a very simple operation. It may be as simple as: apt-get install mysql-server . . . But you cannot use this method to install another MySQL instance. Moreover, if you try to manually install another instance, you may find that some collisions occur.”
Let’s have some SQL Server blogs now. James Luetkehoelter waded into a discussion between Brent Ozar and Brian Kelly on Specialization vs Generalization. James writes, “The basic argument going on is that you if you want to be successful in the future, you should: 1) Specialize in a specific area or 2) Generalize and gain as much exposer to as many technologies as possible[.] . . . My own view is that both viewpoints are incorrect. The key to maintaining your marketability in the future is being both a generalist AND a specialist.”
Adam Machanic has released Who is Active? v7.30, “ . . . a comprehensive DMV-based monitoring script, designed to tell you at a glance what processes are active on your SQL Server and what they’re up to.”
Louis Davidson has some thoughts on business rule enforcement. “I starting thinking that there needed to be some standard way to describe the different levels of business rule enforcement. For example[,] Hard and fast . . . Abnormal range . . . Approval range . . . The question then becomes how to implement these rules. And this is where the battle begins to get ugly doesnâ€™t it? Business rules in the database, the horrors!” Interesting stuff, and a good discussion, too.
The Postgres OnLine Journal provides a detailed HOWTO on using Microsoft SQL Server to Update PostgreSQL data.
Database performance monitoring on the cheap is delivered (stealthily, I assume) by Scott Herbert, the SQL Ninja, complete with a stored procedure and a SQL Server Agent job. “Awesome stuff,” to quote one of the readers.
Speaking of SQL Agent, SQL Batman reports on some SQL Agent Weird Science. He compares an experiment with a standalone SQL Agent to . . . well, as Brent Ozar said on his own blog, “it could be subtitled the SQL Agent Monologues.” If not weird, at least very unexpected!
And that is that. I must thank my colleague Tim Procter for his help this week. Also, I’m happy to say that we can look forward to Sheeri Cabral handling Log Buffer next week. Don’t forget that LB is open to your contributions—leave a comment on your favourite blog items from this week, or publish an edition on your own blog. The first step—send me an email.
Till next time!